{
  "best_global_step": 15264,
  "best_metric": 0.4831865429878235,
  "best_model_checkpoint": "saves_multiple/p-tuning/llama-3-8b-instruct/train_codealpacapy_101112_1770438262/checkpoint-15264",
  "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": 45.76319885253906,
      "learning_rate": 1.0482180293501048e-06,
      "loss": 5.087,
      "num_input_tokens_seen": 3744,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 23.291706085205078,
      "learning_rate": 2.358490566037736e-06,
      "loss": 4.7413,
      "num_input_tokens_seen": 7488,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 23.57452392578125,
      "learning_rate": 3.668763102725367e-06,
      "loss": 4.8611,
      "num_input_tokens_seen": 10464,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 56.69627380371094,
      "learning_rate": 4.979035639412998e-06,
      "loss": 4.1091,
      "num_input_tokens_seen": 13536,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 72.90431213378906,
      "learning_rate": 6.289308176100629e-06,
      "loss": 2.8262,
      "num_input_tokens_seen": 18176,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 30.26812744140625,
      "learning_rate": 7.59958071278826e-06,
      "loss": 1.9392,
      "num_input_tokens_seen": 21184,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 10.97499942779541,
      "learning_rate": 8.90985324947589e-06,
      "loss": 1.1864,
      "num_input_tokens_seen": 23968,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 13.012429237365723,
      "learning_rate": 1.0220125786163522e-05,
      "loss": 0.8954,
      "num_input_tokens_seen": 27712,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 22.863452911376953,
      "learning_rate": 1.1530398322851153e-05,
      "loss": 0.974,
      "num_input_tokens_seen": 30592,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 9.223089218139648,
      "learning_rate": 1.2840670859538784e-05,
      "loss": 0.8576,
      "num_input_tokens_seen": 33440,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 12.941582679748535,
      "learning_rate": 1.4150943396226415e-05,
      "loss": 1.0613,
      "num_input_tokens_seen": 37344,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 3.8042216300964355,
      "learning_rate": 1.5461215932914046e-05,
      "loss": 0.8696,
      "num_input_tokens_seen": 40992,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 1.7685428857803345,
      "learning_rate": 1.6771488469601677e-05,
      "loss": 0.6031,
      "num_input_tokens_seen": 43648,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 1.8687872886657715,
      "learning_rate": 1.8081761006289308e-05,
      "loss": 0.6284,
      "num_input_tokens_seen": 46976,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.227837085723877,
      "learning_rate": 1.9392033542976942e-05,
      "loss": 0.6625,
      "num_input_tokens_seen": 50496,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 1.3158693313598633,
      "learning_rate": 2.070230607966457e-05,
      "loss": 0.6551,
      "num_input_tokens_seen": 53504,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 2.6529858112335205,
      "learning_rate": 2.20125786163522e-05,
      "loss": 0.6627,
      "num_input_tokens_seen": 56032,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 13.987693786621094,
      "learning_rate": 2.3322851153039832e-05,
      "loss": 0.6161,
      "num_input_tokens_seen": 59392,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 1.6513252258300781,
      "learning_rate": 2.4633123689727463e-05,
      "loss": 0.5914,
      "num_input_tokens_seen": 62944,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 1.7314578294754028,
      "learning_rate": 2.5943396226415097e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 66432,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 32.047767639160156,
      "learning_rate": 2.7253668763102725e-05,
      "loss": 0.7286,
      "num_input_tokens_seen": 69280,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 1.6046741008758545,
      "learning_rate": 2.8563941299790356e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 73088,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 2.387032985687256,
      "learning_rate": 2.987421383647799e-05,
      "loss": 0.7644,
      "num_input_tokens_seen": 75840,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 1.9109432697296143,
      "learning_rate": 3.118448637316562e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 79008,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 1.5800665616989136,
      "learning_rate": 3.2494758909853245e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 81856,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.2261016368865967,
      "learning_rate": 3.380503144654088e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 85120,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 2.081563949584961,
      "learning_rate": 3.5115303983228514e-05,
      "loss": 0.6883,
      "num_input_tokens_seen": 88032,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.3736793994903564,
      "learning_rate": 3.642557651991614e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 91104,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 1.7862334251403809,
      "learning_rate": 3.7735849056603776e-05,
      "loss": 0.5975,
      "num_input_tokens_seen": 93984,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 2.1574692726135254,
      "learning_rate": 3.90461215932914e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 96864,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 0.6629489660263062,
      "learning_rate": 4.035639412997904e-05,
      "loss": 0.5665,
      "num_input_tokens_seen": 99648,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 1.39899742603302,
      "learning_rate": 4.1666666666666665e-05,
      "loss": 1.0498,
      "num_input_tokens_seen": 102912,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 1.6714979410171509,
      "learning_rate": 4.29769392033543e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 105568,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 2.73334002494812,
      "learning_rate": 4.4287211740041934e-05,
      "loss": 0.6695,
      "num_input_tokens_seen": 108352,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 2.4047203063964844,
      "learning_rate": 4.5597484276729555e-05,
      "loss": 0.6232,
      "num_input_tokens_seen": 110976,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.8038594722747803,
      "learning_rate": 4.690775681341719e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 114176,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 3.272678852081299,
      "learning_rate": 4.8218029350104823e-05,
      "loss": 0.5218,
      "num_input_tokens_seen": 117056,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 1.33505380153656,
      "learning_rate": 4.952830188679246e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 120384,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 2.102494239807129,
      "learning_rate": 5.0838574423480085e-05,
      "loss": 0.6016,
      "num_input_tokens_seen": 123456,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 2.3866143226623535,
      "learning_rate": 5.214884696016771e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 125984,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.1522237062454224,
      "learning_rate": 5.345911949685535e-05,
      "loss": 0.7034,
      "num_input_tokens_seen": 129280,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.901869535446167,
      "learning_rate": 5.4769392033542975e-05,
      "loss": 0.6468,
      "num_input_tokens_seen": 132096,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 0.7420336604118347,
      "learning_rate": 5.607966457023061e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 134880,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 1.2897765636444092,
      "learning_rate": 5.7389937106918244e-05,
      "loss": 0.5866,
      "num_input_tokens_seen": 137824,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 0.6032651662826538,
      "learning_rate": 5.870020964360587e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 140832,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 0.7148796916007996,
      "learning_rate": 6.00104821802935e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 144704,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 1.2023413181304932,
      "learning_rate": 6.132075471698113e-05,
      "loss": 0.5782,
      "num_input_tokens_seen": 147552,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 1.4720172882080078,
      "learning_rate": 6.263102725366875e-05,
      "loss": 0.6854,
      "num_input_tokens_seen": 150912,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 4.490001678466797,
      "learning_rate": 6.39412997903564e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 153824,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 0.6488051414489746,
      "learning_rate": 6.525157232704402e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 156928,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 1.757938265800476,
      "learning_rate": 6.656184486373166e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 160928,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 0.6757560968399048,
      "learning_rate": 6.787211740041929e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 164096,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.2472119331359863,
      "learning_rate": 6.918238993710691e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 167264,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 0.7289978861808777,
      "learning_rate": 7.049266247379455e-05,
      "loss": 0.5757,
      "num_input_tokens_seen": 170400,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 1.017259955406189,
      "learning_rate": 7.180293501048218e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 173984,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 0.8040484189987183,
      "learning_rate": 7.311320754716982e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 176704,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 0.7746554613113403,
      "learning_rate": 7.442348008385745e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 179616,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 0.3711369037628174,
      "learning_rate": 7.573375262054507e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 182624,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 0.5616642832756042,
      "learning_rate": 7.70440251572327e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 185600,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 0.36720752716064453,
      "learning_rate": 7.835429769392034e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 189728,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 1.0315555334091187,
      "learning_rate": 7.966457023060797e-05,
      "loss": 0.6022,
      "num_input_tokens_seen": 193056,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 2.67647647857666,
      "learning_rate": 8.09748427672956e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 196384,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 0.9196131229400635,
      "learning_rate": 8.228511530398323e-05,
      "loss": 0.5828,
      "num_input_tokens_seen": 200224,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 0.9402783513069153,
      "learning_rate": 8.359538784067086e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 203584,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.479735255241394,
      "learning_rate": 8.490566037735848e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 206208,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.550528883934021,
      "learning_rate": 8.621593291404613e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 210048,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 0.931586503982544,
      "learning_rate": 8.752620545073375e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 213472,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 1.0299196243286133,
      "learning_rate": 8.883647798742137e-05,
      "loss": 0.6853,
      "num_input_tokens_seen": 217312,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 1.0076502561569214,
      "learning_rate": 9.014675052410902e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 220224,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 1.2540391683578491,
      "learning_rate": 9.145702306079664e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 223648,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 1.281694769859314,
      "learning_rate": 9.276729559748428e-05,
      "loss": 0.5687,
      "num_input_tokens_seen": 226336,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.6328620910644531,
      "learning_rate": 9.407756813417191e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 228832,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 0.9139071702957153,
      "learning_rate": 9.538784067085953e-05,
      "loss": 0.7019,
      "num_input_tokens_seen": 231968,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 0.46392232179641724,
      "learning_rate": 9.669811320754718e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 236000,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 0.5050573945045471,
      "learning_rate": 9.80083857442348e-05,
      "loss": 0.7747,
      "num_input_tokens_seen": 239360,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 0.8157622218132019,
      "learning_rate": 9.931865828092243e-05,
      "loss": 0.6216,
      "num_input_tokens_seen": 243168,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 0.46182167530059814,
      "learning_rate": 0.00010062893081761007,
      "loss": 0.4984,
      "num_input_tokens_seen": 247168,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.3443448841571808,
      "learning_rate": 0.0001019392033542977,
      "loss": 0.5427,
      "num_input_tokens_seen": 250400,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.5082590579986572,
      "learning_rate": 0.00010324947589098532,
      "loss": 0.4951,
      "num_input_tokens_seen": 253376,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 0.834794282913208,
      "learning_rate": 0.00010455974842767296,
      "loss": 0.687,
      "num_input_tokens_seen": 256448,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 1.9461393356323242,
      "learning_rate": 0.00010587002096436059,
      "loss": 0.6679,
      "num_input_tokens_seen": 260384,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 0.4176793694496155,
      "learning_rate": 0.00010718029350104821,
      "loss": 0.6014,
      "num_input_tokens_seen": 263264,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 0.3994807004928589,
      "learning_rate": 0.00010849056603773586,
      "loss": 0.4912,
      "num_input_tokens_seen": 266400,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 0.8010081648826599,
      "learning_rate": 0.00010980083857442348,
      "loss": 0.462,
      "num_input_tokens_seen": 269600,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 1.4145066738128662,
      "learning_rate": 0.0001111111111111111,
      "loss": 0.4315,
      "num_input_tokens_seen": 272992,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 2.3304991722106934,
      "learning_rate": 0.00011242138364779875,
      "loss": 0.5666,
      "num_input_tokens_seen": 277056,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 0.4269068241119385,
      "learning_rate": 0.00011373165618448637,
      "loss": 0.553,
      "num_input_tokens_seen": 279968,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 0.6909087300300598,
      "learning_rate": 0.00011504192872117402,
      "loss": 0.4713,
      "num_input_tokens_seen": 282624,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 0.6169834136962891,
      "learning_rate": 0.00011635220125786164,
      "loss": 0.5849,
      "num_input_tokens_seen": 285248,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 1.2057931423187256,
      "learning_rate": 0.00011766247379454926,
      "loss": 0.5078,
      "num_input_tokens_seen": 287648,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 0.6943414211273193,
      "learning_rate": 0.00011897274633123691,
      "loss": 0.8014,
      "num_input_tokens_seen": 290048,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 0.5729154944419861,
      "learning_rate": 0.00012028301886792453,
      "loss": 0.6081,
      "num_input_tokens_seen": 293024,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 0.6483250856399536,
      "learning_rate": 0.00012159329140461216,
      "loss": 0.5142,
      "num_input_tokens_seen": 296960,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 0.43779894709587097,
      "learning_rate": 0.0001229035639412998,
      "loss": 0.4685,
      "num_input_tokens_seen": 299936,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 0.40932586789131165,
      "learning_rate": 0.00012421383647798743,
      "loss": 0.4175,
      "num_input_tokens_seen": 302976,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 0.47719433903694153,
      "learning_rate": 0.00012552410901467507,
      "loss": 0.5388,
      "num_input_tokens_seen": 305792,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 0.364115834236145,
      "learning_rate": 0.00012683438155136267,
      "loss": 0.6229,
      "num_input_tokens_seen": 312960,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 0.6919863820075989,
      "learning_rate": 0.0001281446540880503,
      "loss": 0.6136,
      "num_input_tokens_seen": 316832,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 0.6887328028678894,
      "learning_rate": 0.00012945492662473794,
      "loss": 0.497,
      "num_input_tokens_seen": 319968,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 0.6684688329696655,
      "learning_rate": 0.00013076519916142558,
      "loss": 0.4552,
      "num_input_tokens_seen": 322816,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 0.5425750017166138,
      "learning_rate": 0.0001320754716981132,
      "loss": 0.6453,
      "num_input_tokens_seen": 325664,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 2.849306583404541,
      "learning_rate": 0.00013338574423480085,
      "loss": 0.5846,
      "num_input_tokens_seen": 328960,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 0.26500949263572693,
      "learning_rate": 0.00013469601677148845,
      "loss": 0.4266,
      "num_input_tokens_seen": 332608,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 0.47750523686408997,
      "learning_rate": 0.0001360062893081761,
      "loss": 0.7158,
      "num_input_tokens_seen": 335552,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 0.45497554540634155,
      "learning_rate": 0.00013731656184486375,
      "loss": 0.5216,
      "num_input_tokens_seen": 340256,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.41132524609565735,
      "learning_rate": 0.00013862683438155136,
      "loss": 0.5294,
      "num_input_tokens_seen": 342848,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 0.5809541940689087,
      "learning_rate": 0.000139937106918239,
      "loss": 0.5951,
      "num_input_tokens_seen": 345184,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 1.4757802486419678,
      "learning_rate": 0.00014124737945492663,
      "loss": 0.5517,
      "num_input_tokens_seen": 348608,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 1.2927699089050293,
      "learning_rate": 0.00014255765199161423,
      "loss": 0.4835,
      "num_input_tokens_seen": 351648,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 0.9316359758377075,
      "learning_rate": 0.0001438679245283019,
      "loss": 0.6725,
      "num_input_tokens_seen": 354464,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 0.27764707803726196,
      "learning_rate": 0.00014517819706498953,
      "loss": 0.5991,
      "num_input_tokens_seen": 358240,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 0.43261703848838806,
      "learning_rate": 0.00014648846960167716,
      "loss": 0.4813,
      "num_input_tokens_seen": 361600,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 0.28743383288383484,
      "learning_rate": 0.00014779874213836477,
      "loss": 0.4577,
      "num_input_tokens_seen": 364448,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 0.6802446842193604,
      "learning_rate": 0.00014910901467505243,
      "loss": 0.4637,
      "num_input_tokens_seen": 367424,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 0.5826802253723145,
      "learning_rate": 0.00015041928721174007,
      "loss": 0.5578,
      "num_input_tokens_seen": 370112,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 0.6788904666900635,
      "learning_rate": 0.00015172955974842767,
      "loss": 0.6093,
      "num_input_tokens_seen": 372608,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 0.8397895693778992,
      "learning_rate": 0.0001530398322851153,
      "loss": 0.5404,
      "num_input_tokens_seen": 375584,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 0.41244396567344666,
      "learning_rate": 0.00015435010482180294,
      "loss": 0.5763,
      "num_input_tokens_seen": 378752,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 0.4744315445423126,
      "learning_rate": 0.00015566037735849058,
      "loss": 0.4131,
      "num_input_tokens_seen": 382432,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 0.5054666996002197,
      "learning_rate": 0.0001569706498951782,
      "loss": 0.5704,
      "num_input_tokens_seen": 385344,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 0.40355902910232544,
      "learning_rate": 0.00015828092243186584,
      "loss": 0.4817,
      "num_input_tokens_seen": 388832,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 0.47708660364151,
      "learning_rate": 0.00015959119496855345,
      "loss": 0.5249,
      "num_input_tokens_seen": 392640,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 0.7077701687812805,
      "learning_rate": 0.00016090146750524109,
      "loss": 0.564,
      "num_input_tokens_seen": 397568,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 0.22738002240657806,
      "learning_rate": 0.00016221174004192875,
      "loss": 0.4337,
      "num_input_tokens_seen": 400512,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 0.6836808323860168,
      "learning_rate": 0.00016352201257861635,
      "loss": 0.4983,
      "num_input_tokens_seen": 403520,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.5060572028160095,
      "learning_rate": 0.000164832285115304,
      "loss": 0.4935,
      "num_input_tokens_seen": 407072,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 0.2824517786502838,
      "learning_rate": 0.00016614255765199162,
      "loss": 0.5046,
      "num_input_tokens_seen": 411872,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 0.7582754492759705,
      "learning_rate": 0.00016745283018867923,
      "loss": 0.4818,
      "num_input_tokens_seen": 414784,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 0.417634516954422,
      "learning_rate": 0.0001687631027253669,
      "loss": 0.5976,
      "num_input_tokens_seen": 418656,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 0.6714122295379639,
      "learning_rate": 0.00017007337526205453,
      "loss": 0.6086,
      "num_input_tokens_seen": 421824,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 0.5402345657348633,
      "learning_rate": 0.00017138364779874213,
      "loss": 0.6018,
      "num_input_tokens_seen": 425632,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 0.5077677965164185,
      "learning_rate": 0.00017269392033542977,
      "loss": 0.4711,
      "num_input_tokens_seen": 428864,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 0.7789681553840637,
      "learning_rate": 0.0001740041928721174,
      "loss": 0.402,
      "num_input_tokens_seen": 431104,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 0.4113891124725342,
      "learning_rate": 0.00017531446540880504,
      "loss": 0.5488,
      "num_input_tokens_seen": 434240,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 0.25311338901519775,
      "learning_rate": 0.00017662473794549267,
      "loss": 0.5394,
      "num_input_tokens_seen": 437792,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 0.3483620285987854,
      "learning_rate": 0.0001779350104821803,
      "loss": 0.6147,
      "num_input_tokens_seen": 441280,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 0.29476553201675415,
      "learning_rate": 0.0001792452830188679,
      "loss": 0.4846,
      "num_input_tokens_seen": 445120,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 0.22767210006713867,
      "learning_rate": 0.00018055555555555555,
      "loss": 0.483,
      "num_input_tokens_seen": 449600,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 0.30313536524772644,
      "learning_rate": 0.0001818658280922432,
      "loss": 0.4768,
      "num_input_tokens_seen": 453120,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 0.4735591411590576,
      "learning_rate": 0.00018317610062893082,
      "loss": 0.5398,
      "num_input_tokens_seen": 457216,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 0.3704434037208557,
      "learning_rate": 0.00018448637316561845,
      "loss": 0.4478,
      "num_input_tokens_seen": 461152,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 0.3609219193458557,
      "learning_rate": 0.00018579664570230608,
      "loss": 0.6427,
      "num_input_tokens_seen": 464480,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 0.4563707411289215,
      "learning_rate": 0.0001871069182389937,
      "loss": 0.525,
      "num_input_tokens_seen": 467744,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 5.121705055236816,
      "learning_rate": 0.00018841719077568135,
      "loss": 0.4655,
      "num_input_tokens_seen": 470656,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 0.6383441090583801,
      "learning_rate": 0.000189727463312369,
      "loss": 0.4892,
      "num_input_tokens_seen": 474048,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 0.5397881269454956,
      "learning_rate": 0.00019103773584905662,
      "loss": 0.5115,
      "num_input_tokens_seen": 477184,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 0.3820585012435913,
      "learning_rate": 0.00019234800838574423,
      "loss": 0.4589,
      "num_input_tokens_seen": 480160,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 0.44287750124931335,
      "learning_rate": 0.00019365828092243186,
      "loss": 0.3785,
      "num_input_tokens_seen": 486144,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 0.29693353176116943,
      "learning_rate": 0.00019496855345911953,
      "loss": 0.5479,
      "num_input_tokens_seen": 489920,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.5777443647384644,
      "learning_rate": 0.00019627882599580713,
      "loss": 0.4804,
      "num_input_tokens_seen": 493728,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 0.6860224604606628,
      "learning_rate": 0.00019758909853249477,
      "loss": 0.502,
      "num_input_tokens_seen": 496416,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 0.20663954317569733,
      "learning_rate": 0.0001988993710691824,
      "loss": 0.503,
      "num_input_tokens_seen": 500320,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 0.21799811720848083,
      "learning_rate": 0.00020020964360587,
      "loss": 0.4651,
      "num_input_tokens_seen": 503520,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 0.7602450251579285,
      "learning_rate": 0.00020151991614255767,
      "loss": 0.4948,
      "num_input_tokens_seen": 506656,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 0.41531744599342346,
      "learning_rate": 0.0002028301886792453,
      "loss": 0.545,
      "num_input_tokens_seen": 509568,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 0.7547088265419006,
      "learning_rate": 0.0002041404612159329,
      "loss": 0.4408,
      "num_input_tokens_seen": 512160,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.5219701528549194,
      "learning_rate": 0.00020545073375262055,
      "loss": 0.4051,
      "num_input_tokens_seen": 515872,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 0.35079020261764526,
      "learning_rate": 0.00020676100628930818,
      "loss": 0.5233,
      "num_input_tokens_seen": 519232,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.2940330505371094,
      "learning_rate": 0.00020807127882599581,
      "loss": 0.343,
      "num_input_tokens_seen": 522464,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 0.34664615988731384,
      "learning_rate": 0.00020938155136268345,
      "loss": 0.5028,
      "num_input_tokens_seen": 525664,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 0.26147520542144775,
      "learning_rate": 0.00021069182389937108,
      "loss": 0.6085,
      "num_input_tokens_seen": 529216,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 0.18804827332496643,
      "learning_rate": 0.0002120020964360587,
      "loss": 0.5685,
      "num_input_tokens_seen": 533376,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 0.43761274218559265,
      "learning_rate": 0.00021331236897274632,
      "loss": 0.5844,
      "num_input_tokens_seen": 536192,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 1.2079479694366455,
      "learning_rate": 0.00021462264150943399,
      "loss": 0.472,
      "num_input_tokens_seen": 539392,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 0.6941171288490295,
      "learning_rate": 0.0002159329140461216,
      "loss": 0.4573,
      "num_input_tokens_seen": 542656,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 0.5952037572860718,
      "learning_rate": 0.00021724318658280923,
      "loss": 0.2961,
      "num_input_tokens_seen": 548064,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 0.2720472812652588,
      "learning_rate": 0.00021855345911949686,
      "loss": 0.5164,
      "num_input_tokens_seen": 551104,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 0.28987276554107666,
      "learning_rate": 0.00021986373165618447,
      "loss": 0.7053,
      "num_input_tokens_seen": 555680,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 0.3203318417072296,
      "learning_rate": 0.00022117400419287213,
      "loss": 0.3938,
      "num_input_tokens_seen": 558144,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 0.5189741253852844,
      "learning_rate": 0.00022248427672955977,
      "loss": 0.5376,
      "num_input_tokens_seen": 560384,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 0.5677109360694885,
      "learning_rate": 0.00022379454926624737,
      "loss": 0.502,
      "num_input_tokens_seen": 563072,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 0.6101126074790955,
      "learning_rate": 0.000225104821802935,
      "loss": 0.5916,
      "num_input_tokens_seen": 565792,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 0.49277263879776,
      "learning_rate": 0.00022641509433962264,
      "loss": 0.3766,
      "num_input_tokens_seen": 568160,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 0.30723464488983154,
      "learning_rate": 0.00022772536687631028,
      "loss": 0.5778,
      "num_input_tokens_seen": 571584,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 0.21482914686203003,
      "learning_rate": 0.0002290356394129979,
      "loss": 0.4738,
      "num_input_tokens_seen": 576288,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 0.2245420664548874,
      "learning_rate": 0.00023034591194968554,
      "loss": 0.5048,
      "num_input_tokens_seen": 579456,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 0.2372996211051941,
      "learning_rate": 0.00023165618448637318,
      "loss": 0.6095,
      "num_input_tokens_seen": 583680,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 0.3550943434238434,
      "learning_rate": 0.00023296645702306079,
      "loss": 0.5889,
      "num_input_tokens_seen": 587040,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 2.544424057006836,
      "learning_rate": 0.00023427672955974845,
      "loss": 0.3633,
      "num_input_tokens_seen": 589952,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 0.3836802542209625,
      "learning_rate": 0.00023558700209643608,
      "loss": 0.5763,
      "num_input_tokens_seen": 593600,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 0.34062227606773376,
      "learning_rate": 0.0002368972746331237,
      "loss": 0.4922,
      "num_input_tokens_seen": 597216,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 0.6059651970863342,
      "learning_rate": 0.00023820754716981132,
      "loss": 0.6615,
      "num_input_tokens_seen": 600576,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 0.16177523136138916,
      "learning_rate": 0.00023951781970649896,
      "loss": 0.4774,
      "num_input_tokens_seen": 603936,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 0.2304563820362091,
      "learning_rate": 0.0002408280922431866,
      "loss": 0.5874,
      "num_input_tokens_seen": 606816,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.29322701692581177,
      "learning_rate": 0.00024213836477987423,
      "loss": 0.4446,
      "num_input_tokens_seen": 610112,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 0.3107600808143616,
      "learning_rate": 0.00024344863731656186,
      "loss": 0.4771,
      "num_input_tokens_seen": 613216,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.2034461349248886,
      "learning_rate": 0.00024475890985324947,
      "loss": 0.5798,
      "num_input_tokens_seen": 615904,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 0.4681790769100189,
      "learning_rate": 0.0002460691823899371,
      "loss": 0.6769,
      "num_input_tokens_seen": 619456,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 0.41960281133651733,
      "learning_rate": 0.00024737945492662474,
      "loss": 0.5178,
      "num_input_tokens_seen": 622656,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 0.20612487196922302,
      "learning_rate": 0.00024868972746331237,
      "loss": 0.5536,
      "num_input_tokens_seen": 626016,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 0.3740993142127991,
      "learning_rate": 0.00025,
      "loss": 0.4922,
      "num_input_tokens_seen": 629120,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 0.43691471219062805,
      "learning_rate": 0.00025131027253668764,
      "loss": 0.539,
      "num_input_tokens_seen": 632160,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 0.29898619651794434,
      "learning_rate": 0.0002526205450733753,
      "loss": 0.4622,
      "num_input_tokens_seen": 634784,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 0.25436311960220337,
      "learning_rate": 0.0002539308176100629,
      "loss": 0.5827,
      "num_input_tokens_seen": 637792,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 0.16960865259170532,
      "learning_rate": 0.0002552410901467505,
      "loss": 0.4433,
      "num_input_tokens_seen": 641312,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 0.4151897430419922,
      "learning_rate": 0.0002565513626834381,
      "loss": 0.5475,
      "num_input_tokens_seen": 644832,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 0.29678216576576233,
      "learning_rate": 0.0002578616352201258,
      "loss": 0.6165,
      "num_input_tokens_seen": 647904,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 0.42098554968833923,
      "learning_rate": 0.00025917190775681345,
      "loss": 0.5711,
      "num_input_tokens_seen": 650624,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 0.5828491449356079,
      "learning_rate": 0.0002604821802935011,
      "loss": 0.5352,
      "num_input_tokens_seen": 653248,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 0.4131849706172943,
      "learning_rate": 0.0002617924528301887,
      "loss": 0.567,
      "num_input_tokens_seen": 655936,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 0.19198574125766754,
      "learning_rate": 0.0002631027253668763,
      "loss": 0.438,
      "num_input_tokens_seen": 661888,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 0.3047693371772766,
      "learning_rate": 0.00026441299790356393,
      "loss": 0.5353,
      "num_input_tokens_seen": 664512,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 0.3623755872249603,
      "learning_rate": 0.00026572327044025156,
      "loss": 0.3974,
      "num_input_tokens_seen": 668096,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 0.39255309104919434,
      "learning_rate": 0.0002670335429769392,
      "loss": 0.7355,
      "num_input_tokens_seen": 671744,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 0.2331094741821289,
      "learning_rate": 0.00026834381551362683,
      "loss": 0.4028,
      "num_input_tokens_seen": 674144,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 0.7626714110374451,
      "learning_rate": 0.0002696540880503145,
      "loss": 0.7217,
      "num_input_tokens_seen": 677312,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 0.28120535612106323,
      "learning_rate": 0.0002709643605870021,
      "loss": 0.4206,
      "num_input_tokens_seen": 680256,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 0.7004536390304565,
      "learning_rate": 0.00027227463312368973,
      "loss": 0.4244,
      "num_input_tokens_seen": 683744,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 0.5438204407691956,
      "learning_rate": 0.00027358490566037737,
      "loss": 0.5124,
      "num_input_tokens_seen": 687616,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 0.3403388261795044,
      "learning_rate": 0.000274895178197065,
      "loss": 0.5329,
      "num_input_tokens_seen": 691392,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 0.17482174932956696,
      "learning_rate": 0.00027620545073375264,
      "loss": 0.5519,
      "num_input_tokens_seen": 695904,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.2665926218032837,
      "learning_rate": 0.00027751572327044027,
      "loss": 0.4415,
      "num_input_tokens_seen": 698496,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 2.53566575050354,
      "learning_rate": 0.00027882599580712785,
      "loss": 0.5836,
      "num_input_tokens_seen": 701344,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 0.4268903136253357,
      "learning_rate": 0.0002801362683438155,
      "loss": 0.5231,
      "num_input_tokens_seen": 703968,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 0.291405588388443,
      "learning_rate": 0.0002814465408805031,
      "loss": 0.4473,
      "num_input_tokens_seen": 707040,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.25359758734703064,
      "learning_rate": 0.0002827568134171908,
      "loss": 0.4794,
      "num_input_tokens_seen": 709280,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 0.15286608040332794,
      "learning_rate": 0.00028406708595387844,
      "loss": 0.4386,
      "num_input_tokens_seen": 713792,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.2947215735912323,
      "learning_rate": 0.0002853773584905661,
      "loss": 0.7812,
      "num_input_tokens_seen": 716480,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.28240224719047546,
      "learning_rate": 0.00028668763102725366,
      "loss": 0.476,
      "num_input_tokens_seen": 719136,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 0.18892069160938263,
      "learning_rate": 0.0002879979035639413,
      "loss": 0.446,
      "num_input_tokens_seen": 722752,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 0.23062339425086975,
      "learning_rate": 0.00028930817610062893,
      "loss": 0.6234,
      "num_input_tokens_seen": 725920,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 0.25291651487350464,
      "learning_rate": 0.00029061844863731656,
      "loss": 0.4598,
      "num_input_tokens_seen": 728736,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 0.17565785348415375,
      "learning_rate": 0.0002919287211740042,
      "loss": 0.5175,
      "num_input_tokens_seen": 731776,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 0.24734249711036682,
      "learning_rate": 0.00029323899371069183,
      "loss": 0.493,
      "num_input_tokens_seen": 734784,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 0.203591451048851,
      "learning_rate": 0.00029454926624737946,
      "loss": 0.5541,
      "num_input_tokens_seen": 737888,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 0.1606196165084839,
      "learning_rate": 0.0002958595387840671,
      "loss": 0.3686,
      "num_input_tokens_seen": 741824,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 0.39007025957107544,
      "learning_rate": 0.00029716981132075473,
      "loss": 0.4942,
      "num_input_tokens_seen": 744288,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 0.2888126075267792,
      "learning_rate": 0.00029848008385744237,
      "loss": 0.4932,
      "num_input_tokens_seen": 747648,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.2299097776412964,
      "learning_rate": 0.00029979035639413,
      "loss": 0.5235,
      "num_input_tokens_seen": 751168,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 0.24072030186653137,
      "learning_rate": 0.00030110062893081764,
      "loss": 0.4722,
      "num_input_tokens_seen": 754528,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 0.14319470524787903,
      "learning_rate": 0.00030241090146750527,
      "loss": 0.5785,
      "num_input_tokens_seen": 757152,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 0.274615079164505,
      "learning_rate": 0.00030372117400419285,
      "loss": 0.6139,
      "num_input_tokens_seen": 760064,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 0.29914379119873047,
      "learning_rate": 0.0003050314465408805,
      "loss": 0.4821,
      "num_input_tokens_seen": 762976,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 0.24664218723773956,
      "learning_rate": 0.0003063417190775681,
      "loss": 0.4645,
      "num_input_tokens_seen": 765600,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.32115742564201355,
      "learning_rate": 0.00030765199161425575,
      "loss": 0.5225,
      "num_input_tokens_seen": 768320,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.2984946072101593,
      "learning_rate": 0.00030896226415094344,
      "loss": 0.5035,
      "num_input_tokens_seen": 771392,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 0.2481124997138977,
      "learning_rate": 0.0003102725366876311,
      "loss": 0.4519,
      "num_input_tokens_seen": 774080,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 0.16027294099330902,
      "learning_rate": 0.00031158280922431866,
      "loss": 0.5028,
      "num_input_tokens_seen": 777184,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 0.41977372765541077,
      "learning_rate": 0.0003128930817610063,
      "loss": 0.5188,
      "num_input_tokens_seen": 780768,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 0.2396511286497116,
      "learning_rate": 0.0003142033542976939,
      "loss": 0.6431,
      "num_input_tokens_seen": 784128,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.4337347447872162,
      "learning_rate": 0.00031551362683438156,
      "loss": 0.5443,
      "num_input_tokens_seen": 787200,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 0.25382474064826965,
      "learning_rate": 0.0003168238993710692,
      "loss": 0.4501,
      "num_input_tokens_seen": 790176,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 0.4153142273426056,
      "learning_rate": 0.00031813417190775683,
      "loss": 0.5331,
      "num_input_tokens_seen": 794368,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 0.25574272871017456,
      "learning_rate": 0.0003194444444444444,
      "loss": 0.5395,
      "num_input_tokens_seen": 797120,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.11194293946027756,
      "learning_rate": 0.00032075471698113204,
      "loss": 0.4325,
      "num_input_tokens_seen": 801088,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 0.4542877972126007,
      "learning_rate": 0.00032206498951781973,
      "loss": 0.5181,
      "num_input_tokens_seen": 804704,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 0.49395838379859924,
      "learning_rate": 0.00032337526205450737,
      "loss": 0.6132,
      "num_input_tokens_seen": 808000,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 0.7323117256164551,
      "learning_rate": 0.000324685534591195,
      "loss": 0.5488,
      "num_input_tokens_seen": 810848,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 0.23764409124851227,
      "learning_rate": 0.00032599580712788263,
      "loss": 0.5863,
      "num_input_tokens_seen": 813728,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 0.5278456211090088,
      "learning_rate": 0.0003273060796645702,
      "loss": 0.4873,
      "num_input_tokens_seen": 816992,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 0.3115241229534149,
      "learning_rate": 0.00032861635220125785,
      "loss": 0.4592,
      "num_input_tokens_seen": 819584,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 0.8503628373146057,
      "learning_rate": 0.0003299266247379455,
      "loss": 0.5736,
      "num_input_tokens_seen": 822560,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 0.23614613711833954,
      "learning_rate": 0.0003312368972746331,
      "loss": 0.6482,
      "num_input_tokens_seen": 825792,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 0.21270643174648285,
      "learning_rate": 0.00033254716981132075,
      "loss": 0.4976,
      "num_input_tokens_seen": 829632,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 0.31124362349510193,
      "learning_rate": 0.0003338574423480084,
      "loss": 0.4729,
      "num_input_tokens_seen": 832672,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 0.19799883663654327,
      "learning_rate": 0.0003351677148846961,
      "loss": 0.4756,
      "num_input_tokens_seen": 835840,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.18441246449947357,
      "learning_rate": 0.00033647798742138366,
      "loss": 0.4865,
      "num_input_tokens_seen": 840160,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 0.14041875302791595,
      "learning_rate": 0.0003377882599580713,
      "loss": 0.5699,
      "num_input_tokens_seen": 843392,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 0.20132602751255035,
      "learning_rate": 0.0003390985324947589,
      "loss": 0.4094,
      "num_input_tokens_seen": 847616,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.24104885756969452,
      "learning_rate": 0.00034040880503144656,
      "loss": 0.5101,
      "num_input_tokens_seen": 851392,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 0.18056313693523407,
      "learning_rate": 0.0003417190775681342,
      "loss": 0.543,
      "num_input_tokens_seen": 854336,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 0.14616583287715912,
      "learning_rate": 0.00034302935010482183,
      "loss": 0.4848,
      "num_input_tokens_seen": 857312,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.08481968194246292,
      "learning_rate": 0.0003443396226415094,
      "loss": 0.5321,
      "num_input_tokens_seen": 861152,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 0.19180403649806976,
      "learning_rate": 0.00034564989517819704,
      "loss": 0.6204,
      "num_input_tokens_seen": 864704,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.1961277425289154,
      "learning_rate": 0.0003469601677148847,
      "loss": 0.554,
      "num_input_tokens_seen": 867360,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.43351075053215027,
      "learning_rate": 0.00034827044025157236,
      "loss": 0.598,
      "num_input_tokens_seen": 870464,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 0.1503380537033081,
      "learning_rate": 0.00034958071278826,
      "loss": 0.3874,
      "num_input_tokens_seen": 873792,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 0.18147815763950348,
      "learning_rate": 0.00035089098532494763,
      "loss": 0.6038,
      "num_input_tokens_seen": 878336,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 0.2383013516664505,
      "learning_rate": 0.0003522012578616352,
      "loss": 0.667,
      "num_input_tokens_seen": 881984,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 0.19860410690307617,
      "learning_rate": 0.00035351153039832285,
      "loss": 0.4298,
      "num_input_tokens_seen": 885632,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 0.2838842272758484,
      "learning_rate": 0.0003548218029350105,
      "loss": 0.4324,
      "num_input_tokens_seen": 888544,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 0.25844958424568176,
      "learning_rate": 0.0003561320754716981,
      "loss": 0.4871,
      "num_input_tokens_seen": 891328,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 0.16097503900527954,
      "learning_rate": 0.00035744234800838575,
      "loss": 0.6039,
      "num_input_tokens_seen": 894400,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.22837772965431213,
      "learning_rate": 0.0003587526205450734,
      "loss": 0.462,
      "num_input_tokens_seen": 896704,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.5848563313484192,
      "learning_rate": 0.00036006289308176097,
      "loss": 0.4818,
      "num_input_tokens_seen": 899520,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.28207847476005554,
      "learning_rate": 0.00036137316561844865,
      "loss": 0.53,
      "num_input_tokens_seen": 902304,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 0.19488373398780823,
      "learning_rate": 0.0003626834381551363,
      "loss": 0.5635,
      "num_input_tokens_seen": 905760,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 0.43288591504096985,
      "learning_rate": 0.0003639937106918239,
      "loss": 0.5573,
      "num_input_tokens_seen": 908512,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 0.3609481751918793,
      "learning_rate": 0.00036530398322851156,
      "loss": 0.4697,
      "num_input_tokens_seen": 912736,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 0.16049738228321075,
      "learning_rate": 0.0003666142557651992,
      "loss": 0.8091,
      "num_input_tokens_seen": 916352,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 0.23481504619121552,
      "learning_rate": 0.00036792452830188677,
      "loss": 0.5876,
      "num_input_tokens_seen": 919136,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 0.17881976068019867,
      "learning_rate": 0.0003692348008385744,
      "loss": 0.6209,
      "num_input_tokens_seen": 921760,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 0.194200798869133,
      "learning_rate": 0.00037054507337526204,
      "loss": 0.4471,
      "num_input_tokens_seen": 924096,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 0.16728036105632782,
      "learning_rate": 0.0003718553459119497,
      "loss": 0.5425,
      "num_input_tokens_seen": 926240,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 0.18485739827156067,
      "learning_rate": 0.0003731656184486373,
      "loss": 0.4887,
      "num_input_tokens_seen": 928928,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 0.17098070681095123,
      "learning_rate": 0.000374475890985325,
      "loss": 0.6563,
      "num_input_tokens_seen": 931552,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 0.1552189290523529,
      "learning_rate": 0.00037578616352201263,
      "loss": 0.4741,
      "num_input_tokens_seen": 935648,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.4422528147697449,
      "learning_rate": 0.0003770964360587002,
      "loss": 0.4865,
      "num_input_tokens_seen": 938336,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 0.2602635324001312,
      "learning_rate": 0.00037840670859538785,
      "loss": 0.514,
      "num_input_tokens_seen": 941952,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 0.15377657115459442,
      "learning_rate": 0.0003797169811320755,
      "loss": 0.4452,
      "num_input_tokens_seen": 945248,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 0.1465231329202652,
      "learning_rate": 0.0003810272536687631,
      "loss": 0.5142,
      "num_input_tokens_seen": 948512,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 0.12951184809207916,
      "learning_rate": 0.00038233752620545075,
      "loss": 0.468,
      "num_input_tokens_seen": 951744,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 0.24319617450237274,
      "learning_rate": 0.0003836477987421384,
      "loss": 0.5076,
      "num_input_tokens_seen": 953920,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 0.16555273532867432,
      "learning_rate": 0.00038495807127882596,
      "loss": 0.4649,
      "num_input_tokens_seen": 957376,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 0.24827662110328674,
      "learning_rate": 0.0003862683438155136,
      "loss": 0.5104,
      "num_input_tokens_seen": 960288,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 0.17098593711853027,
      "learning_rate": 0.0003875786163522013,
      "loss": 0.4789,
      "num_input_tokens_seen": 963872,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 0.15776900947093964,
      "learning_rate": 0.0003888888888888889,
      "loss": 0.5305,
      "num_input_tokens_seen": 966368,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.13003361225128174,
      "learning_rate": 0.00039019916142557656,
      "loss": 0.4744,
      "num_input_tokens_seen": 969344,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 0.18352128565311432,
      "learning_rate": 0.0003915094339622642,
      "loss": 0.5631,
      "num_input_tokens_seen": 972320,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 0.1318877786397934,
      "learning_rate": 0.00039281970649895177,
      "loss": 0.5202,
      "num_input_tokens_seen": 976448,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.19467829167842865,
      "learning_rate": 0.0003941299790356394,
      "loss": 0.4556,
      "num_input_tokens_seen": 979648,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 0.13020479679107666,
      "learning_rate": 0.00039544025157232704,
      "loss": 0.3888,
      "num_input_tokens_seen": 983040,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 0.18404820561408997,
      "learning_rate": 0.0003967505241090147,
      "loss": 0.3768,
      "num_input_tokens_seen": 985824,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 0.13270142674446106,
      "learning_rate": 0.0003980607966457023,
      "loss": 0.6914,
      "num_input_tokens_seen": 990368,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 0.11833098530769348,
      "learning_rate": 0.00039937106918238994,
      "loss": 0.5481,
      "num_input_tokens_seen": 993920,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 0.14126750826835632,
      "learning_rate": 0.0004006813417190776,
      "loss": 0.54,
      "num_input_tokens_seen": 997280,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.1914391666650772,
      "learning_rate": 0.0004019916142557652,
      "loss": 0.6919,
      "num_input_tokens_seen": 1000096,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 0.15093828737735748,
      "learning_rate": 0.00040330188679245284,
      "loss": 0.3974,
      "num_input_tokens_seen": 1002624,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 0.36471572518348694,
      "learning_rate": 0.0004046121593291405,
      "loss": 0.5033,
      "num_input_tokens_seen": 1005440,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 0.34295397996902466,
      "learning_rate": 0.0004059224318658281,
      "loss": 0.6111,
      "num_input_tokens_seen": 1009248,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 0.28559166193008423,
      "learning_rate": 0.00040723270440251575,
      "loss": 0.5193,
      "num_input_tokens_seen": 1011744,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.16594013571739197,
      "learning_rate": 0.00040854297693920333,
      "loss": 0.4926,
      "num_input_tokens_seen": 1014432,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 0.12406592071056366,
      "learning_rate": 0.00040985324947589096,
      "loss": 0.5703,
      "num_input_tokens_seen": 1017472,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.15268154442310333,
      "learning_rate": 0.0004111635220125786,
      "loss": 0.6236,
      "num_input_tokens_seen": 1020128,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 0.29445570707321167,
      "learning_rate": 0.00041247379454926623,
      "loss": 0.4076,
      "num_input_tokens_seen": 1023200,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 0.11995675414800644,
      "learning_rate": 0.0004137840670859539,
      "loss": 0.648,
      "num_input_tokens_seen": 1026592,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 0.17187811434268951,
      "learning_rate": 0.00041509433962264155,
      "loss": 0.4616,
      "num_input_tokens_seen": 1029216,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.16137538850307465,
      "learning_rate": 0.00041640461215932913,
      "loss": 0.6045,
      "num_input_tokens_seen": 1032960,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 0.13626788556575775,
      "learning_rate": 0.00041771488469601677,
      "loss": 0.5001,
      "num_input_tokens_seen": 1036224,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 0.0773811861872673,
      "learning_rate": 0.0004190251572327044,
      "loss": 0.4573,
      "num_input_tokens_seen": 1040608,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 0.11352141946554184,
      "learning_rate": 0.00042033542976939204,
      "loss": 0.4586,
      "num_input_tokens_seen": 1043648,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 0.10640938580036163,
      "learning_rate": 0.00042164570230607967,
      "loss": 0.4629,
      "num_input_tokens_seen": 1046400,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.2083621472120285,
      "learning_rate": 0.0004229559748427673,
      "loss": 0.4952,
      "num_input_tokens_seen": 1049568,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 0.32196328043937683,
      "learning_rate": 0.00042426624737945494,
      "loss": 0.5862,
      "num_input_tokens_seen": 1052288,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 0.30359479784965515,
      "learning_rate": 0.0004255765199161425,
      "loss": 0.6671,
      "num_input_tokens_seen": 1056096,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 0.23452332615852356,
      "learning_rate": 0.0004268867924528302,
      "loss": 0.4759,
      "num_input_tokens_seen": 1059968,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 0.3315628468990326,
      "learning_rate": 0.00042819706498951784,
      "loss": 0.452,
      "num_input_tokens_seen": 1063136,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 2.6862130165100098,
      "learning_rate": 0.0004295073375262055,
      "loss": 0.5213,
      "num_input_tokens_seen": 1067616,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 0.22362956404685974,
      "learning_rate": 0.0004308176100628931,
      "loss": 0.5748,
      "num_input_tokens_seen": 1071168,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 0.11870630830526352,
      "learning_rate": 0.00043212788259958075,
      "loss": 0.3935,
      "num_input_tokens_seen": 1075808,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 0.15338659286499023,
      "learning_rate": 0.0004334381551362683,
      "loss": 0.5966,
      "num_input_tokens_seen": 1080224,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 0.22319506108760834,
      "learning_rate": 0.00043474842767295596,
      "loss": 0.6491,
      "num_input_tokens_seen": 1083808,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 0.17444273829460144,
      "learning_rate": 0.0004360587002096436,
      "loss": 0.6254,
      "num_input_tokens_seen": 1086304,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 0.1779547929763794,
      "learning_rate": 0.00043736897274633123,
      "loss": 0.6262,
      "num_input_tokens_seen": 1089856,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 0.3069278299808502,
      "learning_rate": 0.00043867924528301886,
      "loss": 0.4421,
      "num_input_tokens_seen": 1092576,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.12583646178245544,
      "learning_rate": 0.00043998951781970655,
      "loss": 0.453,
      "num_input_tokens_seen": 1096448,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.1204940602183342,
      "learning_rate": 0.00044129979035639413,
      "loss": 0.4301,
      "num_input_tokens_seen": 1100448,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 0.18026916682720184,
      "learning_rate": 0.00044261006289308177,
      "loss": 0.5105,
      "num_input_tokens_seen": 1103520,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 0.3202609717845917,
      "learning_rate": 0.0004439203354297694,
      "loss": 0.3949,
      "num_input_tokens_seen": 1106304,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 0.24762961268424988,
      "learning_rate": 0.00044523060796645704,
      "loss": 0.5114,
      "num_input_tokens_seen": 1110176,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.21976591646671295,
      "learning_rate": 0.00044654088050314467,
      "loss": 0.4924,
      "num_input_tokens_seen": 1112640,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.17099468410015106,
      "learning_rate": 0.0004478511530398323,
      "loss": 0.516,
      "num_input_tokens_seen": 1116736,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 0.31809648871421814,
      "learning_rate": 0.0004491614255765199,
      "loss": 0.4428,
      "num_input_tokens_seen": 1120000,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 0.13155587017536163,
      "learning_rate": 0.0004504716981132075,
      "loss": 0.4131,
      "num_input_tokens_seen": 1122656,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 0.22109392285346985,
      "learning_rate": 0.00045178197064989515,
      "loss": 0.4884,
      "num_input_tokens_seen": 1125856,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.13864944875240326,
      "learning_rate": 0.00045309224318658284,
      "loss": 0.4906,
      "num_input_tokens_seen": 1128288,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 0.1978299468755722,
      "learning_rate": 0.0004544025157232705,
      "loss": 0.5725,
      "num_input_tokens_seen": 1131680,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 0.20579826831817627,
      "learning_rate": 0.0004557127882599581,
      "loss": 0.4744,
      "num_input_tokens_seen": 1134496,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 0.20926207304000854,
      "learning_rate": 0.0004570230607966457,
      "loss": 0.4571,
      "num_input_tokens_seen": 1138176,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 0.09494095295667648,
      "learning_rate": 0.0004583333333333333,
      "loss": 0.3959,
      "num_input_tokens_seen": 1141056,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 0.15310190618038177,
      "learning_rate": 0.00045964360587002096,
      "loss": 0.4691,
      "num_input_tokens_seen": 1144224,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 0.19347332417964935,
      "learning_rate": 0.0004609538784067086,
      "loss": 0.5113,
      "num_input_tokens_seen": 1146720,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 0.24696390330791473,
      "learning_rate": 0.00046226415094339623,
      "loss": 0.6513,
      "num_input_tokens_seen": 1149312,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 0.14347673952579498,
      "learning_rate": 0.00046357442348008386,
      "loss": 0.5014,
      "num_input_tokens_seen": 1152224,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 0.12266577780246735,
      "learning_rate": 0.00046488469601677155,
      "loss": 0.4911,
      "num_input_tokens_seen": 1155456,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 0.12455716729164124,
      "learning_rate": 0.00046619496855345913,
      "loss": 0.4912,
      "num_input_tokens_seen": 1158752,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 0.16605110466480255,
      "learning_rate": 0.00046750524109014677,
      "loss": 0.5525,
      "num_input_tokens_seen": 1161632,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.15769833326339722,
      "learning_rate": 0.0004688155136268344,
      "loss": 0.4952,
      "num_input_tokens_seen": 1165536,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.24186038970947266,
      "learning_rate": 0.00047012578616352203,
      "loss": 0.6191,
      "num_input_tokens_seen": 1168992,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.16551624238491058,
      "learning_rate": 0.00047143605870020967,
      "loss": 0.444,
      "num_input_tokens_seen": 1171648,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 0.43763071298599243,
      "learning_rate": 0.0004727463312368973,
      "loss": 0.6414,
      "num_input_tokens_seen": 1174912,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 0.2200654000043869,
      "learning_rate": 0.0004740566037735849,
      "loss": 0.4039,
      "num_input_tokens_seen": 1178144,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.12735193967819214,
      "learning_rate": 0.0004753668763102725,
      "loss": 0.403,
      "num_input_tokens_seen": 1181856,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 0.19076812267303467,
      "learning_rate": 0.00047667714884696015,
      "loss": 0.6034,
      "num_input_tokens_seen": 1184896,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 0.4735887050628662,
      "learning_rate": 0.00047798742138364784,
      "loss": 0.7188,
      "num_input_tokens_seen": 1187264,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 0.1888461410999298,
      "learning_rate": 0.0004792976939203355,
      "loss": 0.4858,
      "num_input_tokens_seen": 1191168,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 0.21731975674629211,
      "learning_rate": 0.0004806079664570231,
      "loss": 0.4371,
      "num_input_tokens_seen": 1193664,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 0.2055308073759079,
      "learning_rate": 0.0004819182389937107,
      "loss": 0.4543,
      "num_input_tokens_seen": 1200864,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 0.2461959570646286,
      "learning_rate": 0.0004832285115303983,
      "loss": 0.4646,
      "num_input_tokens_seen": 1203680,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 0.0763527974486351,
      "learning_rate": 0.00048453878406708596,
      "loss": 0.5537,
      "num_input_tokens_seen": 1207648,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.10611432045698166,
      "learning_rate": 0.0004858490566037736,
      "loss": 0.4967,
      "num_input_tokens_seen": 1211232,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 0.12363187968730927,
      "learning_rate": 0.0004871593291404612,
      "loss": 0.4525,
      "num_input_tokens_seen": 1214304,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.1393575221300125,
      "learning_rate": 0.0004884696016771489,
      "loss": 0.3894,
      "num_input_tokens_seen": 1217664,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 0.2897704243659973,
      "learning_rate": 0.0004897798742138365,
      "loss": 0.3924,
      "num_input_tokens_seen": 1221184,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 0.16051435470581055,
      "learning_rate": 0.0004910901467505241,
      "loss": 0.4821,
      "num_input_tokens_seen": 1224736,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 0.11409994959831238,
      "learning_rate": 0.0004924004192872118,
      "loss": 0.4982,
      "num_input_tokens_seen": 1228256,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.1683143824338913,
      "learning_rate": 0.0004937106918238993,
      "loss": 0.7353,
      "num_input_tokens_seen": 1231104,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 0.12470951676368713,
      "learning_rate": 0.000495020964360587,
      "loss": 0.4709,
      "num_input_tokens_seen": 1234176,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 0.16363508999347687,
      "learning_rate": 0.0004963312368972746,
      "loss": 0.4626,
      "num_input_tokens_seen": 1237568,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 0.14949628710746765,
      "learning_rate": 0.0004976415094339623,
      "loss": 0.6021,
      "num_input_tokens_seen": 1240896,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 0.1834135502576828,
      "learning_rate": 0.0004989517819706499,
      "loss": 0.6502,
      "num_input_tokens_seen": 1244288,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5047244429588318,
      "eval_runtime": 13.7033,
      "eval_samples_per_second": 61.883,
      "eval_steps_per_second": 15.471,
      "num_input_tokens_seen": 1245424,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 0.1738416850566864,
      "learning_rate": 0.0005002620545073376,
      "loss": 0.4655,
      "num_input_tokens_seen": 1246832,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 0.21822567284107208,
      "learning_rate": 0.0005015723270440253,
      "loss": 0.45,
      "num_input_tokens_seen": 1250000,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.20475144684314728,
      "learning_rate": 0.0005028825995807128,
      "loss": 0.5823,
      "num_input_tokens_seen": 1253488,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 0.09813211858272552,
      "learning_rate": 0.0005041928721174004,
      "loss": 0.527,
      "num_input_tokens_seen": 1257744,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 0.18043328821659088,
      "learning_rate": 0.0005055031446540881,
      "loss": 0.3572,
      "num_input_tokens_seen": 1260688,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.16939203441143036,
      "learning_rate": 0.0005068134171907757,
      "loss": 0.394,
      "num_input_tokens_seen": 1263920,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 0.16247344017028809,
      "learning_rate": 0.0005081236897274634,
      "loss": 0.6,
      "num_input_tokens_seen": 1266992,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 0.13152167201042175,
      "learning_rate": 0.000509433962264151,
      "loss": 0.4211,
      "num_input_tokens_seen": 1270544,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 0.14478784799575806,
      "learning_rate": 0.0005107442348008385,
      "loss": 0.4144,
      "num_input_tokens_seen": 1273840,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.12987084686756134,
      "learning_rate": 0.0005120545073375262,
      "loss": 0.389,
      "num_input_tokens_seen": 1276464,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 0.16224738955497742,
      "learning_rate": 0.0005133647798742138,
      "loss": 0.4894,
      "num_input_tokens_seen": 1280144,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 0.2113373875617981,
      "learning_rate": 0.0005146750524109015,
      "loss": 0.4626,
      "num_input_tokens_seen": 1283088,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 0.08647622168064117,
      "learning_rate": 0.0005159853249475891,
      "loss": 0.5955,
      "num_input_tokens_seen": 1287088,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 0.13488227128982544,
      "learning_rate": 0.0005172955974842768,
      "loss": 0.4859,
      "num_input_tokens_seen": 1290224,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 0.16607345640659332,
      "learning_rate": 0.0005186058700209643,
      "loss": 0.5623,
      "num_input_tokens_seen": 1293296,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.19073045253753662,
      "learning_rate": 0.0005199161425576519,
      "loss": 0.4466,
      "num_input_tokens_seen": 1296048,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 0.21244485676288605,
      "learning_rate": 0.0005212264150943396,
      "loss": 0.5763,
      "num_input_tokens_seen": 1299312,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.18382678925991058,
      "learning_rate": 0.0005225366876310272,
      "loss": 0.5812,
      "num_input_tokens_seen": 1303056,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.2200433909893036,
      "learning_rate": 0.000523846960167715,
      "loss": 0.4543,
      "num_input_tokens_seen": 1306288,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 0.19045689702033997,
      "learning_rate": 0.0005251572327044026,
      "loss": 0.55,
      "num_input_tokens_seen": 1310064,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 0.12094660103321075,
      "learning_rate": 0.0005264675052410901,
      "loss": 0.402,
      "num_input_tokens_seen": 1313936,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 0.11854097992181778,
      "learning_rate": 0.0005277777777777778,
      "loss": 0.4875,
      "num_input_tokens_seen": 1317840,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 0.09729001671075821,
      "learning_rate": 0.0005290880503144654,
      "loss": 0.5265,
      "num_input_tokens_seen": 1320880,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 0.15887197852134705,
      "learning_rate": 0.0005303983228511531,
      "loss": 0.4682,
      "num_input_tokens_seen": 1324400,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.1744476556777954,
      "learning_rate": 0.0005317085953878407,
      "loss": 0.5347,
      "num_input_tokens_seen": 1327216,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 0.16458489000797272,
      "learning_rate": 0.0005330188679245284,
      "loss": 0.5075,
      "num_input_tokens_seen": 1329712,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 0.3277566134929657,
      "learning_rate": 0.000534329140461216,
      "loss": 0.6189,
      "num_input_tokens_seen": 1332784,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 0.2263597548007965,
      "learning_rate": 0.0005356394129979035,
      "loss": 0.482,
      "num_input_tokens_seen": 1335984,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 0.12787699699401855,
      "learning_rate": 0.0005369496855345912,
      "loss": 0.5289,
      "num_input_tokens_seen": 1338960,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 0.3354097604751587,
      "learning_rate": 0.0005382599580712788,
      "loss": 0.4886,
      "num_input_tokens_seen": 1342000,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 0.18017329275608063,
      "learning_rate": 0.0005395702306079665,
      "loss": 0.5154,
      "num_input_tokens_seen": 1344400,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 0.24648654460906982,
      "learning_rate": 0.0005408805031446541,
      "loss": 0.4868,
      "num_input_tokens_seen": 1347440,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 0.12053325772285461,
      "learning_rate": 0.0005421907756813418,
      "loss": 0.5274,
      "num_input_tokens_seen": 1352048,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 0.2913271486759186,
      "learning_rate": 0.0005435010482180293,
      "loss": 0.5403,
      "num_input_tokens_seen": 1354928,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 0.22753329575061798,
      "learning_rate": 0.0005448113207547169,
      "loss": 0.5154,
      "num_input_tokens_seen": 1357968,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 0.15201368927955627,
      "learning_rate": 0.0005461215932914046,
      "loss": 0.4298,
      "num_input_tokens_seen": 1361168,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 0.08677585422992706,
      "learning_rate": 0.0005474318658280922,
      "loss": 0.5769,
      "num_input_tokens_seen": 1365584,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 0.1347944736480713,
      "learning_rate": 0.0005487421383647799,
      "loss": 0.5615,
      "num_input_tokens_seen": 1369424,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 0.1772005259990692,
      "learning_rate": 0.0005500524109014676,
      "loss": 0.5237,
      "num_input_tokens_seen": 1371952,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 0.1421538144350052,
      "learning_rate": 0.0005513626834381551,
      "loss": 0.4379,
      "num_input_tokens_seen": 1374352,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 0.14102020859718323,
      "learning_rate": 0.0005526729559748428,
      "loss": 0.3714,
      "num_input_tokens_seen": 1376848,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 0.12679992616176605,
      "learning_rate": 0.0005539832285115304,
      "loss": 0.5226,
      "num_input_tokens_seen": 1380240,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.16747508943080902,
      "learning_rate": 0.0005552935010482181,
      "loss": 0.4383,
      "num_input_tokens_seen": 1384048,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 0.2467023730278015,
      "learning_rate": 0.0005566037735849057,
      "loss": 0.4217,
      "num_input_tokens_seen": 1387504,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 0.14149312674999237,
      "learning_rate": 0.0005579140461215934,
      "loss": 0.4801,
      "num_input_tokens_seen": 1390960,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 0.12328720092773438,
      "learning_rate": 0.000559224318658281,
      "loss": 0.4325,
      "num_input_tokens_seen": 1393744,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 0.10192029178142548,
      "learning_rate": 0.0005605345911949685,
      "loss": 0.5661,
      "num_input_tokens_seen": 1397040,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 0.172259122133255,
      "learning_rate": 0.0005618448637316562,
      "loss": 0.4695,
      "num_input_tokens_seen": 1400080,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 0.19144777953624725,
      "learning_rate": 0.0005631551362683438,
      "loss": 0.5518,
      "num_input_tokens_seen": 1402864,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 0.18740394711494446,
      "learning_rate": 0.0005644654088050315,
      "loss": 0.588,
      "num_input_tokens_seen": 1405904,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.12202763557434082,
      "learning_rate": 0.0005657756813417191,
      "loss": 0.4603,
      "num_input_tokens_seen": 1408848,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 0.1422017365694046,
      "learning_rate": 0.0005670859538784067,
      "loss": 0.4366,
      "num_input_tokens_seen": 1412208,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 0.10065910220146179,
      "learning_rate": 0.0005683962264150943,
      "loss": 0.5176,
      "num_input_tokens_seen": 1415664,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 0.07838281244039536,
      "learning_rate": 0.0005697064989517819,
      "loss": 0.4427,
      "num_input_tokens_seen": 1419696,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 0.13157251477241516,
      "learning_rate": 0.0005710167714884696,
      "loss": 0.5203,
      "num_input_tokens_seen": 1422672,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 0.13176178932189941,
      "learning_rate": 0.0005723270440251572,
      "loss": 0.5894,
      "num_input_tokens_seen": 1425520,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 0.12074171751737595,
      "learning_rate": 0.0005736373165618449,
      "loss": 0.6136,
      "num_input_tokens_seen": 1428016,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 0.14748550951480865,
      "learning_rate": 0.0005749475890985325,
      "loss": 0.4489,
      "num_input_tokens_seen": 1430480,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 0.283231645822525,
      "learning_rate": 0.0005762578616352201,
      "loss": 0.4657,
      "num_input_tokens_seen": 1432848,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 0.11737137287855148,
      "learning_rate": 0.0005775681341719078,
      "loss": 0.5685,
      "num_input_tokens_seen": 1435728,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 0.12711751461029053,
      "learning_rate": 0.0005788784067085954,
      "loss": 0.5196,
      "num_input_tokens_seen": 1438160,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 0.19242186844348907,
      "learning_rate": 0.0005801886792452831,
      "loss": 0.462,
      "num_input_tokens_seen": 1440848,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.17282123863697052,
      "learning_rate": 0.0005814989517819707,
      "loss": 0.5442,
      "num_input_tokens_seen": 1443728,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 0.11716599017381668,
      "learning_rate": 0.0005828092243186583,
      "loss": 0.4609,
      "num_input_tokens_seen": 1446448,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 0.12512347102165222,
      "learning_rate": 0.000584119496855346,
      "loss": 0.465,
      "num_input_tokens_seen": 1449936,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 0.09274283796548843,
      "learning_rate": 0.0005854297693920335,
      "loss": 0.4845,
      "num_input_tokens_seen": 1452976,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.17883652448654175,
      "learning_rate": 0.0005867400419287212,
      "loss": 0.3964,
      "num_input_tokens_seen": 1456080,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 0.15722347795963287,
      "learning_rate": 0.0005880503144654088,
      "loss": 0.492,
      "num_input_tokens_seen": 1459344,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 0.19936348497867584,
      "learning_rate": 0.0005893605870020965,
      "loss": 0.5026,
      "num_input_tokens_seen": 1462288,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 0.3443396985530853,
      "learning_rate": 0.0005906708595387841,
      "loss": 0.4884,
      "num_input_tokens_seen": 1465072,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 0.09672997146844864,
      "learning_rate": 0.0005919811320754716,
      "loss": 0.4727,
      "num_input_tokens_seen": 1468304,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 0.09018230438232422,
      "learning_rate": 0.0005932914046121593,
      "loss": 0.5812,
      "num_input_tokens_seen": 1471152,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 0.21035481989383698,
      "learning_rate": 0.0005946016771488469,
      "loss": 0.5111,
      "num_input_tokens_seen": 1473744,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 0.09048512578010559,
      "learning_rate": 0.0005959119496855346,
      "loss": 0.6416,
      "num_input_tokens_seen": 1477264,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 0.10320543497800827,
      "learning_rate": 0.0005972222222222222,
      "loss": 0.4179,
      "num_input_tokens_seen": 1481456,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 0.12303782254457474,
      "learning_rate": 0.0005985324947589099,
      "loss": 0.4789,
      "num_input_tokens_seen": 1485200,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.14442922174930573,
      "learning_rate": 0.0005998427672955975,
      "loss": 0.5873,
      "num_input_tokens_seen": 1488656,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 0.11848833411931992,
      "learning_rate": 0.000601153039832285,
      "loss": 0.7214,
      "num_input_tokens_seen": 1492016,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 0.09709551185369492,
      "learning_rate": 0.0006024633123689728,
      "loss": 0.5006,
      "num_input_tokens_seen": 1496144,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 0.1788661777973175,
      "learning_rate": 0.0006037735849056604,
      "loss": 0.527,
      "num_input_tokens_seen": 1498416,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.15020611882209778,
      "learning_rate": 0.0006050838574423481,
      "loss": 0.425,
      "num_input_tokens_seen": 1501264,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.15134499967098236,
      "learning_rate": 0.0006063941299790357,
      "loss": 0.5233,
      "num_input_tokens_seen": 1505040,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 0.33398446440696716,
      "learning_rate": 0.0006077044025157233,
      "loss": 0.5753,
      "num_input_tokens_seen": 1507760,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 0.10379687696695328,
      "learning_rate": 0.000609014675052411,
      "loss": 0.5118,
      "num_input_tokens_seen": 1511216,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 0.08616090565919876,
      "learning_rate": 0.0006103249475890985,
      "loss": 0.4259,
      "num_input_tokens_seen": 1514384,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 0.07489901781082153,
      "learning_rate": 0.0006116352201257862,
      "loss": 0.3829,
      "num_input_tokens_seen": 1518544,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 0.22708088159561157,
      "learning_rate": 0.0006129454926624738,
      "loss": 0.4227,
      "num_input_tokens_seen": 1521936,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 0.1382158249616623,
      "learning_rate": 0.0006142557651991615,
      "loss": 0.4629,
      "num_input_tokens_seen": 1525072,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 0.10671619325876236,
      "learning_rate": 0.0006155660377358491,
      "loss": 0.5717,
      "num_input_tokens_seen": 1529168,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 0.10188738256692886,
      "learning_rate": 0.0006168763102725366,
      "loss": 0.4993,
      "num_input_tokens_seen": 1531984,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 0.16638720035552979,
      "learning_rate": 0.0006181865828092243,
      "loss": 0.6606,
      "num_input_tokens_seen": 1535216,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 0.13342638313770294,
      "learning_rate": 0.0006194968553459119,
      "loss": 0.3638,
      "num_input_tokens_seen": 1538672,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 0.12185689806938171,
      "learning_rate": 0.0006208071278825996,
      "loss": 0.3774,
      "num_input_tokens_seen": 1542352,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 0.11128903925418854,
      "learning_rate": 0.0006221174004192872,
      "loss": 0.5614,
      "num_input_tokens_seen": 1545616,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 0.2209806889295578,
      "learning_rate": 0.0006234276729559748,
      "loss": 0.5404,
      "num_input_tokens_seen": 1548304,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.0611761137843132,
      "learning_rate": 0.0006247379454926625,
      "loss": 0.4123,
      "num_input_tokens_seen": 1552528,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 0.33162322640419006,
      "learning_rate": 0.00062604821802935,
      "loss": 0.3355,
      "num_input_tokens_seen": 1555216,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.12959811091423035,
      "learning_rate": 0.0006273584905660377,
      "loss": 0.5353,
      "num_input_tokens_seen": 1558928,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.14611685276031494,
      "learning_rate": 0.0006286687631027254,
      "loss": 0.3545,
      "num_input_tokens_seen": 1561648,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 0.129911407828331,
      "learning_rate": 0.0006299790356394131,
      "loss": 0.5324,
      "num_input_tokens_seen": 1566640,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 0.1345101296901703,
      "learning_rate": 0.0006312893081761007,
      "loss": 0.5646,
      "num_input_tokens_seen": 1569904,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 0.1624128818511963,
      "learning_rate": 0.0006325995807127883,
      "loss": 0.6761,
      "num_input_tokens_seen": 1572720,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 0.15147995948791504,
      "learning_rate": 0.000633909853249476,
      "loss": 0.5592,
      "num_input_tokens_seen": 1575728,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 0.11109104007482529,
      "learning_rate": 0.0006352201257861635,
      "loss": 0.5293,
      "num_input_tokens_seen": 1578512,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.1617669314146042,
      "learning_rate": 0.0006365303983228512,
      "loss": 0.645,
      "num_input_tokens_seen": 1580912,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 0.1274229735136032,
      "learning_rate": 0.0006378406708595388,
      "loss": 0.4112,
      "num_input_tokens_seen": 1584304,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 0.14213913679122925,
      "learning_rate": 0.0006391509433962265,
      "loss": 0.5042,
      "num_input_tokens_seen": 1587536,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 0.11052165925502777,
      "learning_rate": 0.0006404612159329141,
      "loss": 0.5149,
      "num_input_tokens_seen": 1590544,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 0.15085922181606293,
      "learning_rate": 0.0006417714884696016,
      "loss": 0.5026,
      "num_input_tokens_seen": 1593200,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 0.13586880266666412,
      "learning_rate": 0.0006430817610062893,
      "loss": 0.43,
      "num_input_tokens_seen": 1595952,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 0.07477894425392151,
      "learning_rate": 0.0006443920335429769,
      "loss": 0.547,
      "num_input_tokens_seen": 1598928,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 0.11162329465150833,
      "learning_rate": 0.0006457023060796646,
      "loss": 0.5242,
      "num_input_tokens_seen": 1601936,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.14468061923980713,
      "learning_rate": 0.0006470125786163522,
      "loss": 0.4372,
      "num_input_tokens_seen": 1604688,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 0.12421685457229614,
      "learning_rate": 0.0006483228511530398,
      "loss": 0.5817,
      "num_input_tokens_seen": 1608176,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.17412441968917847,
      "learning_rate": 0.0006496331236897275,
      "loss": 0.5468,
      "num_input_tokens_seen": 1611152,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 0.16579651832580566,
      "learning_rate": 0.000650943396226415,
      "loss": 0.3929,
      "num_input_tokens_seen": 1614096,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.14754638075828552,
      "learning_rate": 0.0006522536687631027,
      "loss": 0.4275,
      "num_input_tokens_seen": 1616560,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.11951287090778351,
      "learning_rate": 0.0006535639412997903,
      "loss": 0.4832,
      "num_input_tokens_seen": 1619408,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.1700424700975418,
      "learning_rate": 0.0006548742138364781,
      "loss": 0.6137,
      "num_input_tokens_seen": 1622384,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 0.13165809214115143,
      "learning_rate": 0.0006561844863731657,
      "loss": 0.5275,
      "num_input_tokens_seen": 1625232,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.20329725742340088,
      "learning_rate": 0.0006574947589098533,
      "loss": 0.3448,
      "num_input_tokens_seen": 1627376,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 0.15315194427967072,
      "learning_rate": 0.000658805031446541,
      "loss": 0.4751,
      "num_input_tokens_seen": 1630384,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.22328101098537445,
      "learning_rate": 0.0006601153039832285,
      "loss": 0.5075,
      "num_input_tokens_seen": 1634192,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 0.09586428850889206,
      "learning_rate": 0.0006614255765199162,
      "loss": 0.5823,
      "num_input_tokens_seen": 1636848,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 0.09945894777774811,
      "learning_rate": 0.0006627358490566038,
      "loss": 0.4161,
      "num_input_tokens_seen": 1639952,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.09630110114812851,
      "learning_rate": 0.0006640461215932914,
      "loss": 0.6436,
      "num_input_tokens_seen": 1642512,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 0.10246512293815613,
      "learning_rate": 0.0006653563941299791,
      "loss": 0.412,
      "num_input_tokens_seen": 1646160,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 0.15637892484664917,
      "learning_rate": 0.0006666666666666666,
      "loss": 0.539,
      "num_input_tokens_seen": 1649840,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 0.08277451992034912,
      "learning_rate": 0.0006679769392033543,
      "loss": 0.5725,
      "num_input_tokens_seen": 1653104,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 0.11832509934902191,
      "learning_rate": 0.0006692872117400419,
      "loss": 0.4875,
      "num_input_tokens_seen": 1656336,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.07250591367483139,
      "learning_rate": 0.0006705974842767296,
      "loss": 0.6059,
      "num_input_tokens_seen": 1659824,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 0.08746378123760223,
      "learning_rate": 0.0006719077568134172,
      "loss": 0.5737,
      "num_input_tokens_seen": 1662992,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 0.10166436433792114,
      "learning_rate": 0.0006732180293501048,
      "loss": 0.4648,
      "num_input_tokens_seen": 1666320,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 0.15495643019676208,
      "learning_rate": 0.0006745283018867925,
      "loss": 0.4105,
      "num_input_tokens_seen": 1669264,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.1125255823135376,
      "learning_rate": 0.00067583857442348,
      "loss": 0.6386,
      "num_input_tokens_seen": 1672112,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 0.11618327349424362,
      "learning_rate": 0.0006771488469601677,
      "loss": 0.4909,
      "num_input_tokens_seen": 1674768,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 0.1186358705163002,
      "learning_rate": 0.0006784591194968553,
      "loss": 0.6396,
      "num_input_tokens_seen": 1678480,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 0.07571376860141754,
      "learning_rate": 0.0006797693920335431,
      "loss": 0.5222,
      "num_input_tokens_seen": 1681744,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 0.17923706769943237,
      "learning_rate": 0.0006810796645702307,
      "loss": 0.5221,
      "num_input_tokens_seen": 1684080,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 0.10977748781442642,
      "learning_rate": 0.0006823899371069183,
      "loss": 0.5392,
      "num_input_tokens_seen": 1688432,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 0.08429865539073944,
      "learning_rate": 0.0006837002096436059,
      "loss": 0.6304,
      "num_input_tokens_seen": 1692144,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 0.20982936024665833,
      "learning_rate": 0.0006850104821802935,
      "loss": 0.4579,
      "num_input_tokens_seen": 1695600,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.0953269824385643,
      "learning_rate": 0.0006863207547169812,
      "loss": 0.4402,
      "num_input_tokens_seen": 1699216,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.24324628710746765,
      "learning_rate": 0.0006876310272536688,
      "loss": 0.5409,
      "num_input_tokens_seen": 1702128,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 0.1296079307794571,
      "learning_rate": 0.0006889412997903564,
      "loss": 0.4763,
      "num_input_tokens_seen": 1705392,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 0.25480979681015015,
      "learning_rate": 0.0006902515723270441,
      "loss": 0.6215,
      "num_input_tokens_seen": 1708272,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.1572587937116623,
      "learning_rate": 0.0006915618448637316,
      "loss": 0.4733,
      "num_input_tokens_seen": 1710736,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 0.1112007349729538,
      "learning_rate": 0.0006928721174004193,
      "loss": 0.546,
      "num_input_tokens_seen": 1713424,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.08677450567483902,
      "learning_rate": 0.0006941823899371069,
      "loss": 0.4833,
      "num_input_tokens_seen": 1716336,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.161203071475029,
      "learning_rate": 0.0006954926624737946,
      "loss": 0.3836,
      "num_input_tokens_seen": 1718960,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 0.10623014718294144,
      "learning_rate": 0.0006968029350104822,
      "loss": 0.516,
      "num_input_tokens_seen": 1721904,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 0.12169389426708221,
      "learning_rate": 0.0006981132075471698,
      "loss": 0.3457,
      "num_input_tokens_seen": 1725552,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 0.32648470997810364,
      "learning_rate": 0.0006994234800838574,
      "loss": 0.5589,
      "num_input_tokens_seen": 1728720,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 0.1445212960243225,
      "learning_rate": 0.000700733752620545,
      "loss": 0.458,
      "num_input_tokens_seen": 1731664,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 0.10059791058301926,
      "learning_rate": 0.0007020440251572327,
      "loss": 0.4907,
      "num_input_tokens_seen": 1734576,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 0.20382173359394073,
      "learning_rate": 0.0007033542976939203,
      "loss": 0.3672,
      "num_input_tokens_seen": 1736784,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.19884270429611206,
      "learning_rate": 0.0007046645702306079,
      "loss": 0.4759,
      "num_input_tokens_seen": 1739472,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 0.1419200897216797,
      "learning_rate": 0.0007059748427672957,
      "loss": 0.4555,
      "num_input_tokens_seen": 1742544,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 0.14318464696407318,
      "learning_rate": 0.0007072851153039833,
      "loss": 0.6462,
      "num_input_tokens_seen": 1745456,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 0.1193319708108902,
      "learning_rate": 0.0007085953878406709,
      "loss": 0.4638,
      "num_input_tokens_seen": 1749360,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 0.07831776887178421,
      "learning_rate": 0.0007099056603773585,
      "loss": 0.4714,
      "num_input_tokens_seen": 1752048,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.1794261634349823,
      "learning_rate": 0.0007112159329140462,
      "loss": 0.4382,
      "num_input_tokens_seen": 1755088,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.16063697636127472,
      "learning_rate": 0.0007125262054507338,
      "loss": 0.5219,
      "num_input_tokens_seen": 1758416,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 0.19130542874336243,
      "learning_rate": 0.0007138364779874214,
      "loss": 0.5258,
      "num_input_tokens_seen": 1760848,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 0.20888450741767883,
      "learning_rate": 0.0007151467505241091,
      "loss": 0.4896,
      "num_input_tokens_seen": 1763024,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 0.0948551595211029,
      "learning_rate": 0.0007164570230607966,
      "loss": 0.4547,
      "num_input_tokens_seen": 1766160,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.17201119661331177,
      "learning_rate": 0.0007177672955974843,
      "loss": 0.4177,
      "num_input_tokens_seen": 1768912,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 0.17617185413837433,
      "learning_rate": 0.0007190775681341719,
      "loss": 0.5176,
      "num_input_tokens_seen": 1771792,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 0.1050964891910553,
      "learning_rate": 0.0007203878406708596,
      "loss": 0.4662,
      "num_input_tokens_seen": 1775088,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 0.09397546947002411,
      "learning_rate": 0.0007216981132075472,
      "loss": 0.4709,
      "num_input_tokens_seen": 1778736,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.09761856496334076,
      "learning_rate": 0.0007230083857442348,
      "loss": 0.4023,
      "num_input_tokens_seen": 1782000,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 0.09593368321657181,
      "learning_rate": 0.0007243186582809224,
      "loss": 0.7024,
      "num_input_tokens_seen": 1785168,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 0.1392102986574173,
      "learning_rate": 0.00072562893081761,
      "loss": 0.5612,
      "num_input_tokens_seen": 1789264,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 0.18503209948539734,
      "learning_rate": 0.0007269392033542977,
      "loss": 0.5072,
      "num_input_tokens_seen": 1792144,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 0.06100712716579437,
      "learning_rate": 0.0007282494758909853,
      "loss": 0.4751,
      "num_input_tokens_seen": 1795952,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 0.09605231881141663,
      "learning_rate": 0.0007295597484276729,
      "loss": 0.439,
      "num_input_tokens_seen": 1799632,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 0.0742773786187172,
      "learning_rate": 0.0007308700209643606,
      "loss": 0.4248,
      "num_input_tokens_seen": 1803088,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 0.13325656950473785,
      "learning_rate": 0.0007321802935010483,
      "loss": 0.4891,
      "num_input_tokens_seen": 1806576,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 0.12158244848251343,
      "learning_rate": 0.0007334905660377359,
      "loss": 0.4808,
      "num_input_tokens_seen": 1809360,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.12915953993797302,
      "learning_rate": 0.0007348008385744235,
      "loss": 0.4297,
      "num_input_tokens_seen": 1812752,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 0.08504823595285416,
      "learning_rate": 0.0007361111111111112,
      "loss": 0.4955,
      "num_input_tokens_seen": 1815952,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 0.0987071841955185,
      "learning_rate": 0.0007374213836477988,
      "loss": 0.4365,
      "num_input_tokens_seen": 1818992,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 0.4330892860889435,
      "learning_rate": 0.0007387316561844864,
      "loss": 0.703,
      "num_input_tokens_seen": 1822320,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.230584979057312,
      "learning_rate": 0.0007400419287211741,
      "loss": 0.5379,
      "num_input_tokens_seen": 1824560,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.08502839505672455,
      "learning_rate": 0.0007413522012578616,
      "loss": 0.3867,
      "num_input_tokens_seen": 1827728,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 0.18322113156318665,
      "learning_rate": 0.0007426624737945493,
      "loss": 0.5337,
      "num_input_tokens_seen": 1830672,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 0.19683511555194855,
      "learning_rate": 0.0007439727463312369,
      "loss": 0.537,
      "num_input_tokens_seen": 1833040,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 0.05732458829879761,
      "learning_rate": 0.0007452830188679245,
      "loss": 0.4736,
      "num_input_tokens_seen": 1837648,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 0.20246246457099915,
      "learning_rate": 0.0007465932914046122,
      "loss": 0.4548,
      "num_input_tokens_seen": 1840688,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 0.12687087059020996,
      "learning_rate": 0.0007479035639412998,
      "loss": 0.5259,
      "num_input_tokens_seen": 1843536,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 0.20518827438354492,
      "learning_rate": 0.0007492138364779874,
      "loss": 0.4096,
      "num_input_tokens_seen": 1845712,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 0.23899465799331665,
      "learning_rate": 0.000750524109014675,
      "loss": 0.4846,
      "num_input_tokens_seen": 1848656,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 0.08778609335422516,
      "learning_rate": 0.0007518343815513627,
      "loss": 0.5029,
      "num_input_tokens_seen": 1852304,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 0.14128153026103973,
      "learning_rate": 0.0007531446540880503,
      "loss": 0.4893,
      "num_input_tokens_seen": 1855728,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.11140955984592438,
      "learning_rate": 0.0007544549266247379,
      "loss": 0.5272,
      "num_input_tokens_seen": 1859344,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 0.1629018932580948,
      "learning_rate": 0.0007557651991614256,
      "loss": 0.5976,
      "num_input_tokens_seen": 1862384,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 0.12769703567028046,
      "learning_rate": 0.0007570754716981131,
      "loss": 0.5666,
      "num_input_tokens_seen": 1865360,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.1172448918223381,
      "learning_rate": 0.0007583857442348009,
      "loss": 0.4735,
      "num_input_tokens_seen": 1868272,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 0.2259131669998169,
      "learning_rate": 0.0007596960167714885,
      "loss": 0.5228,
      "num_input_tokens_seen": 1871984,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 0.11587788164615631,
      "learning_rate": 0.0007610062893081762,
      "loss": 0.5279,
      "num_input_tokens_seen": 1876176,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 0.16513656079769135,
      "learning_rate": 0.0007623165618448638,
      "loss": 0.5516,
      "num_input_tokens_seen": 1878992,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.16630913317203522,
      "learning_rate": 0.0007636268343815514,
      "loss": 0.4287,
      "num_input_tokens_seen": 1882736,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.29936566948890686,
      "learning_rate": 0.0007649371069182391,
      "loss": 0.6979,
      "num_input_tokens_seen": 1885616,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 0.11615362763404846,
      "learning_rate": 0.0007662473794549266,
      "loss": 0.4155,
      "num_input_tokens_seen": 1888848,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 0.15995100140571594,
      "learning_rate": 0.0007675576519916143,
      "loss": 0.5459,
      "num_input_tokens_seen": 1891792,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 0.13439276814460754,
      "learning_rate": 0.0007688679245283019,
      "loss": 0.6448,
      "num_input_tokens_seen": 1894544,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 0.14159497618675232,
      "learning_rate": 0.0007701781970649895,
      "loss": 0.4471,
      "num_input_tokens_seen": 1897264,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.09824809432029724,
      "learning_rate": 0.0007714884696016772,
      "loss": 0.3126,
      "num_input_tokens_seen": 1901904,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 0.10490361601114273,
      "learning_rate": 0.0007727987421383648,
      "loss": 0.5495,
      "num_input_tokens_seen": 1904848,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 0.13674819469451904,
      "learning_rate": 0.0007741090146750524,
      "loss": 0.5259,
      "num_input_tokens_seen": 1907696,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.1405385583639145,
      "learning_rate": 0.00077541928721174,
      "loss": 0.4677,
      "num_input_tokens_seen": 1910128,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 0.1547212451696396,
      "learning_rate": 0.0007767295597484277,
      "loss": 0.4919,
      "num_input_tokens_seen": 1913136,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.14281366765499115,
      "learning_rate": 0.0007780398322851153,
      "loss": 0.5161,
      "num_input_tokens_seen": 1917776,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 0.15818394720554352,
      "learning_rate": 0.0007793501048218029,
      "loss": 0.5325,
      "num_input_tokens_seen": 1920368,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.1691422313451767,
      "learning_rate": 0.0007806603773584906,
      "loss": 0.6414,
      "num_input_tokens_seen": 1922960,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.25800782442092896,
      "learning_rate": 0.0007819706498951781,
      "loss": 0.4745,
      "num_input_tokens_seen": 1925904,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.1569456160068512,
      "learning_rate": 0.0007832809224318658,
      "loss": 0.6032,
      "num_input_tokens_seen": 1933904,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 0.04984861612319946,
      "learning_rate": 0.0007845911949685535,
      "loss": 0.3317,
      "num_input_tokens_seen": 1936656,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.14625917375087738,
      "learning_rate": 0.0007859014675052411,
      "loss": 0.5478,
      "num_input_tokens_seen": 1939440,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.14217457175254822,
      "learning_rate": 0.0007872117400419288,
      "loss": 0.5434,
      "num_input_tokens_seen": 1942416,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.16737870872020721,
      "learning_rate": 0.0007885220125786164,
      "loss": 0.588,
      "num_input_tokens_seen": 1946288,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 0.05276871472597122,
      "learning_rate": 0.0007898322851153041,
      "loss": 0.4643,
      "num_input_tokens_seen": 1951792,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 0.07249964773654938,
      "learning_rate": 0.0007911425576519916,
      "loss": 0.4532,
      "num_input_tokens_seen": 1955504,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 0.12485864013433456,
      "learning_rate": 0.0007924528301886793,
      "loss": 0.5254,
      "num_input_tokens_seen": 1958224,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.10496941208839417,
      "learning_rate": 0.0007937631027253669,
      "loss": 0.3934,
      "num_input_tokens_seen": 1962320,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.3868362605571747,
      "learning_rate": 0.0007950733752620545,
      "loss": 0.396,
      "num_input_tokens_seen": 1965616,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 0.1257205307483673,
      "learning_rate": 0.0007963836477987422,
      "loss": 0.5609,
      "num_input_tokens_seen": 1968272,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 0.07177716493606567,
      "learning_rate": 0.0007976939203354298,
      "loss": 0.4636,
      "num_input_tokens_seen": 1971504,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 0.145207479596138,
      "learning_rate": 0.0007990041928721174,
      "loss": 0.5369,
      "num_input_tokens_seen": 1975312,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 0.0791095718741417,
      "learning_rate": 0.000800314465408805,
      "loss": 0.5006,
      "num_input_tokens_seen": 1977552,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.09369955956935883,
      "learning_rate": 0.0008016247379454927,
      "loss": 0.5248,
      "num_input_tokens_seen": 1980912,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 0.13717837631702423,
      "learning_rate": 0.0008029350104821803,
      "loss": 0.5173,
      "num_input_tokens_seen": 1983920,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 0.08488927036523819,
      "learning_rate": 0.0008042452830188679,
      "loss": 0.4386,
      "num_input_tokens_seen": 1987408,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.10279196500778198,
      "learning_rate": 0.0008055555555555556,
      "loss": 0.4703,
      "num_input_tokens_seen": 1991568,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 0.06373483687639236,
      "learning_rate": 0.0008068658280922431,
      "loss": 0.4625,
      "num_input_tokens_seen": 1995440,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.23222951591014862,
      "learning_rate": 0.0008081761006289308,
      "loss": 0.6685,
      "num_input_tokens_seen": 1999184,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.09610684961080551,
      "learning_rate": 0.0008094863731656184,
      "loss": 0.5055,
      "num_input_tokens_seen": 2002768,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 0.11988063901662827,
      "learning_rate": 0.0008107966457023061,
      "loss": 0.4295,
      "num_input_tokens_seen": 2005552,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 0.13108792901039124,
      "learning_rate": 0.0008121069182389938,
      "loss": 0.5125,
      "num_input_tokens_seen": 2008592,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 0.08842457085847855,
      "learning_rate": 0.0008134171907756814,
      "loss": 0.4978,
      "num_input_tokens_seen": 2012048,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 0.12009960412979126,
      "learning_rate": 0.0008147274633123691,
      "loss": 0.4535,
      "num_input_tokens_seen": 2015312,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 0.10558577626943588,
      "learning_rate": 0.0008160377358490566,
      "loss": 0.2887,
      "num_input_tokens_seen": 2021424,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 0.11667565256357193,
      "learning_rate": 0.0008173480083857443,
      "loss": 0.4864,
      "num_input_tokens_seen": 2024464,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 0.15993629395961761,
      "learning_rate": 0.0008186582809224319,
      "loss": 0.5786,
      "num_input_tokens_seen": 2027728,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 0.09301189333200455,
      "learning_rate": 0.0008199685534591195,
      "loss": 0.4544,
      "num_input_tokens_seen": 2031088,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.08544911444187164,
      "learning_rate": 0.0008212788259958072,
      "loss": 0.4611,
      "num_input_tokens_seen": 2034576,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 0.14519323408603668,
      "learning_rate": 0.0008225890985324948,
      "loss": 0.4867,
      "num_input_tokens_seen": 2039408,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 0.10200848430395126,
      "learning_rate": 0.0008238993710691824,
      "loss": 0.4243,
      "num_input_tokens_seen": 2042384,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.11919158697128296,
      "learning_rate": 0.00082520964360587,
      "loss": 0.5949,
      "num_input_tokens_seen": 2045200,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.07397947460412979,
      "learning_rate": 0.0008265199161425576,
      "loss": 0.3712,
      "num_input_tokens_seen": 2048720,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.08043144643306732,
      "learning_rate": 0.0008278301886792453,
      "loss": 0.4132,
      "num_input_tokens_seen": 2051824,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 0.43013474345207214,
      "learning_rate": 0.0008291404612159329,
      "loss": 0.5982,
      "num_input_tokens_seen": 2055952,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 0.14504198729991913,
      "learning_rate": 0.0008304507337526206,
      "loss": 0.4403,
      "num_input_tokens_seen": 2058512,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.15635667741298676,
      "learning_rate": 0.0008317610062893081,
      "loss": 0.3684,
      "num_input_tokens_seen": 2061232,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.10768473893404007,
      "learning_rate": 0.0008330712788259958,
      "loss": 0.4995,
      "num_input_tokens_seen": 2064336,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.10748642683029175,
      "learning_rate": 0.0008343815513626834,
      "loss": 0.6748,
      "num_input_tokens_seen": 2067984,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.09777948260307312,
      "learning_rate": 0.000835691823899371,
      "loss": 0.4279,
      "num_input_tokens_seen": 2070704,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.10206793993711472,
      "learning_rate": 0.0008370020964360588,
      "loss": 0.4786,
      "num_input_tokens_seen": 2074448,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 0.09777027368545532,
      "learning_rate": 0.0008383123689727464,
      "loss": 0.5169,
      "num_input_tokens_seen": 2077712,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 0.05405806377530098,
      "learning_rate": 0.000839622641509434,
      "loss": 0.4505,
      "num_input_tokens_seen": 2081808,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 0.1526460200548172,
      "learning_rate": 0.0008409329140461216,
      "loss": 0.479,
      "num_input_tokens_seen": 2086352,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 0.22724559903144836,
      "learning_rate": 0.0008422431865828093,
      "loss": 0.4856,
      "num_input_tokens_seen": 2088528,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 0.064881332218647,
      "learning_rate": 0.0008435534591194969,
      "loss": 0.3479,
      "num_input_tokens_seen": 2092688,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 0.12644635140895844,
      "learning_rate": 0.0008448637316561845,
      "loss": 0.4925,
      "num_input_tokens_seen": 2097232,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.08612029254436493,
      "learning_rate": 0.0008461740041928722,
      "loss": 0.5505,
      "num_input_tokens_seen": 2100016,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 0.24328956007957458,
      "learning_rate": 0.0008474842767295598,
      "loss": 0.4839,
      "num_input_tokens_seen": 2103568,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.07780156284570694,
      "learning_rate": 0.0008487945492662474,
      "loss": 0.4973,
      "num_input_tokens_seen": 2107792,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 0.09229113906621933,
      "learning_rate": 0.000850104821802935,
      "loss": 0.5306,
      "num_input_tokens_seen": 2110672,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 0.08931729197502136,
      "learning_rate": 0.0008514150943396226,
      "loss": 0.4757,
      "num_input_tokens_seen": 2113424,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.055585574358701706,
      "learning_rate": 0.0008527253668763103,
      "loss": 0.4548,
      "num_input_tokens_seen": 2116944,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 0.12955044209957123,
      "learning_rate": 0.0008540356394129979,
      "loss": 0.5558,
      "num_input_tokens_seen": 2120304,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 0.07322274148464203,
      "learning_rate": 0.0008553459119496856,
      "loss": 0.4992,
      "num_input_tokens_seen": 2124208,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 0.07956401258707047,
      "learning_rate": 0.0008566561844863731,
      "loss": 0.473,
      "num_input_tokens_seen": 2127536,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 0.08615370094776154,
      "learning_rate": 0.0008579664570230608,
      "loss": 0.3768,
      "num_input_tokens_seen": 2130896,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.0678657591342926,
      "learning_rate": 0.0008592767295597484,
      "loss": 0.5606,
      "num_input_tokens_seen": 2133968,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 0.08449095487594604,
      "learning_rate": 0.000860587002096436,
      "loss": 0.5378,
      "num_input_tokens_seen": 2137648,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 0.12254585325717926,
      "learning_rate": 0.0008618972746331238,
      "loss": 0.5685,
      "num_input_tokens_seen": 2140368,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 0.14766184985637665,
      "learning_rate": 0.0008632075471698114,
      "loss": 0.4027,
      "num_input_tokens_seen": 2143088,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.15608243644237518,
      "learning_rate": 0.000864517819706499,
      "loss": 0.597,
      "num_input_tokens_seen": 2146352,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 0.07854445278644562,
      "learning_rate": 0.0008658280922431866,
      "loss": 0.4695,
      "num_input_tokens_seen": 2150352,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.09818520396947861,
      "learning_rate": 0.0008671383647798742,
      "loss": 0.5415,
      "num_input_tokens_seen": 2153584,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.10152637958526611,
      "learning_rate": 0.0008684486373165619,
      "loss": 0.5263,
      "num_input_tokens_seen": 2157488,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 0.09998765587806702,
      "learning_rate": 0.0008697589098532495,
      "loss": 0.4566,
      "num_input_tokens_seen": 2160144,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 0.23207791149616241,
      "learning_rate": 0.0008710691823899372,
      "loss": 0.7099,
      "num_input_tokens_seen": 2162768,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 0.10398731380701065,
      "learning_rate": 0.0008723794549266247,
      "loss": 0.669,
      "num_input_tokens_seen": 2165616,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 0.22734211385250092,
      "learning_rate": 0.0008736897274633124,
      "loss": 0.4511,
      "num_input_tokens_seen": 2168656,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.10448348522186279,
      "learning_rate": 0.000875,
      "loss": 0.4306,
      "num_input_tokens_seen": 2172048,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 0.13799947500228882,
      "learning_rate": 0.0008763102725366876,
      "loss": 0.3893,
      "num_input_tokens_seen": 2175056,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.14085052907466888,
      "learning_rate": 0.0008776205450733753,
      "loss": 0.5543,
      "num_input_tokens_seen": 2178512,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.15252432227134705,
      "learning_rate": 0.0008789308176100629,
      "loss": 0.4808,
      "num_input_tokens_seen": 2182096,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.09644161909818649,
      "learning_rate": 0.0008802410901467506,
      "loss": 0.4555,
      "num_input_tokens_seen": 2185840,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.14264939725399017,
      "learning_rate": 0.0008815513626834381,
      "loss": 0.5172,
      "num_input_tokens_seen": 2188720,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.1110285297036171,
      "learning_rate": 0.0008828616352201258,
      "loss": 0.5401,
      "num_input_tokens_seen": 2192240,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 0.1219867542386055,
      "learning_rate": 0.0008841719077568134,
      "loss": 0.4379,
      "num_input_tokens_seen": 2196112,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.18457449972629547,
      "learning_rate": 0.000885482180293501,
      "loss": 0.5925,
      "num_input_tokens_seen": 2200272,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 0.07989726215600967,
      "learning_rate": 0.0008867924528301887,
      "loss": 0.5263,
      "num_input_tokens_seen": 2203024,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 0.10969426482915878,
      "learning_rate": 0.0008881027253668763,
      "loss": 0.4141,
      "num_input_tokens_seen": 2206000,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.11037498712539673,
      "learning_rate": 0.000889412997903564,
      "loss": 0.7173,
      "num_input_tokens_seen": 2209392,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 0.08658237755298615,
      "learning_rate": 0.0008907232704402516,
      "loss": 0.4416,
      "num_input_tokens_seen": 2214480,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 0.06048598885536194,
      "learning_rate": 0.0008920335429769392,
      "loss": 0.4896,
      "num_input_tokens_seen": 2217904,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.1213090643286705,
      "learning_rate": 0.0008933438155136269,
      "loss": 0.4756,
      "num_input_tokens_seen": 2221328,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.09207983314990997,
      "learning_rate": 0.0008946540880503145,
      "loss": 0.562,
      "num_input_tokens_seen": 2224304,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.059591539204120636,
      "learning_rate": 0.0008959643605870022,
      "loss": 0.3955,
      "num_input_tokens_seen": 2226928,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.13474611937999725,
      "learning_rate": 0.0008972746331236897,
      "loss": 0.5797,
      "num_input_tokens_seen": 2230032,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 0.07442762702703476,
      "learning_rate": 0.0008985849056603774,
      "loss": 0.6297,
      "num_input_tokens_seen": 2232976,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 0.11629346013069153,
      "learning_rate": 0.000899895178197065,
      "loss": 0.4695,
      "num_input_tokens_seen": 2235664,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 0.10438309609889984,
      "learning_rate": 0.0009012054507337526,
      "loss": 0.4583,
      "num_input_tokens_seen": 2238448,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.11152135580778122,
      "learning_rate": 0.0009025157232704403,
      "loss": 0.4612,
      "num_input_tokens_seen": 2241456,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 0.13446027040481567,
      "learning_rate": 0.0009038259958071279,
      "loss": 0.5867,
      "num_input_tokens_seen": 2244848,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 0.08192357420921326,
      "learning_rate": 0.0009051362683438156,
      "loss": 0.387,
      "num_input_tokens_seen": 2247216,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.09662729501724243,
      "learning_rate": 0.0009064465408805031,
      "loss": 0.6554,
      "num_input_tokens_seen": 2250160,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.1385406255722046,
      "learning_rate": 0.0009077568134171907,
      "loss": 0.45,
      "num_input_tokens_seen": 2252784,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 0.1185065358877182,
      "learning_rate": 0.0009090670859538784,
      "loss": 0.4851,
      "num_input_tokens_seen": 2255504,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.055586427450180054,
      "learning_rate": 0.000910377358490566,
      "loss": 0.4736,
      "num_input_tokens_seen": 2261040,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.12279147654771805,
      "learning_rate": 0.0009116876310272537,
      "loss": 0.4346,
      "num_input_tokens_seen": 2264656,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 0.10028264671564102,
      "learning_rate": 0.0009129979035639413,
      "loss": 0.4021,
      "num_input_tokens_seen": 2268368,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.13441644608974457,
      "learning_rate": 0.000914308176100629,
      "loss": 0.5934,
      "num_input_tokens_seen": 2271056,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 0.1253141611814499,
      "learning_rate": 0.0009156184486373166,
      "loss": 0.8365,
      "num_input_tokens_seen": 2274672,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.12501375377178192,
      "learning_rate": 0.0009169287211740042,
      "loss": 0.4064,
      "num_input_tokens_seen": 2278064,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.20647640526294708,
      "learning_rate": 0.0009182389937106919,
      "loss": 0.5108,
      "num_input_tokens_seen": 2280400,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.10071535408496857,
      "learning_rate": 0.0009195492662473795,
      "loss": 0.544,
      "num_input_tokens_seen": 2283728,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 0.11583755165338516,
      "learning_rate": 0.0009208595387840672,
      "loss": 0.4519,
      "num_input_tokens_seen": 2286448,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 0.12298732250928879,
      "learning_rate": 0.0009221698113207547,
      "loss": 0.548,
      "num_input_tokens_seen": 2289648,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.11077401787042618,
      "learning_rate": 0.0009234800838574424,
      "loss": 0.4397,
      "num_input_tokens_seen": 2292528,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.11848250776529312,
      "learning_rate": 0.00092479035639413,
      "loss": 0.4383,
      "num_input_tokens_seen": 2295472,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 0.10221804678440094,
      "learning_rate": 0.0009261006289308176,
      "loss": 0.4506,
      "num_input_tokens_seen": 2298800,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.24359259009361267,
      "learning_rate": 0.0009274109014675053,
      "loss": 0.5774,
      "num_input_tokens_seen": 2301520,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 0.1302942931652069,
      "learning_rate": 0.0009287211740041929,
      "loss": 0.4118,
      "num_input_tokens_seen": 2304784,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.1918763369321823,
      "learning_rate": 0.0009300314465408806,
      "loss": 0.521,
      "num_input_tokens_seen": 2307728,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 0.16535894572734833,
      "learning_rate": 0.0009313417190775681,
      "loss": 0.481,
      "num_input_tokens_seen": 2311888,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 0.09034889191389084,
      "learning_rate": 0.0009326519916142557,
      "loss": 0.4526,
      "num_input_tokens_seen": 2315536,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.35314929485321045,
      "learning_rate": 0.0009339622641509434,
      "loss": 0.5515,
      "num_input_tokens_seen": 2318096,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.14228682219982147,
      "learning_rate": 0.000935272536687631,
      "loss": 0.4313,
      "num_input_tokens_seen": 2321168,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.15851131081581116,
      "learning_rate": 0.0009365828092243187,
      "loss": 0.4951,
      "num_input_tokens_seen": 2323824,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 0.07727108150720596,
      "learning_rate": 0.0009378930817610063,
      "loss": 0.5668,
      "num_input_tokens_seen": 2326672,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.0743052214384079,
      "learning_rate": 0.0009392033542976939,
      "loss": 0.4095,
      "num_input_tokens_seen": 2330640,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.08158224821090698,
      "learning_rate": 0.0009405136268343816,
      "loss": 0.4044,
      "num_input_tokens_seen": 2333232,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 0.0631183534860611,
      "learning_rate": 0.0009418238993710692,
      "loss": 0.4982,
      "num_input_tokens_seen": 2336560,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.14554853737354279,
      "learning_rate": 0.0009431341719077569,
      "loss": 0.5898,
      "num_input_tokens_seen": 2339632,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.09227868914604187,
      "learning_rate": 0.0009444444444444445,
      "loss": 0.509,
      "num_input_tokens_seen": 2347440,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 0.06756574660539627,
      "learning_rate": 0.0009457547169811322,
      "loss": 0.4609,
      "num_input_tokens_seen": 2350992,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 0.059271249920129776,
      "learning_rate": 0.0009470649895178197,
      "loss": 0.4771,
      "num_input_tokens_seen": 2354384,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.1724841147661209,
      "learning_rate": 0.0009483752620545073,
      "loss": 0.6047,
      "num_input_tokens_seen": 2356752,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 0.11865495145320892,
      "learning_rate": 0.000949685534591195,
      "loss": 0.4002,
      "num_input_tokens_seen": 2359728,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 0.1330760419368744,
      "learning_rate": 0.0009509958071278826,
      "loss": 0.4928,
      "num_input_tokens_seen": 2363728,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.12562058866024017,
      "learning_rate": 0.0009523060796645703,
      "loss": 0.5778,
      "num_input_tokens_seen": 2366480,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.17142583429813385,
      "learning_rate": 0.0009536163522012579,
      "loss": 0.4329,
      "num_input_tokens_seen": 2369232,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 0.14634916186332703,
      "learning_rate": 0.0009549266247379456,
      "loss": 0.6186,
      "num_input_tokens_seen": 2372688,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.0955924540758133,
      "learning_rate": 0.0009562368972746331,
      "loss": 0.5194,
      "num_input_tokens_seen": 2376976,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.19144368171691895,
      "learning_rate": 0.0009575471698113207,
      "loss": 0.4431,
      "num_input_tokens_seen": 2380112,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.22801358997821808,
      "learning_rate": 0.0009588574423480084,
      "loss": 0.4795,
      "num_input_tokens_seen": 2382608,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.10407861322164536,
      "learning_rate": 0.000960167714884696,
      "loss": 0.567,
      "num_input_tokens_seen": 2385424,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.11862916499376297,
      "learning_rate": 0.0009614779874213837,
      "loss": 0.5339,
      "num_input_tokens_seen": 2390032,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.10374821722507477,
      "learning_rate": 0.0009627882599580712,
      "loss": 0.4209,
      "num_input_tokens_seen": 2393328,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.09376797825098038,
      "learning_rate": 0.0009640985324947589,
      "loss": 0.3794,
      "num_input_tokens_seen": 2396432,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.16699697077274323,
      "learning_rate": 0.0009654088050314465,
      "loss": 0.5121,
      "num_input_tokens_seen": 2399952,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.10591643303632736,
      "learning_rate": 0.0009667190775681342,
      "loss": 0.4302,
      "num_input_tokens_seen": 2404240,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 0.1383853703737259,
      "learning_rate": 0.0009680293501048219,
      "loss": 0.3524,
      "num_input_tokens_seen": 2408080,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 0.12838782370090485,
      "learning_rate": 0.0009693396226415095,
      "loss": 0.4824,
      "num_input_tokens_seen": 2412080,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 0.06368566304445267,
      "learning_rate": 0.0009706498951781972,
      "loss": 0.3581,
      "num_input_tokens_seen": 2417136,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.23236505687236786,
      "learning_rate": 0.0009719601677148847,
      "loss": 0.4409,
      "num_input_tokens_seen": 2420720,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.1180746927857399,
      "learning_rate": 0.0009732704402515723,
      "loss": 0.4507,
      "num_input_tokens_seen": 2423632,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.20801745355129242,
      "learning_rate": 0.00097458071278826,
      "loss": 0.5608,
      "num_input_tokens_seen": 2426704,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 0.14209116995334625,
      "learning_rate": 0.0009758909853249476,
      "loss": 0.476,
      "num_input_tokens_seen": 2431056,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 0.1323792189359665,
      "learning_rate": 0.0009772012578616353,
      "loss": 0.35,
      "num_input_tokens_seen": 2433808,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.0921485647559166,
      "learning_rate": 0.0009785115303983228,
      "loss": 0.4395,
      "num_input_tokens_seen": 2437008,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.0820334181189537,
      "learning_rate": 0.0009798218029350107,
      "loss": 0.6408,
      "num_input_tokens_seen": 2441520,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.08448135107755661,
      "learning_rate": 0.0009811320754716981,
      "loss": 0.3493,
      "num_input_tokens_seen": 2444400,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.1544097363948822,
      "learning_rate": 0.0009824423480083858,
      "loss": 0.4838,
      "num_input_tokens_seen": 2447248,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.08867917954921722,
      "learning_rate": 0.0009837526205450735,
      "loss": 0.4921,
      "num_input_tokens_seen": 2450352,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.07707353681325912,
      "learning_rate": 0.000985062893081761,
      "loss": 0.5671,
      "num_input_tokens_seen": 2453072,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 0.10870086401700974,
      "learning_rate": 0.0009863731656184487,
      "loss": 0.4291,
      "num_input_tokens_seen": 2456112,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.09777383506298065,
      "learning_rate": 0.0009876834381551364,
      "loss": 0.4326,
      "num_input_tokens_seen": 2459280,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 0.1475686877965927,
      "learning_rate": 0.0009889937106918238,
      "loss": 0.4166,
      "num_input_tokens_seen": 2462000,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.11835259944200516,
      "learning_rate": 0.0009903039832285115,
      "loss": 0.4925,
      "num_input_tokens_seen": 2464592,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.10018587112426758,
      "learning_rate": 0.0009916142557651992,
      "loss": 0.5555,
      "num_input_tokens_seen": 2468784,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.23387600481510162,
      "learning_rate": 0.000992924528301887,
      "loss": 0.3722,
      "num_input_tokens_seen": 2471824,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 0.10926979035139084,
      "learning_rate": 0.0009942348008385744,
      "loss": 0.7072,
      "num_input_tokens_seen": 2474704,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 0.216945618391037,
      "learning_rate": 0.000995545073375262,
      "loss": 0.567,
      "num_input_tokens_seen": 2477520,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 0.08332747220993042,
      "learning_rate": 0.0009968553459119497,
      "loss": 0.3641,
      "num_input_tokens_seen": 2480528,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.11077765375375748,
      "learning_rate": 0.0009981656184486372,
      "loss": 0.5017,
      "num_input_tokens_seen": 2484560,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.1203056052327156,
      "learning_rate": 0.000999475890985325,
      "loss": 0.5121,
      "num_input_tokens_seen": 2489712,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.496399462223053,
      "eval_runtime": 13.7445,
      "eval_samples_per_second": 61.698,
      "eval_steps_per_second": 15.424,
      "num_input_tokens_seen": 2489840,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.08070757985115051,
      "learning_rate": 0.0009999999811730266,
      "loss": 0.4879,
      "num_input_tokens_seen": 2493040,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.08428430557250977,
      "learning_rate": 0.000999999866119305,
      "loss": 0.5338,
      "num_input_tokens_seen": 2496240,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 0.11376864463090897,
      "learning_rate": 0.000999999646471316,
      "loss": 0.5256,
      "num_input_tokens_seen": 2499216,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 0.10969184339046478,
      "learning_rate": 0.0009999993222291053,
      "loss": 0.5319,
      "num_input_tokens_seen": 2502352,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.11945624649524689,
      "learning_rate": 0.0009999988933927404,
      "loss": 0.5403,
      "num_input_tokens_seen": 2505584,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.07029696553945541,
      "learning_rate": 0.0009999983599623115,
      "loss": 0.4196,
      "num_input_tokens_seen": 2509296,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 0.07300016283988953,
      "learning_rate": 0.0009999977219379299,
      "loss": 0.4609,
      "num_input_tokens_seen": 2512240,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.13056452572345734,
      "learning_rate": 0.000999996979319729,
      "loss": 0.4393,
      "num_input_tokens_seen": 2515728,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.05815855413675308,
      "learning_rate": 0.0009999961321078645,
      "loss": 0.3968,
      "num_input_tokens_seen": 2519184,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 0.1208338513970375,
      "learning_rate": 0.0009999951803025134,
      "loss": 0.4077,
      "num_input_tokens_seen": 2522224,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.0640949234366417,
      "learning_rate": 0.0009999941239038747,
      "loss": 0.3915,
      "num_input_tokens_seen": 2525744,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.13630498945713043,
      "learning_rate": 0.0009999929629121696,
      "loss": 0.4206,
      "num_input_tokens_seen": 2528784,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 0.04225802421569824,
      "learning_rate": 0.000999991697327641,
      "loss": 0.5271,
      "num_input_tokens_seen": 2533040,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.07157709449529648,
      "learning_rate": 0.0009999903271505535,
      "loss": 0.4283,
      "num_input_tokens_seen": 2536848,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.09688005596399307,
      "learning_rate": 0.0009999888523811935,
      "loss": 0.4104,
      "num_input_tokens_seen": 2540112,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.08327209949493408,
      "learning_rate": 0.00099998727301987,
      "loss": 0.4591,
      "num_input_tokens_seen": 2543184,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.13138322532176971,
      "learning_rate": 0.000999985589066913,
      "loss": 0.4169,
      "num_input_tokens_seen": 2545808,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 0.06569751352071762,
      "learning_rate": 0.000999983800522675,
      "loss": 0.3426,
      "num_input_tokens_seen": 2548816,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 0.162979394197464,
      "learning_rate": 0.00099998190738753,
      "loss": 0.5472,
      "num_input_tokens_seen": 2551248,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 0.0828622430562973,
      "learning_rate": 0.000999979909661874,
      "loss": 0.5771,
      "num_input_tokens_seen": 2555152,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 0.12566207349300385,
      "learning_rate": 0.000999977807346125,
      "loss": 0.3919,
      "num_input_tokens_seen": 2558864,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 0.11847173422574997,
      "learning_rate": 0.000999975600440723,
      "loss": 0.5069,
      "num_input_tokens_seen": 2562800,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 0.08273668587207794,
      "learning_rate": 0.000999973288946129,
      "loss": 0.4634,
      "num_input_tokens_seen": 2565904,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 0.0677230954170227,
      "learning_rate": 0.0009999708728628271,
      "loss": 0.4717,
      "num_input_tokens_seen": 2569104,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 0.11968893557786942,
      "learning_rate": 0.0009999683521913227,
      "loss": 0.5233,
      "num_input_tokens_seen": 2571888,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.07807502150535583,
      "learning_rate": 0.000999965726932143,
      "loss": 0.4899,
      "num_input_tokens_seen": 2575184,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 0.15865930914878845,
      "learning_rate": 0.0009999629970858371,
      "loss": 0.5132,
      "num_input_tokens_seen": 2578064,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 0.0903092548251152,
      "learning_rate": 0.000999960162652976,
      "loss": 0.3819,
      "num_input_tokens_seen": 2580848,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.08131546527147293,
      "learning_rate": 0.0009999572236341527,
      "loss": 0.4967,
      "num_input_tokens_seen": 2583920,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 0.14114364981651306,
      "learning_rate": 0.000999954180029982,
      "loss": 0.44,
      "num_input_tokens_seen": 2586992,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 0.09735510498285294,
      "learning_rate": 0.0009999510318411007,
      "loss": 0.4527,
      "num_input_tokens_seen": 2589136,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.11506585031747818,
      "learning_rate": 0.0009999477790681673,
      "loss": 0.5053,
      "num_input_tokens_seen": 2591760,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 0.08320196717977524,
      "learning_rate": 0.000999944421711862,
      "loss": 0.4892,
      "num_input_tokens_seen": 2594576,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.09628415107727051,
      "learning_rate": 0.0009999409597728876,
      "loss": 0.5983,
      "num_input_tokens_seen": 2597872,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 0.252876877784729,
      "learning_rate": 0.0009999373932519679,
      "loss": 0.41,
      "num_input_tokens_seen": 2600752,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.06424782425165176,
      "learning_rate": 0.0009999337221498492,
      "loss": 0.55,
      "num_input_tokens_seen": 2607248,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.09706594049930573,
      "learning_rate": 0.0009999299464672996,
      "loss": 0.3812,
      "num_input_tokens_seen": 2610672,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.09846962988376617,
      "learning_rate": 0.0009999260662051084,
      "loss": 0.4076,
      "num_input_tokens_seen": 2615120,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 0.0930904895067215,
      "learning_rate": 0.0009999220813640877,
      "loss": 0.4056,
      "num_input_tokens_seen": 2618640,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.0676828920841217,
      "learning_rate": 0.000999917991945071,
      "loss": 0.4263,
      "num_input_tokens_seen": 2621744,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.09056296199560165,
      "learning_rate": 0.0009999137979489136,
      "loss": 0.4951,
      "num_input_tokens_seen": 2624496,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 0.07106060534715652,
      "learning_rate": 0.000999909499376493,
      "loss": 0.3394,
      "num_input_tokens_seen": 2627984,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 0.15741220116615295,
      "learning_rate": 0.0009999050962287085,
      "loss": 0.5664,
      "num_input_tokens_seen": 2631024,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.11342083662748337,
      "learning_rate": 0.0009999005885064811,
      "loss": 0.5265,
      "num_input_tokens_seen": 2634000,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.17619171738624573,
      "learning_rate": 0.0009998959762107537,
      "loss": 0.4638,
      "num_input_tokens_seen": 2637488,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.08018229901790619,
      "learning_rate": 0.0009998912593424914,
      "loss": 0.5,
      "num_input_tokens_seen": 2643376,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.06574707478284836,
      "learning_rate": 0.0009998864379026804,
      "loss": 0.4514,
      "num_input_tokens_seen": 2646000,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.1512468159198761,
      "learning_rate": 0.00099988151189233,
      "loss": 0.4615,
      "num_input_tokens_seen": 2649456,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 0.11801764369010925,
      "learning_rate": 0.00099987648131247,
      "loss": 0.4202,
      "num_input_tokens_seen": 2652016,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.08969254046678543,
      "learning_rate": 0.000999871346164153,
      "loss": 0.6427,
      "num_input_tokens_seen": 2654992,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 0.14565709233283997,
      "learning_rate": 0.0009998661064484532,
      "loss": 0.5264,
      "num_input_tokens_seen": 2657648,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 0.09373481571674347,
      "learning_rate": 0.0009998607621664666,
      "loss": 0.4308,
      "num_input_tokens_seen": 2660560,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.07205356657505035,
      "learning_rate": 0.0009998553133193113,
      "loss": 0.4217,
      "num_input_tokens_seen": 2663440,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 0.13177427649497986,
      "learning_rate": 0.0009998497599081272,
      "loss": 0.6687,
      "num_input_tokens_seen": 2666544,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 0.07985863834619522,
      "learning_rate": 0.0009998441019340761,
      "loss": 0.4419,
      "num_input_tokens_seen": 2669616,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 0.1067107692360878,
      "learning_rate": 0.000999838339398341,
      "loss": 0.4817,
      "num_input_tokens_seen": 2673008,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.06537681818008423,
      "learning_rate": 0.000999832472302128,
      "loss": 0.3966,
      "num_input_tokens_seen": 2676016,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.10809481889009476,
      "learning_rate": 0.0009998265006466642,
      "loss": 0.5061,
      "num_input_tokens_seen": 2679920,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 0.07688765227794647,
      "learning_rate": 0.0009998204244331987,
      "loss": 0.5206,
      "num_input_tokens_seen": 2683728,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.11737851053476334,
      "learning_rate": 0.0009998142436630027,
      "loss": 0.5651,
      "num_input_tokens_seen": 2686960,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.07049959897994995,
      "learning_rate": 0.0009998079583373692,
      "loss": 0.4205,
      "num_input_tokens_seen": 2690480,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.14211505651474,
      "learning_rate": 0.0009998015684576128,
      "loss": 0.4428,
      "num_input_tokens_seen": 2694064,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.12269824743270874,
      "learning_rate": 0.0009997950740250703,
      "loss": 0.5007,
      "num_input_tokens_seen": 2696880,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 0.3339681327342987,
      "learning_rate": 0.0009997884750411004,
      "loss": 0.529,
      "num_input_tokens_seen": 2699856,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.10427509993314743,
      "learning_rate": 0.0009997817715070832,
      "loss": 0.4507,
      "num_input_tokens_seen": 2703344,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.16276125609874725,
      "learning_rate": 0.0009997749634244213,
      "loss": 0.5492,
      "num_input_tokens_seen": 2706736,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.12513822317123413,
      "learning_rate": 0.000999768050794539,
      "loss": 0.4897,
      "num_input_tokens_seen": 2709968,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 0.07109865546226501,
      "learning_rate": 0.0009997610336188818,
      "loss": 0.4477,
      "num_input_tokens_seen": 2713456,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.09007564187049866,
      "learning_rate": 0.0009997539118989183,
      "loss": 0.3751,
      "num_input_tokens_seen": 2715952,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.0696859061717987,
      "learning_rate": 0.0009997466856361376,
      "loss": 0.5396,
      "num_input_tokens_seen": 2719120,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.12012071907520294,
      "learning_rate": 0.000999739354832052,
      "loss": 0.4284,
      "num_input_tokens_seen": 2722128,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.11022955179214478,
      "learning_rate": 0.0009997319194881944,
      "loss": 0.5778,
      "num_input_tokens_seen": 2725424,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 0.05947507545351982,
      "learning_rate": 0.0009997243796061204,
      "loss": 0.465,
      "num_input_tokens_seen": 2730640,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 0.16789376735687256,
      "learning_rate": 0.0009997167351874076,
      "loss": 0.6115,
      "num_input_tokens_seen": 2733104,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.12113304436206818,
      "learning_rate": 0.0009997089862336548,
      "loss": 0.4015,
      "num_input_tokens_seen": 2735856,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.08681340515613556,
      "learning_rate": 0.000999701132746483,
      "loss": 0.6048,
      "num_input_tokens_seen": 2738832,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.10473979264497757,
      "learning_rate": 0.0009996931747275352,
      "loss": 0.5141,
      "num_input_tokens_seen": 2742416,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 0.10682865977287292,
      "learning_rate": 0.000999685112178476,
      "loss": 0.4521,
      "num_input_tokens_seen": 2746096,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 0.13714554905891418,
      "learning_rate": 0.0009996769451009922,
      "loss": 0.4118,
      "num_input_tokens_seen": 2748560,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 0.09668846428394318,
      "learning_rate": 0.000999668673496792,
      "loss": 0.4002,
      "num_input_tokens_seen": 2750896,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 0.07350906729698181,
      "learning_rate": 0.0009996602973676057,
      "loss": 0.3923,
      "num_input_tokens_seen": 2753968,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 0.10695020854473114,
      "learning_rate": 0.0009996518167151858,
      "loss": 0.4837,
      "num_input_tokens_seen": 2757840,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.07276348769664764,
      "learning_rate": 0.000999643231541306,
      "loss": 0.5338,
      "num_input_tokens_seen": 2761936,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 0.14672090113162994,
      "learning_rate": 0.0009996345418477625,
      "loss": 0.565,
      "num_input_tokens_seen": 2764592,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.0766868144273758,
      "learning_rate": 0.000999625747636373,
      "loss": 0.4012,
      "num_input_tokens_seen": 2767408,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 0.08475970476865768,
      "learning_rate": 0.000999616848908977,
      "loss": 0.3973,
      "num_input_tokens_seen": 2769808,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 0.10949122160673141,
      "learning_rate": 0.0009996078456674363,
      "loss": 0.452,
      "num_input_tokens_seen": 2772944,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 0.11951322853565216,
      "learning_rate": 0.000999598737913634,
      "loss": 0.456,
      "num_input_tokens_seen": 2775856,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.13504543900489807,
      "learning_rate": 0.0009995895256494755,
      "loss": 0.4896,
      "num_input_tokens_seen": 2778736,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 0.057380154728889465,
      "learning_rate": 0.000999580208876888,
      "loss": 0.4526,
      "num_input_tokens_seen": 2782608,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 0.12863275408744812,
      "learning_rate": 0.0009995707875978198,
      "loss": 0.5986,
      "num_input_tokens_seen": 2785040,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.14852997660636902,
      "learning_rate": 0.0009995612618142428,
      "loss": 0.5072,
      "num_input_tokens_seen": 2788304,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.09200312197208405,
      "learning_rate": 0.000999551631528149,
      "loss": 0.5045,
      "num_input_tokens_seen": 2791248,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 0.09179799258708954,
      "learning_rate": 0.000999541896741553,
      "loss": 0.432,
      "num_input_tokens_seen": 2794736,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 0.10532696545124054,
      "learning_rate": 0.0009995320574564912,
      "loss": 0.4886,
      "num_input_tokens_seen": 2798000,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.12422911822795868,
      "learning_rate": 0.000999522113675022,
      "loss": 0.4156,
      "num_input_tokens_seen": 2800912,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.10101012885570526,
      "learning_rate": 0.0009995120653992255,
      "loss": 0.3996,
      "num_input_tokens_seen": 2803984,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.1211748942732811,
      "learning_rate": 0.0009995019126312035,
      "loss": 0.4243,
      "num_input_tokens_seen": 2806640,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 0.17340560257434845,
      "learning_rate": 0.0009994916553730802,
      "loss": 0.5321,
      "num_input_tokens_seen": 2809360,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 0.09230966866016388,
      "learning_rate": 0.000999481293627001,
      "loss": 0.4649,
      "num_input_tokens_seen": 2811984,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.23413749039173126,
      "learning_rate": 0.0009994708273951337,
      "loss": 0.6067,
      "num_input_tokens_seen": 2815440,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 0.14111419022083282,
      "learning_rate": 0.0009994602566796673,
      "loss": 0.5596,
      "num_input_tokens_seen": 2818192,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 0.09623762965202332,
      "learning_rate": 0.0009994495814828136,
      "loss": 0.3679,
      "num_input_tokens_seen": 2821744,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.10509208589792252,
      "learning_rate": 0.0009994388018068055,
      "loss": 0.6054,
      "num_input_tokens_seen": 2825520,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 0.09654965996742249,
      "learning_rate": 0.0009994279176538977,
      "loss": 0.5158,
      "num_input_tokens_seen": 2829776,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 0.11940756440162659,
      "learning_rate": 0.0009994169290263675,
      "loss": 0.4785,
      "num_input_tokens_seen": 2832624,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.12254112958908081,
      "learning_rate": 0.0009994058359265132,
      "loss": 0.5661,
      "num_input_tokens_seen": 2835536,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.19487079977989197,
      "learning_rate": 0.0009993946383566558,
      "loss": 0.6126,
      "num_input_tokens_seen": 2838384,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 0.09125127643346786,
      "learning_rate": 0.0009993833363191374,
      "loss": 0.49,
      "num_input_tokens_seen": 2842032,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.09150523692369461,
      "learning_rate": 0.0009993719298163222,
      "loss": 0.4134,
      "num_input_tokens_seen": 2844720,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.18292340636253357,
      "learning_rate": 0.0009993604188505965,
      "loss": 0.3871,
      "num_input_tokens_seen": 2847888,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 0.09110405296087265,
      "learning_rate": 0.000999348803424368,
      "loss": 0.7515,
      "num_input_tokens_seen": 2850800,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.07822870463132858,
      "learning_rate": 0.000999337083540067,
      "loss": 0.4686,
      "num_input_tokens_seen": 2853936,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 0.1573552042245865,
      "learning_rate": 0.0009993252592001448,
      "loss": 0.5414,
      "num_input_tokens_seen": 2857040,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 0.055105894804000854,
      "learning_rate": 0.0009993133304070747,
      "loss": 0.4615,
      "num_input_tokens_seen": 2861264,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 0.06494386494159698,
      "learning_rate": 0.0009993012971633527,
      "loss": 0.5137,
      "num_input_tokens_seen": 2864848,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.20083625614643097,
      "learning_rate": 0.0009992891594714952,
      "loss": 0.4293,
      "num_input_tokens_seen": 2867280,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.08145561069250107,
      "learning_rate": 0.0009992769173340422,
      "loss": 0.6236,
      "num_input_tokens_seen": 2870288,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 0.08663763850927353,
      "learning_rate": 0.000999264570753554,
      "loss": 0.4785,
      "num_input_tokens_seen": 2872912,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 0.14614027738571167,
      "learning_rate": 0.0009992521197326135,
      "loss": 0.5301,
      "num_input_tokens_seen": 2877296,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 0.09460891038179398,
      "learning_rate": 0.0009992395642738252,
      "loss": 0.5492,
      "num_input_tokens_seen": 2880464,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 0.18624775111675262,
      "learning_rate": 0.0009992269043798158,
      "loss": 0.4832,
      "num_input_tokens_seen": 2882896,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.07358374446630478,
      "learning_rate": 0.0009992141400532336,
      "loss": 0.417,
      "num_input_tokens_seen": 2886384,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.0825897604227066,
      "learning_rate": 0.0009992012712967484,
      "loss": 0.5384,
      "num_input_tokens_seen": 2890224,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.15516254305839539,
      "learning_rate": 0.0009991882981130525,
      "loss": 0.3691,
      "num_input_tokens_seen": 2893392,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 0.10975851118564606,
      "learning_rate": 0.0009991752205048597,
      "loss": 0.4444,
      "num_input_tokens_seen": 2896240,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.17004717886447906,
      "learning_rate": 0.0009991620384749058,
      "loss": 0.4641,
      "num_input_tokens_seen": 2901520,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 0.13020943105220795,
      "learning_rate": 0.0009991487520259479,
      "loss": 0.545,
      "num_input_tokens_seen": 2903856,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.1272718459367752,
      "learning_rate": 0.000999135361160766,
      "loss": 0.442,
      "num_input_tokens_seen": 2906128,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.07259184867143631,
      "learning_rate": 0.0009991218658821608,
      "loss": 0.5346,
      "num_input_tokens_seen": 2908848,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.0869549959897995,
      "learning_rate": 0.0009991082661929556,
      "loss": 0.4193,
      "num_input_tokens_seen": 2912976,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.13750730454921722,
      "learning_rate": 0.000999094562095995,
      "loss": 0.4325,
      "num_input_tokens_seen": 2915888,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.13342958688735962,
      "learning_rate": 0.0009990807535941461,
      "loss": 0.4018,
      "num_input_tokens_seen": 2918928,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.13482126593589783,
      "learning_rate": 0.0009990668406902976,
      "loss": 0.5756,
      "num_input_tokens_seen": 2921392,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.11364055424928665,
      "learning_rate": 0.0009990528233873594,
      "loss": 0.4396,
      "num_input_tokens_seen": 2924112,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 0.10461170226335526,
      "learning_rate": 0.0009990387016882642,
      "loss": 0.512,
      "num_input_tokens_seen": 2927792,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.06681716442108154,
      "learning_rate": 0.000999024475595966,
      "loss": 0.6341,
      "num_input_tokens_seen": 2931984,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.05492294207215309,
      "learning_rate": 0.0009990101451134406,
      "loss": 0.4606,
      "num_input_tokens_seen": 2935856,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.11444459855556488,
      "learning_rate": 0.0009989957102436858,
      "loss": 0.4489,
      "num_input_tokens_seen": 2939760,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.08454084396362305,
      "learning_rate": 0.0009989811709897212,
      "loss": 0.5855,
      "num_input_tokens_seen": 2944112,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.10599634796380997,
      "learning_rate": 0.0009989665273545884,
      "loss": 0.4066,
      "num_input_tokens_seen": 2947184,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.23932266235351562,
      "learning_rate": 0.0009989517793413507,
      "loss": 0.4287,
      "num_input_tokens_seen": 2950672,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 0.2210950255393982,
      "learning_rate": 0.000998936926953093,
      "loss": 0.5616,
      "num_input_tokens_seen": 2953872,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 0.1265001893043518,
      "learning_rate": 0.0009989219701929224,
      "loss": 0.4965,
      "num_input_tokens_seen": 2956624,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.17207300662994385,
      "learning_rate": 0.0009989069090639675,
      "loss": 0.4106,
      "num_input_tokens_seen": 2959824,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 0.0972542092204094,
      "learning_rate": 0.0009988917435693793,
      "loss": 0.4483,
      "num_input_tokens_seen": 2962736,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 0.11235502362251282,
      "learning_rate": 0.0009988764737123297,
      "loss": 0.5022,
      "num_input_tokens_seen": 2965872,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 0.054667357355356216,
      "learning_rate": 0.0009988610994960134,
      "loss": 0.3298,
      "num_input_tokens_seen": 2969104,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.05735819786787033,
      "learning_rate": 0.0009988456209236464,
      "loss": 0.5718,
      "num_input_tokens_seen": 2973616,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 0.15235044062137604,
      "learning_rate": 0.0009988300379984668,
      "loss": 0.5696,
      "num_input_tokens_seen": 2975888,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 0.08411705493927002,
      "learning_rate": 0.0009988143507237341,
      "loss": 0.5107,
      "num_input_tokens_seen": 2979088,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 0.12426844239234924,
      "learning_rate": 0.00099879855910273,
      "loss": 0.5222,
      "num_input_tokens_seen": 2982576,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 0.16380321979522705,
      "learning_rate": 0.0009987826631387578,
      "loss": 0.4425,
      "num_input_tokens_seen": 2986480,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.07443448901176453,
      "learning_rate": 0.000998766662835143,
      "loss": 0.6338,
      "num_input_tokens_seen": 2990096,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 0.16877596080303192,
      "learning_rate": 0.0009987505581952325,
      "loss": 0.6223,
      "num_input_tokens_seen": 2993040,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.10031921416521072,
      "learning_rate": 0.0009987343492223954,
      "loss": 0.5228,
      "num_input_tokens_seen": 2997008,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 0.0872577652335167,
      "learning_rate": 0.0009987180359200222,
      "loss": 0.4226,
      "num_input_tokens_seen": 3000432,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.06968027353286743,
      "learning_rate": 0.0009987016182915257,
      "loss": 0.4817,
      "num_input_tokens_seen": 3003952,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 0.16465170681476593,
      "learning_rate": 0.0009986850963403398,
      "loss": 0.4419,
      "num_input_tokens_seen": 3006448,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.07302199304103851,
      "learning_rate": 0.0009986684700699214,
      "loss": 0.4217,
      "num_input_tokens_seen": 3009808,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.0976240262389183,
      "learning_rate": 0.000998651739483748,
      "loss": 0.3905,
      "num_input_tokens_seen": 3013392,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.13689006865024567,
      "learning_rate": 0.0009986349045853196,
      "loss": 0.6548,
      "num_input_tokens_seen": 3015888,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 0.11418898403644562,
      "learning_rate": 0.000998617965378158,
      "loss": 0.3772,
      "num_input_tokens_seen": 3019120,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.05342996492981911,
      "learning_rate": 0.0009986009218658064,
      "loss": 0.5878,
      "num_input_tokens_seen": 3023920,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.07270803302526474,
      "learning_rate": 0.0009985837740518306,
      "loss": 0.5846,
      "num_input_tokens_seen": 3027088,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 0.2599889039993286,
      "learning_rate": 0.0009985665219398173,
      "loss": 0.463,
      "num_input_tokens_seen": 3029456,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.10389386862516403,
      "learning_rate": 0.0009985491655333755,
      "loss": 0.3985,
      "num_input_tokens_seen": 3033488,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.06522035598754883,
      "learning_rate": 0.000998531704836136,
      "loss": 0.4048,
      "num_input_tokens_seen": 3039024,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.056806787848472595,
      "learning_rate": 0.0009985141398517513,
      "loss": 0.3618,
      "num_input_tokens_seen": 3043248,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.17264139652252197,
      "learning_rate": 0.000998496470583896,
      "loss": 0.463,
      "num_input_tokens_seen": 3046384,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 0.10053090751171112,
      "learning_rate": 0.0009984786970362663,
      "loss": 0.4275,
      "num_input_tokens_seen": 3049392,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.06823375076055527,
      "learning_rate": 0.00099846081921258,
      "loss": 0.4328,
      "num_input_tokens_seen": 3053392,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.07272344827651978,
      "learning_rate": 0.000998442837116577,
      "loss": 0.4129,
      "num_input_tokens_seen": 3057520,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 0.13846638798713684,
      "learning_rate": 0.0009984247507520193,
      "loss": 0.4497,
      "num_input_tokens_seen": 3059824,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 0.10115277022123337,
      "learning_rate": 0.0009984065601226896,
      "loss": 0.5135,
      "num_input_tokens_seen": 3063120,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 0.10861168056726456,
      "learning_rate": 0.0009983882652323942,
      "loss": 0.4325,
      "num_input_tokens_seen": 3066544,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.14086513221263885,
      "learning_rate": 0.0009983698660849592,
      "loss": 0.381,
      "num_input_tokens_seen": 3070224,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.1200118437409401,
      "learning_rate": 0.0009983513626842342,
      "loss": 0.4234,
      "num_input_tokens_seen": 3073456,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.08878769725561142,
      "learning_rate": 0.0009983327550340893,
      "loss": 0.5259,
      "num_input_tokens_seen": 3076656,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.1038520410656929,
      "learning_rate": 0.0009983140431384177,
      "loss": 0.4537,
      "num_input_tokens_seen": 3078928,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 0.1407124400138855,
      "learning_rate": 0.0009982952270011331,
      "loss": 0.4888,
      "num_input_tokens_seen": 3081520,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 0.11267073452472687,
      "learning_rate": 0.000998276306626172,
      "loss": 0.4702,
      "num_input_tokens_seen": 3084464,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 0.09730887413024902,
      "learning_rate": 0.000998257282017492,
      "loss": 0.6283,
      "num_input_tokens_seen": 3087856,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 0.07610620558261871,
      "learning_rate": 0.0009982381531790732,
      "loss": 0.4824,
      "num_input_tokens_seen": 3091952,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.1278218924999237,
      "learning_rate": 0.0009982189201149167,
      "loss": 0.5683,
      "num_input_tokens_seen": 3095504,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.02252523973584175,
      "learning_rate": 0.0009981995828290465,
      "loss": 0.5922,
      "num_input_tokens_seen": 3101776,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.10510362684726715,
      "learning_rate": 0.0009981801413255068,
      "loss": 0.446,
      "num_input_tokens_seen": 3105200,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 0.14347168803215027,
      "learning_rate": 0.0009981605956083657,
      "loss": 0.5389,
      "num_input_tokens_seen": 3108496,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.11219534277915955,
      "learning_rate": 0.000998140945681711,
      "loss": 0.4775,
      "num_input_tokens_seen": 3111184,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.12336465716362,
      "learning_rate": 0.0009981211915496536,
      "loss": 0.5551,
      "num_input_tokens_seen": 3114320,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.14178994297981262,
      "learning_rate": 0.0009981013332163256,
      "loss": 0.4378,
      "num_input_tokens_seen": 3116880,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 0.07862386852502823,
      "learning_rate": 0.0009980813706858816,
      "loss": 0.3954,
      "num_input_tokens_seen": 3120240,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.08889766037464142,
      "learning_rate": 0.000998061303962497,
      "loss": 0.4825,
      "num_input_tokens_seen": 3124528,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 0.11890481412410736,
      "learning_rate": 0.00099804113305037,
      "loss": 0.4126,
      "num_input_tokens_seen": 3127760,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 0.07587739825248718,
      "learning_rate": 0.0009980208579537199,
      "loss": 0.4361,
      "num_input_tokens_seen": 3131696,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 0.09891029447317123,
      "learning_rate": 0.000998000478676788,
      "loss": 0.4384,
      "num_input_tokens_seen": 3135664,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.1855238676071167,
      "learning_rate": 0.0009979799952238373,
      "loss": 0.5036,
      "num_input_tokens_seen": 3138480,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.12902973592281342,
      "learning_rate": 0.000997959407599153,
      "loss": 0.4498,
      "num_input_tokens_seen": 3141840,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.1066363975405693,
      "learning_rate": 0.000997938715807042,
      "loss": 0.5871,
      "num_input_tokens_seen": 3145168,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 0.07251305878162384,
      "learning_rate": 0.000997917919851832,
      "loss": 0.4954,
      "num_input_tokens_seen": 3148944,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.10367444902658463,
      "learning_rate": 0.0009978970197378736,
      "loss": 0.4993,
      "num_input_tokens_seen": 3151888,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 0.0754481852054596,
      "learning_rate": 0.0009978760154695392,
      "loss": 0.5473,
      "num_input_tokens_seen": 3154864,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 0.07224453240633011,
      "learning_rate": 0.0009978549070512226,
      "loss": 0.3109,
      "num_input_tokens_seen": 3157840,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 0.1287524700164795,
      "learning_rate": 0.000997833694487339,
      "loss": 0.4566,
      "num_input_tokens_seen": 3161264,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 0.13271456956863403,
      "learning_rate": 0.0009978123777823263,
      "loss": 0.414,
      "num_input_tokens_seen": 3164272,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 0.15351057052612305,
      "learning_rate": 0.0009977909569406434,
      "loss": 0.4997,
      "num_input_tokens_seen": 3166256,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.06885002553462982,
      "learning_rate": 0.0009977694319667713,
      "loss": 0.4189,
      "num_input_tokens_seen": 3170800,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.24637337028980255,
      "learning_rate": 0.0009977478028652131,
      "loss": 0.5106,
      "num_input_tokens_seen": 3174544,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.23137016594409943,
      "learning_rate": 0.000997726069640493,
      "loss": 0.4719,
      "num_input_tokens_seen": 3177392,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.082377590239048,
      "learning_rate": 0.0009977042322971577,
      "loss": 0.4364,
      "num_input_tokens_seen": 3180304,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 0.13271930813789368,
      "learning_rate": 0.000997682290839775,
      "loss": 0.4991,
      "num_input_tokens_seen": 3183216,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 0.09753508865833282,
      "learning_rate": 0.0009976602452729348,
      "loss": 0.4842,
      "num_input_tokens_seen": 3186224,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 0.11745236814022064,
      "learning_rate": 0.000997638095601249,
      "loss": 0.427,
      "num_input_tokens_seen": 3188848,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.09204718470573425,
      "learning_rate": 0.000997615841829351,
      "loss": 0.4741,
      "num_input_tokens_seen": 3192176,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 0.0995994582772255,
      "learning_rate": 0.000997593483961896,
      "loss": 0.4923,
      "num_input_tokens_seen": 3196560,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 0.08995363861322403,
      "learning_rate": 0.0009975710220035607,
      "loss": 0.4698,
      "num_input_tokens_seen": 3199600,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.056251805275678635,
      "learning_rate": 0.0009975484559590444,
      "loss": 0.4223,
      "num_input_tokens_seen": 3203600,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.21345925331115723,
      "learning_rate": 0.0009975257858330674,
      "loss": 0.5089,
      "num_input_tokens_seen": 3207984,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.09571608155965805,
      "learning_rate": 0.0009975030116303723,
      "loss": 0.4094,
      "num_input_tokens_seen": 3211504,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.08744937926530838,
      "learning_rate": 0.0009974801333557228,
      "loss": 0.4765,
      "num_input_tokens_seen": 3214608,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.12537862360477448,
      "learning_rate": 0.000997457151013905,
      "loss": 0.6418,
      "num_input_tokens_seen": 3217616,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 0.055254075676202774,
      "learning_rate": 0.0009974340646097264,
      "loss": 0.6092,
      "num_input_tokens_seen": 3221488,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.10011540353298187,
      "learning_rate": 0.0009974108741480166,
      "loss": 0.5225,
      "num_input_tokens_seen": 3224400,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.07848280668258667,
      "learning_rate": 0.0009973875796336267,
      "loss": 0.5521,
      "num_input_tokens_seen": 3227408,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 0.057953841984272,
      "learning_rate": 0.0009973641810714295,
      "loss": 0.362,
      "num_input_tokens_seen": 3230704,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 0.1303715854883194,
      "learning_rate": 0.00099734067846632,
      "loss": 0.4147,
      "num_input_tokens_seen": 3233200,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 0.0897011011838913,
      "learning_rate": 0.0009973170718232144,
      "loss": 0.3913,
      "num_input_tokens_seen": 3236400,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 0.1637910008430481,
      "learning_rate": 0.000997293361147051,
      "loss": 0.3778,
      "num_input_tokens_seen": 3239216,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.06970637291669846,
      "learning_rate": 0.0009972695464427904,
      "loss": 0.3569,
      "num_input_tokens_seen": 3241712,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.05724349990487099,
      "learning_rate": 0.0009972456277154134,
      "loss": 0.4503,
      "num_input_tokens_seen": 3244784,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 0.12126893550157547,
      "learning_rate": 0.000997221604969924,
      "loss": 0.5215,
      "num_input_tokens_seen": 3247568,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 0.06438697874546051,
      "learning_rate": 0.0009971974782113475,
      "loss": 0.5244,
      "num_input_tokens_seen": 3251248,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.08527814596891403,
      "learning_rate": 0.0009971732474447308,
      "loss": 0.5686,
      "num_input_tokens_seen": 3254224,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.11807876825332642,
      "learning_rate": 0.0009971489126751427,
      "loss": 0.4843,
      "num_input_tokens_seen": 3257200,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 0.3132154047489166,
      "learning_rate": 0.0009971244739076742,
      "loss": 0.6243,
      "num_input_tokens_seen": 3260080,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 0.05726611241698265,
      "learning_rate": 0.000997099931147437,
      "loss": 0.6038,
      "num_input_tokens_seen": 3262992,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 0.16444937884807587,
      "learning_rate": 0.0009970752843995654,
      "loss": 0.5652,
      "num_input_tokens_seen": 3265872,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 0.139827162027359,
      "learning_rate": 0.0009970505336692153,
      "loss": 0.5951,
      "num_input_tokens_seen": 3269040,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.1420171707868576,
      "learning_rate": 0.0009970256789615642,
      "loss": 0.4367,
      "num_input_tokens_seen": 3271760,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 0.08075395226478577,
      "learning_rate": 0.0009970007202818115,
      "loss": 0.4973,
      "num_input_tokens_seen": 3274224,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.10211362689733505,
      "learning_rate": 0.000996975657635178,
      "loss": 0.6117,
      "num_input_tokens_seen": 3277392,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.11359323561191559,
      "learning_rate": 0.000996950491026907,
      "loss": 0.447,
      "num_input_tokens_seen": 3280240,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.12106651067733765,
      "learning_rate": 0.0009969252204622624,
      "loss": 0.4833,
      "num_input_tokens_seen": 3282864,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.06930406391620636,
      "learning_rate": 0.0009968998459465312,
      "loss": 0.5112,
      "num_input_tokens_seen": 3286064,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.12237598747015,
      "learning_rate": 0.0009968743674850212,
      "loss": 0.6345,
      "num_input_tokens_seen": 3288656,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 0.06251044571399689,
      "learning_rate": 0.0009968487850830622,
      "loss": 0.5671,
      "num_input_tokens_seen": 3292304,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 0.1064622551202774,
      "learning_rate": 0.0009968230987460055,
      "loss": 0.5992,
      "num_input_tokens_seen": 3295632,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 0.11508874595165253,
      "learning_rate": 0.0009967973084792246,
      "loss": 0.5681,
      "num_input_tokens_seen": 3298608,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 0.15720537304878235,
      "learning_rate": 0.0009967714142881145,
      "loss": 0.5484,
      "num_input_tokens_seen": 3301712,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.2666636109352112,
      "learning_rate": 0.0009967454161780923,
      "loss": 0.448,
      "num_input_tokens_seen": 3304400,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 0.11547107994556427,
      "learning_rate": 0.000996719314154596,
      "loss": 0.4396,
      "num_input_tokens_seen": 3306704,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.08659762144088745,
      "learning_rate": 0.0009966931082230862,
      "loss": 0.4755,
      "num_input_tokens_seen": 3310800,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.07947693765163422,
      "learning_rate": 0.0009966667983890445,
      "loss": 0.6228,
      "num_input_tokens_seen": 3314768,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 0.08551640808582306,
      "learning_rate": 0.000996640384657975,
      "loss": 0.4577,
      "num_input_tokens_seen": 3318768,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 0.18400633335113525,
      "learning_rate": 0.0009966138670354028,
      "loss": 0.4976,
      "num_input_tokens_seen": 3321808,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 0.10357975959777832,
      "learning_rate": 0.0009965872455268755,
      "loss": 0.5636,
      "num_input_tokens_seen": 3325904,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.08117800205945969,
      "learning_rate": 0.0009965605201379616,
      "loss": 0.479,
      "num_input_tokens_seen": 3329008,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 0.07943416386842728,
      "learning_rate": 0.000996533690874252,
      "loss": 0.5476,
      "num_input_tokens_seen": 3333008,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.11976098269224167,
      "learning_rate": 0.0009965067577413593,
      "loss": 0.4662,
      "num_input_tokens_seen": 3335760,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.07926589250564575,
      "learning_rate": 0.0009964797207449173,
      "loss": 0.5788,
      "num_input_tokens_seen": 3339056,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 0.09741771966218948,
      "learning_rate": 0.0009964525798905816,
      "loss": 0.4803,
      "num_input_tokens_seen": 3342064,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 0.10100702941417694,
      "learning_rate": 0.0009964253351840303,
      "loss": 0.5676,
      "num_input_tokens_seen": 3344880,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 0.093464195728302,
      "learning_rate": 0.000996397986630962,
      "loss": 0.481,
      "num_input_tokens_seen": 3347984,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.06763967871665955,
      "learning_rate": 0.0009963705342370982,
      "loss": 0.4639,
      "num_input_tokens_seen": 3351536,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.07995543628931046,
      "learning_rate": 0.000996342978008182,
      "loss": 0.3976,
      "num_input_tokens_seen": 3354160,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 0.07464566826820374,
      "learning_rate": 0.000996315317949977,
      "loss": 0.5743,
      "num_input_tokens_seen": 3356912,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 0.13715477287769318,
      "learning_rate": 0.0009962875540682696,
      "loss": 0.5229,
      "num_input_tokens_seen": 3359984,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 0.11575991660356522,
      "learning_rate": 0.0009962596863688682,
      "loss": 0.4231,
      "num_input_tokens_seen": 3363024,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.11255008727312088,
      "learning_rate": 0.000996231714857602,
      "loss": 0.3899,
      "num_input_tokens_seen": 3366288,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.23120030760765076,
      "learning_rate": 0.000996203639540322,
      "loss": 0.4067,
      "num_input_tokens_seen": 3369104,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 0.10118096321821213,
      "learning_rate": 0.0009961754604229018,
      "loss": 0.4851,
      "num_input_tokens_seen": 3372304,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.08292598277330399,
      "learning_rate": 0.0009961471775112361,
      "loss": 0.4375,
      "num_input_tokens_seen": 3375152,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.16983890533447266,
      "learning_rate": 0.000996118790811241,
      "loss": 0.5582,
      "num_input_tokens_seen": 3377904,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 0.08949913084506989,
      "learning_rate": 0.0009960903003288551,
      "loss": 0.498,
      "num_input_tokens_seen": 3381552,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 0.0711798295378685,
      "learning_rate": 0.0009960617060700378,
      "loss": 0.5425,
      "num_input_tokens_seen": 3384368,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 0.09849511086940765,
      "learning_rate": 0.000996033008040771,
      "loss": 0.2774,
      "num_input_tokens_seen": 3387024,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 0.11209512501955032,
      "learning_rate": 0.0009960042062470583,
      "loss": 0.5178,
      "num_input_tokens_seen": 3389808,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 0.08314228057861328,
      "learning_rate": 0.0009959753006949241,
      "loss": 0.4899,
      "num_input_tokens_seen": 3394160,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.16437944769859314,
      "learning_rate": 0.0009959462913904154,
      "loss": 0.6445,
      "num_input_tokens_seen": 3399664,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 0.10924745351076126,
      "learning_rate": 0.0009959171783396007,
      "loss": 0.5905,
      "num_input_tokens_seen": 3403472,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.06351582705974579,
      "learning_rate": 0.00099588796154857,
      "loss": 0.3945,
      "num_input_tokens_seen": 3406384,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.09460873156785965,
      "learning_rate": 0.000995858641023435,
      "loss": 0.3777,
      "num_input_tokens_seen": 3409584,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.06565311551094055,
      "learning_rate": 0.0009958292167703293,
      "loss": 0.4421,
      "num_input_tokens_seen": 3415152,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 0.09602595120668411,
      "learning_rate": 0.0009957996887954082,
      "loss": 0.3799,
      "num_input_tokens_seen": 3418576,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 0.10471038520336151,
      "learning_rate": 0.0009957700571048486,
      "loss": 0.4556,
      "num_input_tokens_seen": 3421904,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 0.12969934940338135,
      "learning_rate": 0.0009957403217048493,
      "loss": 0.5407,
      "num_input_tokens_seen": 3425584,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 0.08071749657392502,
      "learning_rate": 0.0009957104826016302,
      "loss": 0.4845,
      "num_input_tokens_seen": 3429488,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 0.09547936916351318,
      "learning_rate": 0.0009956805398014337,
      "loss": 0.5401,
      "num_input_tokens_seen": 3433040,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.11165562272071838,
      "learning_rate": 0.0009956504933105231,
      "loss": 0.4936,
      "num_input_tokens_seen": 3436464,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 0.07509937137365341,
      "learning_rate": 0.000995620343135184,
      "loss": 0.3724,
      "num_input_tokens_seen": 3438960,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 0.06490868330001831,
      "learning_rate": 0.0009955900892817235,
      "loss": 0.6373,
      "num_input_tokens_seen": 3442416,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 0.12361502647399902,
      "learning_rate": 0.0009955597317564703,
      "loss": 0.5784,
      "num_input_tokens_seen": 3446128,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.07168909907341003,
      "learning_rate": 0.0009955292705657749,
      "loss": 0.5285,
      "num_input_tokens_seen": 3449168,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 0.0892161875963211,
      "learning_rate": 0.0009954987057160093,
      "loss": 0.6107,
      "num_input_tokens_seen": 3452592,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 0.11779969930648804,
      "learning_rate": 0.0009954680372135675,
      "loss": 0.5164,
      "num_input_tokens_seen": 3455184,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.06708156317472458,
      "learning_rate": 0.000995437265064865,
      "loss": 0.3225,
      "num_input_tokens_seen": 3457968,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.10788837820291519,
      "learning_rate": 0.0009954063892763387,
      "loss": 0.4692,
      "num_input_tokens_seen": 3460400,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 0.127105250954628,
      "learning_rate": 0.0009953754098544479,
      "loss": 0.5228,
      "num_input_tokens_seen": 3463376,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.1551101803779602,
      "learning_rate": 0.0009953443268056726,
      "loss": 0.7371,
      "num_input_tokens_seen": 3465936,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 0.17876674234867096,
      "learning_rate": 0.0009953131401365155,
      "loss": 0.5647,
      "num_input_tokens_seen": 3468912,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.11917408555746078,
      "learning_rate": 0.0009952818498535003,
      "loss": 0.4236,
      "num_input_tokens_seen": 3473200,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.13986578583717346,
      "learning_rate": 0.0009952504559631726,
      "loss": 0.6316,
      "num_input_tokens_seen": 3476048,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 0.11047017574310303,
      "learning_rate": 0.0009952189584720996,
      "loss": 0.4607,
      "num_input_tokens_seen": 3479312,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.08611075580120087,
      "learning_rate": 0.0009951873573868701,
      "loss": 0.5973,
      "num_input_tokens_seen": 3483792,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.06658441573381424,
      "learning_rate": 0.000995155652714095,
      "loss": 0.4523,
      "num_input_tokens_seen": 3487408,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.08061472326517105,
      "learning_rate": 0.0009951238444604064,
      "loss": 0.3841,
      "num_input_tokens_seen": 3490672,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 0.07699026167392731,
      "learning_rate": 0.000995091932632458,
      "loss": 0.4535,
      "num_input_tokens_seen": 3494320,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 0.11291979998350143,
      "learning_rate": 0.000995059917236926,
      "loss": 0.4158,
      "num_input_tokens_seen": 3498256,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.09400998055934906,
      "learning_rate": 0.000995027798280507,
      "loss": 0.5655,
      "num_input_tokens_seen": 3500944,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.09720845520496368,
      "learning_rate": 0.00099499557576992,
      "loss": 0.4895,
      "num_input_tokens_seen": 3504752,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 0.09253627061843872,
      "learning_rate": 0.000994963249711906,
      "loss": 0.4509,
      "num_input_tokens_seen": 3507824,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.1015460342168808,
      "learning_rate": 0.000994930820113227,
      "loss": 0.4701,
      "num_input_tokens_seen": 3510800,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 0.061517611145973206,
      "learning_rate": 0.0009948982869806668,
      "loss": 0.3768,
      "num_input_tokens_seen": 3513808,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 0.07177669554948807,
      "learning_rate": 0.0009948656503210311,
      "loss": 0.5427,
      "num_input_tokens_seen": 3517232,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 0.11602744460105896,
      "learning_rate": 0.000994832910141147,
      "loss": 0.4081,
      "num_input_tokens_seen": 3520016,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.25443214178085327,
      "learning_rate": 0.0009948000664478638,
      "loss": 0.4302,
      "num_input_tokens_seen": 3522992,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.0841495618224144,
      "learning_rate": 0.0009947671192480515,
      "loss": 0.3908,
      "num_input_tokens_seen": 3526384,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.08375395834445953,
      "learning_rate": 0.0009947340685486023,
      "loss": 0.5218,
      "num_input_tokens_seen": 3529840,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 0.07956670224666595,
      "learning_rate": 0.0009947009143564303,
      "loss": 0.4325,
      "num_input_tokens_seen": 3532592,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.09406530857086182,
      "learning_rate": 0.0009946676566784708,
      "loss": 0.4117,
      "num_input_tokens_seen": 3535472,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.07978464663028717,
      "learning_rate": 0.000994634295521681,
      "loss": 0.5141,
      "num_input_tokens_seen": 3538000,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 0.07565433531999588,
      "learning_rate": 0.0009946008308930397,
      "loss": 0.439,
      "num_input_tokens_seen": 3541296,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.06293287128210068,
      "learning_rate": 0.0009945672627995473,
      "loss": 0.6249,
      "num_input_tokens_seen": 3544208,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 0.0753217339515686,
      "learning_rate": 0.0009945335912482256,
      "loss": 0.4491,
      "num_input_tokens_seen": 3547376,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.104660265147686,
      "learning_rate": 0.000994499816246119,
      "loss": 0.4757,
      "num_input_tokens_seen": 3550608,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.0960690900683403,
      "learning_rate": 0.000994465937800292,
      "loss": 0.5574,
      "num_input_tokens_seen": 3553744,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.08358926326036453,
      "learning_rate": 0.0009944319559178321,
      "loss": 0.3776,
      "num_input_tokens_seen": 3556816,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.07604847848415375,
      "learning_rate": 0.0009943978706058478,
      "loss": 0.4782,
      "num_input_tokens_seen": 3560080,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 0.08001116663217545,
      "learning_rate": 0.0009943636818714695,
      "loss": 0.4583,
      "num_input_tokens_seen": 3562544,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.09673517197370529,
      "learning_rate": 0.0009943293897218487,
      "loss": 0.4669,
      "num_input_tokens_seen": 3565264,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 0.1064840704202652,
      "learning_rate": 0.0009942949941641594,
      "loss": 0.6349,
      "num_input_tokens_seen": 3567888,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 0.07339110970497131,
      "learning_rate": 0.0009942604952055964,
      "loss": 0.603,
      "num_input_tokens_seen": 3571440,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.134475976228714,
      "learning_rate": 0.0009942258928533768,
      "loss": 0.5756,
      "num_input_tokens_seen": 3574448,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.09397902339696884,
      "learning_rate": 0.0009941911871147386,
      "loss": 0.4981,
      "num_input_tokens_seen": 3577488,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 0.0584934800863266,
      "learning_rate": 0.000994156377996942,
      "loss": 0.3681,
      "num_input_tokens_seen": 3580432,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.09636905789375305,
      "learning_rate": 0.0009941214655072692,
      "loss": 0.5083,
      "num_input_tokens_seen": 3583504,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 0.12399063259363174,
      "learning_rate": 0.0009940864496530226,
      "loss": 0.531,
      "num_input_tokens_seen": 3586320,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 0.07792597264051437,
      "learning_rate": 0.000994051330441528,
      "loss": 0.3922,
      "num_input_tokens_seen": 3589072,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.16195419430732727,
      "learning_rate": 0.0009940161078801312,
      "loss": 0.6579,
      "num_input_tokens_seen": 3596944,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 0.06381644308567047,
      "learning_rate": 0.0009939807819762008,
      "loss": 0.5067,
      "num_input_tokens_seen": 3600080,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.08779668807983398,
      "learning_rate": 0.0009939453527371262,
      "loss": 0.4533,
      "num_input_tokens_seen": 3603632,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.05304284021258354,
      "learning_rate": 0.0009939098201703193,
      "loss": 0.4969,
      "num_input_tokens_seen": 3605936,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 0.0980013832449913,
      "learning_rate": 0.0009938741842832129,
      "loss": 0.4996,
      "num_input_tokens_seen": 3609488,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 0.0982830598950386,
      "learning_rate": 0.0009938384450832614,
      "loss": 0.4007,
      "num_input_tokens_seen": 3612464,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.1096135750412941,
      "learning_rate": 0.0009938026025779411,
      "loss": 0.6035,
      "num_input_tokens_seen": 3615440,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.04758070036768913,
      "learning_rate": 0.0009937666567747501,
      "loss": 0.488,
      "num_input_tokens_seen": 3619952,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.06616410613059998,
      "learning_rate": 0.0009937306076812076,
      "loss": 0.4611,
      "num_input_tokens_seen": 3623760,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 0.09135220944881439,
      "learning_rate": 0.0009936944553048548,
      "loss": 0.4114,
      "num_input_tokens_seen": 3627088,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.1043570414185524,
      "learning_rate": 0.0009936581996532543,
      "loss": 0.4281,
      "num_input_tokens_seen": 3629520,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.07784772664308548,
      "learning_rate": 0.0009936218407339905,
      "loss": 0.3872,
      "num_input_tokens_seen": 3634064,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.12132707238197327,
      "learning_rate": 0.0009935853785546691,
      "loss": 0.5239,
      "num_input_tokens_seen": 3637296,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.09637764096260071,
      "learning_rate": 0.0009935488131229177,
      "loss": 0.4583,
      "num_input_tokens_seen": 3640400,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.06790746748447418,
      "learning_rate": 0.000993512144446385,
      "loss": 0.565,
      "num_input_tokens_seen": 3643760,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.0986952856183052,
      "learning_rate": 0.000993475372532742,
      "loss": 0.6571,
      "num_input_tokens_seen": 3646352,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.1353045105934143,
      "learning_rate": 0.0009934384973896812,
      "loss": 0.5216,
      "num_input_tokens_seen": 3649904,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 0.09811056405305862,
      "learning_rate": 0.0009934015190249162,
      "loss": 0.4579,
      "num_input_tokens_seen": 3652560,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.06485515832901001,
      "learning_rate": 0.0009933644374461822,
      "loss": 0.4959,
      "num_input_tokens_seen": 3656400,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 0.09027861058712006,
      "learning_rate": 0.0009933272526612366,
      "loss": 0.4054,
      "num_input_tokens_seen": 3659984,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 0.07466865330934525,
      "learning_rate": 0.0009932899646778578,
      "loss": 0.4256,
      "num_input_tokens_seen": 3663152,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.08539976924657822,
      "learning_rate": 0.0009932525735038464,
      "loss": 0.3975,
      "num_input_tokens_seen": 3667472,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 0.09217632561922073,
      "learning_rate": 0.0009932150791470238,
      "loss": 0.5037,
      "num_input_tokens_seen": 3670608,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 0.16137060523033142,
      "learning_rate": 0.0009931774816152334,
      "loss": 0.5797,
      "num_input_tokens_seen": 3674320,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.11102544516324997,
      "learning_rate": 0.0009931397809163406,
      "loss": 0.4136,
      "num_input_tokens_seen": 3677328,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.12410715222358704,
      "learning_rate": 0.0009931019770582316,
      "loss": 0.544,
      "num_input_tokens_seen": 3679664,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.09403685480356216,
      "learning_rate": 0.0009930640700488143,
      "loss": 0.5449,
      "num_input_tokens_seen": 3682704,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.08841120451688766,
      "learning_rate": 0.000993026059896019,
      "loss": 0.6005,
      "num_input_tokens_seen": 3685008,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 0.09292386472225189,
      "learning_rate": 0.0009929879466077968,
      "loss": 0.7366,
      "num_input_tokens_seen": 3688240,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 0.09196802973747253,
      "learning_rate": 0.0009929497301921202,
      "loss": 0.4888,
      "num_input_tokens_seen": 3691472,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 0.06655430793762207,
      "learning_rate": 0.000992911410656984,
      "loss": 0.4314,
      "num_input_tokens_seen": 3694544,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.12190654873847961,
      "learning_rate": 0.000992872988010404,
      "loss": 0.4824,
      "num_input_tokens_seen": 3697968,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.16649948060512543,
      "learning_rate": 0.0009928344622604183,
      "loss": 0.4274,
      "num_input_tokens_seen": 3700944,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 0.0877998098731041,
      "learning_rate": 0.0009927958334150853,
      "loss": 0.5182,
      "num_input_tokens_seen": 3704176,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.25122323632240295,
      "learning_rate": 0.0009927571014824862,
      "loss": 0.5283,
      "num_input_tokens_seen": 3708016,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.07444693148136139,
      "learning_rate": 0.000992718266470723,
      "loss": 0.4161,
      "num_input_tokens_seen": 3711216,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.06570201367139816,
      "learning_rate": 0.00099267932838792,
      "loss": 0.467,
      "num_input_tokens_seen": 3713968,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.10044924914836884,
      "learning_rate": 0.000992640287242222,
      "loss": 0.4784,
      "num_input_tokens_seen": 3717520,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 0.0731368139386177,
      "learning_rate": 0.0009926011430417961,
      "loss": 0.4651,
      "num_input_tokens_seen": 3720528,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 0.10392994433641434,
      "learning_rate": 0.0009925618957948312,
      "loss": 0.5212,
      "num_input_tokens_seen": 3723408,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 0.06443842500448227,
      "learning_rate": 0.0009925225455095373,
      "loss": 0.4938,
      "num_input_tokens_seen": 3727152,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.19915051758289337,
      "learning_rate": 0.0009924830921941455,
      "loss": 0.4634,
      "num_input_tokens_seen": 3730640,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.09892718493938446,
      "learning_rate": 0.0009924435358569096,
      "loss": 0.4514,
      "num_input_tokens_seen": 3733488,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4886251986026764,
      "eval_runtime": 13.7626,
      "eval_samples_per_second": 61.616,
      "eval_steps_per_second": 15.404,
      "num_input_tokens_seen": 3736568,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 0.11308852583169937,
      "learning_rate": 0.000992403876506104,
      "loss": 0.4706,
      "num_input_tokens_seen": 3737048,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.11882027238607407,
      "learning_rate": 0.000992364114150025,
      "loss": 0.4932,
      "num_input_tokens_seen": 3739768,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.13060051202774048,
      "learning_rate": 0.0009923242487969908,
      "loss": 0.4214,
      "num_input_tokens_seen": 3743032,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 0.13310763239860535,
      "learning_rate": 0.0009922842804553403,
      "loss": 0.5284,
      "num_input_tokens_seen": 3746744,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 0.17519646883010864,
      "learning_rate": 0.0009922442091334345,
      "loss": 0.4158,
      "num_input_tokens_seen": 3749272,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.09913211315870285,
      "learning_rate": 0.0009922040348396561,
      "loss": 0.5853,
      "num_input_tokens_seen": 3753048,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 0.10032866895198822,
      "learning_rate": 0.000992163757582409,
      "loss": 0.3777,
      "num_input_tokens_seen": 3755448,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.11748122423887253,
      "learning_rate": 0.0009921233773701188,
      "loss": 0.4742,
      "num_input_tokens_seen": 3758424,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.12382180988788605,
      "learning_rate": 0.0009920828942112322,
      "loss": 0.4369,
      "num_input_tokens_seen": 3762552,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 0.18444214761257172,
      "learning_rate": 0.0009920423081142184,
      "loss": 0.4658,
      "num_input_tokens_seen": 3764920,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.07271473109722137,
      "learning_rate": 0.0009920016190875672,
      "loss": 0.4831,
      "num_input_tokens_seen": 3768056,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.11493515223264694,
      "learning_rate": 0.00099196082713979,
      "loss": 0.5044,
      "num_input_tokens_seen": 3771800,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.14125020802021027,
      "learning_rate": 0.0009919199322794207,
      "loss": 0.4606,
      "num_input_tokens_seen": 3774104,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.05585692822933197,
      "learning_rate": 0.0009918789345150136,
      "loss": 0.4318,
      "num_input_tokens_seen": 3777144,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 0.08513195067644119,
      "learning_rate": 0.000991837833855145,
      "loss": 0.3684,
      "num_input_tokens_seen": 3780248,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.10533401370048523,
      "learning_rate": 0.000991796630308413,
      "loss": 0.429,
      "num_input_tokens_seen": 3783672,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.10733862966299057,
      "learning_rate": 0.0009917553238834363,
      "loss": 0.4645,
      "num_input_tokens_seen": 3787032,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.18803586065769196,
      "learning_rate": 0.0009917139145888562,
      "loss": 0.4391,
      "num_input_tokens_seen": 3789528,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 0.08163569867610931,
      "learning_rate": 0.000991672402433335,
      "loss": 0.502,
      "num_input_tokens_seen": 3792984,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.10404843091964722,
      "learning_rate": 0.0009916307874255565,
      "loss": 0.4268,
      "num_input_tokens_seen": 3795928,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 0.0956476628780365,
      "learning_rate": 0.000991589069574226,
      "loss": 0.4181,
      "num_input_tokens_seen": 3798840,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.09713616967201233,
      "learning_rate": 0.0009915472488880705,
      "loss": 0.4726,
      "num_input_tokens_seen": 3801528,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.13936148583889008,
      "learning_rate": 0.0009915053253758386,
      "loss": 0.5612,
      "num_input_tokens_seen": 3804504,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 0.10578684508800507,
      "learning_rate": 0.0009914632990462998,
      "loss": 0.5452,
      "num_input_tokens_seen": 3808728,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.07825697958469391,
      "learning_rate": 0.0009914211699082458,
      "loss": 0.5687,
      "num_input_tokens_seen": 3811256,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 0.12194161117076874,
      "learning_rate": 0.0009913789379704897,
      "loss": 0.6337,
      "num_input_tokens_seen": 3814360,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.09480904042720795,
      "learning_rate": 0.0009913366032418653,
      "loss": 0.3332,
      "num_input_tokens_seen": 3817016,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 0.06548994779586792,
      "learning_rate": 0.0009912941657312293,
      "loss": 0.5184,
      "num_input_tokens_seen": 3820440,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.1473110020160675,
      "learning_rate": 0.0009912516254474586,
      "loss": 0.4217,
      "num_input_tokens_seen": 3823288,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.06852243840694427,
      "learning_rate": 0.0009912089823994525,
      "loss": 0.4486,
      "num_input_tokens_seen": 3827672,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 0.07888627797365189,
      "learning_rate": 0.0009911662365961313,
      "loss": 0.4169,
      "num_input_tokens_seen": 3830808,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.11308082193136215,
      "learning_rate": 0.000991123388046437,
      "loss": 0.4616,
      "num_input_tokens_seen": 3834616,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.06955573707818985,
      "learning_rate": 0.0009910804367593328,
      "loss": 0.3838,
      "num_input_tokens_seen": 3838232,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 0.08130396157503128,
      "learning_rate": 0.0009910373827438038,
      "loss": 0.5428,
      "num_input_tokens_seen": 3841144,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.12442804872989655,
      "learning_rate": 0.0009909942260088562,
      "loss": 0.4708,
      "num_input_tokens_seen": 3844440,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.09582339972257614,
      "learning_rate": 0.0009909509665635184,
      "loss": 0.5562,
      "num_input_tokens_seen": 3847512,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 0.14874738454818726,
      "learning_rate": 0.0009909076044168394,
      "loss": 0.6975,
      "num_input_tokens_seen": 3850040,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 0.1821829229593277,
      "learning_rate": 0.00099086413957789,
      "loss": 0.5019,
      "num_input_tokens_seen": 3854232,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 0.17132292687892914,
      "learning_rate": 0.0009908205720557627,
      "loss": 0.6112,
      "num_input_tokens_seen": 3857656,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 0.1338268220424652,
      "learning_rate": 0.0009907769018595713,
      "loss": 0.594,
      "num_input_tokens_seen": 3860888,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.12858863174915314,
      "learning_rate": 0.0009907331289984512,
      "loss": 0.6268,
      "num_input_tokens_seen": 3863896,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 0.12227654457092285,
      "learning_rate": 0.000990689253481559,
      "loss": 0.4248,
      "num_input_tokens_seen": 3867000,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.07178083062171936,
      "learning_rate": 0.000990645275318073,
      "loss": 0.5014,
      "num_input_tokens_seen": 3870520,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.14490805566310883,
      "learning_rate": 0.000990601194517193,
      "loss": 0.4418,
      "num_input_tokens_seen": 3872856,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.12417465448379517,
      "learning_rate": 0.0009905570110881402,
      "loss": 0.4978,
      "num_input_tokens_seen": 3877624,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 0.09279169887304306,
      "learning_rate": 0.0009905127250401573,
      "loss": 0.5957,
      "num_input_tokens_seen": 3881752,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.13637594878673553,
      "learning_rate": 0.0009904683363825084,
      "loss": 0.4874,
      "num_input_tokens_seen": 3884408,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 0.08849893510341644,
      "learning_rate": 0.0009904238451244791,
      "loss": 0.4138,
      "num_input_tokens_seen": 3887832,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.07562468945980072,
      "learning_rate": 0.0009903792512753764,
      "loss": 0.4068,
      "num_input_tokens_seen": 3891544,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 0.21361799538135529,
      "learning_rate": 0.0009903345548445289,
      "loss": 0.4022,
      "num_input_tokens_seen": 3894104,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.10452640056610107,
      "learning_rate": 0.0009902897558412864,
      "loss": 0.5006,
      "num_input_tokens_seen": 3897208,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 0.17036443948745728,
      "learning_rate": 0.0009902448542750207,
      "loss": 0.4191,
      "num_input_tokens_seen": 3899960,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 0.08835072815418243,
      "learning_rate": 0.0009901998501551245,
      "loss": 0.4407,
      "num_input_tokens_seen": 3902808,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 0.11831244081258774,
      "learning_rate": 0.000990154743491012,
      "loss": 0.4612,
      "num_input_tokens_seen": 3905336,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.16621868312358856,
      "learning_rate": 0.0009901095342921193,
      "loss": 0.5155,
      "num_input_tokens_seen": 3908376,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.15947051346302032,
      "learning_rate": 0.0009900642225679035,
      "loss": 0.6044,
      "num_input_tokens_seen": 3911256,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 0.10438206046819687,
      "learning_rate": 0.000990018808327843,
      "loss": 0.4703,
      "num_input_tokens_seen": 3914424,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.1507096290588379,
      "learning_rate": 0.0009899732915814386,
      "loss": 0.4087,
      "num_input_tokens_seen": 3917080,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 0.3926956355571747,
      "learning_rate": 0.0009899276723382112,
      "loss": 0.5227,
      "num_input_tokens_seen": 3919800,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.139846071600914,
      "learning_rate": 0.0009898819506077043,
      "loss": 0.412,
      "num_input_tokens_seen": 3922328,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 0.12227805703878403,
      "learning_rate": 0.0009898361263994823,
      "loss": 0.4972,
      "num_input_tokens_seen": 3926488,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 0.0941048413515091,
      "learning_rate": 0.0009897901997231308,
      "loss": 0.4517,
      "num_input_tokens_seen": 3929208,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 0.07378534972667694,
      "learning_rate": 0.0009897441705882576,
      "loss": 0.4906,
      "num_input_tokens_seen": 3932536,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 0.13370713591575623,
      "learning_rate": 0.0009896980390044908,
      "loss": 0.4766,
      "num_input_tokens_seen": 3935832,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.1020093709230423,
      "learning_rate": 0.0009896518049814812,
      "loss": 0.4715,
      "num_input_tokens_seen": 3940600,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 0.09184634685516357,
      "learning_rate": 0.0009896054685289005,
      "loss": 0.6018,
      "num_input_tokens_seen": 3943736,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 0.10634180158376694,
      "learning_rate": 0.0009895590296564412,
      "loss": 0.3439,
      "num_input_tokens_seen": 3947160,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.12710747122764587,
      "learning_rate": 0.000989512488373818,
      "loss": 0.3706,
      "num_input_tokens_seen": 3950360,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.08493848145008087,
      "learning_rate": 0.0009894658446907671,
      "loss": 0.4337,
      "num_input_tokens_seen": 3953240,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.11212906986474991,
      "learning_rate": 0.0009894190986170458,
      "loss": 0.5335,
      "num_input_tokens_seen": 3956792,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 0.0987933874130249,
      "learning_rate": 0.0009893722501624323,
      "loss": 0.6136,
      "num_input_tokens_seen": 3959896,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 0.1429404467344284,
      "learning_rate": 0.0009893252993367272,
      "loss": 0.4773,
      "num_input_tokens_seen": 3963480,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 0.1002204492688179,
      "learning_rate": 0.000989278246149752,
      "loss": 0.5534,
      "num_input_tokens_seen": 3966648,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 0.0896909311413765,
      "learning_rate": 0.0009892310906113497,
      "loss": 0.3977,
      "num_input_tokens_seen": 3969784,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 0.060112372040748596,
      "learning_rate": 0.0009891838327313847,
      "loss": 0.3914,
      "num_input_tokens_seen": 3973624,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 0.07910545915365219,
      "learning_rate": 0.000989136472519743,
      "loss": 0.3956,
      "num_input_tokens_seen": 3978648,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 0.07057569175958633,
      "learning_rate": 0.0009890890099863313,
      "loss": 0.5473,
      "num_input_tokens_seen": 3982296,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.11525354534387589,
      "learning_rate": 0.0009890414451410787,
      "loss": 0.4532,
      "num_input_tokens_seen": 3985944,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.11429453641176224,
      "learning_rate": 0.0009889937779939348,
      "loss": 0.4981,
      "num_input_tokens_seen": 3989240,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.06203731521964073,
      "learning_rate": 0.0009889460085548715,
      "loss": 0.4437,
      "num_input_tokens_seen": 3993688,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.15480190515518188,
      "learning_rate": 0.0009888981368338815,
      "loss": 0.3611,
      "num_input_tokens_seen": 3997432,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 0.1890624463558197,
      "learning_rate": 0.0009888501628409789,
      "loss": 0.4311,
      "num_input_tokens_seen": 4000120,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.10216526687145233,
      "learning_rate": 0.0009888020865861991,
      "loss": 0.4158,
      "num_input_tokens_seen": 4003480,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 0.1573265641927719,
      "learning_rate": 0.0009887539080795996,
      "loss": 0.4792,
      "num_input_tokens_seen": 4006456,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 0.19848144054412842,
      "learning_rate": 0.0009887056273312584,
      "loss": 0.4347,
      "num_input_tokens_seen": 4009048,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 0.08185359835624695,
      "learning_rate": 0.0009886572443512753,
      "loss": 0.4143,
      "num_input_tokens_seen": 4012376,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 0.06020468845963478,
      "learning_rate": 0.0009886087591497717,
      "loss": 0.4258,
      "num_input_tokens_seen": 4015416,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 0.13707588613033295,
      "learning_rate": 0.00098856017173689,
      "loss": 0.5686,
      "num_input_tokens_seen": 4018200,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 0.1115296483039856,
      "learning_rate": 0.0009885114821227942,
      "loss": 0.4667,
      "num_input_tokens_seen": 4020792,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 0.07351770997047424,
      "learning_rate": 0.0009884626903176696,
      "loss": 0.5197,
      "num_input_tokens_seen": 4024280,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 0.12332339584827423,
      "learning_rate": 0.0009884137963317228,
      "loss": 0.4306,
      "num_input_tokens_seen": 4027000,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 0.10331713408231735,
      "learning_rate": 0.000988364800175182,
      "loss": 0.6627,
      "num_input_tokens_seen": 4030008,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 0.05466270074248314,
      "learning_rate": 0.0009883157018582966,
      "loss": 0.4787,
      "num_input_tokens_seen": 4033400,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 0.09854459017515182,
      "learning_rate": 0.0009882665013913373,
      "loss": 0.3559,
      "num_input_tokens_seen": 4036216,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.14910128712654114,
      "learning_rate": 0.0009882171987845962,
      "loss": 0.4954,
      "num_input_tokens_seen": 4040920,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.06934267282485962,
      "learning_rate": 0.000988167794048387,
      "loss": 0.4943,
      "num_input_tokens_seen": 4044344,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 0.11421757936477661,
      "learning_rate": 0.0009881182871930448,
      "loss": 0.4982,
      "num_input_tokens_seen": 4047608,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 0.14515259861946106,
      "learning_rate": 0.0009880686782289256,
      "loss": 0.4963,
      "num_input_tokens_seen": 4055288,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 0.10941985249519348,
      "learning_rate": 0.000988018967166407,
      "loss": 0.4893,
      "num_input_tokens_seen": 4058168,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.08920726180076599,
      "learning_rate": 0.0009879691540158884,
      "loss": 0.4199,
      "num_input_tokens_seen": 4061272,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 0.08837755769491196,
      "learning_rate": 0.0009879192387877895,
      "loss": 0.5016,
      "num_input_tokens_seen": 4064536,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.10641390830278397,
      "learning_rate": 0.0009878692214925523,
      "loss": 0.4664,
      "num_input_tokens_seen": 4068184,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 0.1022506058216095,
      "learning_rate": 0.00098781910214064,
      "loss": 0.6335,
      "num_input_tokens_seen": 4071064,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 0.09889979660511017,
      "learning_rate": 0.0009877688807425368,
      "loss": 0.4385,
      "num_input_tokens_seen": 4073464,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 0.08154711127281189,
      "learning_rate": 0.0009877185573087487,
      "loss": 0.4201,
      "num_input_tokens_seen": 4076696,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 0.14799734950065613,
      "learning_rate": 0.0009876681318498025,
      "loss": 0.6167,
      "num_input_tokens_seen": 4079480,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.10834264010190964,
      "learning_rate": 0.0009876176043762466,
      "loss": 0.4711,
      "num_input_tokens_seen": 4082264,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.0752340778708458,
      "learning_rate": 0.000987566974898651,
      "loss": 0.3271,
      "num_input_tokens_seen": 4085400,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 0.109223872423172,
      "learning_rate": 0.0009875162434276065,
      "loss": 0.4445,
      "num_input_tokens_seen": 4088184,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 0.1386282742023468,
      "learning_rate": 0.000987465409973726,
      "loss": 0.4152,
      "num_input_tokens_seen": 4094072,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.1327780783176422,
      "learning_rate": 0.0009874144745476432,
      "loss": 0.4311,
      "num_input_tokens_seen": 4097912,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.12464430183172226,
      "learning_rate": 0.0009873634371600126,
      "loss": 0.4932,
      "num_input_tokens_seen": 4101048,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.17379072308540344,
      "learning_rate": 0.0009873122978215115,
      "loss": 0.6447,
      "num_input_tokens_seen": 4103480,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.10058514028787613,
      "learning_rate": 0.000987261056542837,
      "loss": 0.4467,
      "num_input_tokens_seen": 4108120,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 0.09631407260894775,
      "learning_rate": 0.0009872097133347085,
      "loss": 0.5008,
      "num_input_tokens_seen": 4110872,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.0650598406791687,
      "learning_rate": 0.0009871582682078664,
      "loss": 0.4058,
      "num_input_tokens_seen": 4114232,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 0.07468733191490173,
      "learning_rate": 0.0009871067211730722,
      "loss": 0.4801,
      "num_input_tokens_seen": 4117592,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 0.18031789362430573,
      "learning_rate": 0.0009870550722411093,
      "loss": 0.4612,
      "num_input_tokens_seen": 4121208,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 0.11002477258443832,
      "learning_rate": 0.000987003321422782,
      "loss": 0.4585,
      "num_input_tokens_seen": 4124568,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 0.1086641475558281,
      "learning_rate": 0.0009869514687289155,
      "loss": 0.5844,
      "num_input_tokens_seen": 4127928,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 0.19375456869602203,
      "learning_rate": 0.0009868995141703576,
      "loss": 0.6313,
      "num_input_tokens_seen": 4130552,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 0.09413613379001617,
      "learning_rate": 0.000986847457757976,
      "loss": 0.4874,
      "num_input_tokens_seen": 4133592,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 0.14632648229599,
      "learning_rate": 0.0009867952995026605,
      "loss": 0.5425,
      "num_input_tokens_seen": 4136728,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.09642650932073593,
      "learning_rate": 0.0009867430394153221,
      "loss": 0.4381,
      "num_input_tokens_seen": 4139736,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 0.08622269332408905,
      "learning_rate": 0.0009866906775068927,
      "loss": 0.4182,
      "num_input_tokens_seen": 4142488,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 0.0837603360414505,
      "learning_rate": 0.0009866382137883262,
      "loss": 0.549,
      "num_input_tokens_seen": 4145400,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 0.05756891146302223,
      "learning_rate": 0.0009865856482705973,
      "loss": 0.631,
      "num_input_tokens_seen": 4149112,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.07444250583648682,
      "learning_rate": 0.0009865329809647019,
      "loss": 0.53,
      "num_input_tokens_seen": 4152184,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.11919073760509491,
      "learning_rate": 0.0009864802118816575,
      "loss": 0.7337,
      "num_input_tokens_seen": 4155288,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.1314268708229065,
      "learning_rate": 0.0009864273410325028,
      "loss": 0.4651,
      "num_input_tokens_seen": 4158488,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 0.048284612596035004,
      "learning_rate": 0.000986374368428298,
      "loss": 0.4323,
      "num_input_tokens_seen": 4162328,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.05948297679424286,
      "learning_rate": 0.000986321294080124,
      "loss": 0.4571,
      "num_input_tokens_seen": 4166232,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 0.14724260568618774,
      "learning_rate": 0.0009862681179990838,
      "loss": 0.5694,
      "num_input_tokens_seen": 4169240,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.0719168558716774,
      "learning_rate": 0.0009862148401963008,
      "loss": 0.5392,
      "num_input_tokens_seen": 4171992,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 0.1434416025876999,
      "learning_rate": 0.0009861614606829201,
      "loss": 0.506,
      "num_input_tokens_seen": 4174488,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.12281232327222824,
      "learning_rate": 0.0009861079794701085,
      "loss": 0.4659,
      "num_input_tokens_seen": 4177880,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.10221896320581436,
      "learning_rate": 0.000986054396569053,
      "loss": 0.551,
      "num_input_tokens_seen": 4181816,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 0.11864496022462845,
      "learning_rate": 0.0009860007119909635,
      "loss": 0.6165,
      "num_input_tokens_seen": 4184824,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 0.11747297644615173,
      "learning_rate": 0.0009859469257470692,
      "loss": 0.5705,
      "num_input_tokens_seen": 4187608,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 0.11304371058940887,
      "learning_rate": 0.0009858930378486223,
      "loss": 0.5647,
      "num_input_tokens_seen": 4190968,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 0.08354206383228302,
      "learning_rate": 0.000985839048306895,
      "loss": 0.4945,
      "num_input_tokens_seen": 4195256,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.14409981667995453,
      "learning_rate": 0.000985784957133182,
      "loss": 0.5022,
      "num_input_tokens_seen": 4199000,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 0.0963921919465065,
      "learning_rate": 0.0009857307643387975,
      "loss": 0.4614,
      "num_input_tokens_seen": 4202104,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 0.14671017229557037,
      "learning_rate": 0.000985676469935079,
      "loss": 0.4546,
      "num_input_tokens_seen": 4204344,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 0.17370691895484924,
      "learning_rate": 0.0009856220739333837,
      "loss": 0.5122,
      "num_input_tokens_seen": 4207544,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 0.07933151721954346,
      "learning_rate": 0.000985567576345091,
      "loss": 0.5841,
      "num_input_tokens_seen": 4210360,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 0.06310591846704483,
      "learning_rate": 0.0009855129771816006,
      "loss": 0.4957,
      "num_input_tokens_seen": 4213272,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.10308241844177246,
      "learning_rate": 0.0009854582764543347,
      "loss": 0.4491,
      "num_input_tokens_seen": 4216536,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.1692546159029007,
      "learning_rate": 0.0009854034741747356,
      "loss": 0.4611,
      "num_input_tokens_seen": 4219256,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 0.06973976641893387,
      "learning_rate": 0.0009853485703542675,
      "loss": 0.6098,
      "num_input_tokens_seen": 4222744,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 0.13713300228118896,
      "learning_rate": 0.0009852935650044158,
      "loss": 0.4397,
      "num_input_tokens_seen": 4226168,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.09761657565832138,
      "learning_rate": 0.0009852384581366866,
      "loss": 0.3083,
      "num_input_tokens_seen": 4229272,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 0.09684581309556961,
      "learning_rate": 0.000985183249762608,
      "loss": 0.48,
      "num_input_tokens_seen": 4232120,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 0.12842509150505066,
      "learning_rate": 0.000985127939893729,
      "loss": 0.5249,
      "num_input_tokens_seen": 4236184,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.13451048731803894,
      "learning_rate": 0.0009850725285416194,
      "loss": 0.537,
      "num_input_tokens_seen": 4240056,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 0.06598196178674698,
      "learning_rate": 0.0009850170157178707,
      "loss": 0.5092,
      "num_input_tokens_seen": 4243640,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.1202792227268219,
      "learning_rate": 0.000984961401434096,
      "loss": 0.5805,
      "num_input_tokens_seen": 4246424,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.14259305596351624,
      "learning_rate": 0.0009849056857019285,
      "loss": 0.4975,
      "num_input_tokens_seen": 4249144,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.10756852477788925,
      "learning_rate": 0.0009848498685330238,
      "loss": 0.5731,
      "num_input_tokens_seen": 4251960,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 0.07928366959095001,
      "learning_rate": 0.000984793949939058,
      "loss": 0.4297,
      "num_input_tokens_seen": 4254712,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.12240723520517349,
      "learning_rate": 0.0009847379299317287,
      "loss": 0.4609,
      "num_input_tokens_seen": 4257656,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.07262878865003586,
      "learning_rate": 0.0009846818085227549,
      "loss": 0.5252,
      "num_input_tokens_seen": 4260696,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.11575257033109665,
      "learning_rate": 0.000984625585723876,
      "loss": 0.464,
      "num_input_tokens_seen": 4263672,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 0.12160696089267731,
      "learning_rate": 0.0009845692615468536,
      "loss": 0.4609,
      "num_input_tokens_seen": 4266488,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 0.068152517080307,
      "learning_rate": 0.00098451283600347,
      "loss": 0.38,
      "num_input_tokens_seen": 4271224,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 0.16959404945373535,
      "learning_rate": 0.0009844563091055286,
      "loss": 0.5558,
      "num_input_tokens_seen": 4274104,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.06830588728189468,
      "learning_rate": 0.0009843996808648542,
      "loss": 0.5113,
      "num_input_tokens_seen": 4277752,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.09669560194015503,
      "learning_rate": 0.000984342951293293,
      "loss": 0.4376,
      "num_input_tokens_seen": 4281080,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 0.0529649443924427,
      "learning_rate": 0.0009842861204027121,
      "loss": 0.4138,
      "num_input_tokens_seen": 4285208,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.11017365753650665,
      "learning_rate": 0.0009842291882049999,
      "loss": 0.4799,
      "num_input_tokens_seen": 4287608,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 0.11837736517190933,
      "learning_rate": 0.0009841721547120658,
      "loss": 0.491,
      "num_input_tokens_seen": 4290872,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.1406104564666748,
      "learning_rate": 0.0009841150199358408,
      "loss": 0.4002,
      "num_input_tokens_seen": 4293912,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 0.0851815938949585,
      "learning_rate": 0.0009840577838882765,
      "loss": 0.474,
      "num_input_tokens_seen": 4297112,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 0.09769383072853088,
      "learning_rate": 0.0009840004465813464,
      "loss": 0.4679,
      "num_input_tokens_seen": 4300248,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 0.14465981721878052,
      "learning_rate": 0.0009839430080270445,
      "loss": 0.5211,
      "num_input_tokens_seen": 4303672,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 0.14025510847568512,
      "learning_rate": 0.0009838854682373865,
      "loss": 0.3768,
      "num_input_tokens_seen": 4311096,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 0.07922389358282089,
      "learning_rate": 0.000983827827224409,
      "loss": 0.4883,
      "num_input_tokens_seen": 4314072,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 0.09342105686664581,
      "learning_rate": 0.0009837700850001698,
      "loss": 0.5806,
      "num_input_tokens_seen": 4317528,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 0.10734878480434418,
      "learning_rate": 0.000983712241576748,
      "loss": 0.4436,
      "num_input_tokens_seen": 4320760,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 0.08619202673435211,
      "learning_rate": 0.000983654296966244,
      "loss": 0.4844,
      "num_input_tokens_seen": 4324440,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.13369277119636536,
      "learning_rate": 0.0009835962511807786,
      "loss": 0.6265,
      "num_input_tokens_seen": 4328056,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 0.07611812651157379,
      "learning_rate": 0.0009835381042324948,
      "loss": 0.4043,
      "num_input_tokens_seen": 4330936,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.11739134788513184,
      "learning_rate": 0.0009834798561335558,
      "loss": 0.443,
      "num_input_tokens_seen": 4334328,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 0.07285176217556,
      "learning_rate": 0.000983421506896147,
      "loss": 0.4433,
      "num_input_tokens_seen": 4337432,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 0.09374696761369705,
      "learning_rate": 0.000983363056532474,
      "loss": 0.5004,
      "num_input_tokens_seen": 4340056,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 0.16975751519203186,
      "learning_rate": 0.000983304505054764,
      "loss": 0.628,
      "num_input_tokens_seen": 4342520,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 0.06322844326496124,
      "learning_rate": 0.0009832458524752655,
      "loss": 0.3872,
      "num_input_tokens_seen": 4346776,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.11457878351211548,
      "learning_rate": 0.0009831870988062476,
      "loss": 0.5483,
      "num_input_tokens_seen": 4350104,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.11976616084575653,
      "learning_rate": 0.0009831282440600014,
      "loss": 0.5541,
      "num_input_tokens_seen": 4352952,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.07534301280975342,
      "learning_rate": 0.0009830692882488383,
      "loss": 0.3978,
      "num_input_tokens_seen": 4355896,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 0.10165025293827057,
      "learning_rate": 0.000983010231385091,
      "loss": 0.5059,
      "num_input_tokens_seen": 4358968,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.07320152968168259,
      "learning_rate": 0.000982951073481114,
      "loss": 0.566,
      "num_input_tokens_seen": 4363000,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 0.1568697988986969,
      "learning_rate": 0.0009828918145492823,
      "loss": 0.6933,
      "num_input_tokens_seen": 4366360,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.21485699713230133,
      "learning_rate": 0.000982832454601992,
      "loss": 0.5421,
      "num_input_tokens_seen": 4369432,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 0.05649883672595024,
      "learning_rate": 0.0009827729936516605,
      "loss": 0.4636,
      "num_input_tokens_seen": 4373432,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.06817199289798737,
      "learning_rate": 0.0009827134317107267,
      "loss": 0.3886,
      "num_input_tokens_seen": 4377528,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.07105287164449692,
      "learning_rate": 0.0009826537687916501,
      "loss": 0.5228,
      "num_input_tokens_seen": 4380632,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.09845425933599472,
      "learning_rate": 0.0009825940049069113,
      "loss": 0.3914,
      "num_input_tokens_seen": 4384504,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 0.1021219938993454,
      "learning_rate": 0.0009825341400690126,
      "loss": 0.4854,
      "num_input_tokens_seen": 4388024,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 0.08979879319667816,
      "learning_rate": 0.0009824741742904767,
      "loss": 0.4313,
      "num_input_tokens_seen": 4390968,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 0.07041601836681366,
      "learning_rate": 0.000982414107583848,
      "loss": 0.468,
      "num_input_tokens_seen": 4394328,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 0.0614919476211071,
      "learning_rate": 0.0009823539399616914,
      "loss": 0.5396,
      "num_input_tokens_seen": 4397976,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 0.08305591344833374,
      "learning_rate": 0.0009822936714365938,
      "loss": 0.4736,
      "num_input_tokens_seen": 4401560,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 0.21717211604118347,
      "learning_rate": 0.0009822333020211623,
      "loss": 0.4682,
      "num_input_tokens_seen": 4404568,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.12055421620607376,
      "learning_rate": 0.0009821728317280256,
      "loss": 0.448,
      "num_input_tokens_seen": 4408472,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 0.012225694954395294,
      "learning_rate": 0.0009821122605698336,
      "loss": 0.416,
      "num_input_tokens_seen": 4415064,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.11008094251155853,
      "learning_rate": 0.0009820515885592567,
      "loss": 0.5441,
      "num_input_tokens_seen": 4417720,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 0.07747987657785416,
      "learning_rate": 0.0009819908157089872,
      "loss": 0.4456,
      "num_input_tokens_seen": 4420536,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 0.11956440657377243,
      "learning_rate": 0.0009819299420317379,
      "loss": 0.4123,
      "num_input_tokens_seen": 4423320,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.1620360016822815,
      "learning_rate": 0.0009818689675402427,
      "loss": 0.526,
      "num_input_tokens_seen": 4425816,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 0.08376801759004593,
      "learning_rate": 0.000981807892247257,
      "loss": 0.4931,
      "num_input_tokens_seen": 4429752,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 0.08325216919183731,
      "learning_rate": 0.0009817467161655571,
      "loss": 0.5378,
      "num_input_tokens_seen": 4432888,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 0.07750184088945389,
      "learning_rate": 0.0009816854393079402,
      "loss": 0.5184,
      "num_input_tokens_seen": 4436568,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 0.19492357969284058,
      "learning_rate": 0.0009816240616872247,
      "loss": 0.4095,
      "num_input_tokens_seen": 4439352,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 0.05929189920425415,
      "learning_rate": 0.0009815625833162502,
      "loss": 0.6442,
      "num_input_tokens_seen": 4442808,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 0.11492802947759628,
      "learning_rate": 0.0009815010042078772,
      "loss": 0.5686,
      "num_input_tokens_seen": 4445816,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 0.08282206207513809,
      "learning_rate": 0.0009814393243749873,
      "loss": 0.5091,
      "num_input_tokens_seen": 4449272,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 0.0863899514079094,
      "learning_rate": 0.0009813775438304835,
      "loss": 0.5304,
      "num_input_tokens_seen": 4453048,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.17935006320476532,
      "learning_rate": 0.0009813156625872893,
      "loss": 0.693,
      "num_input_tokens_seen": 4455832,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.1626591831445694,
      "learning_rate": 0.0009812536806583494,
      "loss": 0.504,
      "num_input_tokens_seen": 4458904,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.059121087193489075,
      "learning_rate": 0.0009811915980566302,
      "loss": 0.4343,
      "num_input_tokens_seen": 4462168,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.06256688386201859,
      "learning_rate": 0.0009811294147951182,
      "loss": 0.4598,
      "num_input_tokens_seen": 4465208,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 0.09376311302185059,
      "learning_rate": 0.000981067130886822,
      "loss": 0.4433,
      "num_input_tokens_seen": 4467448,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.09811857342720032,
      "learning_rate": 0.00098100474634477,
      "loss": 0.4103,
      "num_input_tokens_seen": 4471224,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.08617542684078217,
      "learning_rate": 0.0009809422611820127,
      "loss": 0.3452,
      "num_input_tokens_seen": 4473848,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 0.07948457449674606,
      "learning_rate": 0.0009808796754116212,
      "loss": 0.4282,
      "num_input_tokens_seen": 4477528,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 0.10123014450073242,
      "learning_rate": 0.0009808169890466879,
      "loss": 0.5376,
      "num_input_tokens_seen": 4481784,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 0.07853500545024872,
      "learning_rate": 0.000980754202100326,
      "loss": 0.4545,
      "num_input_tokens_seen": 4485624,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 0.1751125156879425,
      "learning_rate": 0.0009806913145856695,
      "loss": 0.6279,
      "num_input_tokens_seen": 4488120,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 0.08805426955223083,
      "learning_rate": 0.0009806283265158741,
      "loss": 0.5598,
      "num_input_tokens_seen": 4491640,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.11093246936798096,
      "learning_rate": 0.0009805652379041162,
      "loss": 0.4795,
      "num_input_tokens_seen": 4494648,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 0.18126817047595978,
      "learning_rate": 0.000980502048763593,
      "loss": 0.5399,
      "num_input_tokens_seen": 4497176,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 0.0535278357565403,
      "learning_rate": 0.0009804387591075233,
      "loss": 0.4333,
      "num_input_tokens_seen": 4500248,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.08338054269552231,
      "learning_rate": 0.000980375368949146,
      "loss": 0.3394,
      "num_input_tokens_seen": 4503416,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 0.07378730922937393,
      "learning_rate": 0.0009803118783017221,
      "loss": 0.4649,
      "num_input_tokens_seen": 4506904,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 0.05750977620482445,
      "learning_rate": 0.0009802482871785329,
      "loss": 0.3446,
      "num_input_tokens_seen": 4511352,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.0847051739692688,
      "learning_rate": 0.000980184595592881,
      "loss": 0.4602,
      "num_input_tokens_seen": 4514520,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 0.0797916054725647,
      "learning_rate": 0.00098012080355809,
      "loss": 0.4243,
      "num_input_tokens_seen": 4517336,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 0.17022742331027985,
      "learning_rate": 0.0009800569110875043,
      "loss": 0.4115,
      "num_input_tokens_seen": 4520280,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.09837656468153,
      "learning_rate": 0.0009799929181944895,
      "loss": 0.4356,
      "num_input_tokens_seen": 4523960,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 0.09719648212194443,
      "learning_rate": 0.0009799288248924325,
      "loss": 0.5717,
      "num_input_tokens_seen": 4527512,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 0.08258490264415741,
      "learning_rate": 0.0009798646311947404,
      "loss": 0.4149,
      "num_input_tokens_seen": 4531704,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 0.0809776782989502,
      "learning_rate": 0.0009798003371148422,
      "loss": 0.5068,
      "num_input_tokens_seen": 4535192,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 0.10946978628635406,
      "learning_rate": 0.0009797359426661873,
      "loss": 0.3476,
      "num_input_tokens_seen": 4538136,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.08737422525882721,
      "learning_rate": 0.0009796714478622463,
      "loss": 0.3782,
      "num_input_tokens_seen": 4540824,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.12454596906900406,
      "learning_rate": 0.0009796068527165107,
      "loss": 0.5333,
      "num_input_tokens_seen": 4543896,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 0.09442093223333359,
      "learning_rate": 0.0009795421572424935,
      "loss": 0.4968,
      "num_input_tokens_seen": 4547448,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.06413788348436356,
      "learning_rate": 0.0009794773614537275,
      "loss": 0.4728,
      "num_input_tokens_seen": 4549880,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.09405376762151718,
      "learning_rate": 0.0009794124653637677,
      "loss": 0.4631,
      "num_input_tokens_seen": 4552952,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.09247812628746033,
      "learning_rate": 0.00097934746898619,
      "loss": 0.4545,
      "num_input_tokens_seen": 4555384,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.09008338302373886,
      "learning_rate": 0.00097928237233459,
      "loss": 0.4038,
      "num_input_tokens_seen": 4560248,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.11403200775384903,
      "learning_rate": 0.0009792171754225859,
      "loss": 0.3888,
      "num_input_tokens_seen": 4562424,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 0.3010852336883545,
      "learning_rate": 0.0009791518782638158,
      "loss": 0.613,
      "num_input_tokens_seen": 4565656,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 0.09495782852172852,
      "learning_rate": 0.0009790864808719392,
      "loss": 0.4874,
      "num_input_tokens_seen": 4568760,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.09551269561052322,
      "learning_rate": 0.0009790209832606365,
      "loss": 0.6179,
      "num_input_tokens_seen": 4571960,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.0578487366437912,
      "learning_rate": 0.000978955385443609,
      "loss": 0.4832,
      "num_input_tokens_seen": 4575288,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 0.21358799934387207,
      "learning_rate": 0.0009788896874345792,
      "loss": 0.4854,
      "num_input_tokens_seen": 4577848,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 0.11913584917783737,
      "learning_rate": 0.0009788238892472904,
      "loss": 0.4775,
      "num_input_tokens_seen": 4581400,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.11599365621805191,
      "learning_rate": 0.0009787579908955063,
      "loss": 0.4341,
      "num_input_tokens_seen": 4584984,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 0.0682392492890358,
      "learning_rate": 0.0009786919923930127,
      "loss": 0.4147,
      "num_input_tokens_seen": 4587544,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 0.12011174112558365,
      "learning_rate": 0.0009786258937536155,
      "loss": 0.4444,
      "num_input_tokens_seen": 4590680,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 0.08629345893859863,
      "learning_rate": 0.0009785596949911418,
      "loss": 0.4979,
      "num_input_tokens_seen": 4594392,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.10023681819438934,
      "learning_rate": 0.0009784933961194395,
      "loss": 0.378,
      "num_input_tokens_seen": 4597144,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.10347522050142288,
      "learning_rate": 0.0009784269971523777,
      "loss": 0.573,
      "num_input_tokens_seen": 4599736,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 0.07212910801172256,
      "learning_rate": 0.0009783604981038463,
      "loss": 0.3944,
      "num_input_tokens_seen": 4602712,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 0.10759121924638748,
      "learning_rate": 0.000978293898987756,
      "loss": 0.4257,
      "num_input_tokens_seen": 4605752,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.12350412458181381,
      "learning_rate": 0.000978227199818039,
      "loss": 0.4524,
      "num_input_tokens_seen": 4609560,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.06706684082746506,
      "learning_rate": 0.0009781604006086474,
      "loss": 0.494,
      "num_input_tokens_seen": 4612696,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.10608698427677155,
      "learning_rate": 0.0009780935013735553,
      "loss": 0.5166,
      "num_input_tokens_seen": 4615928,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 0.11010009050369263,
      "learning_rate": 0.0009780265021267572,
      "loss": 0.4888,
      "num_input_tokens_seen": 4619160,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 0.12014506757259369,
      "learning_rate": 0.0009779594028822682,
      "loss": 0.5188,
      "num_input_tokens_seen": 4621944,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 0.10775166749954224,
      "learning_rate": 0.0009778922036541252,
      "loss": 0.487,
      "num_input_tokens_seen": 4625400,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 0.049989745020866394,
      "learning_rate": 0.0009778249044563852,
      "loss": 0.3966,
      "num_input_tokens_seen": 4628600,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.10211583226919174,
      "learning_rate": 0.0009777575053031263,
      "loss": 0.4571,
      "num_input_tokens_seen": 4631640,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.04643145948648453,
      "learning_rate": 0.000977690006208448,
      "loss": 0.5708,
      "num_input_tokens_seen": 4635480,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 0.08657078444957733,
      "learning_rate": 0.0009776224071864703,
      "loss": 0.524,
      "num_input_tokens_seen": 4638136,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 0.11531035602092743,
      "learning_rate": 0.000977554708251334,
      "loss": 0.3976,
      "num_input_tokens_seen": 4641176,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.058398764580488205,
      "learning_rate": 0.0009774869094172007,
      "loss": 0.4138,
      "num_input_tokens_seen": 4645144,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 0.10540734231472015,
      "learning_rate": 0.0009774190106982537,
      "loss": 0.5349,
      "num_input_tokens_seen": 4648408,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.1002400740981102,
      "learning_rate": 0.0009773510121086962,
      "loss": 0.5206,
      "num_input_tokens_seen": 4651096,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 0.09329050034284592,
      "learning_rate": 0.0009772829136627528,
      "loss": 0.4159,
      "num_input_tokens_seen": 4654776,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 0.07280603796243668,
      "learning_rate": 0.0009772147153746691,
      "loss": 0.413,
      "num_input_tokens_seen": 4658360,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 0.09835195541381836,
      "learning_rate": 0.0009771464172587112,
      "loss": 0.4699,
      "num_input_tokens_seen": 4661528,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.08874894678592682,
      "learning_rate": 0.0009770780193291667,
      "loss": 0.6372,
      "num_input_tokens_seen": 4664600,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.1167888194322586,
      "learning_rate": 0.000977009521600343,
      "loss": 0.4802,
      "num_input_tokens_seen": 4667416,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 0.09714296460151672,
      "learning_rate": 0.0009769409240865696,
      "loss": 0.4343,
      "num_input_tokens_seen": 4670136,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.07768891006708145,
      "learning_rate": 0.0009768722268021959,
      "loss": 0.4907,
      "num_input_tokens_seen": 4673592,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 0.09877847135066986,
      "learning_rate": 0.000976803429761593,
      "loss": 0.5063,
      "num_input_tokens_seen": 4677272,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 0.0744791105389595,
      "learning_rate": 0.000976734532979152,
      "loss": 0.4672,
      "num_input_tokens_seen": 4681464,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 0.17054738104343414,
      "learning_rate": 0.000976665536469286,
      "loss": 0.4871,
      "num_input_tokens_seen": 4684408,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 0.10277561098337173,
      "learning_rate": 0.0009765964402464276,
      "loss": 0.5426,
      "num_input_tokens_seen": 4687448,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 0.10067476332187653,
      "learning_rate": 0.0009765272443250312,
      "loss": 0.7222,
      "num_input_tokens_seen": 4690904,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 0.08685141056776047,
      "learning_rate": 0.0009764579487195717,
      "loss": 0.5804,
      "num_input_tokens_seen": 4694488,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.11476710438728333,
      "learning_rate": 0.0009763885534445452,
      "loss": 0.4609,
      "num_input_tokens_seen": 4697528,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 0.12841174006462097,
      "learning_rate": 0.0009763190585144682,
      "loss": 0.4498,
      "num_input_tokens_seen": 4701208,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.09383141994476318,
      "learning_rate": 0.0009762494639438783,
      "loss": 0.5184,
      "num_input_tokens_seen": 4705240,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.12744726240634918,
      "learning_rate": 0.0009761797697473339,
      "loss": 0.5224,
      "num_input_tokens_seen": 4708216,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.1629524528980255,
      "learning_rate": 0.0009761099759394142,
      "loss": 0.724,
      "num_input_tokens_seen": 4710712,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 0.09867432713508606,
      "learning_rate": 0.0009760400825347194,
      "loss": 0.5298,
      "num_input_tokens_seen": 4713912,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.08272884786128998,
      "learning_rate": 0.00097597008954787,
      "loss": 0.4567,
      "num_input_tokens_seen": 4716440,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 0.04646630957722664,
      "learning_rate": 0.0009758999969935083,
      "loss": 0.5692,
      "num_input_tokens_seen": 4720184,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 0.058547694236040115,
      "learning_rate": 0.0009758298048862965,
      "loss": 0.5186,
      "num_input_tokens_seen": 4724760,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.10013992339372635,
      "learning_rate": 0.0009757595132409181,
      "loss": 0.461,
      "num_input_tokens_seen": 4727832,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 0.06884495168924332,
      "learning_rate": 0.0009756891220720771,
      "loss": 0.3699,
      "num_input_tokens_seen": 4730584,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 0.1079663410782814,
      "learning_rate": 0.0009756186313944988,
      "loss": 0.484,
      "num_input_tokens_seen": 4733208,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 0.1058480516076088,
      "learning_rate": 0.0009755480412229291,
      "loss": 0.4452,
      "num_input_tokens_seen": 4736088,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 0.07075140625238419,
      "learning_rate": 0.0009754773515721343,
      "loss": 0.5539,
      "num_input_tokens_seen": 4739096,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 0.10414702445268631,
      "learning_rate": 0.0009754065624569022,
      "loss": 0.3699,
      "num_input_tokens_seen": 4742168,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.07108229398727417,
      "learning_rate": 0.000975335673892041,
      "loss": 0.4718,
      "num_input_tokens_seen": 4745432,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.10067420452833176,
      "learning_rate": 0.0009752646858923797,
      "loss": 0.4363,
      "num_input_tokens_seen": 4748536,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 0.08670929819345474,
      "learning_rate": 0.0009751935984727683,
      "loss": 0.5378,
      "num_input_tokens_seen": 4751800,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 0.060213010758161545,
      "learning_rate": 0.0009751224116480772,
      "loss": 0.3583,
      "num_input_tokens_seen": 4755800,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 0.1542588174343109,
      "learning_rate": 0.0009750511254331982,
      "loss": 0.3994,
      "num_input_tokens_seen": 4758904,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 0.08476755023002625,
      "learning_rate": 0.0009749797398430433,
      "loss": 0.3836,
      "num_input_tokens_seen": 4763352,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.11644694954156876,
      "learning_rate": 0.0009749082548925459,
      "loss": 0.4705,
      "num_input_tokens_seen": 4765880,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 0.09210403263568878,
      "learning_rate": 0.0009748366705966593,
      "loss": 0.4204,
      "num_input_tokens_seen": 4769112,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.07697049528360367,
      "learning_rate": 0.0009747649869703588,
      "loss": 0.4081,
      "num_input_tokens_seen": 4772024,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 0.1402280628681183,
      "learning_rate": 0.0009746932040286391,
      "loss": 0.43,
      "num_input_tokens_seen": 4775320,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.14962519705295563,
      "learning_rate": 0.000974621321786517,
      "loss": 0.4906,
      "num_input_tokens_seen": 4778232,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 0.10931897163391113,
      "learning_rate": 0.000974549340259029,
      "loss": 0.4544,
      "num_input_tokens_seen": 4782200,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 0.09509363025426865,
      "learning_rate": 0.000974477259461233,
      "loss": 0.4735,
      "num_input_tokens_seen": 4784888,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.08744565397500992,
      "learning_rate": 0.0009744050794082074,
      "loss": 0.4692,
      "num_input_tokens_seen": 4788344,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 0.07937650382518768,
      "learning_rate": 0.0009743328001150515,
      "loss": 0.393,
      "num_input_tokens_seen": 4791000,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 0.06321968138217926,
      "learning_rate": 0.0009742604215968853,
      "loss": 0.4272,
      "num_input_tokens_seen": 4794392,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 0.13948002457618713,
      "learning_rate": 0.0009741879438688495,
      "loss": 0.6075,
      "num_input_tokens_seen": 4796888,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.0691794827580452,
      "learning_rate": 0.0009741153669461058,
      "loss": 0.4747,
      "num_input_tokens_seen": 4800856,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 0.062173955142498016,
      "learning_rate": 0.0009740426908438362,
      "loss": 0.4809,
      "num_input_tokens_seen": 4804568,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 0.12960118055343628,
      "learning_rate": 0.0009739699155772439,
      "loss": 0.3349,
      "num_input_tokens_seen": 4807384,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.081728994846344,
      "learning_rate": 0.0009738970411615525,
      "loss": 0.4955,
      "num_input_tokens_seen": 4810904,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.1279444396495819,
      "learning_rate": 0.0009738240676120067,
      "loss": 0.4384,
      "num_input_tokens_seen": 4814008,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.054205577820539474,
      "learning_rate": 0.0009737509949438717,
      "loss": 0.5258,
      "num_input_tokens_seen": 4817272,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.0835750475525856,
      "learning_rate": 0.0009736778231724333,
      "loss": 0.4068,
      "num_input_tokens_seen": 4819896,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.1732068806886673,
      "learning_rate": 0.0009736045523129982,
      "loss": 0.4145,
      "num_input_tokens_seen": 4822936,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 0.07802848517894745,
      "learning_rate": 0.0009735311823808938,
      "loss": 0.3612,
      "num_input_tokens_seen": 4826776,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 0.05474779009819031,
      "learning_rate": 0.0009734577133914687,
      "loss": 0.413,
      "num_input_tokens_seen": 4831288,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.10159343481063843,
      "learning_rate": 0.0009733841453600914,
      "loss": 0.5321,
      "num_input_tokens_seen": 4834232,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 0.054498184472322464,
      "learning_rate": 0.0009733104783021515,
      "loss": 0.5185,
      "num_input_tokens_seen": 4837752,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 0.07092639803886414,
      "learning_rate": 0.0009732367122330593,
      "loss": 0.5748,
      "num_input_tokens_seen": 4841304,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 0.18549780547618866,
      "learning_rate": 0.0009731628471682459,
      "loss": 0.4316,
      "num_input_tokens_seen": 4844888,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.1443532407283783,
      "learning_rate": 0.000973088883123163,
      "loss": 0.5517,
      "num_input_tokens_seen": 4848088,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 0.08182500302791595,
      "learning_rate": 0.0009730148201132829,
      "loss": 0.437,
      "num_input_tokens_seen": 4852248,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 0.09600624442100525,
      "learning_rate": 0.0009729406581540991,
      "loss": 0.4426,
      "num_input_tokens_seen": 4855192,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.07812360674142838,
      "learning_rate": 0.0009728663972611251,
      "loss": 0.4174,
      "num_input_tokens_seen": 4858584,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 0.15465040504932404,
      "learning_rate": 0.0009727920374498955,
      "loss": 0.433,
      "num_input_tokens_seen": 4861304,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.04809325933456421,
      "learning_rate": 0.0009727175787359656,
      "loss": 0.5735,
      "num_input_tokens_seen": 4864824,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.18159425258636475,
      "learning_rate": 0.0009726430211349113,
      "loss": 0.5276,
      "num_input_tokens_seen": 4867736,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.1650630086660385,
      "learning_rate": 0.0009725683646623291,
      "loss": 0.4448,
      "num_input_tokens_seen": 4870936,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.05912984535098076,
      "learning_rate": 0.0009724936093338365,
      "loss": 0.61,
      "num_input_tokens_seen": 4875160,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 0.07358112186193466,
      "learning_rate": 0.0009724187551650712,
      "loss": 0.3191,
      "num_input_tokens_seen": 4878072,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 0.11639749258756638,
      "learning_rate": 0.0009723438021716919,
      "loss": 0.547,
      "num_input_tokens_seen": 4882008,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 0.08946248888969421,
      "learning_rate": 0.0009722687503693782,
      "loss": 0.4354,
      "num_input_tokens_seen": 4884952,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 0.10968468338251114,
      "learning_rate": 0.0009721935997738296,
      "loss": 0.4682,
      "num_input_tokens_seen": 4887736,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.09637466818094254,
      "learning_rate": 0.0009721183504007671,
      "loss": 0.4294,
      "num_input_tokens_seen": 4891384,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.0700405165553093,
      "learning_rate": 0.0009720430022659319,
      "loss": 0.3752,
      "num_input_tokens_seen": 4895160,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 0.0997207835316658,
      "learning_rate": 0.000971967555385086,
      "loss": 0.3202,
      "num_input_tokens_seen": 4898456,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.08327487856149673,
      "learning_rate": 0.000971892009774012,
      "loss": 0.432,
      "num_input_tokens_seen": 4901176,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 0.1082632765173912,
      "learning_rate": 0.0009718163654485133,
      "loss": 0.4315,
      "num_input_tokens_seen": 4903608,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 0.09330689162015915,
      "learning_rate": 0.0009717406224244136,
      "loss": 0.4899,
      "num_input_tokens_seen": 4906712,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.09375634789466858,
      "learning_rate": 0.0009716647807175575,
      "loss": 0.4478,
      "num_input_tokens_seen": 4910776,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 0.09534487128257751,
      "learning_rate": 0.0009715888403438105,
      "loss": 0.5187,
      "num_input_tokens_seen": 4913656,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.15556958317756653,
      "learning_rate": 0.0009715128013190581,
      "loss": 0.5821,
      "num_input_tokens_seen": 4916856,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 0.13117778301239014,
      "learning_rate": 0.0009714366636592069,
      "loss": 0.5989,
      "num_input_tokens_seen": 4919960,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 0.10728202015161514,
      "learning_rate": 0.0009713604273801844,
      "loss": 0.3951,
      "num_input_tokens_seen": 4922456,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.08195879310369492,
      "learning_rate": 0.0009712840924979378,
      "loss": 0.6275,
      "num_input_tokens_seen": 4925336,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.13150598108768463,
      "learning_rate": 0.0009712076590284357,
      "loss": 0.3384,
      "num_input_tokens_seen": 4928056,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 0.1387576311826706,
      "learning_rate": 0.0009711311269876674,
      "loss": 0.4087,
      "num_input_tokens_seen": 4931192,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.06739819794893265,
      "learning_rate": 0.0009710544963916421,
      "loss": 0.418,
      "num_input_tokens_seen": 4934872,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.0869663804769516,
      "learning_rate": 0.0009709777672563903,
      "loss": 0.6091,
      "num_input_tokens_seen": 4938104,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 0.06274925172328949,
      "learning_rate": 0.0009709009395979628,
      "loss": 0.3019,
      "num_input_tokens_seen": 4942456,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.06788966804742813,
      "learning_rate": 0.0009708240134324311,
      "loss": 0.541,
      "num_input_tokens_seen": 4946488,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 0.12182977795600891,
      "learning_rate": 0.0009707469887758871,
      "loss": 0.4216,
      "num_input_tokens_seen": 4948920,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.16166676580905914,
      "learning_rate": 0.0009706698656444437,
      "loss": 0.4723,
      "num_input_tokens_seen": 4952152,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.1340913027524948,
      "learning_rate": 0.000970592644054234,
      "loss": 0.3745,
      "num_input_tokens_seen": 4955288,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.22189576923847198,
      "learning_rate": 0.000970515324021412,
      "loss": 0.4249,
      "num_input_tokens_seen": 4958520,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.10610416531562805,
      "learning_rate": 0.0009704379055621523,
      "loss": 0.4022,
      "num_input_tokens_seen": 4961336,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 0.09609129279851913,
      "learning_rate": 0.0009703603886926497,
      "loss": 0.6038,
      "num_input_tokens_seen": 4965944,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 0.09457653760910034,
      "learning_rate": 0.0009702827734291198,
      "loss": 0.4496,
      "num_input_tokens_seen": 4968824,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.07926488667726517,
      "learning_rate": 0.0009702050597877992,
      "loss": 0.3464,
      "num_input_tokens_seen": 4971608,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.14581835269927979,
      "learning_rate": 0.0009701272477849444,
      "loss": 0.4763,
      "num_input_tokens_seen": 4973912,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 0.1351747363805771,
      "learning_rate": 0.0009700493374368327,
      "loss": 0.5707,
      "num_input_tokens_seen": 4977784,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 0.13232837617397308,
      "learning_rate": 0.0009699713287597624,
      "loss": 0.4347,
      "num_input_tokens_seen": 4980344,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 0.12305368483066559,
      "learning_rate": 0.0009698932217700518,
      "loss": 0.4348,
      "num_input_tokens_seen": 4983384,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.49002858996391296,
      "eval_runtime": 13.7136,
      "eval_samples_per_second": 61.836,
      "eval_steps_per_second": 15.459,
      "num_input_tokens_seen": 4984176,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.08135451376438141,
      "learning_rate": 0.0009698150164840399,
      "loss": 0.374,
      "num_input_tokens_seen": 4985968,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.10658412426710129,
      "learning_rate": 0.0009697367129180866,
      "loss": 0.4376,
      "num_input_tokens_seen": 4989232,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 0.10375337302684784,
      "learning_rate": 0.000969658311088572,
      "loss": 0.4629,
      "num_input_tokens_seen": 4992176,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 0.13344572484493256,
      "learning_rate": 0.0009695798110118969,
      "loss": 0.5011,
      "num_input_tokens_seen": 4995024,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.06857980787754059,
      "learning_rate": 0.0009695012127044824,
      "loss": 0.5758,
      "num_input_tokens_seen": 4998224,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 0.06714871525764465,
      "learning_rate": 0.0009694225161827707,
      "loss": 0.4959,
      "num_input_tokens_seen": 5002032,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.1034839004278183,
      "learning_rate": 0.0009693437214632241,
      "loss": 0.4001,
      "num_input_tokens_seen": 5005520,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 0.08647550642490387,
      "learning_rate": 0.0009692648285623256,
      "loss": 0.5459,
      "num_input_tokens_seen": 5008624,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.15439513325691223,
      "learning_rate": 0.0009691858374965784,
      "loss": 0.4594,
      "num_input_tokens_seen": 5010704,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 0.14383268356323242,
      "learning_rate": 0.0009691067482825069,
      "loss": 0.4256,
      "num_input_tokens_seen": 5014416,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 0.10972395539283752,
      "learning_rate": 0.0009690275609366554,
      "loss": 0.486,
      "num_input_tokens_seen": 5018704,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 0.1219710111618042,
      "learning_rate": 0.0009689482754755891,
      "loss": 0.4761,
      "num_input_tokens_seen": 5022768,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.10024676471948624,
      "learning_rate": 0.0009688688919158938,
      "loss": 0.5409,
      "num_input_tokens_seen": 5030096,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 0.10910225659608841,
      "learning_rate": 0.0009687894102741754,
      "loss": 0.512,
      "num_input_tokens_seen": 5033392,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 0.06742015480995178,
      "learning_rate": 0.0009687098305670605,
      "loss": 0.5776,
      "num_input_tokens_seen": 5035920,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.10986873507499695,
      "learning_rate": 0.0009686301528111964,
      "loss": 0.5119,
      "num_input_tokens_seen": 5039568,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.07981524616479874,
      "learning_rate": 0.0009685503770232507,
      "loss": 0.4057,
      "num_input_tokens_seen": 5042992,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.06997290998697281,
      "learning_rate": 0.0009684705032199117,
      "loss": 0.4615,
      "num_input_tokens_seen": 5046608,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 0.11637938767671585,
      "learning_rate": 0.0009683905314178881,
      "loss": 0.4387,
      "num_input_tokens_seen": 5051152,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 0.08628284931182861,
      "learning_rate": 0.000968310461633909,
      "loss": 0.539,
      "num_input_tokens_seen": 5053872,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 0.10606499761343002,
      "learning_rate": 0.0009682302938847238,
      "loss": 0.3981,
      "num_input_tokens_seen": 5056624,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.09799348562955856,
      "learning_rate": 0.0009681500281871031,
      "loss": 0.3864,
      "num_input_tokens_seen": 5059472,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 0.1007499024271965,
      "learning_rate": 0.0009680696645578377,
      "loss": 0.508,
      "num_input_tokens_seen": 5062928,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 0.05121278390288353,
      "learning_rate": 0.0009679892030137382,
      "loss": 0.4537,
      "num_input_tokens_seen": 5066128,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 0.14251674711704254,
      "learning_rate": 0.0009679086435716368,
      "loss": 0.501,
      "num_input_tokens_seen": 5069616,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.10267411172389984,
      "learning_rate": 0.0009678279862483852,
      "loss": 0.5741,
      "num_input_tokens_seen": 5073168,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 0.08451861888170242,
      "learning_rate": 0.0009677472310608561,
      "loss": 0.4663,
      "num_input_tokens_seen": 5076336,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 0.07213269174098969,
      "learning_rate": 0.0009676663780259427,
      "loss": 0.5797,
      "num_input_tokens_seen": 5079088,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 0.11035271733999252,
      "learning_rate": 0.0009675854271605583,
      "loss": 0.6626,
      "num_input_tokens_seen": 5082480,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 0.07864928245544434,
      "learning_rate": 0.0009675043784816371,
      "loss": 0.4057,
      "num_input_tokens_seen": 5085552,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 0.08094184100627899,
      "learning_rate": 0.0009674232320061336,
      "loss": 0.4873,
      "num_input_tokens_seen": 5089232,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.09664618968963623,
      "learning_rate": 0.0009673419877510226,
      "loss": 0.3741,
      "num_input_tokens_seen": 5092080,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.20783299207687378,
      "learning_rate": 0.0009672606457332994,
      "loss": 0.6198,
      "num_input_tokens_seen": 5095152,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 0.11304055154323578,
      "learning_rate": 0.0009671792059699798,
      "loss": 0.5072,
      "num_input_tokens_seen": 5097872,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 0.1375460922718048,
      "learning_rate": 0.0009670976684781003,
      "loss": 0.4656,
      "num_input_tokens_seen": 5100592,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 0.12204957753419876,
      "learning_rate": 0.0009670160332747174,
      "loss": 0.546,
      "num_input_tokens_seen": 5103664,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 0.05934350937604904,
      "learning_rate": 0.0009669343003769085,
      "loss": 0.4428,
      "num_input_tokens_seen": 5107280,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 0.09425344318151474,
      "learning_rate": 0.0009668524698017709,
      "loss": 0.5491,
      "num_input_tokens_seen": 5109968,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.07225166261196136,
      "learning_rate": 0.0009667705415664227,
      "loss": 0.4279,
      "num_input_tokens_seen": 5113040,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.11206669360399246,
      "learning_rate": 0.0009666885156880026,
      "loss": 0.4436,
      "num_input_tokens_seen": 5117072,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 0.12620986998081207,
      "learning_rate": 0.0009666063921836692,
      "loss": 0.4522,
      "num_input_tokens_seen": 5121168,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 0.08056183904409409,
      "learning_rate": 0.0009665241710706019,
      "loss": 0.4493,
      "num_input_tokens_seen": 5124144,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.09124455600976944,
      "learning_rate": 0.0009664418523660003,
      "loss": 0.5441,
      "num_input_tokens_seen": 5126736,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.08706751465797424,
      "learning_rate": 0.0009663594360870847,
      "loss": 0.4411,
      "num_input_tokens_seen": 5130768,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 0.07331693917512894,
      "learning_rate": 0.0009662769222510955,
      "loss": 0.425,
      "num_input_tokens_seen": 5133424,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 0.1418558657169342,
      "learning_rate": 0.0009661943108752939,
      "loss": 0.4359,
      "num_input_tokens_seen": 5136304,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.09048959612846375,
      "learning_rate": 0.0009661116019769609,
      "loss": 0.4892,
      "num_input_tokens_seen": 5139472,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 0.059047549962997437,
      "learning_rate": 0.0009660287955733986,
      "loss": 0.3827,
      "num_input_tokens_seen": 5144144,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.10223119705915451,
      "learning_rate": 0.0009659458916819289,
      "loss": 0.411,
      "num_input_tokens_seen": 5148496,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 0.2119230180978775,
      "learning_rate": 0.0009658628903198945,
      "loss": 0.4463,
      "num_input_tokens_seen": 5151760,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 0.09060840308666229,
      "learning_rate": 0.0009657797915046583,
      "loss": 0.3074,
      "num_input_tokens_seen": 5155920,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 0.12008015811443329,
      "learning_rate": 0.0009656965952536036,
      "loss": 0.7352,
      "num_input_tokens_seen": 5158576,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.08486557006835938,
      "learning_rate": 0.0009656133015841342,
      "loss": 0.4057,
      "num_input_tokens_seen": 5161712,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 0.082997627556324,
      "learning_rate": 0.000965529910513674,
      "loss": 0.478,
      "num_input_tokens_seen": 5165456,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.1042407751083374,
      "learning_rate": 0.0009654464220596676,
      "loss": 0.4755,
      "num_input_tokens_seen": 5168048,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 0.2191908210515976,
      "learning_rate": 0.0009653628362395799,
      "loss": 0.5161,
      "num_input_tokens_seen": 5170928,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 0.10242264717817307,
      "learning_rate": 0.0009652791530708958,
      "loss": 0.468,
      "num_input_tokens_seen": 5173520,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.08374582231044769,
      "learning_rate": 0.0009651953725711212,
      "loss": 0.5017,
      "num_input_tokens_seen": 5176496,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.07590819150209427,
      "learning_rate": 0.0009651114947577818,
      "loss": 0.5138,
      "num_input_tokens_seen": 5180112,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 0.06078317388892174,
      "learning_rate": 0.0009650275196484239,
      "loss": 0.3737,
      "num_input_tokens_seen": 5184048,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.08876938372850418,
      "learning_rate": 0.0009649434472606144,
      "loss": 0.3498,
      "num_input_tokens_seen": 5186960,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 0.13013002276420593,
      "learning_rate": 0.00096485927761194,
      "loss": 0.3831,
      "num_input_tokens_seen": 5190064,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 0.10930291563272476,
      "learning_rate": 0.0009647750107200082,
      "loss": 0.5912,
      "num_input_tokens_seen": 5192912,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.17295919358730316,
      "learning_rate": 0.0009646906466024465,
      "loss": 0.4169,
      "num_input_tokens_seen": 5195152,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 0.12459706515073776,
      "learning_rate": 0.000964606185276903,
      "loss": 0.4632,
      "num_input_tokens_seen": 5198192,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 0.060842983424663544,
      "learning_rate": 0.0009645216267610461,
      "loss": 0.3848,
      "num_input_tokens_seen": 5201296,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 0.05996870994567871,
      "learning_rate": 0.0009644369710725644,
      "loss": 0.4391,
      "num_input_tokens_seen": 5205232,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 0.1577700674533844,
      "learning_rate": 0.0009643522182291669,
      "loss": 0.4036,
      "num_input_tokens_seen": 5208528,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.11983370035886765,
      "learning_rate": 0.000964267368248583,
      "loss": 0.3796,
      "num_input_tokens_seen": 5211824,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 0.11290454119443893,
      "learning_rate": 0.0009641824211485623,
      "loss": 0.5498,
      "num_input_tokens_seen": 5214928,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 0.06315267086029053,
      "learning_rate": 0.0009640973769468747,
      "loss": 0.4937,
      "num_input_tokens_seen": 5218224,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.07977525144815445,
      "learning_rate": 0.0009640122356613105,
      "loss": 0.3905,
      "num_input_tokens_seen": 5221200,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 0.06951248645782471,
      "learning_rate": 0.0009639269973096805,
      "loss": 0.5257,
      "num_input_tokens_seen": 5224368,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 0.15686233341693878,
      "learning_rate": 0.0009638416619098154,
      "loss": 0.5236,
      "num_input_tokens_seen": 5228208,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 0.10845869779586792,
      "learning_rate": 0.0009637562294795663,
      "loss": 0.5807,
      "num_input_tokens_seen": 5230864,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 0.09080027043819427,
      "learning_rate": 0.0009636707000368049,
      "loss": 0.3743,
      "num_input_tokens_seen": 5233360,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 0.09071845561265945,
      "learning_rate": 0.000963585073599423,
      "loss": 0.4628,
      "num_input_tokens_seen": 5236624,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 0.08621454983949661,
      "learning_rate": 0.0009634993501853323,
      "loss": 0.4277,
      "num_input_tokens_seen": 5239696,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 0.0945744439959526,
      "learning_rate": 0.0009634135298124656,
      "loss": 0.4406,
      "num_input_tokens_seen": 5242736,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.09045206755399704,
      "learning_rate": 0.0009633276124987752,
      "loss": 0.5811,
      "num_input_tokens_seen": 5246544,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 0.08195571601390839,
      "learning_rate": 0.0009632415982622342,
      "loss": 0.6741,
      "num_input_tokens_seen": 5249840,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 0.07621531933546066,
      "learning_rate": 0.0009631554871208359,
      "loss": 0.3097,
      "num_input_tokens_seen": 5253456,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 0.09124267101287842,
      "learning_rate": 0.0009630692790925936,
      "loss": 0.4271,
      "num_input_tokens_seen": 5256208,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 0.12620460987091064,
      "learning_rate": 0.0009629829741955411,
      "loss": 0.4765,
      "num_input_tokens_seen": 5258512,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 0.13227632641792297,
      "learning_rate": 0.0009628965724477325,
      "loss": 0.4045,
      "num_input_tokens_seen": 5261520,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 0.124907948076725,
      "learning_rate": 0.0009628100738672419,
      "loss": 0.5729,
      "num_input_tokens_seen": 5264400,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 0.09464811533689499,
      "learning_rate": 0.0009627234784721637,
      "loss": 0.3396,
      "num_input_tokens_seen": 5267472,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 0.12485843151807785,
      "learning_rate": 0.0009626367862806129,
      "loss": 0.4025,
      "num_input_tokens_seen": 5270736,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 0.11573173105716705,
      "learning_rate": 0.0009625499973107246,
      "loss": 0.4772,
      "num_input_tokens_seen": 5274384,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 0.09033893048763275,
      "learning_rate": 0.0009624631115806537,
      "loss": 0.5232,
      "num_input_tokens_seen": 5276912,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 0.11073070019483566,
      "learning_rate": 0.0009623761291085761,
      "loss": 0.5703,
      "num_input_tokens_seen": 5279792,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 0.1036398634314537,
      "learning_rate": 0.0009622890499126873,
      "loss": 0.5247,
      "num_input_tokens_seen": 5282992,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.08775950968265533,
      "learning_rate": 0.0009622018740112032,
      "loss": 0.4775,
      "num_input_tokens_seen": 5285456,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.10190289467573166,
      "learning_rate": 0.0009621146014223603,
      "loss": 0.5832,
      "num_input_tokens_seen": 5288432,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 0.10250365734100342,
      "learning_rate": 0.0009620272321644148,
      "loss": 0.5962,
      "num_input_tokens_seen": 5291248,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 0.13037605583667755,
      "learning_rate": 0.0009619397662556434,
      "loss": 0.3941,
      "num_input_tokens_seen": 5293680,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 0.09084940701723099,
      "learning_rate": 0.000961852203714343,
      "loss": 0.5531,
      "num_input_tokens_seen": 5297040,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 0.12121853977441788,
      "learning_rate": 0.0009617645445588307,
      "loss": 0.4235,
      "num_input_tokens_seen": 5300592,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 0.07427573949098587,
      "learning_rate": 0.0009616767888074438,
      "loss": 0.6537,
      "num_input_tokens_seen": 5303440,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 0.10834221541881561,
      "learning_rate": 0.0009615889364785397,
      "loss": 0.6519,
      "num_input_tokens_seen": 5306384,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 0.16186657547950745,
      "learning_rate": 0.000961500987590496,
      "loss": 0.5354,
      "num_input_tokens_seen": 5309936,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.07185287773609161,
      "learning_rate": 0.0009614129421617111,
      "loss": 0.4844,
      "num_input_tokens_seen": 5314576,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 0.13465380668640137,
      "learning_rate": 0.0009613248002106027,
      "loss": 0.4264,
      "num_input_tokens_seen": 5318128,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 0.0979866087436676,
      "learning_rate": 0.000961236561755609,
      "loss": 0.5099,
      "num_input_tokens_seen": 5320944,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 0.11255344748497009,
      "learning_rate": 0.0009611482268151888,
      "loss": 0.5042,
      "num_input_tokens_seen": 5324208,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 0.08430100232362747,
      "learning_rate": 0.0009610597954078206,
      "loss": 0.4842,
      "num_input_tokens_seen": 5327344,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 0.07278410345315933,
      "learning_rate": 0.0009609712675520031,
      "loss": 0.3827,
      "num_input_tokens_seen": 5330288,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.053863275796175,
      "learning_rate": 0.0009608826432662556,
      "loss": 0.3853,
      "num_input_tokens_seen": 5334192,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 0.0627850592136383,
      "learning_rate": 0.0009607939225691172,
      "loss": 0.545,
      "num_input_tokens_seen": 5337360,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 0.08280535787343979,
      "learning_rate": 0.0009607051054791472,
      "loss": 0.4809,
      "num_input_tokens_seen": 5341072,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.1912153661251068,
      "learning_rate": 0.000960616192014925,
      "loss": 0.5092,
      "num_input_tokens_seen": 5344592,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.10722079128026962,
      "learning_rate": 0.0009605271821950506,
      "loss": 0.6188,
      "num_input_tokens_seen": 5346992,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 0.08487934619188309,
      "learning_rate": 0.0009604380760381434,
      "loss": 0.5129,
      "num_input_tokens_seen": 5349776,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.0757712870836258,
      "learning_rate": 0.0009603488735628439,
      "loss": 0.5376,
      "num_input_tokens_seen": 5353968,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 0.0793408751487732,
      "learning_rate": 0.0009602595747878118,
      "loss": 0.5698,
      "num_input_tokens_seen": 5356560,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 0.08240462839603424,
      "learning_rate": 0.0009601701797317278,
      "loss": 0.3592,
      "num_input_tokens_seen": 5359824,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 0.10300353914499283,
      "learning_rate": 0.0009600806884132917,
      "loss": 0.5704,
      "num_input_tokens_seen": 5364144,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 0.037744902074337006,
      "learning_rate": 0.0009599911008512248,
      "loss": 0.5319,
      "num_input_tokens_seen": 5368336,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 0.20338934659957886,
      "learning_rate": 0.0009599014170642674,
      "loss": 0.4015,
      "num_input_tokens_seen": 5371440,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 0.08042919635772705,
      "learning_rate": 0.0009598116370711805,
      "loss": 0.4096,
      "num_input_tokens_seen": 5374160,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 0.08973535150289536,
      "learning_rate": 0.0009597217608907447,
      "loss": 0.5534,
      "num_input_tokens_seen": 5377680,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.10578598082065582,
      "learning_rate": 0.0009596317885417614,
      "loss": 0.4084,
      "num_input_tokens_seen": 5381488,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 0.10410107672214508,
      "learning_rate": 0.0009595417200430516,
      "loss": 0.4241,
      "num_input_tokens_seen": 5384912,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 0.047178227454423904,
      "learning_rate": 0.0009594515554134568,
      "loss": 0.3824,
      "num_input_tokens_seen": 5387792,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 0.08988776803016663,
      "learning_rate": 0.0009593612946718384,
      "loss": 0.4995,
      "num_input_tokens_seen": 5390608,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 0.08032584190368652,
      "learning_rate": 0.0009592709378370778,
      "loss": 0.4864,
      "num_input_tokens_seen": 5393328,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.09746292233467102,
      "learning_rate": 0.0009591804849280766,
      "loss": 0.4857,
      "num_input_tokens_seen": 5395952,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.11794234067201614,
      "learning_rate": 0.0009590899359637564,
      "loss": 0.4535,
      "num_input_tokens_seen": 5398384,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 0.08797386288642883,
      "learning_rate": 0.0009589992909630594,
      "loss": 0.4287,
      "num_input_tokens_seen": 5401424,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.09705904871225357,
      "learning_rate": 0.0009589085499449471,
      "loss": 0.3516,
      "num_input_tokens_seen": 5404976,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 0.08613159507513046,
      "learning_rate": 0.0009588177129284017,
      "loss": 0.3618,
      "num_input_tokens_seen": 5408464,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 0.11218082904815674,
      "learning_rate": 0.0009587267799324253,
      "loss": 0.4975,
      "num_input_tokens_seen": 5411280,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 0.17990103363990784,
      "learning_rate": 0.0009586357509760399,
      "loss": 0.4441,
      "num_input_tokens_seen": 5415184,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 0.11456860601902008,
      "learning_rate": 0.0009585446260782878,
      "loss": 0.3832,
      "num_input_tokens_seen": 5417456,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 0.11529452353715897,
      "learning_rate": 0.0009584534052582313,
      "loss": 0.4138,
      "num_input_tokens_seen": 5420016,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 0.10042732208967209,
      "learning_rate": 0.0009583620885349527,
      "loss": 0.5575,
      "num_input_tokens_seen": 5423344,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.08849557489156723,
      "learning_rate": 0.0009582706759275546,
      "loss": 0.3978,
      "num_input_tokens_seen": 5426640,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 0.07730532437562943,
      "learning_rate": 0.0009581791674551592,
      "loss": 0.4499,
      "num_input_tokens_seen": 5430256,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 0.11481074243783951,
      "learning_rate": 0.000958087563136909,
      "loss": 0.3036,
      "num_input_tokens_seen": 5433520,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 0.18880659341812134,
      "learning_rate": 0.000957995862991967,
      "loss": 0.4639,
      "num_input_tokens_seen": 5436368,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.09801088273525238,
      "learning_rate": 0.0009579040670395154,
      "loss": 0.4489,
      "num_input_tokens_seen": 5439216,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 0.13306830823421478,
      "learning_rate": 0.000957812175298757,
      "loss": 0.4627,
      "num_input_tokens_seen": 5441776,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 0.062337081879377365,
      "learning_rate": 0.0009577201877889145,
      "loss": 0.5922,
      "num_input_tokens_seen": 5445168,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 0.08845793455839157,
      "learning_rate": 0.0009576281045292308,
      "loss": 0.4819,
      "num_input_tokens_seen": 5450192,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.10130976140499115,
      "learning_rate": 0.0009575359255389686,
      "loss": 0.4032,
      "num_input_tokens_seen": 5453200,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 0.1093606948852539,
      "learning_rate": 0.0009574436508374104,
      "loss": 0.4854,
      "num_input_tokens_seen": 5456144,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 0.08658377081155777,
      "learning_rate": 0.0009573512804438594,
      "loss": 0.6686,
      "num_input_tokens_seen": 5460016,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.22190754115581512,
      "learning_rate": 0.0009572588143776381,
      "loss": 0.42,
      "num_input_tokens_seen": 5462384,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.10858205705881119,
      "learning_rate": 0.0009571662526580897,
      "loss": 0.5937,
      "num_input_tokens_seen": 5466416,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.13720035552978516,
      "learning_rate": 0.0009570735953045768,
      "loss": 0.4089,
      "num_input_tokens_seen": 5470736,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 0.11369368433952332,
      "learning_rate": 0.0009569808423364823,
      "loss": 0.5014,
      "num_input_tokens_seen": 5473648,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 0.0793522372841835,
      "learning_rate": 0.0009568879937732091,
      "loss": 0.4122,
      "num_input_tokens_seen": 5477776,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 0.057230159640312195,
      "learning_rate": 0.0009567950496341802,
      "loss": 0.3702,
      "num_input_tokens_seen": 5481552,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.11604391783475876,
      "learning_rate": 0.0009567020099388382,
      "loss": 0.5333,
      "num_input_tokens_seen": 5484656,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.08682439476251602,
      "learning_rate": 0.0009566088747066459,
      "loss": 0.5016,
      "num_input_tokens_seen": 5487568,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 0.08491957187652588,
      "learning_rate": 0.0009565156439570866,
      "loss": 0.5597,
      "num_input_tokens_seen": 5490928,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 0.08736533671617508,
      "learning_rate": 0.0009564223177096625,
      "loss": 0.4665,
      "num_input_tokens_seen": 5495536,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.09692099690437317,
      "learning_rate": 0.0009563288959838969,
      "loss": 0.4789,
      "num_input_tokens_seen": 5499504,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 0.10563516616821289,
      "learning_rate": 0.0009562353787993321,
      "loss": 0.4256,
      "num_input_tokens_seen": 5502320,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 0.2381521463394165,
      "learning_rate": 0.0009561417661755312,
      "loss": 0.5301,
      "num_input_tokens_seen": 5505552,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 0.09765303879976273,
      "learning_rate": 0.0009560480581320768,
      "loss": 0.4624,
      "num_input_tokens_seen": 5509776,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 0.06388338655233383,
      "learning_rate": 0.0009559542546885714,
      "loss": 0.51,
      "num_input_tokens_seen": 5514640,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 0.10280240327119827,
      "learning_rate": 0.0009558603558646378,
      "loss": 0.4954,
      "num_input_tokens_seen": 5517360,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.08604209125041962,
      "learning_rate": 0.0009557663616799185,
      "loss": 0.419,
      "num_input_tokens_seen": 5520912,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.16463805735111237,
      "learning_rate": 0.0009556722721540759,
      "loss": 0.4592,
      "num_input_tokens_seen": 5524272,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 0.11761453747749329,
      "learning_rate": 0.0009555780873067927,
      "loss": 0.3968,
      "num_input_tokens_seen": 5527280,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 0.1414610594511032,
      "learning_rate": 0.000955483807157771,
      "loss": 0.4137,
      "num_input_tokens_seen": 5529872,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 0.1179981604218483,
      "learning_rate": 0.0009553894317267333,
      "loss": 0.585,
      "num_input_tokens_seen": 5532976,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 0.07843130081892014,
      "learning_rate": 0.0009552949610334219,
      "loss": 0.552,
      "num_input_tokens_seen": 5536400,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.12143892049789429,
      "learning_rate": 0.000955200395097599,
      "loss": 0.5595,
      "num_input_tokens_seen": 5540304,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.06162422522902489,
      "learning_rate": 0.0009551057339390464,
      "loss": 0.4531,
      "num_input_tokens_seen": 5544592,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 0.09328816086053848,
      "learning_rate": 0.0009550109775775666,
      "loss": 0.4367,
      "num_input_tokens_seen": 5547248,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 0.11875202506780624,
      "learning_rate": 0.0009549161260329811,
      "loss": 0.4367,
      "num_input_tokens_seen": 5550096,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 0.30948224663734436,
      "learning_rate": 0.0009548211793251322,
      "loss": 0.456,
      "num_input_tokens_seen": 5553328,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 0.08274311572313309,
      "learning_rate": 0.0009547261374738814,
      "loss": 0.4783,
      "num_input_tokens_seen": 5557264,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 0.09914425015449524,
      "learning_rate": 0.0009546310004991105,
      "loss": 0.449,
      "num_input_tokens_seen": 5560912,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 0.0889965072274208,
      "learning_rate": 0.000954535768420721,
      "loss": 0.4266,
      "num_input_tokens_seen": 5564976,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 0.10291553288698196,
      "learning_rate": 0.0009544404412586343,
      "loss": 0.4117,
      "num_input_tokens_seen": 5568208,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 0.09011854231357574,
      "learning_rate": 0.0009543450190327917,
      "loss": 0.4528,
      "num_input_tokens_seen": 5571696,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 0.07005566358566284,
      "learning_rate": 0.0009542495017631547,
      "loss": 0.4947,
      "num_input_tokens_seen": 5575760,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.18172496557235718,
      "learning_rate": 0.0009541538894697043,
      "loss": 0.4844,
      "num_input_tokens_seen": 5578992,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.10960790514945984,
      "learning_rate": 0.0009540581821724414,
      "loss": 0.5184,
      "num_input_tokens_seen": 5581712,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 0.09496012330055237,
      "learning_rate": 0.000953962379891387,
      "loss": 0.3789,
      "num_input_tokens_seen": 5584688,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 0.05532783642411232,
      "learning_rate": 0.0009538664826465818,
      "loss": 0.5305,
      "num_input_tokens_seen": 5589008,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.11086906492710114,
      "learning_rate": 0.0009537704904580864,
      "loss": 0.4038,
      "num_input_tokens_seen": 5591280,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 0.07927487045526505,
      "learning_rate": 0.0009536744033459815,
      "loss": 0.4154,
      "num_input_tokens_seen": 5596176,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 0.08823686838150024,
      "learning_rate": 0.0009535782213303669,
      "loss": 0.7805,
      "num_input_tokens_seen": 5598832,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 0.08262131363153458,
      "learning_rate": 0.0009534819444313631,
      "loss": 0.4679,
      "num_input_tokens_seen": 5602288,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 0.12472075968980789,
      "learning_rate": 0.0009533855726691103,
      "loss": 0.5709,
      "num_input_tokens_seen": 5608976,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.09955167770385742,
      "learning_rate": 0.0009532891060637681,
      "loss": 0.391,
      "num_input_tokens_seen": 5612688,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 0.09922883659601212,
      "learning_rate": 0.0009531925446355163,
      "loss": 0.5708,
      "num_input_tokens_seen": 5616208,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.07016316056251526,
      "learning_rate": 0.0009530958884045545,
      "loss": 0.4941,
      "num_input_tokens_seen": 5619568,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 0.14405307173728943,
      "learning_rate": 0.000952999137391102,
      "loss": 0.5403,
      "num_input_tokens_seen": 5622864,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 0.0871758759021759,
      "learning_rate": 0.0009529022916153982,
      "loss": 0.5765,
      "num_input_tokens_seen": 5625360,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 0.08887924998998642,
      "learning_rate": 0.0009528053510977017,
      "loss": 0.592,
      "num_input_tokens_seen": 5628400,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 0.11821289360523224,
      "learning_rate": 0.0009527083158582919,
      "loss": 0.5699,
      "num_input_tokens_seen": 5631152,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.09621808677911758,
      "learning_rate": 0.0009526111859174671,
      "loss": 0.4571,
      "num_input_tokens_seen": 5634800,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.09690241515636444,
      "learning_rate": 0.0009525139612955458,
      "loss": 0.3449,
      "num_input_tokens_seen": 5638096,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 0.08049529045820236,
      "learning_rate": 0.0009524166420128664,
      "loss": 0.5244,
      "num_input_tokens_seen": 5641072,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 0.09495560824871063,
      "learning_rate": 0.0009523192280897867,
      "loss": 0.4349,
      "num_input_tokens_seen": 5644944,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 0.11333908140659332,
      "learning_rate": 0.0009522217195466851,
      "loss": 0.473,
      "num_input_tokens_seen": 5647984,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.05844281241297722,
      "learning_rate": 0.0009521241164039589,
      "loss": 0.4361,
      "num_input_tokens_seen": 5651312,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.11907614767551422,
      "learning_rate": 0.0009520264186820258,
      "loss": 0.5068,
      "num_input_tokens_seen": 5653904,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 0.17084097862243652,
      "learning_rate": 0.0009519286264013227,
      "loss": 0.4792,
      "num_input_tokens_seen": 5657552,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 0.09192916750907898,
      "learning_rate": 0.0009518307395823069,
      "loss": 0.4369,
      "num_input_tokens_seen": 5660720,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 0.09608013927936554,
      "learning_rate": 0.0009517327582454551,
      "loss": 0.4715,
      "num_input_tokens_seen": 5663280,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 0.14184918999671936,
      "learning_rate": 0.000951634682411264,
      "loss": 0.5666,
      "num_input_tokens_seen": 5667984,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 0.1562315821647644,
      "learning_rate": 0.0009515365121002498,
      "loss": 0.5075,
      "num_input_tokens_seen": 5670608,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 0.08898278325796127,
      "learning_rate": 0.0009514382473329487,
      "loss": 0.4346,
      "num_input_tokens_seen": 5674000,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 0.14124034345149994,
      "learning_rate": 0.0009513398881299164,
      "loss": 0.4186,
      "num_input_tokens_seen": 5676080,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 0.06962040811777115,
      "learning_rate": 0.0009512414345117289,
      "loss": 0.4539,
      "num_input_tokens_seen": 5679888,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 0.15653353929519653,
      "learning_rate": 0.0009511428864989813,
      "loss": 0.6894,
      "num_input_tokens_seen": 5683344,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 0.060792453587055206,
      "learning_rate": 0.0009510442441122886,
      "loss": 0.4038,
      "num_input_tokens_seen": 5687056,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 0.07261049747467041,
      "learning_rate": 0.0009509455073722859,
      "loss": 0.4349,
      "num_input_tokens_seen": 5690224,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 0.06047463417053223,
      "learning_rate": 0.0009508466762996277,
      "loss": 0.4225,
      "num_input_tokens_seen": 5693840,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.06697635352611542,
      "learning_rate": 0.0009507477509149883,
      "loss": 0.4577,
      "num_input_tokens_seen": 5698896,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.12144966423511505,
      "learning_rate": 0.0009506487312390619,
      "loss": 0.5438,
      "num_input_tokens_seen": 5701744,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 0.06962014734745026,
      "learning_rate": 0.0009505496172925622,
      "loss": 0.4703,
      "num_input_tokens_seen": 5704816,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.1377490907907486,
      "learning_rate": 0.0009504504090962226,
      "loss": 0.4464,
      "num_input_tokens_seen": 5708112,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 0.09651226550340652,
      "learning_rate": 0.0009503511066707966,
      "loss": 0.4888,
      "num_input_tokens_seen": 5711280,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 0.10770486295223236,
      "learning_rate": 0.0009502517100370568,
      "loss": 0.5568,
      "num_input_tokens_seen": 5714576,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 0.13651752471923828,
      "learning_rate": 0.0009501522192157961,
      "loss": 0.4522,
      "num_input_tokens_seen": 5720944,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 0.10852257907390594,
      "learning_rate": 0.0009500526342278266,
      "loss": 0.3977,
      "num_input_tokens_seen": 5723344,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.06488864868879318,
      "learning_rate": 0.0009499529550939807,
      "loss": 0.4843,
      "num_input_tokens_seen": 5726864,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 0.057004835456609726,
      "learning_rate": 0.0009498531818351098,
      "loss": 0.4445,
      "num_input_tokens_seen": 5730288,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 0.10994640737771988,
      "learning_rate": 0.0009497533144720854,
      "loss": 0.4933,
      "num_input_tokens_seen": 5734160,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.10793858766555786,
      "learning_rate": 0.0009496533530257988,
      "loss": 0.6022,
      "num_input_tokens_seen": 5736752,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 0.0762055441737175,
      "learning_rate": 0.0009495532975171605,
      "loss": 0.4962,
      "num_input_tokens_seen": 5740976,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 0.0936499536037445,
      "learning_rate": 0.0009494531479671014,
      "loss": 0.5377,
      "num_input_tokens_seen": 5743376,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 0.1491774469614029,
      "learning_rate": 0.0009493529043965712,
      "loss": 0.4418,
      "num_input_tokens_seen": 5745776,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 0.15138936042785645,
      "learning_rate": 0.00094925256682654,
      "loss": 0.4538,
      "num_input_tokens_seen": 5748784,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 0.12810280919075012,
      "learning_rate": 0.000949152135277997,
      "loss": 0.4697,
      "num_input_tokens_seen": 5751440,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.161020427942276,
      "learning_rate": 0.0009490516097719515,
      "loss": 0.4795,
      "num_input_tokens_seen": 5753648,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.2591994106769562,
      "learning_rate": 0.0009489509903294324,
      "loss": 0.5038,
      "num_input_tokens_seen": 5756016,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 0.07800296694040298,
      "learning_rate": 0.000948850276971488,
      "loss": 0.4454,
      "num_input_tokens_seen": 5759312,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 0.08918770402669907,
      "learning_rate": 0.0009487494697191864,
      "loss": 0.3553,
      "num_input_tokens_seen": 5762640,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 0.07803738862276077,
      "learning_rate": 0.0009486485685936154,
      "loss": 0.5129,
      "num_input_tokens_seen": 5765840,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 0.13334907591342926,
      "learning_rate": 0.0009485475736158822,
      "loss": 0.391,
      "num_input_tokens_seen": 5768208,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 0.1888243407011032,
      "learning_rate": 0.000948446484807114,
      "loss": 0.6517,
      "num_input_tokens_seen": 5770896,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 0.10726070404052734,
      "learning_rate": 0.0009483453021884572,
      "loss": 0.4255,
      "num_input_tokens_seen": 5773488,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 0.12717166543006897,
      "learning_rate": 0.0009482440257810782,
      "loss": 0.3714,
      "num_input_tokens_seen": 5777168,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.14981523156166077,
      "learning_rate": 0.000948142655606163,
      "loss": 0.4872,
      "num_input_tokens_seen": 5779536,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 0.11007840186357498,
      "learning_rate": 0.0009480411916849168,
      "loss": 0.3537,
      "num_input_tokens_seen": 5782992,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 0.1305396854877472,
      "learning_rate": 0.0009479396340385649,
      "loss": 0.5208,
      "num_input_tokens_seen": 5785776,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 0.1309049427509308,
      "learning_rate": 0.0009478379826883519,
      "loss": 0.3788,
      "num_input_tokens_seen": 5788272,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 0.09401501715183258,
      "learning_rate": 0.0009477362376555421,
      "loss": 0.5058,
      "num_input_tokens_seen": 5791248,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 0.07343127578496933,
      "learning_rate": 0.0009476343989614194,
      "loss": 0.3403,
      "num_input_tokens_seen": 5794640,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 0.09840812534093857,
      "learning_rate": 0.0009475324666272873,
      "loss": 0.4523,
      "num_input_tokens_seen": 5797328,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 0.0783202052116394,
      "learning_rate": 0.0009474304406744689,
      "loss": 0.4689,
      "num_input_tokens_seen": 5800144,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 0.07777038216590881,
      "learning_rate": 0.0009473283211243069,
      "loss": 0.5065,
      "num_input_tokens_seen": 5803920,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 0.07898552715778351,
      "learning_rate": 0.0009472261079981637,
      "loss": 0.3768,
      "num_input_tokens_seen": 5806736,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.07295532524585724,
      "learning_rate": 0.0009471238013174206,
      "loss": 0.4451,
      "num_input_tokens_seen": 5809744,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 0.07382342219352722,
      "learning_rate": 0.0009470214011034795,
      "loss": 0.4774,
      "num_input_tokens_seen": 5813712,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 0.0943484753370285,
      "learning_rate": 0.0009469189073777612,
      "loss": 0.4903,
      "num_input_tokens_seen": 5816880,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 0.1374306082725525,
      "learning_rate": 0.0009468163201617061,
      "loss": 0.5351,
      "num_input_tokens_seen": 5819792,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 0.08033180236816406,
      "learning_rate": 0.0009467136394767744,
      "loss": 0.5164,
      "num_input_tokens_seen": 5822768,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.06134646758437157,
      "learning_rate": 0.0009466108653444458,
      "loss": 0.3509,
      "num_input_tokens_seen": 5826032,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.09701204299926758,
      "learning_rate": 0.0009465079977862193,
      "loss": 0.4592,
      "num_input_tokens_seen": 5828656,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 0.12737543880939484,
      "learning_rate": 0.0009464050368236137,
      "loss": 0.4033,
      "num_input_tokens_seen": 5832048,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 0.09398925304412842,
      "learning_rate": 0.0009463019824781674,
      "loss": 0.4104,
      "num_input_tokens_seen": 5836080,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 0.06864964962005615,
      "learning_rate": 0.0009461988347714377,
      "loss": 0.5352,
      "num_input_tokens_seen": 5838928,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 0.07199259102344513,
      "learning_rate": 0.0009460955937250025,
      "loss": 0.4513,
      "num_input_tokens_seen": 5842224,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 0.06439848989248276,
      "learning_rate": 0.0009459922593604584,
      "loss": 0.4406,
      "num_input_tokens_seen": 5845520,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 0.15299005806446075,
      "learning_rate": 0.0009458888316994219,
      "loss": 0.4207,
      "num_input_tokens_seen": 5848368,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 0.09413887560367584,
      "learning_rate": 0.0009457853107635286,
      "loss": 0.4491,
      "num_input_tokens_seen": 5851664,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 0.20253439247608185,
      "learning_rate": 0.0009456816965744342,
      "loss": 0.5211,
      "num_input_tokens_seen": 5854416,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 0.08836973458528519,
      "learning_rate": 0.0009455779891538134,
      "loss": 0.4555,
      "num_input_tokens_seen": 5857744,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 0.06576607376337051,
      "learning_rate": 0.0009454741885233606,
      "loss": 0.3965,
      "num_input_tokens_seen": 5862192,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 0.0754881426692009,
      "learning_rate": 0.0009453702947047899,
      "loss": 0.3798,
      "num_input_tokens_seen": 5865296,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.14898882806301117,
      "learning_rate": 0.0009452663077198347,
      "loss": 0.5521,
      "num_input_tokens_seen": 5868272,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 0.08935975283384323,
      "learning_rate": 0.0009451622275902477,
      "loss": 0.4658,
      "num_input_tokens_seen": 5871408,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 0.10731136798858643,
      "learning_rate": 0.0009450580543378013,
      "loss": 0.4163,
      "num_input_tokens_seen": 5874800,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.10397758334875107,
      "learning_rate": 0.0009449537879842875,
      "loss": 0.5108,
      "num_input_tokens_seen": 5879088,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.10762035846710205,
      "learning_rate": 0.0009448494285515177,
      "loss": 0.5161,
      "num_input_tokens_seen": 5881840,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.09183021634817123,
      "learning_rate": 0.0009447449760613222,
      "loss": 0.4415,
      "num_input_tokens_seen": 5885968,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.12879720330238342,
      "learning_rate": 0.0009446404305355519,
      "loss": 0.5667,
      "num_input_tokens_seen": 5888944,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 0.08403091877698898,
      "learning_rate": 0.0009445357919960762,
      "loss": 0.412,
      "num_input_tokens_seen": 5891920,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 0.08903883397579193,
      "learning_rate": 0.0009444310604647844,
      "loss": 0.4828,
      "num_input_tokens_seen": 5894544,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 0.09309939295053482,
      "learning_rate": 0.000944326235963585,
      "loss": 0.4642,
      "num_input_tokens_seen": 5897648,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.062204744666814804,
      "learning_rate": 0.0009442213185144062,
      "loss": 0.3895,
      "num_input_tokens_seen": 5900848,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 0.08607852458953857,
      "learning_rate": 0.0009441163081391954,
      "loss": 0.388,
      "num_input_tokens_seen": 5903408,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 0.22411371767520905,
      "learning_rate": 0.00094401120485992,
      "loss": 0.6181,
      "num_input_tokens_seen": 5905840,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 0.09020252525806427,
      "learning_rate": 0.0009439060086985658,
      "loss": 0.397,
      "num_input_tokens_seen": 5909968,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 0.1779738813638687,
      "learning_rate": 0.000943800719677139,
      "loss": 0.3636,
      "num_input_tokens_seen": 5912624,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 0.0754212811589241,
      "learning_rate": 0.0009436953378176649,
      "loss": 0.4183,
      "num_input_tokens_seen": 5915824,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.09458637237548828,
      "learning_rate": 0.0009435898631421879,
      "loss": 0.4792,
      "num_input_tokens_seen": 5920048,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.11796688288450241,
      "learning_rate": 0.0009434842956727724,
      "loss": 0.5703,
      "num_input_tokens_seen": 5923088,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 0.07205184549093246,
      "learning_rate": 0.0009433786354315017,
      "loss": 0.5059,
      "num_input_tokens_seen": 5925872,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 0.3014639914035797,
      "learning_rate": 0.0009432728824404789,
      "loss": 0.5468,
      "num_input_tokens_seen": 5928464,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 0.07082657516002655,
      "learning_rate": 0.0009431670367218262,
      "loss": 0.2995,
      "num_input_tokens_seen": 5931344,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.1937992423772812,
      "learning_rate": 0.0009430610982976852,
      "loss": 0.5674,
      "num_input_tokens_seen": 5934224,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 0.11174140870571136,
      "learning_rate": 0.0009429550671902171,
      "loss": 0.4035,
      "num_input_tokens_seen": 5937424,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.09428904950618744,
      "learning_rate": 0.0009428489434216028,
      "loss": 0.4042,
      "num_input_tokens_seen": 5940624,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 0.25063684582710266,
      "learning_rate": 0.0009427427270140415,
      "loss": 0.4398,
      "num_input_tokens_seen": 5943184,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 0.15282374620437622,
      "learning_rate": 0.0009426364179897529,
      "loss": 0.4656,
      "num_input_tokens_seen": 5945680,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.09049132466316223,
      "learning_rate": 0.0009425300163709756,
      "loss": 0.3865,
      "num_input_tokens_seen": 5948816,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 0.1062132939696312,
      "learning_rate": 0.0009424235221799673,
      "loss": 0.4289,
      "num_input_tokens_seen": 5951984,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 0.08258180320262909,
      "learning_rate": 0.0009423169354390058,
      "loss": 0.5561,
      "num_input_tokens_seen": 5954864,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 0.08428171277046204,
      "learning_rate": 0.0009422102561703875,
      "loss": 0.444,
      "num_input_tokens_seen": 5957744,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 0.12187477946281433,
      "learning_rate": 0.0009421034843964287,
      "loss": 0.4988,
      "num_input_tokens_seen": 5961008,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 0.12947233021259308,
      "learning_rate": 0.0009419966201394646,
      "loss": 0.3293,
      "num_input_tokens_seen": 5964336,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 0.0716022476553917,
      "learning_rate": 0.0009418896634218503,
      "loss": 0.4264,
      "num_input_tokens_seen": 5967920,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.05720067769289017,
      "learning_rate": 0.0009417826142659596,
      "loss": 0.4554,
      "num_input_tokens_seen": 5971440,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.09647823125123978,
      "learning_rate": 0.0009416754726941863,
      "loss": 0.4621,
      "num_input_tokens_seen": 5974096,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.10781451314687729,
      "learning_rate": 0.0009415682387289428,
      "loss": 0.5267,
      "num_input_tokens_seen": 5976624,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 0.12571127712726593,
      "learning_rate": 0.0009414609123926616,
      "loss": 0.5588,
      "num_input_tokens_seen": 5980336,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.10373854637145996,
      "learning_rate": 0.000941353493707794,
      "loss": 0.5216,
      "num_input_tokens_seen": 5982992,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 0.11722561717033386,
      "learning_rate": 0.0009412459826968107,
      "loss": 0.5732,
      "num_input_tokens_seen": 5986320,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 0.10892565548419952,
      "learning_rate": 0.0009411383793822017,
      "loss": 0.338,
      "num_input_tokens_seen": 5990032,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 0.10923690348863602,
      "learning_rate": 0.0009410306837864766,
      "loss": 0.5469,
      "num_input_tokens_seen": 5993136,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 0.11073914170265198,
      "learning_rate": 0.000940922895932164,
      "loss": 0.5557,
      "num_input_tokens_seen": 5996304,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.12242121249437332,
      "learning_rate": 0.0009408150158418119,
      "loss": 0.4584,
      "num_input_tokens_seen": 5998992,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 0.1423073261976242,
      "learning_rate": 0.0009407070435379876,
      "loss": 0.4241,
      "num_input_tokens_seen": 6001904,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 0.1255839467048645,
      "learning_rate": 0.0009405989790432776,
      "loss": 0.4763,
      "num_input_tokens_seen": 6004912,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 0.10692989826202393,
      "learning_rate": 0.0009404908223802877,
      "loss": 0.548,
      "num_input_tokens_seen": 6007696,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 0.12972769141197205,
      "learning_rate": 0.0009403825735716433,
      "loss": 0.473,
      "num_input_tokens_seen": 6010544,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 0.13744360208511353,
      "learning_rate": 0.0009402742326399887,
      "loss": 0.4895,
      "num_input_tokens_seen": 6013136,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 0.16402077674865723,
      "learning_rate": 0.0009401657996079873,
      "loss": 0.425,
      "num_input_tokens_seen": 6016336,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.10169900953769684,
      "learning_rate": 0.0009400572744983224,
      "loss": 0.5385,
      "num_input_tokens_seen": 6019120,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 0.10504340380430222,
      "learning_rate": 0.0009399486573336963,
      "loss": 0.7082,
      "num_input_tokens_seen": 6021616,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 0.07910323888063431,
      "learning_rate": 0.00093983994813683,
      "loss": 0.4318,
      "num_input_tokens_seen": 6024272,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.07686159759759903,
      "learning_rate": 0.0009397311469304648,
      "loss": 0.4526,
      "num_input_tokens_seen": 6027440,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 0.17803166806697845,
      "learning_rate": 0.0009396222537373602,
      "loss": 0.3895,
      "num_input_tokens_seen": 6032784,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.06253755837678909,
      "learning_rate": 0.0009395132685802956,
      "loss": 0.4847,
      "num_input_tokens_seen": 6035600,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.1398393213748932,
      "learning_rate": 0.0009394041914820695,
      "loss": 0.6629,
      "num_input_tokens_seen": 6039408,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 0.14055708050727844,
      "learning_rate": 0.0009392950224654994,
      "loss": 0.4246,
      "num_input_tokens_seen": 6042448,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 0.08977120369672775,
      "learning_rate": 0.0009391857615534225,
      "loss": 0.5344,
      "num_input_tokens_seen": 6046032,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.09591087698936462,
      "learning_rate": 0.0009390764087686948,
      "loss": 0.5089,
      "num_input_tokens_seen": 6048656,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 0.087856225669384,
      "learning_rate": 0.0009389669641341916,
      "loss": 0.704,
      "num_input_tokens_seen": 6052176,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 0.24499863386154175,
      "learning_rate": 0.0009388574276728075,
      "loss": 0.4414,
      "num_input_tokens_seen": 6054704,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 0.1522250473499298,
      "learning_rate": 0.0009387477994074562,
      "loss": 0.5621,
      "num_input_tokens_seen": 6057040,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 0.09575341641902924,
      "learning_rate": 0.0009386380793610707,
      "loss": 0.4349,
      "num_input_tokens_seen": 6060272,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 0.11457666754722595,
      "learning_rate": 0.0009385282675566034,
      "loss": 0.4657,
      "num_input_tokens_seen": 6063792,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 0.07134880125522614,
      "learning_rate": 0.0009384183640170255,
      "loss": 0.3357,
      "num_input_tokens_seen": 6067312,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.10709720104932785,
      "learning_rate": 0.0009383083687653275,
      "loss": 0.5249,
      "num_input_tokens_seen": 6070544,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 0.09447930008172989,
      "learning_rate": 0.0009381982818245193,
      "loss": 0.3973,
      "num_input_tokens_seen": 6073904,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 0.08622512221336365,
      "learning_rate": 0.0009380881032176299,
      "loss": 0.4625,
      "num_input_tokens_seen": 6077392,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.17023198306560516,
      "learning_rate": 0.0009379778329677071,
      "loss": 0.4369,
      "num_input_tokens_seen": 6080880,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.08757133036851883,
      "learning_rate": 0.0009378674710978184,
      "loss": 0.4491,
      "num_input_tokens_seen": 6084144,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 0.08732865005731583,
      "learning_rate": 0.0009377570176310503,
      "loss": 0.4101,
      "num_input_tokens_seen": 6087792,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 0.08897802978754044,
      "learning_rate": 0.0009376464725905082,
      "loss": 0.4615,
      "num_input_tokens_seen": 6092240,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 0.0667528286576271,
      "learning_rate": 0.000937535835999317,
      "loss": 0.5572,
      "num_input_tokens_seen": 6095920,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.0421193428337574,
      "learning_rate": 0.0009374251078806206,
      "loss": 0.4483,
      "num_input_tokens_seen": 6102960,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 0.069926418364048,
      "learning_rate": 0.000937314288257582,
      "loss": 0.4874,
      "num_input_tokens_seen": 6106320,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.07470256835222244,
      "learning_rate": 0.0009372033771533835,
      "loss": 0.5013,
      "num_input_tokens_seen": 6109104,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.11432788521051407,
      "learning_rate": 0.0009370923745912264,
      "loss": 0.5138,
      "num_input_tokens_seen": 6112368,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 0.15509644150733948,
      "learning_rate": 0.000936981280594331,
      "loss": 0.4453,
      "num_input_tokens_seen": 6114800,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 0.474692702293396,
      "learning_rate": 0.0009368700951859371,
      "loss": 0.5098,
      "num_input_tokens_seen": 6117232,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 0.18524473905563354,
      "learning_rate": 0.0009367588183893035,
      "loss": 0.5369,
      "num_input_tokens_seen": 6120112,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 0.07875817269086838,
      "learning_rate": 0.0009366474502277076,
      "loss": 0.4261,
      "num_input_tokens_seen": 6123984,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 0.11118149012327194,
      "learning_rate": 0.0009365359907244469,
      "loss": 0.3763,
      "num_input_tokens_seen": 6127568,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 0.0801849290728569,
      "learning_rate": 0.0009364244399028372,
      "loss": 0.3541,
      "num_input_tokens_seen": 6130800,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.09583701193332672,
      "learning_rate": 0.0009363127977862136,
      "loss": 0.5279,
      "num_input_tokens_seen": 6134224,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 0.1143789142370224,
      "learning_rate": 0.0009362010643979304,
      "loss": 0.5962,
      "num_input_tokens_seen": 6136912,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 0.1600050926208496,
      "learning_rate": 0.000936089239761361,
      "loss": 0.4571,
      "num_input_tokens_seen": 6141552,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 0.1581294983625412,
      "learning_rate": 0.0009359773238998978,
      "loss": 0.5087,
      "num_input_tokens_seen": 6144336,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.09121610224246979,
      "learning_rate": 0.0009358653168369524,
      "loss": 0.4275,
      "num_input_tokens_seen": 6147472,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 0.0841415598988533,
      "learning_rate": 0.0009357532185959554,
      "loss": 0.5455,
      "num_input_tokens_seen": 6150544,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 0.07831835001707077,
      "learning_rate": 0.0009356410292003562,
      "loss": 0.4983,
      "num_input_tokens_seen": 6154192,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.12066932022571564,
      "learning_rate": 0.0009355287486736239,
      "loss": 0.5044,
      "num_input_tokens_seen": 6156880,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 0.08687509596347809,
      "learning_rate": 0.0009354163770392461,
      "loss": 0.4157,
      "num_input_tokens_seen": 6160240,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.0721890777349472,
      "learning_rate": 0.0009353039143207295,
      "loss": 0.4332,
      "num_input_tokens_seen": 6164048,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 0.12261973321437836,
      "learning_rate": 0.0009351913605416007,
      "loss": 0.4615,
      "num_input_tokens_seen": 6166576,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 0.060763027518987656,
      "learning_rate": 0.000935078715725404,
      "loss": 0.484,
      "num_input_tokens_seen": 6169808,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 0.14470039308071136,
      "learning_rate": 0.0009349659798957034,
      "loss": 0.45,
      "num_input_tokens_seen": 6172176,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 0.07221272587776184,
      "learning_rate": 0.0009348531530760823,
      "loss": 0.4605,
      "num_input_tokens_seen": 6175120,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 0.1339171826839447,
      "learning_rate": 0.0009347402352901426,
      "loss": 0.3572,
      "num_input_tokens_seen": 6178832,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 0.1263502538204193,
      "learning_rate": 0.0009346272265615057,
      "loss": 0.4784,
      "num_input_tokens_seen": 6182224,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.13164521753787994,
      "learning_rate": 0.0009345141269138112,
      "loss": 0.4827,
      "num_input_tokens_seen": 6185392,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.18879276514053345,
      "learning_rate": 0.0009344009363707186,
      "loss": 0.5156,
      "num_input_tokens_seen": 6188848,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 0.12320993095636368,
      "learning_rate": 0.000934287654955906,
      "loss": 0.4994,
      "num_input_tokens_seen": 6191760,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 0.16802580654621124,
      "learning_rate": 0.0009341742826930708,
      "loss": 0.6476,
      "num_input_tokens_seen": 6194864,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 0.11917976289987564,
      "learning_rate": 0.0009340608196059289,
      "loss": 0.4756,
      "num_input_tokens_seen": 6197648,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.1478823721408844,
      "learning_rate": 0.0009339472657182155,
      "loss": 0.4437,
      "num_input_tokens_seen": 6200816,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.1454077512025833,
      "learning_rate": 0.0009338336210536848,
      "loss": 0.4466,
      "num_input_tokens_seen": 6203344,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.07423311471939087,
      "learning_rate": 0.0009337198856361102,
      "loss": 0.3441,
      "num_input_tokens_seen": 6206768,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 0.11782034486532211,
      "learning_rate": 0.0009336060594892834,
      "loss": 0.4892,
      "num_input_tokens_seen": 6209168,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 0.0839124470949173,
      "learning_rate": 0.000933492142637016,
      "loss": 0.5183,
      "num_input_tokens_seen": 6214320,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.11690998077392578,
      "learning_rate": 0.0009333781351031379,
      "loss": 0.4572,
      "num_input_tokens_seen": 6219472,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.10882049053907394,
      "learning_rate": 0.0009332640369114981,
      "loss": 0.4907,
      "num_input_tokens_seen": 6222800,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 0.10355263948440552,
      "learning_rate": 0.0009331498480859647,
      "loss": 0.5792,
      "num_input_tokens_seen": 6226416,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2704574763774872,
      "learning_rate": 0.0009330355686504247,
      "loss": 0.4158,
      "num_input_tokens_seen": 6229240,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.48624399304389954,
      "eval_runtime": 13.7136,
      "eval_samples_per_second": 61.836,
      "eval_steps_per_second": 15.459,
      "num_input_tokens_seen": 6229240,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 0.06732863187789917,
      "learning_rate": 0.0009329211986287842,
      "loss": 0.4261,
      "num_input_tokens_seen": 6233752,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 0.15119659900665283,
      "learning_rate": 0.0009328067380449678,
      "loss": 0.4472,
      "num_input_tokens_seen": 6236568,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 0.0828176736831665,
      "learning_rate": 0.0009326921869229197,
      "loss": 0.4915,
      "num_input_tokens_seen": 6239096,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.1470690220594406,
      "learning_rate": 0.0009325775452866023,
      "loss": 0.4063,
      "num_input_tokens_seen": 6241688,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.1089417040348053,
      "learning_rate": 0.0009324628131599977,
      "loss": 0.4026,
      "num_input_tokens_seen": 6244888,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.08379697799682617,
      "learning_rate": 0.0009323479905671064,
      "loss": 0.3987,
      "num_input_tokens_seen": 6247352,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.0802951529622078,
      "learning_rate": 0.0009322330775319479,
      "loss": 0.5504,
      "num_input_tokens_seen": 6250136,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 0.12214938551187515,
      "learning_rate": 0.0009321180740785607,
      "loss": 0.5302,
      "num_input_tokens_seen": 6252920,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.1426679491996765,
      "learning_rate": 0.0009320029802310023,
      "loss": 0.3786,
      "num_input_tokens_seen": 6255416,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.06734108179807663,
      "learning_rate": 0.0009318877960133489,
      "loss": 0.4558,
      "num_input_tokens_seen": 6258584,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.07881653308868408,
      "learning_rate": 0.0009317725214496959,
      "loss": 0.5084,
      "num_input_tokens_seen": 6265528,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 0.09008628875017166,
      "learning_rate": 0.0009316571565641574,
      "loss": 0.4801,
      "num_input_tokens_seen": 6269016,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 0.10606305301189423,
      "learning_rate": 0.0009315417013808663,
      "loss": 0.4384,
      "num_input_tokens_seen": 6272216,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 0.05260324850678444,
      "learning_rate": 0.0009314261559239745,
      "loss": 0.4578,
      "num_input_tokens_seen": 6275960,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 0.08493319153785706,
      "learning_rate": 0.000931310520217653,
      "loss": 0.4425,
      "num_input_tokens_seen": 6278744,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 0.13593606650829315,
      "learning_rate": 0.0009311947942860912,
      "loss": 0.5357,
      "num_input_tokens_seen": 6281880,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 0.10671301186084747,
      "learning_rate": 0.0009310789781534979,
      "loss": 0.5327,
      "num_input_tokens_seen": 6285336,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 0.10498816519975662,
      "learning_rate": 0.0009309630718441003,
      "loss": 0.5292,
      "num_input_tokens_seen": 6289176,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 0.11669747531414032,
      "learning_rate": 0.0009308470753821445,
      "loss": 0.4859,
      "num_input_tokens_seen": 6292824,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 0.10722268372774124,
      "learning_rate": 0.0009307309887918962,
      "loss": 0.3726,
      "num_input_tokens_seen": 6296760,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 0.050820332020521164,
      "learning_rate": 0.000930614812097639,
      "loss": 0.5158,
      "num_input_tokens_seen": 6300280,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 0.1293872743844986,
      "learning_rate": 0.0009304985453236758,
      "loss": 0.4459,
      "num_input_tokens_seen": 6304056,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 0.07122402638196945,
      "learning_rate": 0.0009303821884943284,
      "loss": 0.3201,
      "num_input_tokens_seen": 6307000,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.08267150074243546,
      "learning_rate": 0.0009302657416339371,
      "loss": 0.4337,
      "num_input_tokens_seen": 6311032,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 0.11965586990118027,
      "learning_rate": 0.0009301492047668615,
      "loss": 0.5017,
      "num_input_tokens_seen": 6314680,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 0.11258068680763245,
      "learning_rate": 0.0009300325779174796,
      "loss": 0.3925,
      "num_input_tokens_seen": 6317560,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 0.11886180192232132,
      "learning_rate": 0.0009299158611101885,
      "loss": 0.5882,
      "num_input_tokens_seen": 6321336,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 0.09489763528108597,
      "learning_rate": 0.000929799054369404,
      "loss": 0.6108,
      "num_input_tokens_seen": 6324376,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.11839800328016281,
      "learning_rate": 0.0009296821577195606,
      "loss": 0.4614,
      "num_input_tokens_seen": 6327256,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 0.0849372074007988,
      "learning_rate": 0.000929565171185112,
      "loss": 0.3533,
      "num_input_tokens_seen": 6330680,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 0.26634180545806885,
      "learning_rate": 0.0009294480947905304,
      "loss": 0.4809,
      "num_input_tokens_seen": 6334872,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.11414965242147446,
      "learning_rate": 0.0009293309285603067,
      "loss": 0.4432,
      "num_input_tokens_seen": 6337752,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 0.22319675981998444,
      "learning_rate": 0.0009292136725189506,
      "loss": 0.4081,
      "num_input_tokens_seen": 6340376,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 0.1016557365655899,
      "learning_rate": 0.0009290963266909912,
      "loss": 0.6092,
      "num_input_tokens_seen": 6343768,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 0.0855480208992958,
      "learning_rate": 0.0009289788911009756,
      "loss": 0.4441,
      "num_input_tokens_seen": 6346968,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.06769488006830215,
      "learning_rate": 0.0009288613657734699,
      "loss": 0.4687,
      "num_input_tokens_seen": 6352280,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.12471736967563629,
      "learning_rate": 0.0009287437507330594,
      "loss": 0.4435,
      "num_input_tokens_seen": 6355288,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 0.09602309763431549,
      "learning_rate": 0.0009286260460043475,
      "loss": 0.5776,
      "num_input_tokens_seen": 6358040,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 0.07968825101852417,
      "learning_rate": 0.0009285082516119567,
      "loss": 0.3595,
      "num_input_tokens_seen": 6360984,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 0.09892599284648895,
      "learning_rate": 0.0009283903675805285,
      "loss": 0.4192,
      "num_input_tokens_seen": 6363736,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 0.08448895066976547,
      "learning_rate": 0.0009282723939347227,
      "loss": 0.4745,
      "num_input_tokens_seen": 6365944,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 0.10742855817079544,
      "learning_rate": 0.0009281543306992181,
      "loss": 0.3449,
      "num_input_tokens_seen": 6370488,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 0.11907149851322174,
      "learning_rate": 0.0009280361778987121,
      "loss": 0.4838,
      "num_input_tokens_seen": 6374008,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 0.08528224378824234,
      "learning_rate": 0.0009279179355579211,
      "loss": 0.5512,
      "num_input_tokens_seen": 6377688,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 0.05633434280753136,
      "learning_rate": 0.0009277996037015798,
      "loss": 0.403,
      "num_input_tokens_seen": 6381752,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 0.12166518718004227,
      "learning_rate": 0.000927681182354442,
      "loss": 0.4527,
      "num_input_tokens_seen": 6384600,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 0.07476352155208588,
      "learning_rate": 0.0009275626715412802,
      "loss": 0.6011,
      "num_input_tokens_seen": 6388888,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 0.0888824388384819,
      "learning_rate": 0.0009274440712868853,
      "loss": 0.4759,
      "num_input_tokens_seen": 6392248,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.07946870476007462,
      "learning_rate": 0.0009273253816160673,
      "loss": 0.4852,
      "num_input_tokens_seen": 6395160,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 0.10146526247262955,
      "learning_rate": 0.0009272066025536545,
      "loss": 0.5621,
      "num_input_tokens_seen": 6397912,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 0.13522911071777344,
      "learning_rate": 0.0009270877341244945,
      "loss": 0.5703,
      "num_input_tokens_seen": 6400280,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 0.10818690806627274,
      "learning_rate": 0.0009269687763534529,
      "loss": 0.463,
      "num_input_tokens_seen": 6404440,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 0.08078264445066452,
      "learning_rate": 0.0009268497292654143,
      "loss": 0.7031,
      "num_input_tokens_seen": 6406872,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 0.09443633258342743,
      "learning_rate": 0.0009267305928852823,
      "loss": 0.4429,
      "num_input_tokens_seen": 6411288,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 0.10097821801900864,
      "learning_rate": 0.0009266113672379786,
      "loss": 0.4672,
      "num_input_tokens_seen": 6414360,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 0.1298568844795227,
      "learning_rate": 0.0009264920523484437,
      "loss": 0.4818,
      "num_input_tokens_seen": 6417624,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 0.11802022904157639,
      "learning_rate": 0.0009263726482416374,
      "loss": 0.548,
      "num_input_tokens_seen": 6421080,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 0.09001997113227844,
      "learning_rate": 0.0009262531549425372,
      "loss": 0.4735,
      "num_input_tokens_seen": 6425752,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.10447648167610168,
      "learning_rate": 0.0009261335724761402,
      "loss": 0.4318,
      "num_input_tokens_seen": 6429016,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 0.09717103838920593,
      "learning_rate": 0.0009260139008674612,
      "loss": 0.5319,
      "num_input_tokens_seen": 6432056,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 0.06727085262537003,
      "learning_rate": 0.0009258941401415344,
      "loss": 0.4592,
      "num_input_tokens_seen": 6435448,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 0.11120820045471191,
      "learning_rate": 0.0009257742903234123,
      "loss": 0.3822,
      "num_input_tokens_seen": 6438520,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 0.1059800535440445,
      "learning_rate": 0.0009256543514381664,
      "loss": 0.4497,
      "num_input_tokens_seen": 6440984,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.17931801080703735,
      "learning_rate": 0.0009255343235108859,
      "loss": 0.5196,
      "num_input_tokens_seen": 6443992,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 0.14405842125415802,
      "learning_rate": 0.00092541420656668,
      "loss": 0.4055,
      "num_input_tokens_seen": 6446552,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 0.13487114012241364,
      "learning_rate": 0.0009252940006306753,
      "loss": 0.4551,
      "num_input_tokens_seen": 6449080,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 0.07035711407661438,
      "learning_rate": 0.0009251737057280179,
      "loss": 0.4532,
      "num_input_tokens_seen": 6452856,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 0.10513897240161896,
      "learning_rate": 0.0009250533218838717,
      "loss": 0.4998,
      "num_input_tokens_seen": 6456184,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 0.04795841872692108,
      "learning_rate": 0.0009249328491234199,
      "loss": 0.3471,
      "num_input_tokens_seen": 6459512,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 0.09078606963157654,
      "learning_rate": 0.0009248122874718638,
      "loss": 0.5506,
      "num_input_tokens_seen": 6462360,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 0.11631342023611069,
      "learning_rate": 0.0009246916369544238,
      "loss": 0.4987,
      "num_input_tokens_seen": 6465144,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.1090327799320221,
      "learning_rate": 0.0009245708975963386,
      "loss": 0.577,
      "num_input_tokens_seen": 6467864,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.14132563769817352,
      "learning_rate": 0.0009244500694228653,
      "loss": 0.3328,
      "num_input_tokens_seen": 6471512,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 0.07426527142524719,
      "learning_rate": 0.0009243291524592799,
      "loss": 0.3467,
      "num_input_tokens_seen": 6474808,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 0.11507809907197952,
      "learning_rate": 0.0009242081467308766,
      "loss": 0.4129,
      "num_input_tokens_seen": 6477208,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 0.08951404690742493,
      "learning_rate": 0.0009240870522629688,
      "loss": 0.6253,
      "num_input_tokens_seen": 6481144,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 0.07095246762037277,
      "learning_rate": 0.0009239658690808879,
      "loss": 0.415,
      "num_input_tokens_seen": 6486200,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 0.1519165337085724,
      "learning_rate": 0.000923844597209984,
      "loss": 0.6064,
      "num_input_tokens_seen": 6489656,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 0.13230009377002716,
      "learning_rate": 0.0009237232366756258,
      "loss": 0.4977,
      "num_input_tokens_seen": 6492440,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.08161315321922302,
      "learning_rate": 0.0009236017875032007,
      "loss": 0.3635,
      "num_input_tokens_seen": 6495096,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 0.09048278629779816,
      "learning_rate": 0.0009234802497181143,
      "loss": 0.4362,
      "num_input_tokens_seen": 6499384,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 0.10535068809986115,
      "learning_rate": 0.0009233586233457909,
      "loss": 0.5211,
      "num_input_tokens_seen": 6502584,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.15870241820812225,
      "learning_rate": 0.0009232369084116736,
      "loss": 0.4431,
      "num_input_tokens_seen": 6505944,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 0.06919056922197342,
      "learning_rate": 0.0009231151049412234,
      "loss": 0.4969,
      "num_input_tokens_seen": 6508696,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 0.1688111126422882,
      "learning_rate": 0.0009229932129599205,
      "loss": 0.3665,
      "num_input_tokens_seen": 6511768,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.10531267523765564,
      "learning_rate": 0.0009228712324932634,
      "loss": 0.4543,
      "num_input_tokens_seen": 6515480,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 0.045882098376750946,
      "learning_rate": 0.0009227491635667685,
      "loss": 0.3893,
      "num_input_tokens_seen": 6519448,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 0.16476038098335266,
      "learning_rate": 0.0009226270062059717,
      "loss": 0.4689,
      "num_input_tokens_seen": 6522584,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 0.0629366859793663,
      "learning_rate": 0.0009225047604364267,
      "loss": 0.3149,
      "num_input_tokens_seen": 6527576,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 0.08485660701990128,
      "learning_rate": 0.0009223824262837062,
      "loss": 0.4459,
      "num_input_tokens_seen": 6530520,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 0.09995540976524353,
      "learning_rate": 0.0009222600037734008,
      "loss": 0.3584,
      "num_input_tokens_seen": 6533560,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 0.11434417217969894,
      "learning_rate": 0.00092213749293112,
      "loss": 0.4315,
      "num_input_tokens_seen": 6535992,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 0.117324598133564,
      "learning_rate": 0.0009220148937824917,
      "loss": 0.4457,
      "num_input_tokens_seen": 6539416,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 0.07872005552053452,
      "learning_rate": 0.0009218922063531623,
      "loss": 0.4206,
      "num_input_tokens_seen": 6542968,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 0.072165347635746,
      "learning_rate": 0.0009217694306687963,
      "loss": 0.4583,
      "num_input_tokens_seen": 6546488,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 0.10350536555051804,
      "learning_rate": 0.0009216465667550774,
      "loss": 0.4309,
      "num_input_tokens_seen": 6549400,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 0.19803594052791595,
      "learning_rate": 0.0009215236146377071,
      "loss": 0.3653,
      "num_input_tokens_seen": 6551736,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 0.07327276468276978,
      "learning_rate": 0.0009214005743424056,
      "loss": 0.3758,
      "num_input_tokens_seen": 6555256,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 0.26033279299736023,
      "learning_rate": 0.0009212774458949116,
      "loss": 0.6313,
      "num_input_tokens_seen": 6557624,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 0.09929520636796951,
      "learning_rate": 0.000921154229320982,
      "loss": 0.4393,
      "num_input_tokens_seen": 6562008,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 0.1453232616186142,
      "learning_rate": 0.0009210309246463924,
      "loss": 0.4068,
      "num_input_tokens_seen": 6564376,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 0.1434216946363449,
      "learning_rate": 0.0009209075318969369,
      "loss": 0.5323,
      "num_input_tokens_seen": 6567192,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 0.19245532155036926,
      "learning_rate": 0.0009207840510984276,
      "loss": 0.5932,
      "num_input_tokens_seen": 6570712,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.11380471289157867,
      "learning_rate": 0.0009206604822766953,
      "loss": 0.4839,
      "num_input_tokens_seen": 6574424,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 0.10530159622430801,
      "learning_rate": 0.0009205368254575892,
      "loss": 0.5865,
      "num_input_tokens_seen": 6576920,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.12488938122987747,
      "learning_rate": 0.000920413080666977,
      "loss": 0.4653,
      "num_input_tokens_seen": 6580408,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 0.07252698391675949,
      "learning_rate": 0.0009202892479307448,
      "loss": 0.4813,
      "num_input_tokens_seen": 6583928,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.15294207632541656,
      "learning_rate": 0.0009201653272747967,
      "loss": 0.4509,
      "num_input_tokens_seen": 6587064,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 0.06764619797468185,
      "learning_rate": 0.0009200413187250558,
      "loss": 0.4531,
      "num_input_tokens_seen": 6591096,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 0.13181300461292267,
      "learning_rate": 0.000919917222307463,
      "loss": 0.4938,
      "num_input_tokens_seen": 6594168,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 0.14285261929035187,
      "learning_rate": 0.000919793038047978,
      "loss": 0.4878,
      "num_input_tokens_seen": 6597432,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 0.16393645107746124,
      "learning_rate": 0.0009196687659725787,
      "loss": 0.4187,
      "num_input_tokens_seen": 6601464,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 0.07914752513170242,
      "learning_rate": 0.0009195444061072612,
      "loss": 0.3815,
      "num_input_tokens_seen": 6604376,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 0.0897737443447113,
      "learning_rate": 0.0009194199584780405,
      "loss": 0.4704,
      "num_input_tokens_seen": 6608184,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 0.1193312332034111,
      "learning_rate": 0.0009192954231109496,
      "loss": 0.4824,
      "num_input_tokens_seen": 6612152,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 0.1735643446445465,
      "learning_rate": 0.0009191708000320396,
      "loss": 0.5781,
      "num_input_tokens_seen": 6615448,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.09630147367715836,
      "learning_rate": 0.0009190460892673805,
      "loss": 0.4489,
      "num_input_tokens_seen": 6619000,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 0.04133350029587746,
      "learning_rate": 0.0009189212908430601,
      "loss": 0.3415,
      "num_input_tokens_seen": 6623416,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 0.1704641878604889,
      "learning_rate": 0.0009187964047851851,
      "loss": 0.4512,
      "num_input_tokens_seen": 6626968,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 0.08165574818849564,
      "learning_rate": 0.0009186714311198801,
      "loss": 0.4003,
      "num_input_tokens_seen": 6629912,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 0.05999147519469261,
      "learning_rate": 0.000918546369873288,
      "loss": 0.4407,
      "num_input_tokens_seen": 6633560,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 0.09288433194160461,
      "learning_rate": 0.0009184212210715704,
      "loss": 0.4284,
      "num_input_tokens_seen": 6636760,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 0.10125935822725296,
      "learning_rate": 0.0009182959847409072,
      "loss": 0.3926,
      "num_input_tokens_seen": 6639992,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 0.1068793535232544,
      "learning_rate": 0.0009181706609074959,
      "loss": 0.4852,
      "num_input_tokens_seen": 6643096,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 0.1151815727353096,
      "learning_rate": 0.0009180452495975531,
      "loss": 0.4444,
      "num_input_tokens_seen": 6646616,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 0.19535322487354279,
      "learning_rate": 0.0009179197508373134,
      "loss": 0.5052,
      "num_input_tokens_seen": 6649848,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 0.15114283561706543,
      "learning_rate": 0.0009177941646530299,
      "loss": 0.4064,
      "num_input_tokens_seen": 6652696,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 0.10436709970235825,
      "learning_rate": 0.0009176684910709733,
      "loss": 0.4521,
      "num_input_tokens_seen": 6655640,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 0.11216067522764206,
      "learning_rate": 0.0009175427301174338,
      "loss": 0.396,
      "num_input_tokens_seen": 6658776,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 0.09128962457180023,
      "learning_rate": 0.0009174168818187183,
      "loss": 0.5795,
      "num_input_tokens_seen": 6661464,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 0.07859195023775101,
      "learning_rate": 0.0009172909462011536,
      "loss": 0.4151,
      "num_input_tokens_seen": 6664280,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 0.1058889627456665,
      "learning_rate": 0.0009171649232910835,
      "loss": 0.5149,
      "num_input_tokens_seen": 6667672,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 0.1372610479593277,
      "learning_rate": 0.0009170388131148707,
      "loss": 0.5786,
      "num_input_tokens_seen": 6670648,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.10170814394950867,
      "learning_rate": 0.000916912615698896,
      "loss": 0.5549,
      "num_input_tokens_seen": 6673368,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 0.06506555527448654,
      "learning_rate": 0.0009167863310695585,
      "loss": 0.4429,
      "num_input_tokens_seen": 6677016,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 0.0852765366435051,
      "learning_rate": 0.0009166599592532756,
      "loss": 0.5635,
      "num_input_tokens_seen": 6680152,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 0.10822812467813492,
      "learning_rate": 0.0009165335002764828,
      "loss": 0.4059,
      "num_input_tokens_seen": 6686040,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 0.056766193360090256,
      "learning_rate": 0.0009164069541656337,
      "loss": 0.5144,
      "num_input_tokens_seen": 6689400,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 0.10055996477603912,
      "learning_rate": 0.0009162803209472004,
      "loss": 0.4517,
      "num_input_tokens_seen": 6692376,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 0.09138453751802444,
      "learning_rate": 0.000916153600647673,
      "loss": 0.4127,
      "num_input_tokens_seen": 6695480,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 0.07481683045625687,
      "learning_rate": 0.0009160267932935602,
      "loss": 0.4711,
      "num_input_tokens_seen": 6698680,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 0.08010625839233398,
      "learning_rate": 0.0009158998989113885,
      "loss": 0.4617,
      "num_input_tokens_seen": 6701784,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.06891977787017822,
      "learning_rate": 0.0009157729175277028,
      "loss": 0.5377,
      "num_input_tokens_seen": 6705976,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 0.08946734666824341,
      "learning_rate": 0.0009156458491690662,
      "loss": 0.3522,
      "num_input_tokens_seen": 6708920,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 0.1371614634990692,
      "learning_rate": 0.0009155186938620599,
      "loss": 0.4509,
      "num_input_tokens_seen": 6711864,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 0.13593429327011108,
      "learning_rate": 0.0009153914516332833,
      "loss": 0.4908,
      "num_input_tokens_seen": 6714904,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 0.08115419000387192,
      "learning_rate": 0.000915264122509354,
      "loss": 0.4165,
      "num_input_tokens_seen": 6717752,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 0.09076663106679916,
      "learning_rate": 0.0009151367065169078,
      "loss": 0.4897,
      "num_input_tokens_seen": 6720568,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 0.11379750072956085,
      "learning_rate": 0.0009150092036825989,
      "loss": 0.4085,
      "num_input_tokens_seen": 6723480,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 0.1258917599916458,
      "learning_rate": 0.0009148816140330991,
      "loss": 0.3847,
      "num_input_tokens_seen": 6729400,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 0.10251623392105103,
      "learning_rate": 0.000914753937595099,
      "loss": 0.4325,
      "num_input_tokens_seen": 6731960,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 0.11152804642915726,
      "learning_rate": 0.0009146261743953068,
      "loss": 0.4409,
      "num_input_tokens_seen": 6734648,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 0.09468875825405121,
      "learning_rate": 0.0009144983244604493,
      "loss": 0.4219,
      "num_input_tokens_seen": 6737368,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 0.08392684906721115,
      "learning_rate": 0.0009143703878172712,
      "loss": 0.6237,
      "num_input_tokens_seen": 6740952,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 0.0970705971121788,
      "learning_rate": 0.0009142423644925353,
      "loss": 0.5113,
      "num_input_tokens_seen": 6743640,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 0.08819543570280075,
      "learning_rate": 0.0009141142545130228,
      "loss": 0.4414,
      "num_input_tokens_seen": 6746264,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.07343883067369461,
      "learning_rate": 0.0009139860579055326,
      "loss": 0.4236,
      "num_input_tokens_seen": 6749592,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 0.1546536087989807,
      "learning_rate": 0.0009138577746968821,
      "loss": 0.5489,
      "num_input_tokens_seen": 6752280,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.13539378345012665,
      "learning_rate": 0.0009137294049139066,
      "loss": 0.3772,
      "num_input_tokens_seen": 6755960,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 0.09717545658349991,
      "learning_rate": 0.0009136009485834598,
      "loss": 0.4885,
      "num_input_tokens_seen": 6760312,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 0.11390107870101929,
      "learning_rate": 0.0009134724057324131,
      "loss": 0.4416,
      "num_input_tokens_seen": 6762616,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.11000930517911911,
      "learning_rate": 0.0009133437763876562,
      "loss": 0.4981,
      "num_input_tokens_seen": 6766040,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 0.14826416969299316,
      "learning_rate": 0.000913215060576097,
      "loss": 0.4289,
      "num_input_tokens_seen": 6768568,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 0.23617565631866455,
      "learning_rate": 0.0009130862583246613,
      "loss": 0.476,
      "num_input_tokens_seen": 6771288,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 0.09631948918104172,
      "learning_rate": 0.0009129573696602932,
      "loss": 0.5702,
      "num_input_tokens_seen": 6774104,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 0.11940010637044907,
      "learning_rate": 0.0009128283946099546,
      "loss": 0.4235,
      "num_input_tokens_seen": 6777240,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.08633536100387573,
      "learning_rate": 0.0009126993332006256,
      "loss": 0.364,
      "num_input_tokens_seen": 6780856,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 0.12036531418561935,
      "learning_rate": 0.0009125701854593045,
      "loss": 0.6181,
      "num_input_tokens_seen": 6783736,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 0.07584112882614136,
      "learning_rate": 0.0009124409514130073,
      "loss": 0.4893,
      "num_input_tokens_seen": 6788472,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 0.1303047239780426,
      "learning_rate": 0.0009123116310887686,
      "loss": 0.6174,
      "num_input_tokens_seen": 6791032,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 0.08860243856906891,
      "learning_rate": 0.0009121822245136404,
      "loss": 0.5302,
      "num_input_tokens_seen": 6794680,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 0.08010479807853699,
      "learning_rate": 0.0009120527317146934,
      "loss": 0.4262,
      "num_input_tokens_seen": 6798072,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 0.10055126249790192,
      "learning_rate": 0.0009119231527190158,
      "loss": 0.7979,
      "num_input_tokens_seen": 6802296,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.09755154699087143,
      "learning_rate": 0.0009117934875537141,
      "loss": 0.5096,
      "num_input_tokens_seen": 6807000,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 0.29010042548179626,
      "learning_rate": 0.0009116637362459129,
      "loss": 0.5165,
      "num_input_tokens_seen": 6809944,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 0.09870494902133942,
      "learning_rate": 0.0009115338988227546,
      "loss": 0.399,
      "num_input_tokens_seen": 6812984,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 0.10432706773281097,
      "learning_rate": 0.0009114039753113997,
      "loss": 0.4316,
      "num_input_tokens_seen": 6816728,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 0.09884367883205414,
      "learning_rate": 0.0009112739657390265,
      "loss": 0.6604,
      "num_input_tokens_seen": 6819352,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 0.13746744394302368,
      "learning_rate": 0.0009111438701328319,
      "loss": 0.6169,
      "num_input_tokens_seen": 6822776,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.10110170394182205,
      "learning_rate": 0.0009110136885200302,
      "loss": 0.4206,
      "num_input_tokens_seen": 6826040,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 0.1365027278661728,
      "learning_rate": 0.000910883420927854,
      "loss": 0.4876,
      "num_input_tokens_seen": 6829752,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 0.09477417916059494,
      "learning_rate": 0.0009107530673835536,
      "loss": 0.4001,
      "num_input_tokens_seen": 6832728,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 0.08953817188739777,
      "learning_rate": 0.0009106226279143979,
      "loss": 0.4229,
      "num_input_tokens_seen": 6835096,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 0.11018422245979309,
      "learning_rate": 0.0009104921025476728,
      "loss": 0.3768,
      "num_input_tokens_seen": 6837592,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 0.16449664533138275,
      "learning_rate": 0.0009103614913106832,
      "loss": 0.5051,
      "num_input_tokens_seen": 6840664,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 0.15594078600406647,
      "learning_rate": 0.0009102307942307511,
      "loss": 0.5292,
      "num_input_tokens_seen": 6843992,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 0.2662055492401123,
      "learning_rate": 0.000910100011335217,
      "loss": 0.4982,
      "num_input_tokens_seen": 6846296,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 0.1664292961359024,
      "learning_rate": 0.0009099691426514392,
      "loss": 0.5054,
      "num_input_tokens_seen": 6849784,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 0.07705123722553253,
      "learning_rate": 0.0009098381882067941,
      "loss": 0.6087,
      "num_input_tokens_seen": 6854296,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.1325729340314865,
      "learning_rate": 0.0009097071480286756,
      "loss": 0.5438,
      "num_input_tokens_seen": 6857528,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 0.14085586369037628,
      "learning_rate": 0.0009095760221444959,
      "loss": 0.4728,
      "num_input_tokens_seen": 6861112,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 0.10447070002555847,
      "learning_rate": 0.000909444810581685,
      "loss": 0.406,
      "num_input_tokens_seen": 6865144,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 0.06934870779514313,
      "learning_rate": 0.000909313513367691,
      "loss": 0.5627,
      "num_input_tokens_seen": 6869720,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 0.15199586749076843,
      "learning_rate": 0.0009091821305299798,
      "loss": 0.4775,
      "num_input_tokens_seen": 6872696,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 0.09225229918956757,
      "learning_rate": 0.000909050662096035,
      "loss": 0.4248,
      "num_input_tokens_seen": 6875384,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 0.08929488062858582,
      "learning_rate": 0.0009089191080933583,
      "loss": 0.3664,
      "num_input_tokens_seen": 6879352,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 0.06872999668121338,
      "learning_rate": 0.0009087874685494695,
      "loss": 0.4417,
      "num_input_tokens_seen": 6883128,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 0.0828493982553482,
      "learning_rate": 0.0009086557434919059,
      "loss": 0.4209,
      "num_input_tokens_seen": 6886200,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.0856042206287384,
      "learning_rate": 0.0009085239329482231,
      "loss": 0.3855,
      "num_input_tokens_seen": 6889336,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 0.10531501471996307,
      "learning_rate": 0.0009083920369459941,
      "loss": 0.5176,
      "num_input_tokens_seen": 6893208,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 0.09625986218452454,
      "learning_rate": 0.00090826005551281,
      "loss": 0.4688,
      "num_input_tokens_seen": 6899896,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 0.12107489258050919,
      "learning_rate": 0.0009081279886762802,
      "loss": 0.4948,
      "num_input_tokens_seen": 6903416,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 0.07335690408945084,
      "learning_rate": 0.0009079958364640313,
      "loss": 0.4657,
      "num_input_tokens_seen": 6906776,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 0.10067755728960037,
      "learning_rate": 0.0009078635989037081,
      "loss": 0.4397,
      "num_input_tokens_seen": 6909784,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 0.11394631117582321,
      "learning_rate": 0.000907731276022973,
      "loss": 0.5912,
      "num_input_tokens_seen": 6912856,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 0.3284432888031006,
      "learning_rate": 0.0009075988678495066,
      "loss": 0.7328,
      "num_input_tokens_seen": 6915896,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 0.08276743441820145,
      "learning_rate": 0.0009074663744110073,
      "loss": 0.444,
      "num_input_tokens_seen": 6918744,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 0.10523199290037155,
      "learning_rate": 0.0009073337957351911,
      "loss": 0.3911,
      "num_input_tokens_seen": 6921528,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.08631580322980881,
      "learning_rate": 0.0009072011318497919,
      "loss": 0.5298,
      "num_input_tokens_seen": 6924504,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 0.07614728808403015,
      "learning_rate": 0.0009070683827825615,
      "loss": 0.4389,
      "num_input_tokens_seen": 6927832,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 0.0938650518655777,
      "learning_rate": 0.0009069355485612695,
      "loss": 0.5439,
      "num_input_tokens_seen": 6930520,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.08085216581821442,
      "learning_rate": 0.0009068026292137034,
      "loss": 0.4478,
      "num_input_tokens_seen": 6935736,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 0.08631741255521774,
      "learning_rate": 0.0009066696247676682,
      "loss": 0.4281,
      "num_input_tokens_seen": 6938744,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 0.12075670808553696,
      "learning_rate": 0.0009065365352509871,
      "loss": 0.4111,
      "num_input_tokens_seen": 6941848,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 0.28741082549095154,
      "learning_rate": 0.0009064033606915008,
      "loss": 0.4999,
      "num_input_tokens_seen": 6944504,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 0.07448124885559082,
      "learning_rate": 0.000906270101117068,
      "loss": 0.5342,
      "num_input_tokens_seen": 6948632,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 0.10294031351804733,
      "learning_rate": 0.0009061367565555649,
      "loss": 0.4253,
      "num_input_tokens_seen": 6951672,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 0.1311008781194687,
      "learning_rate": 0.000906003327034886,
      "loss": 0.5009,
      "num_input_tokens_seen": 6955544,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 0.14436182379722595,
      "learning_rate": 0.0009058698125829428,
      "loss": 0.4987,
      "num_input_tokens_seen": 6958712,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 0.11498524993658066,
      "learning_rate": 0.0009057362132276653,
      "loss": 0.4114,
      "num_input_tokens_seen": 6961304,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 0.10267475247383118,
      "learning_rate": 0.0009056025289970009,
      "loss": 0.5046,
      "num_input_tokens_seen": 6963704,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 0.09744368493556976,
      "learning_rate": 0.0009054687599189148,
      "loss": 0.5066,
      "num_input_tokens_seen": 6966488,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 0.13656213879585266,
      "learning_rate": 0.0009053349060213899,
      "loss": 0.7508,
      "num_input_tokens_seen": 6970360,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 0.15893732011318207,
      "learning_rate": 0.0009052009673324269,
      "loss": 0.5078,
      "num_input_tokens_seen": 6972952,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 0.10384492576122284,
      "learning_rate": 0.0009050669438800445,
      "loss": 0.3782,
      "num_input_tokens_seen": 6976888,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 0.09350047260522842,
      "learning_rate": 0.0009049328356922786,
      "loss": 0.576,
      "num_input_tokens_seen": 6979608,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 0.06839235872030258,
      "learning_rate": 0.0009047986427971831,
      "loss": 0.5047,
      "num_input_tokens_seen": 6983704,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.09283656626939774,
      "learning_rate": 0.0009046643652228297,
      "loss": 0.6341,
      "num_input_tokens_seen": 6986808,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.10927046090364456,
      "learning_rate": 0.0009045300029973079,
      "loss": 0.5021,
      "num_input_tokens_seen": 6989784,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 0.14143070578575134,
      "learning_rate": 0.0009043955561487245,
      "loss": 0.3977,
      "num_input_tokens_seen": 6992952,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.08575772494077682,
      "learning_rate": 0.0009042610247052044,
      "loss": 0.4012,
      "num_input_tokens_seen": 6996056,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 0.086884506046772,
      "learning_rate": 0.0009041264086948899,
      "loss": 0.4507,
      "num_input_tokens_seen": 6999960,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 0.11773667484521866,
      "learning_rate": 0.0009039917081459414,
      "loss": 0.4946,
      "num_input_tokens_seen": 7003608,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 0.07352627068758011,
      "learning_rate": 0.0009038569230865363,
      "loss": 0.4232,
      "num_input_tokens_seen": 7006680,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 0.09681133180856705,
      "learning_rate": 0.0009037220535448705,
      "loss": 0.4802,
      "num_input_tokens_seen": 7009592,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 0.11478819698095322,
      "learning_rate": 0.0009035870995491571,
      "loss": 0.4115,
      "num_input_tokens_seen": 7011960,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 0.11669623106718063,
      "learning_rate": 0.0009034520611276265,
      "loss": 0.4254,
      "num_input_tokens_seen": 7015384,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 0.07868927717208862,
      "learning_rate": 0.0009033169383085278,
      "loss": 0.5356,
      "num_input_tokens_seen": 7019928,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 0.06445638835430145,
      "learning_rate": 0.0009031817311201268,
      "loss": 0.4037,
      "num_input_tokens_seen": 7023384,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 0.08504360169172287,
      "learning_rate": 0.0009030464395907074,
      "loss": 0.4207,
      "num_input_tokens_seen": 7026296,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.07013989239931107,
      "learning_rate": 0.0009029110637485711,
      "loss": 0.3764,
      "num_input_tokens_seen": 7029304,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 0.13872985541820526,
      "learning_rate": 0.0009027756036220368,
      "loss": 0.3777,
      "num_input_tokens_seen": 7031960,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 0.11961892992258072,
      "learning_rate": 0.0009026400592394415,
      "loss": 0.4087,
      "num_input_tokens_seen": 7035000,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 0.1709447205066681,
      "learning_rate": 0.0009025044306291393,
      "loss": 0.46,
      "num_input_tokens_seen": 7038392,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 0.10095938295125961,
      "learning_rate": 0.0009023687178195022,
      "loss": 0.4294,
      "num_input_tokens_seen": 7041528,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 0.07598400115966797,
      "learning_rate": 0.0009022329208389196,
      "loss": 0.3869,
      "num_input_tokens_seen": 7045304,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.05881975591182709,
      "learning_rate": 0.000902097039715799,
      "loss": 0.4966,
      "num_input_tokens_seen": 7049880,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 0.07636416703462601,
      "learning_rate": 0.0009019610744785651,
      "loss": 0.3368,
      "num_input_tokens_seen": 7052312,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 0.05169909819960594,
      "learning_rate": 0.0009018250251556603,
      "loss": 0.3918,
      "num_input_tokens_seen": 7055448,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 0.06506919115781784,
      "learning_rate": 0.0009016888917755445,
      "loss": 0.4923,
      "num_input_tokens_seen": 7058520,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 0.09526734799146652,
      "learning_rate": 0.0009015526743666951,
      "loss": 0.3836,
      "num_input_tokens_seen": 7061208,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.09724501520395279,
      "learning_rate": 0.0009014163729576074,
      "loss": 0.3714,
      "num_input_tokens_seen": 7065080,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 0.08776222914457321,
      "learning_rate": 0.0009012799875767943,
      "loss": 0.4087,
      "num_input_tokens_seen": 7067864,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 0.14254871010780334,
      "learning_rate": 0.0009011435182527856,
      "loss": 0.4081,
      "num_input_tokens_seen": 7070200,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 0.10079045593738556,
      "learning_rate": 0.0009010069650141295,
      "loss": 0.4394,
      "num_input_tokens_seen": 7073304,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.10168685764074326,
      "learning_rate": 0.0009008703278893913,
      "loss": 0.3678,
      "num_input_tokens_seen": 7076824,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 0.08463882654905319,
      "learning_rate": 0.0009007336069071537,
      "loss": 0.6282,
      "num_input_tokens_seen": 7080504,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.1291484534740448,
      "learning_rate": 0.0009005968020960175,
      "loss": 0.4127,
      "num_input_tokens_seen": 7082840,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 0.08261959254741669,
      "learning_rate": 0.0009004599134846004,
      "loss": 0.514,
      "num_input_tokens_seen": 7086008,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 0.09540143609046936,
      "learning_rate": 0.0009003229411015382,
      "loss": 0.456,
      "num_input_tokens_seen": 7088760,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 0.14244593679904938,
      "learning_rate": 0.0009001858849754838,
      "loss": 0.5384,
      "num_input_tokens_seen": 7091672,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 0.0671730563044548,
      "learning_rate": 0.0009000487451351078,
      "loss": 0.4719,
      "num_input_tokens_seen": 7095608,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 0.10934364795684814,
      "learning_rate": 0.0008999115216090985,
      "loss": 0.4353,
      "num_input_tokens_seen": 7098456,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 0.10564924776554108,
      "learning_rate": 0.0008997742144261612,
      "loss": 0.4764,
      "num_input_tokens_seen": 7101304,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.07544930279254913,
      "learning_rate": 0.000899636823615019,
      "loss": 0.4442,
      "num_input_tokens_seen": 7105176,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 0.07799098640680313,
      "learning_rate": 0.0008994993492044127,
      "loss": 0.7117,
      "num_input_tokens_seen": 7109080,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 0.08230659365653992,
      "learning_rate": 0.0008993617912231003,
      "loss": 0.5641,
      "num_input_tokens_seen": 7112504,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.07929889857769012,
      "learning_rate": 0.0008992241496998572,
      "loss": 0.478,
      "num_input_tokens_seen": 7115768,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 0.09427642077207565,
      "learning_rate": 0.0008990864246634767,
      "loss": 0.5118,
      "num_input_tokens_seen": 7118712,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 0.09102937579154968,
      "learning_rate": 0.0008989486161427691,
      "loss": 0.3243,
      "num_input_tokens_seen": 7121624,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 0.1397014707326889,
      "learning_rate": 0.0008988107241665624,
      "loss": 0.663,
      "num_input_tokens_seen": 7124696,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 0.12630152702331543,
      "learning_rate": 0.0008986727487637022,
      "loss": 0.6733,
      "num_input_tokens_seen": 7128216,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 0.1673492044210434,
      "learning_rate": 0.0008985346899630513,
      "loss": 0.5192,
      "num_input_tokens_seen": 7131064,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 0.168153315782547,
      "learning_rate": 0.0008983965477934899,
      "loss": 0.5687,
      "num_input_tokens_seen": 7133784,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 0.10850607603788376,
      "learning_rate": 0.0008982583222839158,
      "loss": 0.4093,
      "num_input_tokens_seen": 7137496,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 0.06000450998544693,
      "learning_rate": 0.0008981200134632444,
      "loss": 0.5077,
      "num_input_tokens_seen": 7140792,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 0.09767128527164459,
      "learning_rate": 0.000897981621360408,
      "loss": 0.3682,
      "num_input_tokens_seen": 7143768,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.11465223133563995,
      "learning_rate": 0.0008978431460043569,
      "loss": 0.432,
      "num_input_tokens_seen": 7146904,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 0.10578522086143494,
      "learning_rate": 0.0008977045874240585,
      "loss": 0.4631,
      "num_input_tokens_seen": 7149816,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.09027665853500366,
      "learning_rate": 0.0008975659456484977,
      "loss": 0.4463,
      "num_input_tokens_seen": 7152760,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 0.07528757303953171,
      "learning_rate": 0.0008974272207066767,
      "loss": 0.5189,
      "num_input_tokens_seen": 7155576,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 0.11593978106975555,
      "learning_rate": 0.0008972884126276152,
      "loss": 0.5425,
      "num_input_tokens_seen": 7157688,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 0.07352189719676971,
      "learning_rate": 0.0008971495214403502,
      "loss": 0.4439,
      "num_input_tokens_seen": 7160696,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 0.17755481600761414,
      "learning_rate": 0.0008970105471739364,
      "loss": 0.3907,
      "num_input_tokens_seen": 7165720,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 0.07355675846338272,
      "learning_rate": 0.0008968714898574454,
      "loss": 0.4039,
      "num_input_tokens_seen": 7169496,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.11750837415456772,
      "learning_rate": 0.0008967323495199665,
      "loss": 0.5426,
      "num_input_tokens_seen": 7172984,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.2597258985042572,
      "learning_rate": 0.0008965931261906061,
      "loss": 0.3941,
      "num_input_tokens_seen": 7176664,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 0.07344093918800354,
      "learning_rate": 0.0008964538198984885,
      "loss": 0.426,
      "num_input_tokens_seen": 7179544,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 0.12553972005844116,
      "learning_rate": 0.0008963144306727547,
      "loss": 0.4529,
      "num_input_tokens_seen": 7182680,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 0.1387602686882019,
      "learning_rate": 0.0008961749585425634,
      "loss": 0.6472,
      "num_input_tokens_seen": 7185880,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 0.0995573103427887,
      "learning_rate": 0.0008960354035370905,
      "loss": 0.4404,
      "num_input_tokens_seen": 7188760,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 0.07474534958600998,
      "learning_rate": 0.0008958957656855294,
      "loss": 0.4819,
      "num_input_tokens_seen": 7192632,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 0.19805487990379333,
      "learning_rate": 0.0008957560450170907,
      "loss": 0.4353,
      "num_input_tokens_seen": 7195576,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 0.07453661412000656,
      "learning_rate": 0.0008956162415610025,
      "loss": 0.5292,
      "num_input_tokens_seen": 7199512,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 0.12060149013996124,
      "learning_rate": 0.00089547635534651,
      "loss": 0.5609,
      "num_input_tokens_seen": 7202168,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 0.14301329851150513,
      "learning_rate": 0.0008953363864028758,
      "loss": 0.4451,
      "num_input_tokens_seen": 7206200,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 0.10205152630805969,
      "learning_rate": 0.0008951963347593796,
      "loss": 0.5304,
      "num_input_tokens_seen": 7209848,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 0.05952974036335945,
      "learning_rate": 0.0008950562004453191,
      "loss": 0.4836,
      "num_input_tokens_seen": 7213848,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 0.09240876138210297,
      "learning_rate": 0.0008949159834900082,
      "loss": 0.2964,
      "num_input_tokens_seen": 7217048,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 0.08455701917409897,
      "learning_rate": 0.0008947756839227792,
      "loss": 0.4228,
      "num_input_tokens_seen": 7220344,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 0.07589063048362732,
      "learning_rate": 0.0008946353017729809,
      "loss": 0.4384,
      "num_input_tokens_seen": 7222936,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 0.16429142653942108,
      "learning_rate": 0.0008944948370699798,
      "loss": 0.4354,
      "num_input_tokens_seen": 7226040,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 0.214950829744339,
      "learning_rate": 0.0008943542898431593,
      "loss": 0.4812,
      "num_input_tokens_seen": 7229752,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 0.27906545996665955,
      "learning_rate": 0.0008942136601219205,
      "loss": 0.4029,
      "num_input_tokens_seen": 7232504,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 0.0819150060415268,
      "learning_rate": 0.0008940729479356813,
      "loss": 0.4121,
      "num_input_tokens_seen": 7235928,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 0.1122041642665863,
      "learning_rate": 0.0008939321533138774,
      "loss": 0.45,
      "num_input_tokens_seen": 7238616,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 0.13678692281246185,
      "learning_rate": 0.0008937912762859611,
      "loss": 0.5843,
      "num_input_tokens_seen": 7242200,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 0.10860881954431534,
      "learning_rate": 0.0008936503168814024,
      "loss": 0.422,
      "num_input_tokens_seen": 7245400,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 0.16965866088867188,
      "learning_rate": 0.0008935092751296885,
      "loss": 0.4229,
      "num_input_tokens_seen": 7247960,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 0.06923017650842667,
      "learning_rate": 0.0008933681510603235,
      "loss": 0.4178,
      "num_input_tokens_seen": 7251512,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 0.10530035942792892,
      "learning_rate": 0.0008932269447028292,
      "loss": 0.4262,
      "num_input_tokens_seen": 7253784,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 0.08991211652755737,
      "learning_rate": 0.0008930856560867442,
      "loss": 0.5427,
      "num_input_tokens_seen": 7256824,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 0.14967623353004456,
      "learning_rate": 0.0008929442852416245,
      "loss": 0.5621,
      "num_input_tokens_seen": 7259928,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 0.0887584537267685,
      "learning_rate": 0.0008928028321970433,
      "loss": 0.5909,
      "num_input_tokens_seen": 7262680,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 0.08523688465356827,
      "learning_rate": 0.0008926612969825909,
      "loss": 0.5574,
      "num_input_tokens_seen": 7265432,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.11695300787687302,
      "learning_rate": 0.000892519679627875,
      "loss": 0.3929,
      "num_input_tokens_seen": 7268120,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 0.15001019835472107,
      "learning_rate": 0.00089237798016252,
      "loss": 0.5152,
      "num_input_tokens_seen": 7271160,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.2547321915626526,
      "learning_rate": 0.0008922361986161682,
      "loss": 0.6891,
      "num_input_tokens_seen": 7275416,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 0.142013281583786,
      "learning_rate": 0.0008920943350184785,
      "loss": 0.4436,
      "num_input_tokens_seen": 7278360,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 0.16079741716384888,
      "learning_rate": 0.0008919523893991271,
      "loss": 0.5912,
      "num_input_tokens_seen": 7281752,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 0.08587338030338287,
      "learning_rate": 0.0008918103617878075,
      "loss": 0.5289,
      "num_input_tokens_seen": 7284472,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 0.07746488600969315,
      "learning_rate": 0.0008916682522142302,
      "loss": 0.4683,
      "num_input_tokens_seen": 7287448,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 0.12051030993461609,
      "learning_rate": 0.000891526060708123,
      "loss": 0.4345,
      "num_input_tokens_seen": 7290776,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.11903196573257446,
      "learning_rate": 0.0008913837872992306,
      "loss": 0.4834,
      "num_input_tokens_seen": 7293752,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 0.12204211950302124,
      "learning_rate": 0.000891241432017315,
      "loss": 0.6157,
      "num_input_tokens_seen": 7296856,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 0.12317273020744324,
      "learning_rate": 0.0008910989948921555,
      "loss": 0.5377,
      "num_input_tokens_seen": 7299992,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 0.1569110006093979,
      "learning_rate": 0.000890956475953548,
      "loss": 0.582,
      "num_input_tokens_seen": 7302808,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 0.13246987760066986,
      "learning_rate": 0.0008908138752313061,
      "loss": 0.6008,
      "num_input_tokens_seen": 7306040,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 0.08754006773233414,
      "learning_rate": 0.0008906711927552601,
      "loss": 0.4098,
      "num_input_tokens_seen": 7308760,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 0.10294555872678757,
      "learning_rate": 0.0008905284285552575,
      "loss": 0.4474,
      "num_input_tokens_seen": 7311128,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 0.06482075899839401,
      "learning_rate": 0.0008903855826611631,
      "loss": 0.51,
      "num_input_tokens_seen": 7314424,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 0.1114492416381836,
      "learning_rate": 0.0008902426551028586,
      "loss": 0.4654,
      "num_input_tokens_seen": 7317656,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 0.23101980984210968,
      "learning_rate": 0.0008900996459102428,
      "loss": 0.5562,
      "num_input_tokens_seen": 7321432,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 0.06867347657680511,
      "learning_rate": 0.0008899565551132315,
      "loss": 0.5058,
      "num_input_tokens_seen": 7325016,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 0.12421515583992004,
      "learning_rate": 0.0008898133827417578,
      "loss": 0.4113,
      "num_input_tokens_seen": 7327704,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 0.09423677623271942,
      "learning_rate": 0.0008896701288257716,
      "loss": 0.3972,
      "num_input_tokens_seen": 7331288,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 0.13073858618736267,
      "learning_rate": 0.00088952679339524,
      "loss": 0.4082,
      "num_input_tokens_seen": 7333784,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 0.06238872557878494,
      "learning_rate": 0.0008893833764801473,
      "loss": 0.5052,
      "num_input_tokens_seen": 7336440,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 0.07216870039701462,
      "learning_rate": 0.0008892398781104945,
      "loss": 0.5028,
      "num_input_tokens_seen": 7342360,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 0.16241487860679626,
      "learning_rate": 0.0008890962983162997,
      "loss": 0.4241,
      "num_input_tokens_seen": 7344664,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.17440053820610046,
      "learning_rate": 0.0008889526371275987,
      "loss": 0.4587,
      "num_input_tokens_seen": 7347736,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 0.07145164161920547,
      "learning_rate": 0.0008888088945744432,
      "loss": 0.5503,
      "num_input_tokens_seen": 7350584,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 0.11326996982097626,
      "learning_rate": 0.0008886650706869027,
      "loss": 0.4344,
      "num_input_tokens_seen": 7353368,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 0.06537839770317078,
      "learning_rate": 0.0008885211654950636,
      "loss": 0.437,
      "num_input_tokens_seen": 7356952,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 0.16032376885414124,
      "learning_rate": 0.0008883771790290291,
      "loss": 0.6152,
      "num_input_tokens_seen": 7360280,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 0.08328321576118469,
      "learning_rate": 0.0008882331113189197,
      "loss": 0.424,
      "num_input_tokens_seen": 7363352,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 0.21396002173423767,
      "learning_rate": 0.0008880889623948724,
      "loss": 0.3943,
      "num_input_tokens_seen": 7366104,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 0.13397395610809326,
      "learning_rate": 0.000887944732287042,
      "loss": 0.4557,
      "num_input_tokens_seen": 7368952,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 0.13380491733551025,
      "learning_rate": 0.0008878004210255995,
      "loss": 0.4602,
      "num_input_tokens_seen": 7372120,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 0.1268850564956665,
      "learning_rate": 0.0008876560286407329,
      "loss": 0.3098,
      "num_input_tokens_seen": 7375768,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 0.09603214263916016,
      "learning_rate": 0.000887511555162648,
      "loss": 0.4483,
      "num_input_tokens_seen": 7378520,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 0.14765340089797974,
      "learning_rate": 0.0008873670006215666,
      "loss": 0.4288,
      "num_input_tokens_seen": 7381304,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 0.08308751881122589,
      "learning_rate": 0.0008872223650477281,
      "loss": 0.3978,
      "num_input_tokens_seen": 7384632,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 0.12303397804498672,
      "learning_rate": 0.0008870776484713882,
      "loss": 0.526,
      "num_input_tokens_seen": 7387320,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.11311947554349899,
      "learning_rate": 0.0008869328509228205,
      "loss": 0.4415,
      "num_input_tokens_seen": 7390328,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 0.09910401701927185,
      "learning_rate": 0.0008867879724323147,
      "loss": 0.4505,
      "num_input_tokens_seen": 7393976,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.06779944896697998,
      "learning_rate": 0.0008866430130301777,
      "loss": 0.5664,
      "num_input_tokens_seen": 7396984,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.1340721696615219,
      "learning_rate": 0.0008864979727467335,
      "loss": 0.5457,
      "num_input_tokens_seen": 7399832,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.1205047070980072,
      "learning_rate": 0.000886352851612323,
      "loss": 0.5273,
      "num_input_tokens_seen": 7403064,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 0.10352329164743423,
      "learning_rate": 0.0008862076496573034,
      "loss": 0.5641,
      "num_input_tokens_seen": 7405432,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 0.08774352073669434,
      "learning_rate": 0.0008860623669120496,
      "loss": 0.3942,
      "num_input_tokens_seen": 7407800,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 0.11956840008497238,
      "learning_rate": 0.0008859170034069533,
      "loss": 0.4601,
      "num_input_tokens_seen": 7410552,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 0.09840361773967743,
      "learning_rate": 0.0008857715591724225,
      "loss": 0.5205,
      "num_input_tokens_seen": 7413912,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 0.09204123914241791,
      "learning_rate": 0.0008856260342388827,
      "loss": 0.5179,
      "num_input_tokens_seen": 7417400,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.10142338275909424,
      "learning_rate": 0.0008854804286367758,
      "loss": 0.4656,
      "num_input_tokens_seen": 7420344,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 0.09678857773542404,
      "learning_rate": 0.0008853347423965612,
      "loss": 0.4086,
      "num_input_tokens_seen": 7423288,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.0500115342438221,
      "learning_rate": 0.0008851889755487146,
      "loss": 0.4152,
      "num_input_tokens_seen": 7428536,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 0.21973153948783875,
      "learning_rate": 0.0008850431281237288,
      "loss": 0.5324,
      "num_input_tokens_seen": 7431736,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 0.16019666194915771,
      "learning_rate": 0.0008848972001521134,
      "loss": 0.5942,
      "num_input_tokens_seen": 7434136,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 0.13992443680763245,
      "learning_rate": 0.0008847511916643948,
      "loss": 0.5001,
      "num_input_tokens_seen": 7437112,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.07219807803630829,
      "learning_rate": 0.0008846051026911164,
      "loss": 0.4516,
      "num_input_tokens_seen": 7440120,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 0.12565337121486664,
      "learning_rate": 0.0008844589332628383,
      "loss": 0.6278,
      "num_input_tokens_seen": 7443608,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 0.12288446724414825,
      "learning_rate": 0.0008843126834101376,
      "loss": 0.4145,
      "num_input_tokens_seen": 7446808,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 0.14408531785011292,
      "learning_rate": 0.000884166353163608,
      "loss": 0.3683,
      "num_input_tokens_seen": 7449880,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 0.11110653728246689,
      "learning_rate": 0.0008840199425538599,
      "loss": 0.5175,
      "num_input_tokens_seen": 7454040,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 0.0918063223361969,
      "learning_rate": 0.0008838734516115213,
      "loss": 0.5049,
      "num_input_tokens_seen": 7456984,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 0.0788426473736763,
      "learning_rate": 0.0008837268803672359,
      "loss": 0.4272,
      "num_input_tokens_seen": 7460440,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 0.10682480782270432,
      "learning_rate": 0.0008835802288516647,
      "loss": 0.4139,
      "num_input_tokens_seen": 7465336,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 0.41057831048965454,
      "learning_rate": 0.0008834334970954861,
      "loss": 0.4585,
      "num_input_tokens_seen": 7468536,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 0.09381679445505142,
      "learning_rate": 0.0008832866851293941,
      "loss": 0.3776,
      "num_input_tokens_seen": 7471928,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 0.07130415737628937,
      "learning_rate": 0.0008831397929841005,
      "loss": 0.4036,
      "num_input_tokens_seen": 7475320,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.49057719111442566,
      "eval_runtime": 13.737,
      "eval_samples_per_second": 61.731,
      "eval_steps_per_second": 15.433,
      "num_input_tokens_seen": 7476480,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.17346985638141632,
      "learning_rate": 0.0008829928206903333,
      "loss": 0.3927,
      "num_input_tokens_seen": 7477440,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.0897599458694458,
      "learning_rate": 0.0008828457682788373,
      "loss": 0.4461,
      "num_input_tokens_seen": 7480224,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 0.06086263433098793,
      "learning_rate": 0.0008826986357803743,
      "loss": 0.3373,
      "num_input_tokens_seen": 7483392,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 0.09286234527826309,
      "learning_rate": 0.0008825514232257228,
      "loss": 0.3575,
      "num_input_tokens_seen": 7486688,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 0.07866925746202469,
      "learning_rate": 0.000882404130645678,
      "loss": 0.4384,
      "num_input_tokens_seen": 7489216,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 0.0912725180387497,
      "learning_rate": 0.0008822567580710518,
      "loss": 0.3793,
      "num_input_tokens_seen": 7491968,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 0.07423844188451767,
      "learning_rate": 0.0008821093055326728,
      "loss": 0.4231,
      "num_input_tokens_seen": 7495648,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 0.15199652314186096,
      "learning_rate": 0.0008819617730613863,
      "loss": 0.3416,
      "num_input_tokens_seen": 7497792,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 0.08263679593801498,
      "learning_rate": 0.0008818141606880545,
      "loss": 0.4165,
      "num_input_tokens_seen": 7500736,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 0.08035746961832047,
      "learning_rate": 0.0008816664684435564,
      "loss": 0.4544,
      "num_input_tokens_seen": 7504480,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 0.07508910447359085,
      "learning_rate": 0.0008815186963587873,
      "loss": 0.555,
      "num_input_tokens_seen": 7507392,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 0.13630785048007965,
      "learning_rate": 0.0008813708444646596,
      "loss": 0.4647,
      "num_input_tokens_seen": 7512896,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 0.10332755744457245,
      "learning_rate": 0.0008812229127921021,
      "loss": 0.5681,
      "num_input_tokens_seen": 7516064,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.1264263540506363,
      "learning_rate": 0.0008810749013720605,
      "loss": 0.49,
      "num_input_tokens_seen": 7518560,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.11625564098358154,
      "learning_rate": 0.0008809268102354969,
      "loss": 0.3788,
      "num_input_tokens_seen": 7523168,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 0.1635240614414215,
      "learning_rate": 0.0008807786394133905,
      "loss": 0.4558,
      "num_input_tokens_seen": 7526144,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 0.05916918069124222,
      "learning_rate": 0.0008806303889367371,
      "loss": 0.4494,
      "num_input_tokens_seen": 7530240,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 0.06137445196509361,
      "learning_rate": 0.0008804820588365485,
      "loss": 0.3604,
      "num_input_tokens_seen": 7534304,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 0.2211591750383377,
      "learning_rate": 0.0008803336491438543,
      "loss": 0.6082,
      "num_input_tokens_seen": 7537088,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 0.08991304039955139,
      "learning_rate": 0.0008801851598896995,
      "loss": 0.4887,
      "num_input_tokens_seen": 7540768,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 0.1074845939874649,
      "learning_rate": 0.0008800365911051467,
      "loss": 0.4509,
      "num_input_tokens_seen": 7543968,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 0.1042802631855011,
      "learning_rate": 0.0008798879428212747,
      "loss": 0.3844,
      "num_input_tokens_seen": 7547296,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 0.14062751829624176,
      "learning_rate": 0.0008797392150691792,
      "loss": 0.4765,
      "num_input_tokens_seen": 7550720,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.07633409649133682,
      "learning_rate": 0.000879590407879972,
      "loss": 0.4565,
      "num_input_tokens_seen": 7554208,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 0.10607584565877914,
      "learning_rate": 0.0008794415212847822,
      "loss": 0.4437,
      "num_input_tokens_seen": 7560288,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 0.12954291701316833,
      "learning_rate": 0.000879292555314755,
      "loss": 0.3973,
      "num_input_tokens_seen": 7563264,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 0.11192799359560013,
      "learning_rate": 0.0008791435100010524,
      "loss": 0.3984,
      "num_input_tokens_seen": 7566624,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 0.08960049599409103,
      "learning_rate": 0.000878994385374853,
      "loss": 0.5022,
      "num_input_tokens_seen": 7569952,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 0.06586234271526337,
      "learning_rate": 0.0008788451814673521,
      "loss": 0.4367,
      "num_input_tokens_seen": 7574240,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 0.19564352929592133,
      "learning_rate": 0.0008786958983097613,
      "loss": 0.4496,
      "num_input_tokens_seen": 7577792,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 0.08575217425823212,
      "learning_rate": 0.0008785465359333088,
      "loss": 0.4815,
      "num_input_tokens_seen": 7581440,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 0.13684651255607605,
      "learning_rate": 0.0008783970943692398,
      "loss": 0.3073,
      "num_input_tokens_seen": 7584640,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 0.1105891540646553,
      "learning_rate": 0.0008782475736488156,
      "loss": 0.4496,
      "num_input_tokens_seen": 7587776,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 0.05281759798526764,
      "learning_rate": 0.0008780979738033142,
      "loss": 0.4005,
      "num_input_tokens_seen": 7591520,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 0.17497988045215607,
      "learning_rate": 0.0008779482948640305,
      "loss": 0.4348,
      "num_input_tokens_seen": 7595040,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 0.09551442414522171,
      "learning_rate": 0.0008777985368622752,
      "loss": 0.5604,
      "num_input_tokens_seen": 7598048,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 0.1222761869430542,
      "learning_rate": 0.0008776486998293762,
      "loss": 0.4277,
      "num_input_tokens_seen": 7601344,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 0.16365112364292145,
      "learning_rate": 0.0008774987837966775,
      "loss": 0.4449,
      "num_input_tokens_seen": 7604256,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.13512013852596283,
      "learning_rate": 0.0008773487887955401,
      "loss": 0.4104,
      "num_input_tokens_seen": 7607456,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 0.11332128196954727,
      "learning_rate": 0.000877198714857341,
      "loss": 0.5189,
      "num_input_tokens_seen": 7611456,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 0.1114249974489212,
      "learning_rate": 0.000877048562013474,
      "loss": 0.4501,
      "num_input_tokens_seen": 7614912,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 0.21657857298851013,
      "learning_rate": 0.0008768983302953498,
      "loss": 0.4868,
      "num_input_tokens_seen": 7618688,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.10057055950164795,
      "learning_rate": 0.0008767480197343945,
      "loss": 0.5174,
      "num_input_tokens_seen": 7621280,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 0.07309242337942123,
      "learning_rate": 0.0008765976303620515,
      "loss": 0.5224,
      "num_input_tokens_seen": 7625600,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 0.06542923301458359,
      "learning_rate": 0.0008764471622097809,
      "loss": 0.5304,
      "num_input_tokens_seen": 7628992,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 0.07117654383182526,
      "learning_rate": 0.0008762966153090588,
      "loss": 0.3745,
      "num_input_tokens_seen": 7631904,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 0.05438242107629776,
      "learning_rate": 0.0008761459896913776,
      "loss": 0.4421,
      "num_input_tokens_seen": 7635424,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 0.07853037863969803,
      "learning_rate": 0.0008759952853882466,
      "loss": 0.6252,
      "num_input_tokens_seen": 7638240,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 0.09557481110095978,
      "learning_rate": 0.0008758445024311917,
      "loss": 0.4187,
      "num_input_tokens_seen": 7640768,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 0.09160564839839935,
      "learning_rate": 0.0008756936408517547,
      "loss": 0.3826,
      "num_input_tokens_seen": 7643904,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 0.10382909327745438,
      "learning_rate": 0.000875542700681494,
      "loss": 0.439,
      "num_input_tokens_seen": 7646720,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.1281897872686386,
      "learning_rate": 0.0008753916819519849,
      "loss": 0.3769,
      "num_input_tokens_seen": 7648928,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 0.10560503602027893,
      "learning_rate": 0.0008752405846948186,
      "loss": 0.5477,
      "num_input_tokens_seen": 7652320,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 0.08865197002887726,
      "learning_rate": 0.0008750894089416029,
      "loss": 0.5974,
      "num_input_tokens_seen": 7656096,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 0.08963274210691452,
      "learning_rate": 0.0008749381547239621,
      "loss": 0.4389,
      "num_input_tokens_seen": 7659328,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 0.13067764043807983,
      "learning_rate": 0.000874786822073537,
      "loss": 0.3313,
      "num_input_tokens_seen": 7661984,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 0.08556395024061203,
      "learning_rate": 0.0008746354110219843,
      "loss": 0.5802,
      "num_input_tokens_seen": 7665728,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 0.04585501179099083,
      "learning_rate": 0.0008744839216009779,
      "loss": 0.4687,
      "num_input_tokens_seen": 7669760,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 0.08759548515081406,
      "learning_rate": 0.0008743323538422072,
      "loss": 0.3484,
      "num_input_tokens_seen": 7673120,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.059804484248161316,
      "learning_rate": 0.0008741807077773789,
      "loss": 0.4865,
      "num_input_tokens_seen": 7676384,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 0.08628079295158386,
      "learning_rate": 0.0008740289834382153,
      "loss": 0.4923,
      "num_input_tokens_seen": 7679424,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 0.11109688878059387,
      "learning_rate": 0.0008738771808564555,
      "loss": 0.4166,
      "num_input_tokens_seen": 7682816,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 0.07401067018508911,
      "learning_rate": 0.000873725300063855,
      "loss": 0.439,
      "num_input_tokens_seen": 7687488,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 0.08540881425142288,
      "learning_rate": 0.0008735733410921853,
      "loss": 0.4401,
      "num_input_tokens_seen": 7691648,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 0.02498738467693329,
      "learning_rate": 0.0008734213039732346,
      "loss": 0.2647,
      "num_input_tokens_seen": 7696928,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 0.05478015914559364,
      "learning_rate": 0.0008732691887388073,
      "loss": 0.3419,
      "num_input_tokens_seen": 7699744,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.1437426060438156,
      "learning_rate": 0.0008731169954207243,
      "loss": 0.5722,
      "num_input_tokens_seen": 7702912,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 0.11735054850578308,
      "learning_rate": 0.0008729647240508225,
      "loss": 0.5086,
      "num_input_tokens_seen": 7706560,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 0.09272880852222443,
      "learning_rate": 0.0008728123746609555,
      "loss": 0.6377,
      "num_input_tokens_seen": 7710304,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.07878055423498154,
      "learning_rate": 0.000872659947282993,
      "loss": 0.4569,
      "num_input_tokens_seen": 7713184,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 0.1423899531364441,
      "learning_rate": 0.000872507441948821,
      "loss": 0.6832,
      "num_input_tokens_seen": 7716192,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 0.120905302464962,
      "learning_rate": 0.0008723548586903418,
      "loss": 0.477,
      "num_input_tokens_seen": 7720256,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 0.10948173701763153,
      "learning_rate": 0.0008722021975394745,
      "loss": 0.5111,
      "num_input_tokens_seen": 7723392,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.12659738957881927,
      "learning_rate": 0.0008720494585281536,
      "loss": 0.4829,
      "num_input_tokens_seen": 7726720,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 0.11294083297252655,
      "learning_rate": 0.0008718966416883306,
      "loss": 0.4152,
      "num_input_tokens_seen": 7729248,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 0.15422464907169342,
      "learning_rate": 0.0008717437470519729,
      "loss": 0.517,
      "num_input_tokens_seen": 7732800,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 0.1259642243385315,
      "learning_rate": 0.0008715907746510645,
      "loss": 0.4187,
      "num_input_tokens_seen": 7736032,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 0.10354246944189072,
      "learning_rate": 0.0008714377245176054,
      "loss": 0.4445,
      "num_input_tokens_seen": 7738624,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 0.16654382646083832,
      "learning_rate": 0.0008712845966836119,
      "loss": 0.4904,
      "num_input_tokens_seen": 7741728,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 0.10071665793657303,
      "learning_rate": 0.0008711313911811165,
      "loss": 0.4621,
      "num_input_tokens_seen": 7744288,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 0.10123591125011444,
      "learning_rate": 0.0008709781080421681,
      "loss": 0.5106,
      "num_input_tokens_seen": 7748448,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 0.07078272849321365,
      "learning_rate": 0.0008708247472988319,
      "loss": 0.4623,
      "num_input_tokens_seen": 7752096,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 0.07776940613985062,
      "learning_rate": 0.0008706713089831893,
      "loss": 0.444,
      "num_input_tokens_seen": 7755104,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 0.08471556007862091,
      "learning_rate": 0.0008705177931273377,
      "loss": 0.4078,
      "num_input_tokens_seen": 7758592,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 0.10344825685024261,
      "learning_rate": 0.0008703641997633907,
      "loss": 0.4699,
      "num_input_tokens_seen": 7762880,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 0.1309724748134613,
      "learning_rate": 0.0008702105289234786,
      "loss": 0.5627,
      "num_input_tokens_seen": 7766944,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.10599230229854584,
      "learning_rate": 0.0008700567806397473,
      "loss": 0.3582,
      "num_input_tokens_seen": 7770656,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 0.2617606222629547,
      "learning_rate": 0.0008699029549443595,
      "loss": 0.5546,
      "num_input_tokens_seen": 7773504,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 0.1173219159245491,
      "learning_rate": 0.0008697490518694935,
      "loss": 0.5279,
      "num_input_tokens_seen": 7777216,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 0.15978524088859558,
      "learning_rate": 0.0008695950714473442,
      "loss": 0.5441,
      "num_input_tokens_seen": 7780384,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 0.1698368787765503,
      "learning_rate": 0.0008694410137101224,
      "loss": 0.491,
      "num_input_tokens_seen": 7783776,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 0.13837330043315887,
      "learning_rate": 0.0008692868786900554,
      "loss": 0.5042,
      "num_input_tokens_seen": 7787296,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 0.12470237910747528,
      "learning_rate": 0.0008691326664193865,
      "loss": 0.5633,
      "num_input_tokens_seen": 7789888,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 0.07049357146024704,
      "learning_rate": 0.000868978376930375,
      "loss": 0.4247,
      "num_input_tokens_seen": 7793280,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 0.14574085175991058,
      "learning_rate": 0.0008688240102552965,
      "loss": 0.452,
      "num_input_tokens_seen": 7796224,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.09869998693466187,
      "learning_rate": 0.000868669566426443,
      "loss": 0.4967,
      "num_input_tokens_seen": 7799424,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 0.09139006584882736,
      "learning_rate": 0.0008685150454761222,
      "loss": 0.4614,
      "num_input_tokens_seen": 7802528,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 0.11759073287248611,
      "learning_rate": 0.000868360447436658,
      "loss": 0.4914,
      "num_input_tokens_seen": 7806272,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 0.10928219556808472,
      "learning_rate": 0.0008682057723403907,
      "loss": 0.5016,
      "num_input_tokens_seen": 7809248,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 0.09482371062040329,
      "learning_rate": 0.0008680510202196765,
      "loss": 0.4546,
      "num_input_tokens_seen": 7813536,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 0.13442175090312958,
      "learning_rate": 0.0008678961911068879,
      "loss": 0.4248,
      "num_input_tokens_seen": 7816768,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 0.07391441613435745,
      "learning_rate": 0.0008677412850344133,
      "loss": 0.3611,
      "num_input_tokens_seen": 7820224,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 0.09171653538942337,
      "learning_rate": 0.0008675863020346573,
      "loss": 0.6978,
      "num_input_tokens_seen": 7823104,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 0.09628558158874512,
      "learning_rate": 0.0008674312421400405,
      "loss": 0.4793,
      "num_input_tokens_seen": 7826688,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 0.11736822128295898,
      "learning_rate": 0.0008672761053829997,
      "loss": 0.4137,
      "num_input_tokens_seen": 7830272,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.10498352348804474,
      "learning_rate": 0.0008671208917959878,
      "loss": 0.4209,
      "num_input_tokens_seen": 7833824,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 0.1192079707980156,
      "learning_rate": 0.0008669656014114737,
      "loss": 0.4305,
      "num_input_tokens_seen": 7837632,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.0763007253408432,
      "learning_rate": 0.0008668102342619424,
      "loss": 0.4343,
      "num_input_tokens_seen": 7840928,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 0.15052054822444916,
      "learning_rate": 0.0008666547903798947,
      "loss": 0.6079,
      "num_input_tokens_seen": 7847680,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 0.11007712036371231,
      "learning_rate": 0.000866499269797848,
      "loss": 0.4067,
      "num_input_tokens_seen": 7853984,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.07516677677631378,
      "learning_rate": 0.0008663436725483354,
      "loss": 0.6126,
      "num_input_tokens_seen": 7857600,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 0.1362943798303604,
      "learning_rate": 0.0008661879986639058,
      "loss": 0.5287,
      "num_input_tokens_seen": 7860448,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 0.1370057910680771,
      "learning_rate": 0.0008660322481771246,
      "loss": 0.4203,
      "num_input_tokens_seen": 7863424,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 0.07062871009111404,
      "learning_rate": 0.0008658764211205731,
      "loss": 0.5744,
      "num_input_tokens_seen": 7866400,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 0.05696363002061844,
      "learning_rate": 0.0008657205175268484,
      "loss": 0.3236,
      "num_input_tokens_seen": 7870240,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 0.17277593910694122,
      "learning_rate": 0.0008655645374285637,
      "loss": 0.4501,
      "num_input_tokens_seen": 7873280,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 0.15936051309108734,
      "learning_rate": 0.0008654084808583485,
      "loss": 0.5623,
      "num_input_tokens_seen": 7876640,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 0.12612122297286987,
      "learning_rate": 0.0008652523478488478,
      "loss": 0.4479,
      "num_input_tokens_seen": 7880352,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 0.14103806018829346,
      "learning_rate": 0.000865096138432723,
      "loss": 0.497,
      "num_input_tokens_seen": 7883040,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 0.07314707338809967,
      "learning_rate": 0.0008649398526426512,
      "loss": 0.5114,
      "num_input_tokens_seen": 7886112,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 0.11924866586923599,
      "learning_rate": 0.0008647834905113256,
      "loss": 0.4872,
      "num_input_tokens_seen": 7888512,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 0.21670717000961304,
      "learning_rate": 0.0008646270520714556,
      "loss": 0.446,
      "num_input_tokens_seen": 7891296,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 0.132796049118042,
      "learning_rate": 0.0008644705373557663,
      "loss": 0.5716,
      "num_input_tokens_seen": 7893696,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 0.07225551456212997,
      "learning_rate": 0.0008643139463969985,
      "loss": 0.3955,
      "num_input_tokens_seen": 7897632,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 0.1876266896724701,
      "learning_rate": 0.0008641572792279095,
      "loss": 0.4058,
      "num_input_tokens_seen": 7899936,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 0.10602270066738129,
      "learning_rate": 0.0008640005358812723,
      "loss": 0.4464,
      "num_input_tokens_seen": 7902848,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 0.08418064564466476,
      "learning_rate": 0.0008638437163898756,
      "loss": 0.3901,
      "num_input_tokens_seen": 7906176,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 0.11767416447401047,
      "learning_rate": 0.0008636868207865244,
      "loss": 0.3824,
      "num_input_tokens_seen": 7909408,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 0.09420856088399887,
      "learning_rate": 0.0008635298491040394,
      "loss": 0.3359,
      "num_input_tokens_seen": 7912576,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 0.09603141248226166,
      "learning_rate": 0.0008633728013752574,
      "loss": 0.6443,
      "num_input_tokens_seen": 7915552,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 0.07973173260688782,
      "learning_rate": 0.000863215677633031,
      "loss": 0.3865,
      "num_input_tokens_seen": 7920384,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 0.10947740823030472,
      "learning_rate": 0.0008630584779102285,
      "loss": 0.4193,
      "num_input_tokens_seen": 7923552,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 0.129250168800354,
      "learning_rate": 0.0008629012022397344,
      "loss": 0.5047,
      "num_input_tokens_seen": 7926080,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 0.13106365501880646,
      "learning_rate": 0.0008627438506544491,
      "loss": 0.5841,
      "num_input_tokens_seen": 7928992,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 0.09388386458158493,
      "learning_rate": 0.0008625864231872886,
      "loss": 0.5475,
      "num_input_tokens_seen": 7931936,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 0.08258006721735,
      "learning_rate": 0.000862428919871185,
      "loss": 0.4309,
      "num_input_tokens_seen": 7934368,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 0.164352685213089,
      "learning_rate": 0.0008622713407390861,
      "loss": 0.4049,
      "num_input_tokens_seen": 7937088,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.07843374460935593,
      "learning_rate": 0.0008621136858239557,
      "loss": 0.4146,
      "num_input_tokens_seen": 7939776,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 0.0747799500823021,
      "learning_rate": 0.0008619559551587736,
      "loss": 0.4455,
      "num_input_tokens_seen": 7942944,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 0.09744968265295029,
      "learning_rate": 0.0008617981487765349,
      "loss": 0.5282,
      "num_input_tokens_seen": 7946336,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 0.1025528758764267,
      "learning_rate": 0.0008616402667102512,
      "loss": 0.4731,
      "num_input_tokens_seen": 7950880,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 0.1673612743616104,
      "learning_rate": 0.0008614823089929495,
      "loss": 0.5522,
      "num_input_tokens_seen": 7953856,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 0.10768971592187881,
      "learning_rate": 0.0008613242756576728,
      "loss": 0.4072,
      "num_input_tokens_seen": 7956672,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 0.08468055725097656,
      "learning_rate": 0.0008611661667374798,
      "loss": 0.4206,
      "num_input_tokens_seen": 7959552,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 0.13213953375816345,
      "learning_rate": 0.0008610079822654451,
      "loss": 0.4959,
      "num_input_tokens_seen": 7962816,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 0.22794613242149353,
      "learning_rate": 0.0008608497222746591,
      "loss": 0.5339,
      "num_input_tokens_seen": 7965664,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.1175820603966713,
      "learning_rate": 0.000860691386798228,
      "loss": 0.5819,
      "num_input_tokens_seen": 7969632,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 0.11255897581577301,
      "learning_rate": 0.000860532975869274,
      "loss": 0.4403,
      "num_input_tokens_seen": 7973216,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.09706003218889236,
      "learning_rate": 0.0008603744895209342,
      "loss": 0.4187,
      "num_input_tokens_seen": 7975936,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 0.07686977088451385,
      "learning_rate": 0.0008602159277863628,
      "loss": 0.5019,
      "num_input_tokens_seen": 7979456,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 0.12251576036214828,
      "learning_rate": 0.0008600572906987289,
      "loss": 0.618,
      "num_input_tokens_seen": 7982528,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 0.06799769401550293,
      "learning_rate": 0.0008598985782912174,
      "loss": 0.5315,
      "num_input_tokens_seen": 7985344,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 0.0999397486448288,
      "learning_rate": 0.0008597397905970293,
      "loss": 0.4276,
      "num_input_tokens_seen": 7989088,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 0.17530345916748047,
      "learning_rate": 0.0008595809276493812,
      "loss": 0.5094,
      "num_input_tokens_seen": 7992736,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 0.12582527101039886,
      "learning_rate": 0.0008594219894815054,
      "loss": 0.5078,
      "num_input_tokens_seen": 7995072,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 0.09695494174957275,
      "learning_rate": 0.0008592629761266497,
      "loss": 0.4947,
      "num_input_tokens_seen": 7998240,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 0.1303030550479889,
      "learning_rate": 0.0008591038876180784,
      "loss": 0.4595,
      "num_input_tokens_seen": 8001600,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 0.10923199355602264,
      "learning_rate": 0.0008589447239890705,
      "loss": 0.3454,
      "num_input_tokens_seen": 8004128,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.18138903379440308,
      "learning_rate": 0.0008587854852729213,
      "loss": 0.4818,
      "num_input_tokens_seen": 8007936,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 0.1000678539276123,
      "learning_rate": 0.000858626171502942,
      "loss": 0.4537,
      "num_input_tokens_seen": 8011488,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 0.09863263368606567,
      "learning_rate": 0.0008584667827124589,
      "loss": 0.4891,
      "num_input_tokens_seen": 8014656,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 0.11383878439664841,
      "learning_rate": 0.0008583073189348145,
      "loss": 0.3967,
      "num_input_tokens_seen": 8017504,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 0.09830217808485031,
      "learning_rate": 0.0008581477802033668,
      "loss": 0.3873,
      "num_input_tokens_seen": 8021728,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 0.10955876111984253,
      "learning_rate": 0.0008579881665514895,
      "loss": 0.4365,
      "num_input_tokens_seen": 8024832,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 0.09074734896421432,
      "learning_rate": 0.0008578284780125718,
      "loss": 0.4667,
      "num_input_tokens_seen": 8028192,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 0.2710210084915161,
      "learning_rate": 0.0008576687146200189,
      "loss": 0.4987,
      "num_input_tokens_seen": 8031168,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 0.0648871660232544,
      "learning_rate": 0.0008575088764072511,
      "loss": 0.475,
      "num_input_tokens_seen": 8034208,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 0.13013266026973724,
      "learning_rate": 0.0008573489634077054,
      "loss": 0.4589,
      "num_input_tokens_seen": 8037088,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 0.10725747793912888,
      "learning_rate": 0.0008571889756548333,
      "loss": 0.6793,
      "num_input_tokens_seen": 8039968,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 0.14474350214004517,
      "learning_rate": 0.0008570289131821025,
      "loss": 0.4268,
      "num_input_tokens_seen": 8042720,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 0.07094388455152512,
      "learning_rate": 0.0008568687760229963,
      "loss": 0.4571,
      "num_input_tokens_seen": 8045792,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.11499614268541336,
      "learning_rate": 0.0008567085642110135,
      "loss": 0.4297,
      "num_input_tokens_seen": 8048512,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 0.13547590374946594,
      "learning_rate": 0.0008565482777796687,
      "loss": 0.4903,
      "num_input_tokens_seen": 8051168,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 0.12917183339595795,
      "learning_rate": 0.0008563879167624919,
      "loss": 0.4204,
      "num_input_tokens_seen": 8053376,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 0.16448304057121277,
      "learning_rate": 0.0008562274811930287,
      "loss": 0.5587,
      "num_input_tokens_seen": 8056128,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 0.1767899990081787,
      "learning_rate": 0.0008560669711048406,
      "loss": 0.4942,
      "num_input_tokens_seen": 8059232,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 0.1202191561460495,
      "learning_rate": 0.0008559063865315042,
      "loss": 0.4686,
      "num_input_tokens_seen": 8062688,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 0.10992155969142914,
      "learning_rate": 0.0008557457275066122,
      "loss": 0.4893,
      "num_input_tokens_seen": 8065952,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 0.09947189688682556,
      "learning_rate": 0.0008555849940637725,
      "loss": 0.4974,
      "num_input_tokens_seen": 8068960,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 0.13454942405223846,
      "learning_rate": 0.000855424186236609,
      "loss": 0.4267,
      "num_input_tokens_seen": 8071776,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 0.1597214639186859,
      "learning_rate": 0.0008552633040587605,
      "loss": 0.5787,
      "num_input_tokens_seen": 8073856,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 0.09528917819261551,
      "learning_rate": 0.0008551023475638819,
      "loss": 0.5713,
      "num_input_tokens_seen": 8077248,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 0.09213234484195709,
      "learning_rate": 0.0008549413167856432,
      "loss": 0.3526,
      "num_input_tokens_seen": 8080992,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 0.12012866139411926,
      "learning_rate": 0.0008547802117577307,
      "loss": 0.526,
      "num_input_tokens_seen": 8084096,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.1337207406759262,
      "learning_rate": 0.0008546190325138454,
      "loss": 0.5003,
      "num_input_tokens_seen": 8087072,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 0.07185932248830795,
      "learning_rate": 0.000854457779087704,
      "loss": 0.4288,
      "num_input_tokens_seen": 8090912,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 0.15595781803131104,
      "learning_rate": 0.0008542964515130392,
      "loss": 0.439,
      "num_input_tokens_seen": 8093408,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 0.08853012323379517,
      "learning_rate": 0.0008541350498235989,
      "loss": 0.455,
      "num_input_tokens_seen": 8096480,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 0.12404504418373108,
      "learning_rate": 0.0008539735740531462,
      "loss": 0.4421,
      "num_input_tokens_seen": 8099200,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 0.07275819778442383,
      "learning_rate": 0.0008538120242354602,
      "loss": 0.5148,
      "num_input_tokens_seen": 8102272,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 0.08105175942182541,
      "learning_rate": 0.0008536504004043353,
      "loss": 0.3485,
      "num_input_tokens_seen": 8105440,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 0.08242668211460114,
      "learning_rate": 0.0008534887025935811,
      "loss": 0.4898,
      "num_input_tokens_seen": 8108320,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 0.12166956067085266,
      "learning_rate": 0.0008533269308370233,
      "loss": 0.5698,
      "num_input_tokens_seen": 8111744,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 0.10953070968389511,
      "learning_rate": 0.0008531650851685025,
      "loss": 0.4451,
      "num_input_tokens_seen": 8114848,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 0.158122718334198,
      "learning_rate": 0.000853003165621875,
      "loss": 0.4037,
      "num_input_tokens_seen": 8117504,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 0.0754537582397461,
      "learning_rate": 0.0008528411722310126,
      "loss": 0.3773,
      "num_input_tokens_seen": 8121344,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 0.20168761909008026,
      "learning_rate": 0.0008526791050298022,
      "loss": 0.5454,
      "num_input_tokens_seen": 8124288,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 0.07905388623476028,
      "learning_rate": 0.0008525169640521467,
      "loss": 0.4847,
      "num_input_tokens_seen": 8127552,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.07469148933887482,
      "learning_rate": 0.0008523547493319641,
      "loss": 0.5407,
      "num_input_tokens_seen": 8131264,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 0.10376733541488647,
      "learning_rate": 0.0008521924609031877,
      "loss": 0.4146,
      "num_input_tokens_seen": 8134432,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 0.11802135407924652,
      "learning_rate": 0.0008520300987997667,
      "loss": 0.5407,
      "num_input_tokens_seen": 8137152,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 0.06719963997602463,
      "learning_rate": 0.000851867663055665,
      "loss": 0.4049,
      "num_input_tokens_seen": 8140512,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 0.1584746092557907,
      "learning_rate": 0.0008517051537048625,
      "loss": 0.5617,
      "num_input_tokens_seen": 8144160,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 0.09327813982963562,
      "learning_rate": 0.0008515425707813544,
      "loss": 0.3608,
      "num_input_tokens_seen": 8149184,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 0.055183324962854385,
      "learning_rate": 0.0008513799143191512,
      "loss": 0.6247,
      "num_input_tokens_seen": 8152224,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 0.14792391657829285,
      "learning_rate": 0.0008512171843522786,
      "loss": 0.4639,
      "num_input_tokens_seen": 8155232,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 0.13452932238578796,
      "learning_rate": 0.000851054380914778,
      "loss": 0.426,
      "num_input_tokens_seen": 8157536,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 0.09727153927087784,
      "learning_rate": 0.0008508915040407057,
      "loss": 0.4337,
      "num_input_tokens_seen": 8159872,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 0.07498008012771606,
      "learning_rate": 0.0008507285537641342,
      "loss": 0.4741,
      "num_input_tokens_seen": 8163872,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 0.13122457265853882,
      "learning_rate": 0.0008505655301191504,
      "loss": 0.4351,
      "num_input_tokens_seen": 8167168,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 0.05940939486026764,
      "learning_rate": 0.0008504024331398572,
      "loss": 0.3674,
      "num_input_tokens_seen": 8171040,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 0.10270952433347702,
      "learning_rate": 0.0008502392628603727,
      "loss": 0.3588,
      "num_input_tokens_seen": 8174336,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 0.07250943779945374,
      "learning_rate": 0.0008500760193148301,
      "loss": 0.4093,
      "num_input_tokens_seen": 8178144,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 0.11773917078971863,
      "learning_rate": 0.0008499127025373781,
      "loss": 0.4664,
      "num_input_tokens_seen": 8181664,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 0.2009725719690323,
      "learning_rate": 0.0008497493125621807,
      "loss": 0.6175,
      "num_input_tokens_seen": 8185344,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 0.09355459362268448,
      "learning_rate": 0.0008495858494234173,
      "loss": 0.4386,
      "num_input_tokens_seen": 8187840,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 0.14053528010845184,
      "learning_rate": 0.0008494223131552825,
      "loss": 0.3959,
      "num_input_tokens_seen": 8191520,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 0.11690589040517807,
      "learning_rate": 0.0008492587037919864,
      "loss": 0.3352,
      "num_input_tokens_seen": 8194976,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 0.09854595363140106,
      "learning_rate": 0.0008490950213677537,
      "loss": 0.4572,
      "num_input_tokens_seen": 8197600,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 0.11843733489513397,
      "learning_rate": 0.0008489312659168254,
      "loss": 0.4097,
      "num_input_tokens_seen": 8200448,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 0.08965648710727692,
      "learning_rate": 0.0008487674374734569,
      "loss": 0.4976,
      "num_input_tokens_seen": 8203328,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 0.10339942574501038,
      "learning_rate": 0.0008486035360719197,
      "loss": 0.4853,
      "num_input_tokens_seen": 8208480,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 0.1047523096203804,
      "learning_rate": 0.0008484395617464997,
      "loss": 0.3859,
      "num_input_tokens_seen": 8211424,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 0.056317638605833054,
      "learning_rate": 0.0008482755145314986,
      "loss": 0.4486,
      "num_input_tokens_seen": 8214848,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 0.14321279525756836,
      "learning_rate": 0.0008481113944612331,
      "loss": 0.4242,
      "num_input_tokens_seen": 8217568,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 0.10520225018262863,
      "learning_rate": 0.0008479472015700356,
      "loss": 0.5092,
      "num_input_tokens_seen": 8220672,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 0.14880679547786713,
      "learning_rate": 0.0008477829358922529,
      "loss": 0.4371,
      "num_input_tokens_seen": 8224032,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 0.2119201421737671,
      "learning_rate": 0.0008476185974622477,
      "loss": 0.5699,
      "num_input_tokens_seen": 8227872,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 0.21688687801361084,
      "learning_rate": 0.0008474541863143978,
      "loss": 0.5244,
      "num_input_tokens_seen": 8230720,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 0.0645579919219017,
      "learning_rate": 0.0008472897024830961,
      "loss": 0.4564,
      "num_input_tokens_seen": 8234080,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.14689423143863678,
      "learning_rate": 0.0008471251460027508,
      "loss": 0.534,
      "num_input_tokens_seen": 8237568,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 0.10765025019645691,
      "learning_rate": 0.0008469605169077852,
      "loss": 0.4218,
      "num_input_tokens_seen": 8240544,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 0.15275129675865173,
      "learning_rate": 0.0008467958152326377,
      "loss": 0.4703,
      "num_input_tokens_seen": 8243488,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 0.2275949865579605,
      "learning_rate": 0.0008466310410117622,
      "loss": 0.4656,
      "num_input_tokens_seen": 8246208,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.11928866803646088,
      "learning_rate": 0.0008464661942796275,
      "loss": 0.5373,
      "num_input_tokens_seen": 8249728,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 0.12395045161247253,
      "learning_rate": 0.0008463012750707175,
      "loss": 0.5101,
      "num_input_tokens_seen": 8252512,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 0.09323512017726898,
      "learning_rate": 0.0008461362834195316,
      "loss": 0.4263,
      "num_input_tokens_seen": 8255136,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 0.08070236444473267,
      "learning_rate": 0.0008459712193605842,
      "loss": 0.8086,
      "num_input_tokens_seen": 8258112,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 0.11620331555604935,
      "learning_rate": 0.0008458060829284048,
      "loss": 0.5408,
      "num_input_tokens_seen": 8261312,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 0.07524947822093964,
      "learning_rate": 0.0008456408741575378,
      "loss": 0.3626,
      "num_input_tokens_seen": 8264608,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 0.09155817329883575,
      "learning_rate": 0.0008454755930825435,
      "loss": 0.5325,
      "num_input_tokens_seen": 8267328,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 0.08740652352571487,
      "learning_rate": 0.0008453102397379963,
      "loss": 0.5578,
      "num_input_tokens_seen": 8269984,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 0.13035669922828674,
      "learning_rate": 0.0008451448141584868,
      "loss": 0.485,
      "num_input_tokens_seen": 8273472,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.09887368977069855,
      "learning_rate": 0.0008449793163786194,
      "loss": 0.4392,
      "num_input_tokens_seen": 8276320,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 0.10333661735057831,
      "learning_rate": 0.0008448137464330149,
      "loss": 0.5848,
      "num_input_tokens_seen": 8279200,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.08375509828329086,
      "learning_rate": 0.0008446481043563086,
      "loss": 0.5031,
      "num_input_tokens_seen": 8282400,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 0.11817033588886261,
      "learning_rate": 0.0008444823901831507,
      "loss": 0.3477,
      "num_input_tokens_seen": 8285728,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 0.11211778223514557,
      "learning_rate": 0.0008443166039482067,
      "loss": 0.3524,
      "num_input_tokens_seen": 8288704,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 0.09685435146093369,
      "learning_rate": 0.0008441507456861575,
      "loss": 0.4862,
      "num_input_tokens_seen": 8291904,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.0915321409702301,
      "learning_rate": 0.0008439848154316985,
      "loss": 0.4708,
      "num_input_tokens_seen": 8295488,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 0.12965375185012817,
      "learning_rate": 0.0008438188132195406,
      "loss": 0.4571,
      "num_input_tokens_seen": 8298784,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 0.15480229258537292,
      "learning_rate": 0.0008436527390844094,
      "loss": 0.4955,
      "num_input_tokens_seen": 8301376,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 0.11898434907197952,
      "learning_rate": 0.0008434865930610458,
      "loss": 0.6671,
      "num_input_tokens_seen": 8304544,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 0.13466157019138336,
      "learning_rate": 0.0008433203751842057,
      "loss": 0.5489,
      "num_input_tokens_seen": 8307232,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.09593378752470016,
      "learning_rate": 0.0008431540854886598,
      "loss": 0.5288,
      "num_input_tokens_seen": 8310816,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 0.11784105002880096,
      "learning_rate": 0.0008429877240091941,
      "loss": 0.4737,
      "num_input_tokens_seen": 8313632,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 0.16350698471069336,
      "learning_rate": 0.0008428212907806097,
      "loss": 0.3869,
      "num_input_tokens_seen": 8316416,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 0.08730915188789368,
      "learning_rate": 0.0008426547858377222,
      "loss": 0.4943,
      "num_input_tokens_seen": 8321056,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 0.39511412382125854,
      "learning_rate": 0.0008424882092153629,
      "loss": 0.3601,
      "num_input_tokens_seen": 8324064,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 0.11209419369697571,
      "learning_rate": 0.0008423215609483774,
      "loss": 0.519,
      "num_input_tokens_seen": 8326592,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 0.14293360710144043,
      "learning_rate": 0.0008421548410716269,
      "loss": 0.3981,
      "num_input_tokens_seen": 8328992,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 0.08181741088628769,
      "learning_rate": 0.0008419880496199869,
      "loss": 0.5012,
      "num_input_tokens_seen": 8332448,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 0.1396515667438507,
      "learning_rate": 0.0008418211866283488,
      "loss": 0.4498,
      "num_input_tokens_seen": 8335136,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.10732109844684601,
      "learning_rate": 0.0008416542521316182,
      "loss": 0.5852,
      "num_input_tokens_seen": 8338688,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 0.07166848331689835,
      "learning_rate": 0.0008414872461647156,
      "loss": 0.5659,
      "num_input_tokens_seen": 8342208,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 0.11974062770605087,
      "learning_rate": 0.0008413201687625772,
      "loss": 0.455,
      "num_input_tokens_seen": 8345088,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 0.1090938001871109,
      "learning_rate": 0.0008411530199601534,
      "loss": 0.4646,
      "num_input_tokens_seen": 8348192,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 0.12525059282779694,
      "learning_rate": 0.0008409857997924097,
      "loss": 0.4273,
      "num_input_tokens_seen": 8351488,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 0.15144643187522888,
      "learning_rate": 0.0008408185082943271,
      "loss": 0.4284,
      "num_input_tokens_seen": 8353856,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 0.08351729810237885,
      "learning_rate": 0.0008406511455009007,
      "loss": 0.3141,
      "num_input_tokens_seen": 8357568,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 0.11373569816350937,
      "learning_rate": 0.000840483711447141,
      "loss": 0.4492,
      "num_input_tokens_seen": 8359808,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 0.10861584544181824,
      "learning_rate": 0.0008403162061680736,
      "loss": 0.369,
      "num_input_tokens_seen": 8363424,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 0.09802230447530746,
      "learning_rate": 0.0008401486296987382,
      "loss": 0.4346,
      "num_input_tokens_seen": 8365696,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 0.04370921850204468,
      "learning_rate": 0.0008399809820741901,
      "loss": 0.4805,
      "num_input_tokens_seen": 8370240,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 0.04052837938070297,
      "learning_rate": 0.0008398132633294993,
      "loss": 0.4243,
      "num_input_tokens_seen": 8377792,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 0.13940191268920898,
      "learning_rate": 0.0008396454734997506,
      "loss": 0.4334,
      "num_input_tokens_seen": 8380448,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 0.08152610063552856,
      "learning_rate": 0.0008394776126200436,
      "loss": 0.3984,
      "num_input_tokens_seen": 8385248,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 0.1014631912112236,
      "learning_rate": 0.0008393096807254931,
      "loss": 0.3701,
      "num_input_tokens_seen": 8387968,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 0.10551749169826508,
      "learning_rate": 0.0008391416778512284,
      "loss": 0.4932,
      "num_input_tokens_seen": 8391008,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 0.15533380210399628,
      "learning_rate": 0.0008389736040323939,
      "loss": 0.5,
      "num_input_tokens_seen": 8395008,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 0.17736366391181946,
      "learning_rate": 0.0008388054593041485,
      "loss": 0.4612,
      "num_input_tokens_seen": 8397376,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 0.0916401743888855,
      "learning_rate": 0.0008386372437016665,
      "loss": 0.3939,
      "num_input_tokens_seen": 8400704,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 0.11000484973192215,
      "learning_rate": 0.0008384689572601364,
      "loss": 0.5059,
      "num_input_tokens_seen": 8404384,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 0.1498555690050125,
      "learning_rate": 0.000838300600014762,
      "loss": 0.5303,
      "num_input_tokens_seen": 8407072,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 0.06150829419493675,
      "learning_rate": 0.0008381321720007613,
      "loss": 0.4351,
      "num_input_tokens_seen": 8410816,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 0.09895196557044983,
      "learning_rate": 0.0008379636732533681,
      "loss": 0.4234,
      "num_input_tokens_seen": 8413632,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 0.1717565804719925,
      "learning_rate": 0.0008377951038078302,
      "loss": 0.4237,
      "num_input_tokens_seen": 8416864,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 0.20137131214141846,
      "learning_rate": 0.0008376264636994102,
      "loss": 0.5016,
      "num_input_tokens_seen": 8420800,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 0.07737871259450912,
      "learning_rate": 0.000837457752963386,
      "loss": 0.378,
      "num_input_tokens_seen": 8423520,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 0.09521850943565369,
      "learning_rate": 0.0008372889716350496,
      "loss": 0.5011,
      "num_input_tokens_seen": 8426496,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.08464586734771729,
      "learning_rate": 0.0008371201197497084,
      "loss": 0.4165,
      "num_input_tokens_seen": 8430048,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 0.1190241277217865,
      "learning_rate": 0.0008369511973426842,
      "loss": 0.5303,
      "num_input_tokens_seen": 8432960,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 0.05775552615523338,
      "learning_rate": 0.0008367822044493138,
      "loss": 0.438,
      "num_input_tokens_seen": 8436544,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 0.08474480360746384,
      "learning_rate": 0.0008366131411049483,
      "loss": 0.3973,
      "num_input_tokens_seen": 8439840,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 0.1370745599269867,
      "learning_rate": 0.000836444007344954,
      "loss": 0.3551,
      "num_input_tokens_seen": 8443104,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 0.12640473246574402,
      "learning_rate": 0.0008362748032047119,
      "loss": 0.4947,
      "num_input_tokens_seen": 8446528,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 0.16929012537002563,
      "learning_rate": 0.0008361055287196171,
      "loss": 0.3979,
      "num_input_tokens_seen": 8449024,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 0.06423585116863251,
      "learning_rate": 0.0008359361839250803,
      "loss": 0.453,
      "num_input_tokens_seen": 8453184,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 0.09028072655200958,
      "learning_rate": 0.0008357667688565265,
      "loss": 0.4247,
      "num_input_tokens_seen": 8455776,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 0.11568942666053772,
      "learning_rate": 0.0008355972835493951,
      "loss": 0.4501,
      "num_input_tokens_seen": 8458528,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 0.062219686806201935,
      "learning_rate": 0.0008354277280391408,
      "loss": 0.4506,
      "num_input_tokens_seen": 8462240,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 0.06295282393693924,
      "learning_rate": 0.0008352581023612324,
      "loss": 0.5828,
      "num_input_tokens_seen": 8465568,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.0989118367433548,
      "learning_rate": 0.0008350884065511541,
      "loss": 0.4564,
      "num_input_tokens_seen": 8468384,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 0.17807121574878693,
      "learning_rate": 0.0008349186406444038,
      "loss": 0.4886,
      "num_input_tokens_seen": 8471232,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 0.11610735952854156,
      "learning_rate": 0.0008347488046764949,
      "loss": 0.4386,
      "num_input_tokens_seen": 8474624,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 0.078083835542202,
      "learning_rate": 0.0008345788986829551,
      "loss": 0.4101,
      "num_input_tokens_seen": 8478240,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 0.08101742714643478,
      "learning_rate": 0.0008344089226993268,
      "loss": 0.3734,
      "num_input_tokens_seen": 8481408,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 0.10417918115854263,
      "learning_rate": 0.000834238876761167,
      "loss": 0.4287,
      "num_input_tokens_seen": 8484672,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 0.09860898554325104,
      "learning_rate": 0.0008340687609040473,
      "loss": 0.3833,
      "num_input_tokens_seen": 8488032,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 0.1957663744688034,
      "learning_rate": 0.0008338985751635542,
      "loss": 0.366,
      "num_input_tokens_seen": 8491648,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 0.14590609073638916,
      "learning_rate": 0.0008337283195752883,
      "loss": 0.5727,
      "num_input_tokens_seen": 8494336,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 0.141069233417511,
      "learning_rate": 0.0008335579941748654,
      "loss": 0.3979,
      "num_input_tokens_seen": 8497888,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 0.11004926264286041,
      "learning_rate": 0.0008333875989979156,
      "loss": 0.4999,
      "num_input_tokens_seen": 8500704,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 0.09282206743955612,
      "learning_rate": 0.0008332171340800835,
      "loss": 0.587,
      "num_input_tokens_seen": 8503872,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 0.1075824424624443,
      "learning_rate": 0.0008330465994570287,
      "loss": 0.5176,
      "num_input_tokens_seen": 8506816,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 0.08577787131071091,
      "learning_rate": 0.0008328759951644247,
      "loss": 0.5645,
      "num_input_tokens_seen": 8510272,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 0.07962977886199951,
      "learning_rate": 0.0008327053212379603,
      "loss": 0.5247,
      "num_input_tokens_seen": 8514784,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 0.0679922103881836,
      "learning_rate": 0.0008325345777133382,
      "loss": 0.4952,
      "num_input_tokens_seen": 8517728,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 0.13247841596603394,
      "learning_rate": 0.0008323637646262763,
      "loss": 0.4174,
      "num_input_tokens_seen": 8520608,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 0.11017640680074692,
      "learning_rate": 0.0008321928820125067,
      "loss": 0.5401,
      "num_input_tokens_seen": 8523360,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 0.041003357619047165,
      "learning_rate": 0.000832021929907776,
      "loss": 0.4711,
      "num_input_tokens_seen": 8526976,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 0.05711822584271431,
      "learning_rate": 0.0008318509083478454,
      "loss": 0.5583,
      "num_input_tokens_seen": 8530624,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 0.10485420376062393,
      "learning_rate": 0.0008316798173684907,
      "loss": 0.4108,
      "num_input_tokens_seen": 8535424,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 0.07721754163503647,
      "learning_rate": 0.0008315086570055024,
      "loss": 0.4234,
      "num_input_tokens_seen": 8538112,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 0.09115638583898544,
      "learning_rate": 0.000831337427294685,
      "loss": 0.6007,
      "num_input_tokens_seen": 8541696,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 0.08801106363534927,
      "learning_rate": 0.0008311661282718579,
      "loss": 0.4002,
      "num_input_tokens_seen": 8544832,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 0.09498363733291626,
      "learning_rate": 0.0008309947599728549,
      "loss": 0.471,
      "num_input_tokens_seen": 8547328,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 0.06289753317832947,
      "learning_rate": 0.0008308233224335243,
      "loss": 0.3947,
      "num_input_tokens_seen": 8550592,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 0.13135941326618195,
      "learning_rate": 0.0008306518156897289,
      "loss": 0.4411,
      "num_input_tokens_seen": 8553408,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 0.10954257100820541,
      "learning_rate": 0.0008304802397773459,
      "loss": 0.487,
      "num_input_tokens_seen": 8556768,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 0.06855899095535278,
      "learning_rate": 0.0008303085947322672,
      "loss": 0.4347,
      "num_input_tokens_seen": 8560416,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 0.07340344041585922,
      "learning_rate": 0.0008301368805903988,
      "loss": 0.4715,
      "num_input_tokens_seen": 8563424,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 0.09204958379268646,
      "learning_rate": 0.0008299650973876611,
      "loss": 0.3709,
      "num_input_tokens_seen": 8566560,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 0.1056034043431282,
      "learning_rate": 0.0008297932451599898,
      "loss": 0.3944,
      "num_input_tokens_seen": 8570080,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 0.1061817854642868,
      "learning_rate": 0.0008296213239433339,
      "loss": 0.5727,
      "num_input_tokens_seen": 8572704,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 0.10117878764867783,
      "learning_rate": 0.0008294493337736577,
      "loss": 0.549,
      "num_input_tokens_seen": 8575872,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 0.07485666871070862,
      "learning_rate": 0.0008292772746869392,
      "loss": 0.4289,
      "num_input_tokens_seen": 8579392,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 0.2763795852661133,
      "learning_rate": 0.0008291051467191717,
      "loss": 0.5037,
      "num_input_tokens_seen": 8582144,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 0.09131979197263718,
      "learning_rate": 0.000828932949906362,
      "loss": 0.3473,
      "num_input_tokens_seen": 8585184,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 0.23520687222480774,
      "learning_rate": 0.0008287606842845319,
      "loss": 0.7568,
      "num_input_tokens_seen": 8588128,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 0.09905411303043365,
      "learning_rate": 0.0008285883498897174,
      "loss": 0.4803,
      "num_input_tokens_seen": 8591008,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 0.15538696944713593,
      "learning_rate": 0.0008284159467579686,
      "loss": 0.614,
      "num_input_tokens_seen": 8594592,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.10679858177900314,
      "learning_rate": 0.0008282434749253507,
      "loss": 0.6831,
      "num_input_tokens_seen": 8598112,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 0.09123674780130386,
      "learning_rate": 0.0008280709344279426,
      "loss": 0.5119,
      "num_input_tokens_seen": 8602272,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 0.13341741263866425,
      "learning_rate": 0.0008278983253018377,
      "loss": 0.3943,
      "num_input_tokens_seen": 8605760,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 0.09430970251560211,
      "learning_rate": 0.0008277256475831443,
      "loss": 0.408,
      "num_input_tokens_seen": 8609056,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 0.08829453587532043,
      "learning_rate": 0.0008275529013079841,
      "loss": 0.541,
      "num_input_tokens_seen": 8612480,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 0.12748321890830994,
      "learning_rate": 0.0008273800865124941,
      "loss": 0.3311,
      "num_input_tokens_seen": 8615584,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 0.0709468200802803,
      "learning_rate": 0.0008272072032328246,
      "loss": 0.4079,
      "num_input_tokens_seen": 8618816,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 0.10073421895503998,
      "learning_rate": 0.0008270342515051414,
      "loss": 0.4603,
      "num_input_tokens_seen": 8622176,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 0.07289375364780426,
      "learning_rate": 0.0008268612313656238,
      "loss": 0.4483,
      "num_input_tokens_seen": 8625984,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 0.09890294075012207,
      "learning_rate": 0.0008266881428504655,
      "loss": 0.6036,
      "num_input_tokens_seen": 8629056,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 0.08092275261878967,
      "learning_rate": 0.000826514985995875,
      "loss": 0.4053,
      "num_input_tokens_seen": 8633312,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 0.11019667237997055,
      "learning_rate": 0.0008263417608380743,
      "loss": 0.5571,
      "num_input_tokens_seen": 8637376,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 0.11494884639978409,
      "learning_rate": 0.0008261684674133003,
      "loss": 0.4371,
      "num_input_tokens_seen": 8640544,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 0.08772212266921997,
      "learning_rate": 0.0008259951057578041,
      "loss": 0.5609,
      "num_input_tokens_seen": 8644032,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 0.0970267578959465,
      "learning_rate": 0.000825821675907851,
      "loss": 0.384,
      "num_input_tokens_seen": 8646432,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 0.16254323720932007,
      "learning_rate": 0.0008256481778997205,
      "loss": 0.4962,
      "num_input_tokens_seen": 8649472,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 0.10939892381429672,
      "learning_rate": 0.0008254746117697062,
      "loss": 0.4478,
      "num_input_tokens_seen": 8652864,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 0.07743914425373077,
      "learning_rate": 0.0008253009775541165,
      "loss": 0.4661,
      "num_input_tokens_seen": 8657024,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.08419858664274216,
      "learning_rate": 0.0008251272752892733,
      "loss": 0.4481,
      "num_input_tokens_seen": 8659872,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 0.1617799550294876,
      "learning_rate": 0.0008249535050115135,
      "loss": 0.4459,
      "num_input_tokens_seen": 8662528,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 0.13207276165485382,
      "learning_rate": 0.0008247796667571879,
      "loss": 0.457,
      "num_input_tokens_seen": 8665920,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 0.15814250707626343,
      "learning_rate": 0.000824605760562661,
      "loss": 0.5489,
      "num_input_tokens_seen": 8668480,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 0.15973912179470062,
      "learning_rate": 0.0008244317864643125,
      "loss": 0.4582,
      "num_input_tokens_seen": 8672224,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.08775033801794052,
      "learning_rate": 0.0008242577444985355,
      "loss": 0.4802,
      "num_input_tokens_seen": 8675296,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 0.09170124679803848,
      "learning_rate": 0.0008240836347017375,
      "loss": 0.4226,
      "num_input_tokens_seen": 8678272,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 0.21360301971435547,
      "learning_rate": 0.0008239094571103408,
      "loss": 0.575,
      "num_input_tokens_seen": 8681568,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 0.10359099507331848,
      "learning_rate": 0.0008237352117607808,
      "loss": 0.5061,
      "num_input_tokens_seen": 8684544,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 0.18841221928596497,
      "learning_rate": 0.000823560898689508,
      "loss": 0.4313,
      "num_input_tokens_seen": 8687072,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 0.11358564347028732,
      "learning_rate": 0.0008233865179329865,
      "loss": 0.4524,
      "num_input_tokens_seen": 8690496,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 0.07852023839950562,
      "learning_rate": 0.0008232120695276948,
      "loss": 0.3862,
      "num_input_tokens_seen": 8693376,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 0.10005715489387512,
      "learning_rate": 0.0008230375535101255,
      "loss": 0.5463,
      "num_input_tokens_seen": 8696384,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 0.13825277984142303,
      "learning_rate": 0.0008228629699167855,
      "loss": 0.5672,
      "num_input_tokens_seen": 8699264,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.1583670973777771,
      "learning_rate": 0.0008226883187841957,
      "loss": 0.4689,
      "num_input_tokens_seen": 8701760,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 0.13257169723510742,
      "learning_rate": 0.0008225136001488908,
      "loss": 0.3358,
      "num_input_tokens_seen": 8704352,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 0.21287378668785095,
      "learning_rate": 0.0008223388140474205,
      "loss": 0.4029,
      "num_input_tokens_seen": 8706560,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 0.07680748403072357,
      "learning_rate": 0.0008221639605163476,
      "loss": 0.4408,
      "num_input_tokens_seen": 8709408,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 0.08827900141477585,
      "learning_rate": 0.0008219890395922495,
      "loss": 0.4355,
      "num_input_tokens_seen": 8713440,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 0.16466861963272095,
      "learning_rate": 0.0008218140513117178,
      "loss": 0.4241,
      "num_input_tokens_seen": 8716352,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 0.1319967359304428,
      "learning_rate": 0.0008216389957113582,
      "loss": 0.4044,
      "num_input_tokens_seen": 8718944,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 0.20639945566654205,
      "learning_rate": 0.0008214638728277899,
      "loss": 0.425,
      "num_input_tokens_seen": 8722304,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.4845752716064453,
      "eval_runtime": 13.8095,
      "eval_samples_per_second": 61.407,
      "eval_steps_per_second": 15.352,
      "num_input_tokens_seen": 8722408,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 0.09615267813205719,
      "learning_rate": 0.0008212886826976469,
      "loss": 0.3979,
      "num_input_tokens_seen": 8725256,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 0.0814061164855957,
      "learning_rate": 0.0008211134253575769,
      "loss": 0.5,
      "num_input_tokens_seen": 8729704,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 0.09058725088834763,
      "learning_rate": 0.000820938100844242,
      "loss": 0.4194,
      "num_input_tokens_seen": 8734888,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 0.09735600650310516,
      "learning_rate": 0.0008207627091943177,
      "loss": 0.4624,
      "num_input_tokens_seen": 8738728,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 0.20223942399024963,
      "learning_rate": 0.000820587250444494,
      "loss": 0.4605,
      "num_input_tokens_seen": 8742248,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 0.1180163100361824,
      "learning_rate": 0.0008204117246314752,
      "loss": 0.454,
      "num_input_tokens_seen": 8745672,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 0.1274401694536209,
      "learning_rate": 0.0008202361317919788,
      "loss": 0.3518,
      "num_input_tokens_seen": 8749768,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 0.05591064319014549,
      "learning_rate": 0.0008200604719627373,
      "loss": 0.3888,
      "num_input_tokens_seen": 8753864,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 0.09386216849088669,
      "learning_rate": 0.0008198847451804964,
      "loss": 0.3908,
      "num_input_tokens_seen": 8757256,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 0.12459103763103485,
      "learning_rate": 0.0008197089514820162,
      "loss": 0.3507,
      "num_input_tokens_seen": 8760808,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 0.08667583763599396,
      "learning_rate": 0.0008195330909040708,
      "loss": 0.3022,
      "num_input_tokens_seen": 8765896,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 0.1457834094762802,
      "learning_rate": 0.0008193571634834483,
      "loss": 0.56,
      "num_input_tokens_seen": 8768488,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 0.11751123517751694,
      "learning_rate": 0.0008191811692569506,
      "loss": 0.3781,
      "num_input_tokens_seen": 8770664,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 0.07359790056943893,
      "learning_rate": 0.0008190051082613936,
      "loss": 0.4088,
      "num_input_tokens_seen": 8773768,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 0.15768449008464813,
      "learning_rate": 0.0008188289805336074,
      "loss": 0.5192,
      "num_input_tokens_seen": 8778280,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 0.14384059607982635,
      "learning_rate": 0.0008186527861104359,
      "loss": 0.5352,
      "num_input_tokens_seen": 8781256,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 0.11161612719297409,
      "learning_rate": 0.0008184765250287369,
      "loss": 0.387,
      "num_input_tokens_seen": 8783976,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 0.10871171206235886,
      "learning_rate": 0.0008183001973253823,
      "loss": 0.4082,
      "num_input_tokens_seen": 8786184,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 0.09417814761400223,
      "learning_rate": 0.0008181238030372576,
      "loss": 0.4156,
      "num_input_tokens_seen": 8789512,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 0.14746174216270447,
      "learning_rate": 0.0008179473422012628,
      "loss": 0.4668,
      "num_input_tokens_seen": 8792680,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 0.10368698835372925,
      "learning_rate": 0.0008177708148543114,
      "loss": 0.3294,
      "num_input_tokens_seen": 8796104,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.13730238378047943,
      "learning_rate": 0.0008175942210333307,
      "loss": 0.5598,
      "num_input_tokens_seen": 8799112,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 0.07247272878885269,
      "learning_rate": 0.0008174175607752626,
      "loss": 0.5034,
      "num_input_tokens_seen": 8803272,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 0.08374092727899551,
      "learning_rate": 0.0008172408341170617,
      "loss": 0.523,
      "num_input_tokens_seen": 8807304,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 0.12371095269918442,
      "learning_rate": 0.0008170640410956976,
      "loss": 0.4976,
      "num_input_tokens_seen": 8810024,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 0.09024769067764282,
      "learning_rate": 0.0008168871817481536,
      "loss": 0.5882,
      "num_input_tokens_seen": 8812936,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 0.09000518918037415,
      "learning_rate": 0.0008167102561114261,
      "loss": 0.4028,
      "num_input_tokens_seen": 8815912,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 0.11815245449542999,
      "learning_rate": 0.0008165332642225265,
      "loss": 0.3742,
      "num_input_tokens_seen": 8819528,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.15431149303913116,
      "learning_rate": 0.0008163562061184791,
      "loss": 0.4625,
      "num_input_tokens_seen": 8822152,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 0.12629765272140503,
      "learning_rate": 0.0008161790818363227,
      "loss": 0.422,
      "num_input_tokens_seen": 8825160,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 0.1261257380247116,
      "learning_rate": 0.0008160018914131094,
      "loss": 0.6775,
      "num_input_tokens_seen": 8828264,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 0.08787170052528381,
      "learning_rate": 0.0008158246348859057,
      "loss": 0.4619,
      "num_input_tokens_seen": 8831560,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 0.1179998517036438,
      "learning_rate": 0.0008156473122917913,
      "loss": 0.4193,
      "num_input_tokens_seen": 8834408,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 0.107111856341362,
      "learning_rate": 0.0008154699236678604,
      "loss": 0.4431,
      "num_input_tokens_seen": 8837448,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 0.13015522062778473,
      "learning_rate": 0.0008152924690512205,
      "loss": 0.3974,
      "num_input_tokens_seen": 8840392,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 0.19276826083660126,
      "learning_rate": 0.0008151149484789932,
      "loss": 0.3343,
      "num_input_tokens_seen": 8843816,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 0.1344432234764099,
      "learning_rate": 0.0008149373619883136,
      "loss": 0.6094,
      "num_input_tokens_seen": 8847368,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 0.08650229126214981,
      "learning_rate": 0.0008147597096163308,
      "loss": 0.5352,
      "num_input_tokens_seen": 8850536,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 0.09096166491508484,
      "learning_rate": 0.0008145819914002079,
      "loss": 0.4685,
      "num_input_tokens_seen": 8853672,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 0.15084119141101837,
      "learning_rate": 0.0008144042073771214,
      "loss": 0.449,
      "num_input_tokens_seen": 8856712,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 0.1252797245979309,
      "learning_rate": 0.0008142263575842615,
      "loss": 0.3542,
      "num_input_tokens_seen": 8859656,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 0.12871776521205902,
      "learning_rate": 0.0008140484420588323,
      "loss": 0.4207,
      "num_input_tokens_seen": 8863144,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 0.0953788012266159,
      "learning_rate": 0.000813870460838052,
      "loss": 0.4047,
      "num_input_tokens_seen": 8866216,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 0.15444818139076233,
      "learning_rate": 0.0008136924139591522,
      "loss": 0.5419,
      "num_input_tokens_seen": 8869640,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 0.11631589382886887,
      "learning_rate": 0.0008135143014593782,
      "loss": 0.4117,
      "num_input_tokens_seen": 8872392,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 0.13395777344703674,
      "learning_rate": 0.000813336123375989,
      "loss": 0.3064,
      "num_input_tokens_seen": 8874728,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 0.16462759673595428,
      "learning_rate": 0.0008131578797462575,
      "loss": 0.5211,
      "num_input_tokens_seen": 8877256,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 0.09808733314275742,
      "learning_rate": 0.0008129795706074703,
      "loss": 0.5817,
      "num_input_tokens_seen": 8880392,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 0.08477244526147842,
      "learning_rate": 0.0008128011959969277,
      "loss": 0.3781,
      "num_input_tokens_seen": 8883656,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 0.024838387966156006,
      "learning_rate": 0.0008126227559519434,
      "loss": 0.3675,
      "num_input_tokens_seen": 8889320,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 0.08938594162464142,
      "learning_rate": 0.0008124442505098452,
      "loss": 0.4786,
      "num_input_tokens_seen": 8892424,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 0.07409331947565079,
      "learning_rate": 0.0008122656797079744,
      "loss": 0.4225,
      "num_input_tokens_seen": 8895784,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 0.08978909254074097,
      "learning_rate": 0.0008120870435836858,
      "loss": 0.5608,
      "num_input_tokens_seen": 8899016,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 0.16463898122310638,
      "learning_rate": 0.0008119083421743481,
      "loss": 0.578,
      "num_input_tokens_seen": 8901224,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 0.08905558288097382,
      "learning_rate": 0.0008117295755173437,
      "loss": 0.4202,
      "num_input_tokens_seen": 8904392,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 0.10559272766113281,
      "learning_rate": 0.0008115507436500687,
      "loss": 0.4185,
      "num_input_tokens_seen": 8908264,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 0.23503732681274414,
      "learning_rate": 0.0008113718466099322,
      "loss": 0.5655,
      "num_input_tokens_seen": 8911208,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 0.12770695984363556,
      "learning_rate": 0.0008111928844343579,
      "loss": 0.4688,
      "num_input_tokens_seen": 8914792,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 0.09653659909963608,
      "learning_rate": 0.0008110138571607823,
      "loss": 0.417,
      "num_input_tokens_seen": 8917640,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 0.23515084385871887,
      "learning_rate": 0.0008108347648266563,
      "loss": 0.4516,
      "num_input_tokens_seen": 8920616,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 0.11607223749160767,
      "learning_rate": 0.0008106556074694434,
      "loss": 0.385,
      "num_input_tokens_seen": 8924360,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 0.10453499853610992,
      "learning_rate": 0.0008104763851266216,
      "loss": 0.4403,
      "num_input_tokens_seen": 8927752,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 0.06833043694496155,
      "learning_rate": 0.0008102970978356822,
      "loss": 0.3781,
      "num_input_tokens_seen": 8930568,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 0.14824944734573364,
      "learning_rate": 0.00081011774563413,
      "loss": 0.4839,
      "num_input_tokens_seen": 8933000,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.08078756183385849,
      "learning_rate": 0.0008099383285594835,
      "loss": 0.5104,
      "num_input_tokens_seen": 8936264,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 0.1449861079454422,
      "learning_rate": 0.0008097588466492746,
      "loss": 0.3367,
      "num_input_tokens_seen": 8939656,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 0.11332891136407852,
      "learning_rate": 0.0008095792999410487,
      "loss": 0.6069,
      "num_input_tokens_seen": 8942408,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 0.07995792478322983,
      "learning_rate": 0.0008093996884723653,
      "loss": 0.4029,
      "num_input_tokens_seen": 8946088,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 0.1022791787981987,
      "learning_rate": 0.0008092200122807969,
      "loss": 0.4336,
      "num_input_tokens_seen": 8948680,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 0.13452452421188354,
      "learning_rate": 0.0008090402714039295,
      "loss": 0.4216,
      "num_input_tokens_seen": 8951080,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 0.2171112596988678,
      "learning_rate": 0.0008088604658793632,
      "loss": 0.4107,
      "num_input_tokens_seen": 8954248,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 0.13244926929473877,
      "learning_rate": 0.0008086805957447111,
      "loss": 0.503,
      "num_input_tokens_seen": 8957000,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 0.08360367268323898,
      "learning_rate": 0.0008085006610376,
      "loss": 0.3878,
      "num_input_tokens_seen": 8959944,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 0.0812518447637558,
      "learning_rate": 0.0008083206617956702,
      "loss": 0.4422,
      "num_input_tokens_seen": 8962984,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 0.1980350911617279,
      "learning_rate": 0.0008081405980565755,
      "loss": 0.5427,
      "num_input_tokens_seen": 8965672,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 0.1655881255865097,
      "learning_rate": 0.0008079604698579829,
      "loss": 0.6026,
      "num_input_tokens_seen": 8968680,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 0.18828831613063812,
      "learning_rate": 0.0008077802772375736,
      "loss": 0.373,
      "num_input_tokens_seen": 8972136,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 0.2306460589170456,
      "learning_rate": 0.0008076000202330416,
      "loss": 0.5153,
      "num_input_tokens_seen": 8975688,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 0.07013064622879028,
      "learning_rate": 0.0008074196988820945,
      "loss": 0.335,
      "num_input_tokens_seen": 8978792,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 0.11342135816812515,
      "learning_rate": 0.0008072393132224539,
      "loss": 0.5204,
      "num_input_tokens_seen": 8981768,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 0.2083585113286972,
      "learning_rate": 0.0008070588632918541,
      "loss": 0.6547,
      "num_input_tokens_seen": 8984328,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 0.1144132912158966,
      "learning_rate": 0.000806878349128043,
      "loss": 0.4848,
      "num_input_tokens_seen": 8987016,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 0.08590017259120941,
      "learning_rate": 0.0008066977707687826,
      "loss": 0.5999,
      "num_input_tokens_seen": 8992264,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 0.1252715140581131,
      "learning_rate": 0.0008065171282518473,
      "loss": 0.5009,
      "num_input_tokens_seen": 8995368,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.10035571455955505,
      "learning_rate": 0.0008063364216150257,
      "loss": 0.4163,
      "num_input_tokens_seen": 8998120,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 0.1437293291091919,
      "learning_rate": 0.0008061556508961199,
      "loss": 0.376,
      "num_input_tokens_seen": 9001768,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 0.13015690445899963,
      "learning_rate": 0.0008059748161329443,
      "loss": 0.6137,
      "num_input_tokens_seen": 9004328,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 0.23719976842403412,
      "learning_rate": 0.0008057939173633282,
      "loss": 0.6264,
      "num_input_tokens_seen": 9007688,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 0.11213093250989914,
      "learning_rate": 0.0008056129546251132,
      "loss": 0.5459,
      "num_input_tokens_seen": 9011112,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 0.13461405038833618,
      "learning_rate": 0.0008054319279561546,
      "loss": 0.4775,
      "num_input_tokens_seen": 9013864,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 0.11657533794641495,
      "learning_rate": 0.0008052508373943214,
      "loss": 0.5129,
      "num_input_tokens_seen": 9017000,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 0.061104971915483475,
      "learning_rate": 0.0008050696829774954,
      "loss": 0.415,
      "num_input_tokens_seen": 9020744,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 0.17849107086658478,
      "learning_rate": 0.0008048884647435721,
      "loss": 0.5997,
      "num_input_tokens_seen": 9025064,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 0.18797241151332855,
      "learning_rate": 0.0008047071827304604,
      "loss": 0.3856,
      "num_input_tokens_seen": 9027688,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 0.24432174861431122,
      "learning_rate": 0.0008045258369760824,
      "loss": 0.5433,
      "num_input_tokens_seen": 9030472,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 0.08075349032878876,
      "learning_rate": 0.0008043444275183735,
      "loss": 0.3711,
      "num_input_tokens_seen": 9033992,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 0.10444798320531845,
      "learning_rate": 0.0008041629543952824,
      "loss": 0.5021,
      "num_input_tokens_seen": 9037000,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 0.08977051824331284,
      "learning_rate": 0.0008039814176447714,
      "loss": 0.4431,
      "num_input_tokens_seen": 9040232,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 0.11065960675477982,
      "learning_rate": 0.0008037998173048157,
      "loss": 0.4879,
      "num_input_tokens_seen": 9043112,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 0.17182742059230804,
      "learning_rate": 0.0008036181534134044,
      "loss": 0.4535,
      "num_input_tokens_seen": 9045672,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 0.18194563686847687,
      "learning_rate": 0.0008034364260085391,
      "loss": 0.4151,
      "num_input_tokens_seen": 9048072,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 0.11098221689462662,
      "learning_rate": 0.0008032546351282353,
      "loss": 0.5385,
      "num_input_tokens_seen": 9051080,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 0.13142499327659607,
      "learning_rate": 0.0008030727808105215,
      "loss": 0.5325,
      "num_input_tokens_seen": 9053864,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 0.08617260307073593,
      "learning_rate": 0.0008028908630934397,
      "loss": 0.3619,
      "num_input_tokens_seen": 9057544,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 0.10544094443321228,
      "learning_rate": 0.0008027088820150447,
      "loss": 0.4853,
      "num_input_tokens_seen": 9060776,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 0.056105900555849075,
      "learning_rate": 0.0008025268376134054,
      "loss": 0.3885,
      "num_input_tokens_seen": 9064648,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 0.12049004435539246,
      "learning_rate": 0.0008023447299266027,
      "loss": 0.387,
      "num_input_tokens_seen": 9067688,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 0.11130666732788086,
      "learning_rate": 0.0008021625589927321,
      "loss": 0.5866,
      "num_input_tokens_seen": 9071368,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 0.1365925520658493,
      "learning_rate": 0.0008019803248499013,
      "loss": 0.4685,
      "num_input_tokens_seen": 9074120,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 0.05293969064950943,
      "learning_rate": 0.0008017980275362318,
      "loss": 0.3884,
      "num_input_tokens_seen": 9077896,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 0.1563268005847931,
      "learning_rate": 0.000801615667089858,
      "loss": 0.4161,
      "num_input_tokens_seen": 9080968,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 0.09555726498365402,
      "learning_rate": 0.0008014332435489275,
      "loss": 0.3435,
      "num_input_tokens_seen": 9084456,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 0.14009718596935272,
      "learning_rate": 0.0008012507569516016,
      "loss": 0.452,
      "num_input_tokens_seen": 9086856,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 0.14296157658100128,
      "learning_rate": 0.0008010682073360541,
      "loss": 0.5027,
      "num_input_tokens_seen": 9089768,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 0.13279949128627777,
      "learning_rate": 0.0008008855947404724,
      "loss": 0.3939,
      "num_input_tokens_seen": 9092424,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 0.12304141372442245,
      "learning_rate": 0.000800702919203057,
      "loss": 0.668,
      "num_input_tokens_seen": 9095304,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 0.10582614690065384,
      "learning_rate": 0.0008005201807620214,
      "loss": 0.5238,
      "num_input_tokens_seen": 9098248,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 0.10713765770196915,
      "learning_rate": 0.0008003373794555926,
      "loss": 0.5002,
      "num_input_tokens_seen": 9101768,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 0.06578894704580307,
      "learning_rate": 0.0008001545153220104,
      "loss": 0.3494,
      "num_input_tokens_seen": 9105128,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 0.10596881806850433,
      "learning_rate": 0.000799971588399528,
      "loss": 0.3947,
      "num_input_tokens_seen": 9108936,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 0.09680864214897156,
      "learning_rate": 0.0007997885987264115,
      "loss": 0.4747,
      "num_input_tokens_seen": 9111784,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 0.1679941713809967,
      "learning_rate": 0.0007996055463409403,
      "loss": 0.4762,
      "num_input_tokens_seen": 9115464,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 0.1540524661540985,
      "learning_rate": 0.000799422431281407,
      "loss": 0.4112,
      "num_input_tokens_seen": 9118280,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 0.08700738102197647,
      "learning_rate": 0.000799239253586117,
      "loss": 0.4459,
      "num_input_tokens_seen": 9122728,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 0.10293945670127869,
      "learning_rate": 0.0007990560132933891,
      "loss": 0.3985,
      "num_input_tokens_seen": 9126856,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 0.12459903210401535,
      "learning_rate": 0.0007988727104415549,
      "loss": 0.3756,
      "num_input_tokens_seen": 9129864,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 0.0898512601852417,
      "learning_rate": 0.0007986893450689594,
      "loss": 0.4402,
      "num_input_tokens_seen": 9133064,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 0.11939819902181625,
      "learning_rate": 0.0007985059172139606,
      "loss": 0.3877,
      "num_input_tokens_seen": 9136264,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.2239948958158493,
      "learning_rate": 0.0007983224269149296,
      "loss": 0.4467,
      "num_input_tokens_seen": 9139016,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 0.08654364943504333,
      "learning_rate": 0.00079813887421025,
      "loss": 0.5598,
      "num_input_tokens_seen": 9141704,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 0.9091293215751648,
      "learning_rate": 0.0007979552591383195,
      "loss": 0.3992,
      "num_input_tokens_seen": 9144616,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 0.08085007965564728,
      "learning_rate": 0.0007977715817375481,
      "loss": 0.4889,
      "num_input_tokens_seen": 9148456,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 0.16664238274097443,
      "learning_rate": 0.0007975878420463588,
      "loss": 0.5465,
      "num_input_tokens_seen": 9151016,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 0.1269255131483078,
      "learning_rate": 0.0007974040401031882,
      "loss": 0.5864,
      "num_input_tokens_seen": 9155016,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 0.13436661660671234,
      "learning_rate": 0.0007972201759464851,
      "loss": 0.454,
      "num_input_tokens_seen": 9157800,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 0.10495325922966003,
      "learning_rate": 0.0007970362496147121,
      "loss": 0.4266,
      "num_input_tokens_seen": 9161736,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.11760729551315308,
      "learning_rate": 0.0007968522611463447,
      "loss": 0.4099,
      "num_input_tokens_seen": 9164872,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 0.09800946712493896,
      "learning_rate": 0.0007966682105798708,
      "loss": 0.4867,
      "num_input_tokens_seen": 9168712,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 0.10470658540725708,
      "learning_rate": 0.0007964840979537918,
      "loss": 0.4604,
      "num_input_tokens_seen": 9171592,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 0.1196318194270134,
      "learning_rate": 0.0007962999233066219,
      "loss": 0.4576,
      "num_input_tokens_seen": 9175688,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 0.10377860069274902,
      "learning_rate": 0.0007961156866768885,
      "loss": 0.3578,
      "num_input_tokens_seen": 9180232,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 0.15669795870780945,
      "learning_rate": 0.0007959313881031317,
      "loss": 0.5708,
      "num_input_tokens_seen": 9183304,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 0.31148210167884827,
      "learning_rate": 0.0007957470276239048,
      "loss": 0.4826,
      "num_input_tokens_seen": 9185832,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 0.10414770245552063,
      "learning_rate": 0.0007955626052777735,
      "loss": 0.4624,
      "num_input_tokens_seen": 9188872,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 0.16088256239891052,
      "learning_rate": 0.0007953781211033173,
      "loss": 0.4817,
      "num_input_tokens_seen": 9192072,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 0.10064107924699783,
      "learning_rate": 0.000795193575139128,
      "loss": 0.3604,
      "num_input_tokens_seen": 9195848,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 0.12445864826440811,
      "learning_rate": 0.0007950089674238106,
      "loss": 0.5585,
      "num_input_tokens_seen": 9198472,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 0.09773962944746017,
      "learning_rate": 0.0007948242979959828,
      "loss": 0.4458,
      "num_input_tokens_seen": 9201576,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 0.14106951653957367,
      "learning_rate": 0.0007946395668942754,
      "loss": 0.4058,
      "num_input_tokens_seen": 9204712,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 0.08995738625526428,
      "learning_rate": 0.0007944547741573319,
      "loss": 0.4595,
      "num_input_tokens_seen": 9207816,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 0.0766063928604126,
      "learning_rate": 0.0007942699198238091,
      "loss": 0.4377,
      "num_input_tokens_seen": 9211432,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 0.12539944052696228,
      "learning_rate": 0.0007940850039323763,
      "loss": 0.4214,
      "num_input_tokens_seen": 9214248,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 0.13901478052139282,
      "learning_rate": 0.0007939000265217156,
      "loss": 0.4728,
      "num_input_tokens_seen": 9216904,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 0.1259300410747528,
      "learning_rate": 0.0007937149876305226,
      "loss": 0.446,
      "num_input_tokens_seen": 9219880,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 0.09369131177663803,
      "learning_rate": 0.000793529887297505,
      "loss": 0.5436,
      "num_input_tokens_seen": 9224808,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 0.09675969928503036,
      "learning_rate": 0.0007933447255613835,
      "loss": 0.5174,
      "num_input_tokens_seen": 9228552,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 0.08054115623235703,
      "learning_rate": 0.0007931595024608924,
      "loss": 0.3931,
      "num_input_tokens_seen": 9232680,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 0.08340208232402802,
      "learning_rate": 0.0007929742180347776,
      "loss": 0.4198,
      "num_input_tokens_seen": 9235400,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 0.11381560564041138,
      "learning_rate": 0.0007927888723217991,
      "loss": 0.4846,
      "num_input_tokens_seen": 9238056,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 0.09461326897144318,
      "learning_rate": 0.0007926034653607288,
      "loss": 0.3454,
      "num_input_tokens_seen": 9241288,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 0.10095126926898956,
      "learning_rate": 0.0007924179971903516,
      "loss": 0.5136,
      "num_input_tokens_seen": 9244872,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 0.12522725760936737,
      "learning_rate": 0.0007922324678494655,
      "loss": 0.4677,
      "num_input_tokens_seen": 9247784,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 0.10546691715717316,
      "learning_rate": 0.0007920468773768811,
      "loss": 0.425,
      "num_input_tokens_seen": 9252904,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 0.16400575637817383,
      "learning_rate": 0.0007918612258114217,
      "loss": 0.395,
      "num_input_tokens_seen": 9255944,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 0.14500831067562103,
      "learning_rate": 0.0007916755131919238,
      "loss": 0.4965,
      "num_input_tokens_seen": 9258376,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 0.10448005795478821,
      "learning_rate": 0.000791489739557236,
      "loss": 0.4408,
      "num_input_tokens_seen": 9261640,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 0.13146427273750305,
      "learning_rate": 0.0007913039049462203,
      "loss": 0.506,
      "num_input_tokens_seen": 9264712,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 0.10834785550832748,
      "learning_rate": 0.0007911180093977511,
      "loss": 0.425,
      "num_input_tokens_seen": 9268936,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 0.20315396785736084,
      "learning_rate": 0.0007909320529507154,
      "loss": 0.6974,
      "num_input_tokens_seen": 9272328,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 0.14895367622375488,
      "learning_rate": 0.0007907460356440134,
      "loss": 0.3399,
      "num_input_tokens_seen": 9275592,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 0.10042289644479752,
      "learning_rate": 0.0007905599575165577,
      "loss": 0.4793,
      "num_input_tokens_seen": 9279336,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 0.10882191359996796,
      "learning_rate": 0.0007903738186072739,
      "loss": 0.3435,
      "num_input_tokens_seen": 9282280,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 0.11404050141572952,
      "learning_rate": 0.0007901876189550999,
      "loss": 0.457,
      "num_input_tokens_seen": 9285960,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 0.11478503048419952,
      "learning_rate": 0.0007900013585989867,
      "loss": 0.3265,
      "num_input_tokens_seen": 9291464,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 0.10619615018367767,
      "learning_rate": 0.0007898150375778979,
      "loss": 0.4238,
      "num_input_tokens_seen": 9295432,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 0.05256584659218788,
      "learning_rate": 0.0007896286559308095,
      "loss": 0.4215,
      "num_input_tokens_seen": 9298600,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 0.08313100039958954,
      "learning_rate": 0.0007894422136967105,
      "loss": 0.3796,
      "num_input_tokens_seen": 9301416,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 0.195240780711174,
      "learning_rate": 0.0007892557109146026,
      "loss": 0.4368,
      "num_input_tokens_seen": 9304328,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 0.08619768917560577,
      "learning_rate": 0.0007890691476234999,
      "loss": 0.3973,
      "num_input_tokens_seen": 9307016,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 0.07236132770776749,
      "learning_rate": 0.0007888825238624294,
      "loss": 0.4734,
      "num_input_tokens_seen": 9310504,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 0.08458784967660904,
      "learning_rate": 0.0007886958396704307,
      "loss": 0.4044,
      "num_input_tokens_seen": 9313512,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 0.09885953366756439,
      "learning_rate": 0.0007885090950865559,
      "loss": 0.4804,
      "num_input_tokens_seen": 9317128,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 0.12995973229408264,
      "learning_rate": 0.0007883222901498701,
      "loss": 0.7895,
      "num_input_tokens_seen": 9319560,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 0.08772021532058716,
      "learning_rate": 0.0007881354248994503,
      "loss": 0.4505,
      "num_input_tokens_seen": 9322504,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 0.09922874718904495,
      "learning_rate": 0.0007879484993743869,
      "loss": 0.4168,
      "num_input_tokens_seen": 9325992,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 0.11061937361955643,
      "learning_rate": 0.0007877615136137827,
      "loss": 0.4474,
      "num_input_tokens_seen": 9328520,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 0.14413869380950928,
      "learning_rate": 0.0007875744676567527,
      "loss": 0.4135,
      "num_input_tokens_seen": 9330760,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 0.11818405985832214,
      "learning_rate": 0.0007873873615424248,
      "loss": 0.5899,
      "num_input_tokens_seen": 9333256,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 0.12080354243516922,
      "learning_rate": 0.0007872001953099396,
      "loss": 0.5415,
      "num_input_tokens_seen": 9337352,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 0.15494631230831146,
      "learning_rate": 0.0007870129689984501,
      "loss": 0.5921,
      "num_input_tokens_seen": 9340104,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.1065947487950325,
      "learning_rate": 0.000786825682647122,
      "loss": 0.4686,
      "num_input_tokens_seen": 9342888,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 0.150186225771904,
      "learning_rate": 0.0007866383362951332,
      "loss": 0.4156,
      "num_input_tokens_seen": 9345320,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 0.09178438037633896,
      "learning_rate": 0.0007864509299816746,
      "loss": 0.4879,
      "num_input_tokens_seen": 9348648,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 0.31308335065841675,
      "learning_rate": 0.0007862634637459496,
      "loss": 0.5179,
      "num_input_tokens_seen": 9351112,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 0.1974375694990158,
      "learning_rate": 0.0007860759376271737,
      "loss": 0.5272,
      "num_input_tokens_seen": 9354280,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 0.09613283723592758,
      "learning_rate": 0.0007858883516645755,
      "loss": 0.5919,
      "num_input_tokens_seen": 9357224,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 0.18595902621746063,
      "learning_rate": 0.0007857007058973957,
      "loss": 0.4844,
      "num_input_tokens_seen": 9359592,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 0.07936245948076248,
      "learning_rate": 0.0007855130003648876,
      "loss": 0.4105,
      "num_input_tokens_seen": 9363400,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 0.08266755938529968,
      "learning_rate": 0.0007853252351063171,
      "loss": 0.3673,
      "num_input_tokens_seen": 9365832,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 0.10920806974172592,
      "learning_rate": 0.0007851374101609627,
      "loss": 0.4456,
      "num_input_tokens_seen": 9369256,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 0.10467202216386795,
      "learning_rate": 0.000784949525568115,
      "loss": 0.5105,
      "num_input_tokens_seen": 9372392,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 0.12456003576517105,
      "learning_rate": 0.0007847615813670776,
      "loss": 0.5253,
      "num_input_tokens_seen": 9375752,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 0.10984178632497787,
      "learning_rate": 0.000784573577597166,
      "loss": 0.4187,
      "num_input_tokens_seen": 9379816,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 0.08648858964443207,
      "learning_rate": 0.0007843855142977086,
      "loss": 0.5662,
      "num_input_tokens_seen": 9383976,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 0.13563162088394165,
      "learning_rate": 0.000784197391508046,
      "loss": 0.5117,
      "num_input_tokens_seen": 9387400,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 0.15084296464920044,
      "learning_rate": 0.0007840092092675313,
      "loss": 0.4547,
      "num_input_tokens_seen": 9390216,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 0.11422592401504517,
      "learning_rate": 0.0007838209676155302,
      "loss": 0.4336,
      "num_input_tokens_seen": 9393960,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 0.11184369027614594,
      "learning_rate": 0.0007836326665914209,
      "loss": 0.3266,
      "num_input_tokens_seen": 9397416,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.12303527444601059,
      "learning_rate": 0.0007834443062345932,
      "loss": 0.4449,
      "num_input_tokens_seen": 9400232,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 0.23327027261257172,
      "learning_rate": 0.0007832558865844507,
      "loss": 0.456,
      "num_input_tokens_seen": 9402568,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 0.11514489352703094,
      "learning_rate": 0.0007830674076804083,
      "loss": 0.434,
      "num_input_tokens_seen": 9404904,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 0.15515531599521637,
      "learning_rate": 0.0007828788695618934,
      "loss": 0.4495,
      "num_input_tokens_seen": 9407720,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 0.14975208044052124,
      "learning_rate": 0.0007826902722683462,
      "loss": 0.5566,
      "num_input_tokens_seen": 9410280,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 0.08402874320745468,
      "learning_rate": 0.0007825016158392193,
      "loss": 0.4576,
      "num_input_tokens_seen": 9413032,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 0.08616756647825241,
      "learning_rate": 0.0007823129003139773,
      "loss": 0.4294,
      "num_input_tokens_seen": 9416488,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 0.11597710102796555,
      "learning_rate": 0.0007821241257320972,
      "loss": 0.4144,
      "num_input_tokens_seen": 9419720,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 0.1407928168773651,
      "learning_rate": 0.0007819352921330689,
      "loss": 0.5558,
      "num_input_tokens_seen": 9422824,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 0.11431095004081726,
      "learning_rate": 0.0007817463995563938,
      "loss": 0.4359,
      "num_input_tokens_seen": 9426248,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 0.1733442097902298,
      "learning_rate": 0.0007815574480415864,
      "loss": 0.4692,
      "num_input_tokens_seen": 9429032,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 0.12048695981502533,
      "learning_rate": 0.0007813684376281729,
      "loss": 0.4284,
      "num_input_tokens_seen": 9432520,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 0.1462780088186264,
      "learning_rate": 0.0007811793683556922,
      "loss": 0.5969,
      "num_input_tokens_seen": 9435592,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 0.07652163505554199,
      "learning_rate": 0.0007809902402636957,
      "loss": 0.6081,
      "num_input_tokens_seen": 9439784,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 0.17687734961509705,
      "learning_rate": 0.0007808010533917464,
      "loss": 0.4261,
      "num_input_tokens_seen": 9442408,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 0.13440951704978943,
      "learning_rate": 0.0007806118077794205,
      "loss": 0.418,
      "num_input_tokens_seen": 9445096,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 0.12521709501743317,
      "learning_rate": 0.0007804225034663058,
      "loss": 0.4032,
      "num_input_tokens_seen": 9449224,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 0.11764611303806305,
      "learning_rate": 0.0007802331404920024,
      "loss": 0.446,
      "num_input_tokens_seen": 9455784,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 0.06358097493648529,
      "learning_rate": 0.0007800437188961232,
      "loss": 0.424,
      "num_input_tokens_seen": 9461704,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 0.0716201514005661,
      "learning_rate": 0.0007798542387182929,
      "loss": 0.3536,
      "num_input_tokens_seen": 9464744,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 0.12451714277267456,
      "learning_rate": 0.0007796646999981488,
      "loss": 0.5024,
      "num_input_tokens_seen": 9468168,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 0.06993639469146729,
      "learning_rate": 0.0007794751027753397,
      "loss": 0.4426,
      "num_input_tokens_seen": 9470600,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 0.12085995823144913,
      "learning_rate": 0.0007792854470895278,
      "loss": 0.4476,
      "num_input_tokens_seen": 9473608,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 0.31463316082954407,
      "learning_rate": 0.0007790957329803865,
      "loss": 0.4809,
      "num_input_tokens_seen": 9475976,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 0.0988176241517067,
      "learning_rate": 0.0007789059604876019,
      "loss": 0.4107,
      "num_input_tokens_seen": 9479624,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 0.1269955188035965,
      "learning_rate": 0.0007787161296508724,
      "loss": 0.5807,
      "num_input_tokens_seen": 9482216,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 0.12786924839019775,
      "learning_rate": 0.0007785262405099083,
      "loss": 0.4659,
      "num_input_tokens_seen": 9484744,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 0.08929409831762314,
      "learning_rate": 0.0007783362931044322,
      "loss": 0.3513,
      "num_input_tokens_seen": 9487528,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 0.15610164403915405,
      "learning_rate": 0.0007781462874741793,
      "loss": 0.3818,
      "num_input_tokens_seen": 9490856,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 0.11476784944534302,
      "learning_rate": 0.000777956223658896,
      "loss": 0.4826,
      "num_input_tokens_seen": 9493832,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 0.09843695908784866,
      "learning_rate": 0.000777766101698342,
      "loss": 0.3662,
      "num_input_tokens_seen": 9496872,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 0.09369064122438431,
      "learning_rate": 0.0007775759216322882,
      "loss": 0.5816,
      "num_input_tokens_seen": 9500072,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 0.1207735687494278,
      "learning_rate": 0.0007773856835005187,
      "loss": 0.4558,
      "num_input_tokens_seen": 9502792,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 0.16702033579349518,
      "learning_rate": 0.0007771953873428285,
      "loss": 0.3356,
      "num_input_tokens_seen": 9505320,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 0.12104016542434692,
      "learning_rate": 0.0007770050331990259,
      "loss": 0.429,
      "num_input_tokens_seen": 9508232,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 0.05642099305987358,
      "learning_rate": 0.0007768146211089304,
      "loss": 0.4504,
      "num_input_tokens_seen": 9511304,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 0.09083728492259979,
      "learning_rate": 0.0007766241511123744,
      "loss": 0.519,
      "num_input_tokens_seen": 9514952,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 0.06679273396730423,
      "learning_rate": 0.0007764336232492018,
      "loss": 0.404,
      "num_input_tokens_seen": 9519272,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 0.08562625199556351,
      "learning_rate": 0.0007762430375592689,
      "loss": 0.494,
      "num_input_tokens_seen": 9523112,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 0.1307375431060791,
      "learning_rate": 0.0007760523940824441,
      "loss": 0.4281,
      "num_input_tokens_seen": 9526792,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 0.07490961253643036,
      "learning_rate": 0.0007758616928586077,
      "loss": 0.5249,
      "num_input_tokens_seen": 9529736,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 0.10060933232307434,
      "learning_rate": 0.0007756709339276527,
      "loss": 0.4032,
      "num_input_tokens_seen": 9532968,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 0.12907744944095612,
      "learning_rate": 0.0007754801173294831,
      "loss": 0.4101,
      "num_input_tokens_seen": 9537320,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 0.16829709708690643,
      "learning_rate": 0.0007752892431040158,
      "loss": 0.5201,
      "num_input_tokens_seen": 9540680,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 0.09517624974250793,
      "learning_rate": 0.0007750983112911796,
      "loss": 0.6766,
      "num_input_tokens_seen": 9543944,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.09744280576705933,
      "learning_rate": 0.0007749073219309151,
      "loss": 0.3875,
      "num_input_tokens_seen": 9546728,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 0.09857358783483505,
      "learning_rate": 0.0007747162750631751,
      "loss": 0.4825,
      "num_input_tokens_seen": 9550280,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 0.2089889496564865,
      "learning_rate": 0.0007745251707279246,
      "loss": 0.5362,
      "num_input_tokens_seen": 9552488,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 0.10288634151220322,
      "learning_rate": 0.0007743340089651403,
      "loss": 0.366,
      "num_input_tokens_seen": 9555304,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 0.1050984337925911,
      "learning_rate": 0.0007741427898148111,
      "loss": 0.5468,
      "num_input_tokens_seen": 9560008,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 0.11563657224178314,
      "learning_rate": 0.0007739515133169379,
      "loss": 0.3873,
      "num_input_tokens_seen": 9562792,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 0.12055163830518723,
      "learning_rate": 0.0007737601795115334,
      "loss": 0.3965,
      "num_input_tokens_seen": 9565416,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 0.1408175677061081,
      "learning_rate": 0.0007735687884386226,
      "loss": 0.4329,
      "num_input_tokens_seen": 9568296,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 0.10857395082712173,
      "learning_rate": 0.0007733773401382424,
      "loss": 0.5181,
      "num_input_tokens_seen": 9571624,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 0.11854881048202515,
      "learning_rate": 0.0007731858346504414,
      "loss": 0.5434,
      "num_input_tokens_seen": 9574728,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 0.08989092707633972,
      "learning_rate": 0.0007729942720152805,
      "loss": 0.3496,
      "num_input_tokens_seen": 9577192,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 0.16719357669353485,
      "learning_rate": 0.0007728026522728324,
      "loss": 0.404,
      "num_input_tokens_seen": 9581928,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 0.1703069806098938,
      "learning_rate": 0.0007726109754631817,
      "loss": 0.7131,
      "num_input_tokens_seen": 9585096,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 0.151450514793396,
      "learning_rate": 0.000772419241626425,
      "loss": 0.4912,
      "num_input_tokens_seen": 9588360,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 0.19992820918560028,
      "learning_rate": 0.000772227450802671,
      "loss": 0.503,
      "num_input_tokens_seen": 9591304,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 0.11859012395143509,
      "learning_rate": 0.0007720356030320399,
      "loss": 0.4175,
      "num_input_tokens_seen": 9593704,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 0.06724441796541214,
      "learning_rate": 0.0007718436983546642,
      "loss": 0.6132,
      "num_input_tokens_seen": 9597416,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 0.07968366146087646,
      "learning_rate": 0.0007716517368106882,
      "loss": 0.4834,
      "num_input_tokens_seen": 9600264,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 0.07503511011600494,
      "learning_rate": 0.000771459718440268,
      "loss": 0.5582,
      "num_input_tokens_seen": 9603880,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 0.13851772248744965,
      "learning_rate": 0.0007712676432835717,
      "loss": 0.5076,
      "num_input_tokens_seen": 9607144,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.08049949258565903,
      "learning_rate": 0.0007710755113807794,
      "loss": 0.392,
      "num_input_tokens_seen": 9610312,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 0.0995284840464592,
      "learning_rate": 0.0007708833227720824,
      "loss": 0.4196,
      "num_input_tokens_seen": 9613224,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 0.09048798680305481,
      "learning_rate": 0.0007706910774976848,
      "loss": 0.5381,
      "num_input_tokens_seen": 9616456,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.08836483955383301,
      "learning_rate": 0.0007704987755978021,
      "loss": 0.5206,
      "num_input_tokens_seen": 9620328,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 0.1729525327682495,
      "learning_rate": 0.0007703064171126615,
      "loss": 0.4048,
      "num_input_tokens_seen": 9623016,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 0.2271309494972229,
      "learning_rate": 0.0007701140020825022,
      "loss": 0.5056,
      "num_input_tokens_seen": 9625800,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 0.10265235602855682,
      "learning_rate": 0.0007699215305475753,
      "loss": 0.6275,
      "num_input_tokens_seen": 9629032,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.14519953727722168,
      "learning_rate": 0.0007697290025481436,
      "loss": 0.4863,
      "num_input_tokens_seen": 9633160,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 0.10326956957578659,
      "learning_rate": 0.0007695364181244819,
      "loss": 0.565,
      "num_input_tokens_seen": 9637160,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 0.26198574900627136,
      "learning_rate": 0.0007693437773168764,
      "loss": 0.5996,
      "num_input_tokens_seen": 9640264,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 0.09827445447444916,
      "learning_rate": 0.0007691510801656256,
      "loss": 0.6013,
      "num_input_tokens_seen": 9643816,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 0.11381252110004425,
      "learning_rate": 0.0007689583267110395,
      "loss": 0.3735,
      "num_input_tokens_seen": 9647112,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 0.09232404083013535,
      "learning_rate": 0.0007687655169934398,
      "loss": 0.3795,
      "num_input_tokens_seen": 9650248,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 0.17015598714351654,
      "learning_rate": 0.0007685726510531603,
      "loss": 0.5907,
      "num_input_tokens_seen": 9652680,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 0.0840507224202156,
      "learning_rate": 0.0007683797289305463,
      "loss": 0.3552,
      "num_input_tokens_seen": 9655976,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 0.15885503590106964,
      "learning_rate": 0.0007681867506659548,
      "loss": 0.5961,
      "num_input_tokens_seen": 9658664,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 0.11426253616809845,
      "learning_rate": 0.0007679937162997546,
      "loss": 0.5973,
      "num_input_tokens_seen": 9661736,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 0.082461878657341,
      "learning_rate": 0.0007678006258723264,
      "loss": 0.3508,
      "num_input_tokens_seen": 9664904,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 0.1566765159368515,
      "learning_rate": 0.0007676074794240626,
      "loss": 0.4612,
      "num_input_tokens_seen": 9667816,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 0.10164382308721542,
      "learning_rate": 0.000767414276995367,
      "loss": 0.3764,
      "num_input_tokens_seen": 9671208,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 0.09810273349285126,
      "learning_rate": 0.0007672210186266555,
      "loss": 0.3575,
      "num_input_tokens_seen": 9673864,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 0.1477145403623581,
      "learning_rate": 0.0007670277043583556,
      "loss": 0.4257,
      "num_input_tokens_seen": 9677768,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 0.12053689360618591,
      "learning_rate": 0.0007668343342309063,
      "loss": 0.4373,
      "num_input_tokens_seen": 9681640,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.12462883442640305,
      "learning_rate": 0.0007666409082847586,
      "loss": 0.3746,
      "num_input_tokens_seen": 9685192,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 0.11312952637672424,
      "learning_rate": 0.0007664474265603747,
      "loss": 0.4786,
      "num_input_tokens_seen": 9688392,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 0.08503203094005585,
      "learning_rate": 0.0007662538890982291,
      "loss": 0.4205,
      "num_input_tokens_seen": 9691528,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 0.1512094885110855,
      "learning_rate": 0.0007660602959388075,
      "loss": 0.3919,
      "num_input_tokens_seen": 9694248,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 0.08120337873697281,
      "learning_rate": 0.0007658666471226073,
      "loss": 0.5012,
      "num_input_tokens_seen": 9697288,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 0.1480705738067627,
      "learning_rate": 0.0007656729426901377,
      "loss": 0.6352,
      "num_input_tokens_seen": 9701000,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 0.17511500418186188,
      "learning_rate": 0.0007654791826819194,
      "loss": 0.4089,
      "num_input_tokens_seen": 9704008,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 0.1279667168855667,
      "learning_rate": 0.0007652853671384847,
      "loss": 0.3573,
      "num_input_tokens_seen": 9707336,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 0.17718490958213806,
      "learning_rate": 0.0007650914961003781,
      "loss": 0.2923,
      "num_input_tokens_seen": 9709480,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 0.10648729652166367,
      "learning_rate": 0.0007648975696081546,
      "loss": 0.5359,
      "num_input_tokens_seen": 9712936,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 0.08483392000198364,
      "learning_rate": 0.0007647035877023816,
      "loss": 0.4165,
      "num_input_tokens_seen": 9715432,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 0.19000816345214844,
      "learning_rate": 0.0007645095504236381,
      "loss": 0.5161,
      "num_input_tokens_seen": 9718280,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 0.13634027540683746,
      "learning_rate": 0.0007643154578125142,
      "loss": 0.3823,
      "num_input_tokens_seen": 9721736,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 0.2029729038476944,
      "learning_rate": 0.0007641213099096121,
      "loss": 0.5444,
      "num_input_tokens_seen": 9724616,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 0.2172134965658188,
      "learning_rate": 0.0007639271067555452,
      "loss": 0.5207,
      "num_input_tokens_seen": 9728648,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 0.08500788360834122,
      "learning_rate": 0.0007637328483909385,
      "loss": 0.6143,
      "num_input_tokens_seen": 9732168,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 0.07936105877161026,
      "learning_rate": 0.000763538534856429,
      "loss": 0.4861,
      "num_input_tokens_seen": 9735432,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 0.08209620416164398,
      "learning_rate": 0.0007633441661926643,
      "loss": 0.5795,
      "num_input_tokens_seen": 9738088,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 0.07326769828796387,
      "learning_rate": 0.0007631497424403046,
      "loss": 0.4821,
      "num_input_tokens_seen": 9743816,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 0.11827938258647919,
      "learning_rate": 0.000762955263640021,
      "loss": 0.4123,
      "num_input_tokens_seen": 9747272,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 0.0921197459101677,
      "learning_rate": 0.0007627607298324961,
      "loss": 0.4136,
      "num_input_tokens_seen": 9751016,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 0.12198736518621445,
      "learning_rate": 0.0007625661410584244,
      "loss": 0.5296,
      "num_input_tokens_seen": 9753992,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 0.09764767438173294,
      "learning_rate": 0.0007623714973585113,
      "loss": 0.4042,
      "num_input_tokens_seen": 9756456,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 0.06040094047784805,
      "learning_rate": 0.0007621767987734743,
      "loss": 0.4797,
      "num_input_tokens_seen": 9760264,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 0.05970176309347153,
      "learning_rate": 0.000761982045344042,
      "loss": 0.559,
      "num_input_tokens_seen": 9763848,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 0.11282450705766678,
      "learning_rate": 0.0007617872371109549,
      "loss": 0.4204,
      "num_input_tokens_seen": 9766824,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 0.08184246718883514,
      "learning_rate": 0.0007615923741149643,
      "loss": 0.3736,
      "num_input_tokens_seen": 9771368,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 0.07130981236696243,
      "learning_rate": 0.0007613974563968333,
      "loss": 0.3151,
      "num_input_tokens_seen": 9774248,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 0.10575234889984131,
      "learning_rate": 0.0007612024839973368,
      "loss": 0.4397,
      "num_input_tokens_seen": 9776936,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 0.10172237455844879,
      "learning_rate": 0.0007610074569572605,
      "loss": 0.3944,
      "num_input_tokens_seen": 9780104,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 0.23069359362125397,
      "learning_rate": 0.0007608123753174019,
      "loss": 0.3567,
      "num_input_tokens_seen": 9782600,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 0.13351887464523315,
      "learning_rate": 0.00076061723911857,
      "loss": 0.4605,
      "num_input_tokens_seen": 9785544,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 0.07497552782297134,
      "learning_rate": 0.0007604220484015849,
      "loss": 0.3478,
      "num_input_tokens_seen": 9788936,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 0.09292350709438324,
      "learning_rate": 0.0007602268032072784,
      "loss": 0.5385,
      "num_input_tokens_seen": 9792136,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 0.19492293894290924,
      "learning_rate": 0.0007600315035764933,
      "loss": 0.5787,
      "num_input_tokens_seen": 9795208,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 0.12920838594436646,
      "learning_rate": 0.0007598361495500844,
      "loss": 0.4236,
      "num_input_tokens_seen": 9799016,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 0.14590637385845184,
      "learning_rate": 0.0007596407411689173,
      "loss": 0.5325,
      "num_input_tokens_seen": 9802248,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 0.09034955501556396,
      "learning_rate": 0.0007594452784738695,
      "loss": 0.5021,
      "num_input_tokens_seen": 9806184,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 0.1456579864025116,
      "learning_rate": 0.000759249761505829,
      "loss": 0.4973,
      "num_input_tokens_seen": 9808776,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 0.11872096359729767,
      "learning_rate": 0.0007590541903056963,
      "loss": 0.4626,
      "num_input_tokens_seen": 9812008,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 0.18695172667503357,
      "learning_rate": 0.0007588585649143825,
      "loss": 0.4387,
      "num_input_tokens_seen": 9815144,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 0.08811063319444656,
      "learning_rate": 0.0007586628853728099,
      "loss": 0.4566,
      "num_input_tokens_seen": 9817672,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 0.11009518057107925,
      "learning_rate": 0.0007584671517219128,
      "loss": 0.6592,
      "num_input_tokens_seen": 9821128,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 0.08734402060508728,
      "learning_rate": 0.0007582713640026364,
      "loss": 0.4255,
      "num_input_tokens_seen": 9824040,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 0.07828835397958755,
      "learning_rate": 0.000758075522255937,
      "loss": 0.5073,
      "num_input_tokens_seen": 9827496,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 0.2824850082397461,
      "learning_rate": 0.0007578796265227828,
      "loss": 0.4884,
      "num_input_tokens_seen": 9830728,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 0.0864325761795044,
      "learning_rate": 0.0007576836768441525,
      "loss": 0.4722,
      "num_input_tokens_seen": 9834600,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 0.09244420379400253,
      "learning_rate": 0.0007574876732610369,
      "loss": 0.3636,
      "num_input_tokens_seen": 9837992,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 0.1599326878786087,
      "learning_rate": 0.0007572916158144379,
      "loss": 0.4652,
      "num_input_tokens_seen": 9841384,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 0.17239730060100555,
      "learning_rate": 0.0007570955045453679,
      "loss": 0.4622,
      "num_input_tokens_seen": 9843976,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 0.1669914722442627,
      "learning_rate": 0.0007568993394948516,
      "loss": 0.5002,
      "num_input_tokens_seen": 9847144,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 0.11583609133958817,
      "learning_rate": 0.0007567031207039243,
      "loss": 0.4654,
      "num_input_tokens_seen": 9850536,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 0.0814276784658432,
      "learning_rate": 0.0007565068482136328,
      "loss": 0.3047,
      "num_input_tokens_seen": 9854024,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 0.0988997220993042,
      "learning_rate": 0.000756310522065035,
      "loss": 0.3969,
      "num_input_tokens_seen": 9857640,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.21785448491573334,
      "learning_rate": 0.0007561141422992002,
      "loss": 0.5025,
      "num_input_tokens_seen": 9860264,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 0.06454785168170929,
      "learning_rate": 0.0007559177089572086,
      "loss": 0.358,
      "num_input_tokens_seen": 9863912,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.15675801038742065,
      "learning_rate": 0.0007557212220801521,
      "loss": 0.5095,
      "num_input_tokens_seen": 9870408,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 0.09256961941719055,
      "learning_rate": 0.0007555246817091332,
      "loss": 0.6273,
      "num_input_tokens_seen": 9874152,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 0.16559241712093353,
      "learning_rate": 0.0007553280878852663,
      "loss": 0.4722,
      "num_input_tokens_seen": 9877480,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 0.1522400677204132,
      "learning_rate": 0.0007551314406496762,
      "loss": 0.4456,
      "num_input_tokens_seen": 9881032,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 0.08231532573699951,
      "learning_rate": 0.0007549347400434994,
      "loss": 0.3996,
      "num_input_tokens_seen": 9884680,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 0.08633976429700851,
      "learning_rate": 0.0007547379861078835,
      "loss": 0.4667,
      "num_input_tokens_seen": 9888392,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 0.09539724886417389,
      "learning_rate": 0.0007545411788839871,
      "loss": 0.4619,
      "num_input_tokens_seen": 9890952,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 0.10121933370828629,
      "learning_rate": 0.0007543443184129799,
      "loss": 0.3332,
      "num_input_tokens_seen": 9893704,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 0.11450332403182983,
      "learning_rate": 0.000754147404736043,
      "loss": 0.4339,
      "num_input_tokens_seen": 9897256,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 0.06952736526727676,
      "learning_rate": 0.0007539504378943686,
      "loss": 0.6201,
      "num_input_tokens_seen": 9901160,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 0.05915765464305878,
      "learning_rate": 0.0007537534179291599,
      "loss": 0.4044,
      "num_input_tokens_seen": 9904872,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 0.08008833229541779,
      "learning_rate": 0.000753556344881631,
      "loss": 0.4659,
      "num_input_tokens_seen": 9908232,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 0.0943683534860611,
      "learning_rate": 0.0007533592187930076,
      "loss": 0.4356,
      "num_input_tokens_seen": 9911368,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 0.1297512948513031,
      "learning_rate": 0.000753162039704526,
      "loss": 0.5555,
      "num_input_tokens_seen": 9914312,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 0.11146537959575653,
      "learning_rate": 0.000752964807657434,
      "loss": 0.5729,
      "num_input_tokens_seen": 9917640,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 0.09582405537366867,
      "learning_rate": 0.0007527675226929902,
      "loss": 0.4466,
      "num_input_tokens_seen": 9920424,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 0.10294517874717712,
      "learning_rate": 0.0007525701848524643,
      "loss": 0.4929,
      "num_input_tokens_seen": 9922856,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 0.08428806066513062,
      "learning_rate": 0.0007523727941771372,
      "loss": 0.4493,
      "num_input_tokens_seen": 9926024,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 0.12312342971563339,
      "learning_rate": 0.0007521753507083009,
      "loss": 0.4954,
      "num_input_tokens_seen": 9929416,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 0.16323596239089966,
      "learning_rate": 0.0007519778544872581,
      "loss": 0.6218,
      "num_input_tokens_seen": 9933000,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 0.14288753271102905,
      "learning_rate": 0.000751780305555323,
      "loss": 0.3807,
      "num_input_tokens_seen": 9935464,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 0.11204807460308075,
      "learning_rate": 0.0007515827039538202,
      "loss": 0.494,
      "num_input_tokens_seen": 9938760,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 0.15835554897785187,
      "learning_rate": 0.0007513850497240861,
      "loss": 0.4017,
      "num_input_tokens_seen": 9943016,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 0.16102539002895355,
      "learning_rate": 0.0007511873429074676,
      "loss": 0.3163,
      "num_input_tokens_seen": 9945640,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 0.07593416422605515,
      "learning_rate": 0.0007509895835453224,
      "loss": 0.5696,
      "num_input_tokens_seen": 9949672,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 0.12269578129053116,
      "learning_rate": 0.0007507917716790201,
      "loss": 0.4931,
      "num_input_tokens_seen": 9954120,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 0.08637267351150513,
      "learning_rate": 0.0007505939073499401,
      "loss": 0.445,
      "num_input_tokens_seen": 9958344,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 0.14549237489700317,
      "learning_rate": 0.0007503959905994737,
      "loss": 0.521,
      "num_input_tokens_seen": 9961320,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 0.16800908744335175,
      "learning_rate": 0.0007501980214690227,
      "loss": 0.4785,
      "num_input_tokens_seen": 9965160,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.4831865429878235,
      "eval_runtime": 13.7099,
      "eval_samples_per_second": 61.853,
      "eval_steps_per_second": 15.463,
      "num_input_tokens_seen": 9967176,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 0.08149902522563934,
      "learning_rate": 0.00075,
      "loss": 0.6209,
      "num_input_tokens_seen": 9968520,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 0.15595051646232605,
      "learning_rate": 0.0007498019262338295,
      "loss": 0.3794,
      "num_input_tokens_seen": 9972328,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.0860937312245369,
      "learning_rate": 0.0007496038002119459,
      "loss": 0.5855,
      "num_input_tokens_seen": 9975688,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 0.1112118661403656,
      "learning_rate": 0.000749405621975795,
      "loss": 0.4945,
      "num_input_tokens_seen": 9977960,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 0.09994781017303467,
      "learning_rate": 0.0007492073915668334,
      "loss": 0.4301,
      "num_input_tokens_seen": 9981192,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 0.14398157596588135,
      "learning_rate": 0.0007490091090265283,
      "loss": 0.4121,
      "num_input_tokens_seen": 9983624,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 0.09712312370538712,
      "learning_rate": 0.0007488107743963587,
      "loss": 0.5182,
      "num_input_tokens_seen": 9986728,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 0.12519845366477966,
      "learning_rate": 0.0007486123877178136,
      "loss": 0.4628,
      "num_input_tokens_seen": 9990792,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 0.10450451076030731,
      "learning_rate": 0.0007484139490323932,
      "loss": 0.4581,
      "num_input_tokens_seen": 9994056,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 0.10048707574605942,
      "learning_rate": 0.000748215458381609,
      "loss": 0.4087,
      "num_input_tokens_seen": 9997480,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 0.07087961584329605,
      "learning_rate": 0.0007480169158069826,
      "loss": 0.5442,
      "num_input_tokens_seen": 10000296,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 0.13189589977264404,
      "learning_rate": 0.0007478183213500468,
      "loss": 0.44,
      "num_input_tokens_seen": 10002888,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 0.09244448691606522,
      "learning_rate": 0.0007476196750523456,
      "loss": 0.4862,
      "num_input_tokens_seen": 10006216,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 0.1426077038049698,
      "learning_rate": 0.000747420976955433,
      "loss": 0.4556,
      "num_input_tokens_seen": 10009928,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 0.11329376697540283,
      "learning_rate": 0.0007472222271008752,
      "loss": 0.4258,
      "num_input_tokens_seen": 10012072,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 0.0930766835808754,
      "learning_rate": 0.0007470234255302478,
      "loss": 0.465,
      "num_input_tokens_seen": 10016168,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 0.19217868149280548,
      "learning_rate": 0.0007468245722851379,
      "loss": 0.4841,
      "num_input_tokens_seen": 10018664,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 0.06695034354925156,
      "learning_rate": 0.0007466256674071434,
      "loss": 0.3235,
      "num_input_tokens_seen": 10022088,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 0.06964343041181564,
      "learning_rate": 0.000746426710937873,
      "loss": 0.4337,
      "num_input_tokens_seen": 10024456,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 0.25964367389678955,
      "learning_rate": 0.0007462277029189461,
      "loss": 0.4122,
      "num_input_tokens_seen": 10027464,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 0.1286257952451706,
      "learning_rate": 0.0007460286433919928,
      "loss": 0.4893,
      "num_input_tokens_seen": 10030504,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.17721784114837646,
      "learning_rate": 0.0007458295323986542,
      "loss": 0.5919,
      "num_input_tokens_seen": 10033512,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 0.12057245522737503,
      "learning_rate": 0.0007456303699805817,
      "loss": 0.4559,
      "num_input_tokens_seen": 10036456,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 0.08994871377944946,
      "learning_rate": 0.0007454311561794386,
      "loss": 0.4119,
      "num_input_tokens_seen": 10039528,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 0.2090422511100769,
      "learning_rate": 0.0007452318910368972,
      "loss": 0.6434,
      "num_input_tokens_seen": 10042152,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 0.12872736155986786,
      "learning_rate": 0.0007450325745946422,
      "loss": 0.3756,
      "num_input_tokens_seen": 10044328,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 0.1339959055185318,
      "learning_rate": 0.0007448332068943681,
      "loss": 0.3869,
      "num_input_tokens_seen": 10046856,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 0.12323048710823059,
      "learning_rate": 0.0007446337879777802,
      "loss": 0.4307,
      "num_input_tokens_seen": 10049576,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 0.08842746913433075,
      "learning_rate": 0.0007444343178865948,
      "loss": 0.3692,
      "num_input_tokens_seen": 10052648,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 0.11646448075771332,
      "learning_rate": 0.0007442347966625389,
      "loss": 0.4333,
      "num_input_tokens_seen": 10055272,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 0.14209923148155212,
      "learning_rate": 0.0007440352243473496,
      "loss": 0.3926,
      "num_input_tokens_seen": 10058056,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 0.061379969120025635,
      "learning_rate": 0.0007438356009827758,
      "loss": 0.4644,
      "num_input_tokens_seen": 10062696,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 0.1611187756061554,
      "learning_rate": 0.0007436359266105759,
      "loss": 0.5034,
      "num_input_tokens_seen": 10065960,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 0.14045237004756927,
      "learning_rate": 0.0007434362012725198,
      "loss": 0.5353,
      "num_input_tokens_seen": 10069000,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 0.1511681079864502,
      "learning_rate": 0.0007432364250103877,
      "loss": 0.3678,
      "num_input_tokens_seen": 10075624,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 0.07054904103279114,
      "learning_rate": 0.0007430365978659704,
      "loss": 0.3361,
      "num_input_tokens_seen": 10078600,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 0.09577517211437225,
      "learning_rate": 0.0007428367198810694,
      "loss": 0.5137,
      "num_input_tokens_seen": 10081736,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.1473720669746399,
      "learning_rate": 0.0007426367910974973,
      "loss": 0.444,
      "num_input_tokens_seen": 10085544,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 0.1379392147064209,
      "learning_rate": 0.0007424368115570763,
      "loss": 0.411,
      "num_input_tokens_seen": 10089480,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 0.07191670686006546,
      "learning_rate": 0.0007422367813016405,
      "loss": 0.4239,
      "num_input_tokens_seen": 10092392,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 0.12007862329483032,
      "learning_rate": 0.0007420367003730335,
      "loss": 0.45,
      "num_input_tokens_seen": 10095720,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 0.15179602801799774,
      "learning_rate": 0.0007418365688131103,
      "loss": 0.4406,
      "num_input_tokens_seen": 10099272,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 0.14635999500751495,
      "learning_rate": 0.0007416363866637356,
      "loss": 0.418,
      "num_input_tokens_seen": 10106504,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.08635161817073822,
      "learning_rate": 0.0007414361539667857,
      "loss": 0.4964,
      "num_input_tokens_seen": 10109672,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 0.10813538730144501,
      "learning_rate": 0.0007412358707641468,
      "loss": 0.4812,
      "num_input_tokens_seen": 10113192,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 0.09454327076673508,
      "learning_rate": 0.0007410355370977158,
      "loss": 0.4541,
      "num_input_tokens_seen": 10116264,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 0.11606746166944504,
      "learning_rate": 0.0007408351530094005,
      "loss": 0.4878,
      "num_input_tokens_seen": 10119496,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 0.1296033263206482,
      "learning_rate": 0.0007406347185411187,
      "loss": 0.68,
      "num_input_tokens_seen": 10122888,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 0.09857670962810516,
      "learning_rate": 0.0007404342337347992,
      "loss": 0.5411,
      "num_input_tokens_seen": 10125288,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 0.09779363125562668,
      "learning_rate": 0.000740233698632381,
      "loss": 0.4326,
      "num_input_tokens_seen": 10128712,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 0.1001986563205719,
      "learning_rate": 0.0007400331132758136,
      "loss": 0.3991,
      "num_input_tokens_seen": 10132072,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 0.18044081330299377,
      "learning_rate": 0.0007398324777070576,
      "loss": 0.4111,
      "num_input_tokens_seen": 10135592,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 0.10615423321723938,
      "learning_rate": 0.0007396317919680832,
      "loss": 0.3723,
      "num_input_tokens_seen": 10138600,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 0.09960363060235977,
      "learning_rate": 0.0007394310561008717,
      "loss": 0.4405,
      "num_input_tokens_seen": 10141896,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 0.17214961349964142,
      "learning_rate": 0.0007392302701474151,
      "loss": 0.4751,
      "num_input_tokens_seen": 10146024,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 0.12080950289964676,
      "learning_rate": 0.0007390294341497151,
      "loss": 0.3869,
      "num_input_tokens_seen": 10149000,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 0.12494078278541565,
      "learning_rate": 0.0007388285481497845,
      "loss": 0.5363,
      "num_input_tokens_seen": 10151976,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 0.10454198718070984,
      "learning_rate": 0.0007386276121896465,
      "loss": 0.3916,
      "num_input_tokens_seen": 10156104,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 0.12601305544376373,
      "learning_rate": 0.0007384266263113342,
      "loss": 0.3551,
      "num_input_tokens_seen": 10158664,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 0.0855683982372284,
      "learning_rate": 0.0007382255905568919,
      "loss": 0.5179,
      "num_input_tokens_seen": 10161768,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 0.10300906747579575,
      "learning_rate": 0.000738024504968374,
      "loss": 0.4961,
      "num_input_tokens_seen": 10165288,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 0.12151257693767548,
      "learning_rate": 0.0007378233695878449,
      "loss": 0.5998,
      "num_input_tokens_seen": 10167848,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 0.10150381922721863,
      "learning_rate": 0.0007376221844573803,
      "loss": 0.476,
      "num_input_tokens_seen": 10171144,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 0.07775898277759552,
      "learning_rate": 0.0007374209496190656,
      "loss": 0.4065,
      "num_input_tokens_seen": 10174792,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 0.1415693312883377,
      "learning_rate": 0.0007372196651149968,
      "loss": 0.4691,
      "num_input_tokens_seen": 10177864,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 0.13089339435100555,
      "learning_rate": 0.0007370183309872805,
      "loss": 0.552,
      "num_input_tokens_seen": 10180776,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 0.20890399813652039,
      "learning_rate": 0.0007368169472780335,
      "loss": 0.4623,
      "num_input_tokens_seen": 10183528,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 0.09943410009145737,
      "learning_rate": 0.0007366155140293828,
      "loss": 0.366,
      "num_input_tokens_seen": 10186248,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 0.08970876038074493,
      "learning_rate": 0.000736414031283466,
      "loss": 0.5928,
      "num_input_tokens_seen": 10188968,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 0.10329515486955643,
      "learning_rate": 0.000736212499082431,
      "loss": 0.4208,
      "num_input_tokens_seen": 10192328,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 0.08268310874700546,
      "learning_rate": 0.0007360109174684359,
      "loss": 0.3611,
      "num_input_tokens_seen": 10195592,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 0.1260008066892624,
      "learning_rate": 0.0007358092864836497,
      "loss": 0.4956,
      "num_input_tokens_seen": 10199016,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 0.09633652865886688,
      "learning_rate": 0.0007356076061702508,
      "loss": 0.4745,
      "num_input_tokens_seen": 10202152,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 0.22922687232494354,
      "learning_rate": 0.0007354058765704288,
      "loss": 0.4911,
      "num_input_tokens_seen": 10204968,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 0.18008635938167572,
      "learning_rate": 0.000735204097726383,
      "loss": 0.527,
      "num_input_tokens_seen": 10208328,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 0.1916513442993164,
      "learning_rate": 0.0007350022696803233,
      "loss": 0.3167,
      "num_input_tokens_seen": 10210888,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 0.13648071885108948,
      "learning_rate": 0.0007348003924744699,
      "loss": 0.3774,
      "num_input_tokens_seen": 10213544,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 0.07260682433843613,
      "learning_rate": 0.000734598466151053,
      "loss": 0.388,
      "num_input_tokens_seen": 10217096,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 0.12185773253440857,
      "learning_rate": 0.0007343964907523135,
      "loss": 0.4081,
      "num_input_tokens_seen": 10221160,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 0.11083182692527771,
      "learning_rate": 0.0007341944663205023,
      "loss": 0.5196,
      "num_input_tokens_seen": 10223720,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 0.08706362545490265,
      "learning_rate": 0.0007339923928978805,
      "loss": 0.4527,
      "num_input_tokens_seen": 10226760,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 0.1467556357383728,
      "learning_rate": 0.0007337902705267197,
      "loss": 0.4832,
      "num_input_tokens_seen": 10229416,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 0.07909098267555237,
      "learning_rate": 0.0007335880992493016,
      "loss": 0.4444,
      "num_input_tokens_seen": 10232488,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 0.09307213872671127,
      "learning_rate": 0.000733385879107918,
      "loss": 0.4196,
      "num_input_tokens_seen": 10235560,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 0.14649799466133118,
      "learning_rate": 0.000733183610144871,
      "loss": 0.3684,
      "num_input_tokens_seen": 10238952,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 0.0868058130145073,
      "learning_rate": 0.0007329812924024732,
      "loss": 0.4444,
      "num_input_tokens_seen": 10242664,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 0.08291427791118622,
      "learning_rate": 0.0007327789259230468,
      "loss": 0.5615,
      "num_input_tokens_seen": 10246536,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 0.1379786729812622,
      "learning_rate": 0.0007325765107489249,
      "loss": 0.4283,
      "num_input_tokens_seen": 10249160,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 0.1952410340309143,
      "learning_rate": 0.0007323740469224505,
      "loss": 0.3738,
      "num_input_tokens_seen": 10252328,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 0.12650223076343536,
      "learning_rate": 0.0007321715344859763,
      "loss": 0.5195,
      "num_input_tokens_seen": 10256584,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 0.22379349172115326,
      "learning_rate": 0.000731968973481866,
      "loss": 0.3527,
      "num_input_tokens_seen": 10259336,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 0.0826609656214714,
      "learning_rate": 0.0007317663639524928,
      "loss": 0.4696,
      "num_input_tokens_seen": 10262440,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 0.10002267360687256,
      "learning_rate": 0.0007315637059402404,
      "loss": 0.4077,
      "num_input_tokens_seen": 10265384,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 0.109841488301754,
      "learning_rate": 0.0007313609994875024,
      "loss": 0.4782,
      "num_input_tokens_seen": 10268936,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 0.16850297152996063,
      "learning_rate": 0.0007311582446366831,
      "loss": 0.3288,
      "num_input_tokens_seen": 10271304,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 0.15473175048828125,
      "learning_rate": 0.000730955441430196,
      "loss": 0.6439,
      "num_input_tokens_seen": 10274536,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 0.10333991050720215,
      "learning_rate": 0.0007307525899104655,
      "loss": 0.294,
      "num_input_tokens_seen": 10278280,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 0.17595866322517395,
      "learning_rate": 0.0007305496901199259,
      "loss": 0.4055,
      "num_input_tokens_seen": 10280872,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 0.1077604815363884,
      "learning_rate": 0.0007303467421010211,
      "loss": 0.3126,
      "num_input_tokens_seen": 10284680,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 0.17496459186077118,
      "learning_rate": 0.0007301437458962061,
      "loss": 0.5077,
      "num_input_tokens_seen": 10287528,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 0.07684222608804703,
      "learning_rate": 0.0007299407015479448,
      "loss": 0.3121,
      "num_input_tokens_seen": 10291112,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 0.13728414475917816,
      "learning_rate": 0.0007297376090987122,
      "loss": 0.3278,
      "num_input_tokens_seen": 10294248,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 0.23110154271125793,
      "learning_rate": 0.0007295344685909926,
      "loss": 0.5824,
      "num_input_tokens_seen": 10297064,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 0.12177728861570358,
      "learning_rate": 0.000729331280067281,
      "loss": 0.4577,
      "num_input_tokens_seen": 10301000,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 0.18947389721870422,
      "learning_rate": 0.0007291280435700819,
      "loss": 0.4707,
      "num_input_tokens_seen": 10303816,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 0.13671083748340607,
      "learning_rate": 0.0007289247591419102,
      "loss": 0.379,
      "num_input_tokens_seen": 10306440,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 0.1706368327140808,
      "learning_rate": 0.0007287214268252904,
      "loss": 0.5128,
      "num_input_tokens_seen": 10310152,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 0.1129646897315979,
      "learning_rate": 0.0007285180466627577,
      "loss": 0.4132,
      "num_input_tokens_seen": 10313896,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 0.12821368873119354,
      "learning_rate": 0.0007283146186968565,
      "loss": 0.4093,
      "num_input_tokens_seen": 10317320,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 0.10586117208003998,
      "learning_rate": 0.000728111142970142,
      "loss": 0.3491,
      "num_input_tokens_seen": 10320456,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 0.07385362684726715,
      "learning_rate": 0.0007279076195251788,
      "loss": 0.5561,
      "num_input_tokens_seen": 10325224,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 0.1199074387550354,
      "learning_rate": 0.0007277040484045416,
      "loss": 0.4519,
      "num_input_tokens_seen": 10328296,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 0.0864974856376648,
      "learning_rate": 0.0007275004296508154,
      "loss": 0.4569,
      "num_input_tokens_seen": 10331240,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 0.08200880140066147,
      "learning_rate": 0.0007272967633065947,
      "loss": 0.5104,
      "num_input_tokens_seen": 10334824,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 0.07831785827875137,
      "learning_rate": 0.0007270930494144843,
      "loss": 0.4339,
      "num_input_tokens_seen": 10338472,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.07829607278108597,
      "learning_rate": 0.0007268892880170988,
      "loss": 0.3763,
      "num_input_tokens_seen": 10342184,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 0.11461328715085983,
      "learning_rate": 0.0007266854791570625,
      "loss": 0.7057,
      "num_input_tokens_seen": 10345128,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 0.10909399390220642,
      "learning_rate": 0.0007264816228770103,
      "loss": 0.5134,
      "num_input_tokens_seen": 10348168,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 0.0953967273235321,
      "learning_rate": 0.0007262777192195866,
      "loss": 0.5348,
      "num_input_tokens_seen": 10350824,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 0.07803580164909363,
      "learning_rate": 0.0007260737682274452,
      "loss": 0.3486,
      "num_input_tokens_seen": 10355464,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 0.1347765177488327,
      "learning_rate": 0.0007258697699432511,
      "loss": 0.4205,
      "num_input_tokens_seen": 10358088,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 0.12271750718355179,
      "learning_rate": 0.0007256657244096775,
      "loss": 0.4048,
      "num_input_tokens_seen": 10360808,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 0.18014873564243317,
      "learning_rate": 0.0007254616316694091,
      "loss": 0.474,
      "num_input_tokens_seen": 10364616,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 0.09347748756408691,
      "learning_rate": 0.0007252574917651398,
      "loss": 0.3674,
      "num_input_tokens_seen": 10368008,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 0.16181151568889618,
      "learning_rate": 0.0007250533047395728,
      "loss": 0.4329,
      "num_input_tokens_seen": 10371144,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 0.11200971156358719,
      "learning_rate": 0.0007248490706354221,
      "loss": 0.4677,
      "num_input_tokens_seen": 10374952,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 0.11182751506567001,
      "learning_rate": 0.0007246447894954109,
      "loss": 0.5339,
      "num_input_tokens_seen": 10378664,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.12273169308900833,
      "learning_rate": 0.0007244404613622727,
      "loss": 0.4704,
      "num_input_tokens_seen": 10382216,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 0.11806275695562363,
      "learning_rate": 0.0007242360862787507,
      "loss": 0.4688,
      "num_input_tokens_seen": 10385448,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 0.1186332106590271,
      "learning_rate": 0.0007240316642875975,
      "loss": 0.6443,
      "num_input_tokens_seen": 10388008,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 0.17524389922618866,
      "learning_rate": 0.0007238271954315759,
      "loss": 0.5531,
      "num_input_tokens_seen": 10392328,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 0.076512411236763,
      "learning_rate": 0.0007236226797534589,
      "loss": 0.3277,
      "num_input_tokens_seen": 10395400,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 0.12983307242393494,
      "learning_rate": 0.0007234181172960283,
      "loss": 0.3593,
      "num_input_tokens_seen": 10397896,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 0.10178903490304947,
      "learning_rate": 0.0007232135081020766,
      "loss": 0.5167,
      "num_input_tokens_seen": 10401032,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 0.18091097474098206,
      "learning_rate": 0.0007230088522144054,
      "loss": 0.4508,
      "num_input_tokens_seen": 10404616,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 0.10147642344236374,
      "learning_rate": 0.0007228041496758265,
      "loss": 0.4143,
      "num_input_tokens_seen": 10407432,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 0.0859404131770134,
      "learning_rate": 0.0007225994005291615,
      "loss": 0.5091,
      "num_input_tokens_seen": 10411560,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 0.15461649000644684,
      "learning_rate": 0.0007223946048172414,
      "loss": 0.5546,
      "num_input_tokens_seen": 10414248,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 0.13320140540599823,
      "learning_rate": 0.0007221897625829072,
      "loss": 0.6098,
      "num_input_tokens_seen": 10417608,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 0.1488102376461029,
      "learning_rate": 0.0007219848738690095,
      "loss": 0.4056,
      "num_input_tokens_seen": 10419912,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 0.12159451842308044,
      "learning_rate": 0.0007217799387184089,
      "loss": 0.4066,
      "num_input_tokens_seen": 10422920,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 0.13855281472206116,
      "learning_rate": 0.0007215749571739752,
      "loss": 0.4588,
      "num_input_tokens_seen": 10426952,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 0.13619676232337952,
      "learning_rate": 0.0007213699292785883,
      "loss": 0.4139,
      "num_input_tokens_seen": 10430568,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 0.0984937772154808,
      "learning_rate": 0.0007211648550751377,
      "loss": 0.4454,
      "num_input_tokens_seen": 10433832,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 0.14919164776802063,
      "learning_rate": 0.0007209597346065226,
      "loss": 0.4467,
      "num_input_tokens_seen": 10436584,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 0.12219024449586868,
      "learning_rate": 0.000720754567915652,
      "loss": 0.5331,
      "num_input_tokens_seen": 10439816,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 0.07525265216827393,
      "learning_rate": 0.000720549355045444,
      "loss": 0.4603,
      "num_input_tokens_seen": 10442920,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 0.10725235939025879,
      "learning_rate": 0.0007203440960388273,
      "loss": 0.3178,
      "num_input_tokens_seen": 10445800,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 0.18058301508426666,
      "learning_rate": 0.0007201387909387396,
      "loss": 0.4725,
      "num_input_tokens_seen": 10448424,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 0.08124116063117981,
      "learning_rate": 0.0007199334397881282,
      "loss": 0.5509,
      "num_input_tokens_seen": 10452104,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 0.1073259562253952,
      "learning_rate": 0.0007197280426299502,
      "loss": 0.5387,
      "num_input_tokens_seen": 10454792,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 0.09522485733032227,
      "learning_rate": 0.0007195225995071727,
      "loss": 0.4741,
      "num_input_tokens_seen": 10457896,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 0.0797736868262291,
      "learning_rate": 0.0007193171104627718,
      "loss": 0.4448,
      "num_input_tokens_seen": 10461416,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 0.098525270819664,
      "learning_rate": 0.0007191115755397333,
      "loss": 0.4888,
      "num_input_tokens_seen": 10464936,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 0.1273089349269867,
      "learning_rate": 0.0007189059947810531,
      "loss": 0.4346,
      "num_input_tokens_seen": 10468744,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 0.23561197519302368,
      "learning_rate": 0.0007187003682297363,
      "loss": 0.4692,
      "num_input_tokens_seen": 10471528,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 0.17027857899665833,
      "learning_rate": 0.0007184946959287972,
      "loss": 0.4902,
      "num_input_tokens_seen": 10474344,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 0.22207029163837433,
      "learning_rate": 0.0007182889779212609,
      "loss": 0.4365,
      "num_input_tokens_seen": 10477256,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 0.11591248214244843,
      "learning_rate": 0.0007180832142501605,
      "loss": 0.6122,
      "num_input_tokens_seen": 10480104,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 0.17055481672286987,
      "learning_rate": 0.0007178774049585397,
      "loss": 0.4917,
      "num_input_tokens_seen": 10484680,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 0.0950188860297203,
      "learning_rate": 0.0007176715500894515,
      "loss": 0.3988,
      "num_input_tokens_seen": 10488008,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 0.13862161338329315,
      "learning_rate": 0.0007174656496859584,
      "loss": 0.392,
      "num_input_tokens_seen": 10490792,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 0.07139939814805984,
      "learning_rate": 0.0007172597037911323,
      "loss": 0.45,
      "num_input_tokens_seen": 10496200,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 0.13372789323329926,
      "learning_rate": 0.0007170537124480548,
      "loss": 0.4253,
      "num_input_tokens_seen": 10498472,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 0.08793877065181732,
      "learning_rate": 0.0007168476756998169,
      "loss": 0.4889,
      "num_input_tokens_seen": 10501064,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 0.12247894704341888,
      "learning_rate": 0.0007166415935895194,
      "loss": 0.5078,
      "num_input_tokens_seen": 10504616,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 0.18308772146701813,
      "learning_rate": 0.0007164354661602718,
      "loss": 0.5802,
      "num_input_tokens_seen": 10506792,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 0.17691412568092346,
      "learning_rate": 0.000716229293455194,
      "loss": 0.3515,
      "num_input_tokens_seen": 10509384,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 0.09184219688177109,
      "learning_rate": 0.000716023075517415,
      "loss": 0.4128,
      "num_input_tokens_seen": 10512328,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 0.16273698210716248,
      "learning_rate": 0.000715816812390073,
      "loss": 0.5808,
      "num_input_tokens_seen": 10516136,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 0.18120339512825012,
      "learning_rate": 0.0007156105041163161,
      "loss": 0.5335,
      "num_input_tokens_seen": 10520040,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 0.12198659777641296,
      "learning_rate": 0.0007154041507393015,
      "loss": 0.5671,
      "num_input_tokens_seen": 10523144,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 0.18207180500030518,
      "learning_rate": 0.000715197752302196,
      "loss": 0.4421,
      "num_input_tokens_seen": 10526600,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 0.1763375997543335,
      "learning_rate": 0.0007149913088481759,
      "loss": 0.5237,
      "num_input_tokens_seen": 10531560,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 0.09760147333145142,
      "learning_rate": 0.0007147848204204266,
      "loss": 0.4237,
      "num_input_tokens_seen": 10535592,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 0.1727578043937683,
      "learning_rate": 0.0007145782870621435,
      "loss": 0.4198,
      "num_input_tokens_seen": 10537992,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 0.11126714199781418,
      "learning_rate": 0.0007143717088165304,
      "loss": 0.4349,
      "num_input_tokens_seen": 10542152,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 0.20965638756752014,
      "learning_rate": 0.0007141650857268019,
      "loss": 0.4189,
      "num_input_tokens_seen": 10545064,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 0.0740964263677597,
      "learning_rate": 0.0007139584178361807,
      "loss": 0.5031,
      "num_input_tokens_seen": 10548008,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 0.20256279408931732,
      "learning_rate": 0.0007137517051878994,
      "loss": 0.5159,
      "num_input_tokens_seen": 10551048,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 0.22447547316551208,
      "learning_rate": 0.0007135449478251998,
      "loss": 0.4605,
      "num_input_tokens_seen": 10554312,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 0.3101308047771454,
      "learning_rate": 0.0007133381457913336,
      "loss": 0.4649,
      "num_input_tokens_seen": 10556840,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 0.1463085412979126,
      "learning_rate": 0.0007131312991295611,
      "loss": 0.4221,
      "num_input_tokens_seen": 10562376,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 0.14313960075378418,
      "learning_rate": 0.0007129244078831525,
      "loss": 0.4798,
      "num_input_tokens_seen": 10564776,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 0.07638902962207794,
      "learning_rate": 0.0007127174720953866,
      "loss": 0.4264,
      "num_input_tokens_seen": 10567944,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 0.10684222728013992,
      "learning_rate": 0.0007125104918095526,
      "loss": 0.3159,
      "num_input_tokens_seen": 10570952,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 0.07242812216281891,
      "learning_rate": 0.0007123034670689483,
      "loss": 0.2921,
      "num_input_tokens_seen": 10573544,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.11503328382968903,
      "learning_rate": 0.0007120963979168804,
      "loss": 0.3643,
      "num_input_tokens_seen": 10579112,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 0.07096611708402634,
      "learning_rate": 0.000711889284396666,
      "loss": 0.3262,
      "num_input_tokens_seen": 10583080,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 0.10943715274333954,
      "learning_rate": 0.0007116821265516307,
      "loss": 0.3927,
      "num_input_tokens_seen": 10585960,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 0.1181199923157692,
      "learning_rate": 0.0007114749244251094,
      "loss": 0.3923,
      "num_input_tokens_seen": 10588584,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 0.18842582404613495,
      "learning_rate": 0.0007112676780604468,
      "loss": 0.3193,
      "num_input_tokens_seen": 10591720,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 0.22882255911827087,
      "learning_rate": 0.000711060387500996,
      "loss": 0.4303,
      "num_input_tokens_seen": 10594760,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 0.08571963012218475,
      "learning_rate": 0.0007108530527901199,
      "loss": 0.4872,
      "num_input_tokens_seen": 10598600,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 0.08849386870861053,
      "learning_rate": 0.000710645673971191,
      "loss": 0.3659,
      "num_input_tokens_seen": 10601864,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 0.09585703909397125,
      "learning_rate": 0.00071043825108759,
      "loss": 0.3944,
      "num_input_tokens_seen": 10604488,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 0.10695316642522812,
      "learning_rate": 0.0007102307841827079,
      "loss": 0.4488,
      "num_input_tokens_seen": 10607400,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 0.23216651380062103,
      "learning_rate": 0.0007100232732999443,
      "loss": 0.4575,
      "num_input_tokens_seen": 10610184,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 0.09865345805883408,
      "learning_rate": 0.0007098157184827076,
      "loss": 0.4986,
      "num_input_tokens_seen": 10613096,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 0.07058024406433105,
      "learning_rate": 0.0007096081197744166,
      "loss": 0.3648,
      "num_input_tokens_seen": 10616872,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 0.11313457041978836,
      "learning_rate": 0.0007094004772184981,
      "loss": 0.426,
      "num_input_tokens_seen": 10620104,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 0.12855134904384613,
      "learning_rate": 0.0007091927908583889,
      "loss": 0.3524,
      "num_input_tokens_seen": 10623400,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 0.12755274772644043,
      "learning_rate": 0.0007089850607375343,
      "loss": 0.4523,
      "num_input_tokens_seen": 10626344,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 0.10835407674312592,
      "learning_rate": 0.0007087772868993894,
      "loss": 0.5826,
      "num_input_tokens_seen": 10629256,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 0.14283454418182373,
      "learning_rate": 0.0007085694693874178,
      "loss": 0.5367,
      "num_input_tokens_seen": 10631624,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 0.16176430881023407,
      "learning_rate": 0.0007083616082450928,
      "loss": 0.5173,
      "num_input_tokens_seen": 10633928,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 0.08493222296237946,
      "learning_rate": 0.0007081537035158962,
      "loss": 0.5625,
      "num_input_tokens_seen": 10637800,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 0.12738679349422455,
      "learning_rate": 0.0007079457552433198,
      "loss": 0.4415,
      "num_input_tokens_seen": 10641096,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 0.1380312740802765,
      "learning_rate": 0.0007077377634708637,
      "loss": 0.4719,
      "num_input_tokens_seen": 10643848,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 0.19552063941955566,
      "learning_rate": 0.0007075297282420375,
      "loss": 0.423,
      "num_input_tokens_seen": 10646408,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 0.14352624118328094,
      "learning_rate": 0.0007073216496003598,
      "loss": 0.5563,
      "num_input_tokens_seen": 10650056,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 0.09129060059785843,
      "learning_rate": 0.0007071135275893584,
      "loss": 0.4976,
      "num_input_tokens_seen": 10653384,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 0.07830371707677841,
      "learning_rate": 0.0007069053622525696,
      "loss": 0.5123,
      "num_input_tokens_seen": 10656936,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 0.22028405964374542,
      "learning_rate": 0.0007066971536335395,
      "loss": 0.3755,
      "num_input_tokens_seen": 10659400,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 0.10859061032533646,
      "learning_rate": 0.0007064889017758234,
      "loss": 0.4376,
      "num_input_tokens_seen": 10662376,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 0.13072766363620758,
      "learning_rate": 0.0007062806067229845,
      "loss": 0.4062,
      "num_input_tokens_seen": 10665288,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 0.04015851393342018,
      "learning_rate": 0.0007060722685185961,
      "loss": 0.4501,
      "num_input_tokens_seen": 10670280,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 0.11656428873538971,
      "learning_rate": 0.00070586388720624,
      "loss": 0.3068,
      "num_input_tokens_seen": 10672872,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 0.12064273655414581,
      "learning_rate": 0.0007056554628295076,
      "loss": 0.4667,
      "num_input_tokens_seen": 10676232,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 0.08479231595993042,
      "learning_rate": 0.0007054469954319984,
      "loss": 0.5032,
      "num_input_tokens_seen": 10679336,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 0.1569456309080124,
      "learning_rate": 0.0007052384850573217,
      "loss": 0.3939,
      "num_input_tokens_seen": 10682024,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 0.14140738546848297,
      "learning_rate": 0.0007050299317490952,
      "loss": 0.4376,
      "num_input_tokens_seen": 10685608,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 0.08508200198411942,
      "learning_rate": 0.0007048213355509463,
      "loss": 0.3681,
      "num_input_tokens_seen": 10688744,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 0.09799184650182724,
      "learning_rate": 0.0007046126965065107,
      "loss": 0.4193,
      "num_input_tokens_seen": 10692104,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 0.09747081249952316,
      "learning_rate": 0.0007044040146594332,
      "loss": 0.4808,
      "num_input_tokens_seen": 10695400,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 0.09796178340911865,
      "learning_rate": 0.0007041952900533681,
      "loss": 0.3634,
      "num_input_tokens_seen": 10698632,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 0.08048585057258606,
      "learning_rate": 0.0007039865227319777,
      "loss": 0.4988,
      "num_input_tokens_seen": 10701672,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 0.11316604912281036,
      "learning_rate": 0.0007037777127389339,
      "loss": 0.4518,
      "num_input_tokens_seen": 10704936,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 0.1413871943950653,
      "learning_rate": 0.0007035688601179177,
      "loss": 0.6561,
      "num_input_tokens_seen": 10707720,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 0.11817767471075058,
      "learning_rate": 0.000703359964912618,
      "loss": 0.6468,
      "num_input_tokens_seen": 10710440,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 0.11527110636234283,
      "learning_rate": 0.000703151027166734,
      "loss": 0.4958,
      "num_input_tokens_seen": 10713896,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 0.13908955454826355,
      "learning_rate": 0.0007029420469239726,
      "loss": 0.4112,
      "num_input_tokens_seen": 10717032,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 0.1008792519569397,
      "learning_rate": 0.0007027330242280503,
      "loss": 0.4158,
      "num_input_tokens_seen": 10719784,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.13096413016319275,
      "learning_rate": 0.0007025239591226923,
      "loss": 0.4991,
      "num_input_tokens_seen": 10722216,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 0.13685745000839233,
      "learning_rate": 0.0007023148516516326,
      "loss": 0.4704,
      "num_input_tokens_seen": 10725160,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 0.12295655161142349,
      "learning_rate": 0.0007021057018586139,
      "loss": 0.4313,
      "num_input_tokens_seen": 10728136,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 0.11785206198692322,
      "learning_rate": 0.0007018965097873883,
      "loss": 0.4124,
      "num_input_tokens_seen": 10730920,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 0.09823919832706451,
      "learning_rate": 0.0007016872754817161,
      "loss": 0.5328,
      "num_input_tokens_seen": 10733896,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 0.12477331608533859,
      "learning_rate": 0.0007014779989853668,
      "loss": 0.5628,
      "num_input_tokens_seen": 10736616,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 0.08038084954023361,
      "learning_rate": 0.0007012686803421189,
      "loss": 0.3811,
      "num_input_tokens_seen": 10739560,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 0.13389994204044342,
      "learning_rate": 0.000701059319595759,
      "loss": 0.437,
      "num_input_tokens_seen": 10742472,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 0.07293672114610672,
      "learning_rate": 0.0007008499167900833,
      "loss": 0.5292,
      "num_input_tokens_seen": 10746120,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 0.0963696837425232,
      "learning_rate": 0.0007006404719688966,
      "loss": 0.4213,
      "num_input_tokens_seen": 10749608,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 0.15080247819423676,
      "learning_rate": 0.000700430985176012,
      "loss": 0.6415,
      "num_input_tokens_seen": 10753544,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 0.0871458351612091,
      "learning_rate": 0.0007002214564552521,
      "loss": 0.4962,
      "num_input_tokens_seen": 10756744,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 0.06637705862522125,
      "learning_rate": 0.0007000118858504476,
      "loss": 0.3003,
      "num_input_tokens_seen": 10760424,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 0.1318361759185791,
      "learning_rate": 0.0006998022734054386,
      "loss": 0.4722,
      "num_input_tokens_seen": 10763432,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 0.1240222305059433,
      "learning_rate": 0.0006995926191640734,
      "loss": 0.4912,
      "num_input_tokens_seen": 10766792,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 0.056581512093544006,
      "learning_rate": 0.0006993829231702092,
      "loss": 0.3889,
      "num_input_tokens_seen": 10770888,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 0.1374060958623886,
      "learning_rate": 0.0006991731854677124,
      "loss": 0.3888,
      "num_input_tokens_seen": 10775432,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 0.10908229649066925,
      "learning_rate": 0.0006989634061004572,
      "loss": 0.6009,
      "num_input_tokens_seen": 10777544,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 0.05401119589805603,
      "learning_rate": 0.0006987535851123276,
      "loss": 0.4929,
      "num_input_tokens_seen": 10780840,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 0.0995161160826683,
      "learning_rate": 0.0006985437225472155,
      "loss": 0.4366,
      "num_input_tokens_seen": 10786216,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 0.09143802523612976,
      "learning_rate": 0.0006983338184490215,
      "loss": 0.3113,
      "num_input_tokens_seen": 10789160,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 0.09196839481592178,
      "learning_rate": 0.0006981238728616557,
      "loss": 0.4066,
      "num_input_tokens_seen": 10793032,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 0.1032867357134819,
      "learning_rate": 0.0006979138858290358,
      "loss": 0.495,
      "num_input_tokens_seen": 10796008,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 0.08458910137414932,
      "learning_rate": 0.0006977038573950889,
      "loss": 0.4271,
      "num_input_tokens_seen": 10799016,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 0.12405505031347275,
      "learning_rate": 0.0006974937876037508,
      "loss": 0.6033,
      "num_input_tokens_seen": 10802120,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 0.11858722567558289,
      "learning_rate": 0.0006972836764989652,
      "loss": 0.4006,
      "num_input_tokens_seen": 10805832,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.08689028769731522,
      "learning_rate": 0.0006970735241246853,
      "loss": 0.3993,
      "num_input_tokens_seen": 10808744,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 0.09026987850666046,
      "learning_rate": 0.0006968633305248724,
      "loss": 0.3901,
      "num_input_tokens_seen": 10812072,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 0.2003176212310791,
      "learning_rate": 0.0006966530957434968,
      "loss": 0.6024,
      "num_input_tokens_seen": 10814440,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 0.12511591613292694,
      "learning_rate": 0.0006964428198245372,
      "loss": 0.462,
      "num_input_tokens_seen": 10817192,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 0.15407945215702057,
      "learning_rate": 0.0006962325028119806,
      "loss": 0.3784,
      "num_input_tokens_seen": 10819720,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 0.13279971480369568,
      "learning_rate": 0.0006960221447498232,
      "loss": 0.4141,
      "num_input_tokens_seen": 10822600,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 0.09093176573514938,
      "learning_rate": 0.0006958117456820696,
      "loss": 0.3727,
      "num_input_tokens_seen": 10825640,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 0.10595210641622543,
      "learning_rate": 0.0006956013056527326,
      "loss": 0.468,
      "num_input_tokens_seen": 10832648,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 0.1435445249080658,
      "learning_rate": 0.0006953908247058341,
      "loss": 0.4285,
      "num_input_tokens_seen": 10835816,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 0.07074780762195587,
      "learning_rate": 0.0006951803028854041,
      "loss": 0.5071,
      "num_input_tokens_seen": 10839496,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 0.0776825100183487,
      "learning_rate": 0.0006949697402354816,
      "loss": 0.4937,
      "num_input_tokens_seen": 10843432,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 0.124705471098423,
      "learning_rate": 0.0006947591368001138,
      "loss": 0.3695,
      "num_input_tokens_seen": 10846152,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 0.13767977058887482,
      "learning_rate": 0.0006945484926233563,
      "loss": 0.4312,
      "num_input_tokens_seen": 10849896,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 0.14652155339717865,
      "learning_rate": 0.0006943378077492737,
      "loss": 0.4668,
      "num_input_tokens_seen": 10852904,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 0.1217976063489914,
      "learning_rate": 0.000694127082221939,
      "loss": 0.4817,
      "num_input_tokens_seen": 10855624,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 0.07765846699476242,
      "learning_rate": 0.0006939163160854334,
      "loss": 0.4121,
      "num_input_tokens_seen": 10859432,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 0.058987706899642944,
      "learning_rate": 0.0006937055093838467,
      "loss": 0.3515,
      "num_input_tokens_seen": 10862472,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 0.13708600401878357,
      "learning_rate": 0.0006934946621612774,
      "loss": 0.4401,
      "num_input_tokens_seen": 10865128,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 0.1085156723856926,
      "learning_rate": 0.0006932837744618322,
      "loss": 0.5261,
      "num_input_tokens_seen": 10868520,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 0.07597745209932327,
      "learning_rate": 0.0006930728463296265,
      "loss": 0.5376,
      "num_input_tokens_seen": 10872232,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 0.0736202746629715,
      "learning_rate": 0.0006928618778087842,
      "loss": 0.5506,
      "num_input_tokens_seen": 10877288,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 0.1288158893585205,
      "learning_rate": 0.0006926508689434371,
      "loss": 0.3492,
      "num_input_tokens_seen": 10880424,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 0.06665214896202087,
      "learning_rate": 0.0006924398197777263,
      "loss": 0.3497,
      "num_input_tokens_seen": 10884072,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 0.11489395797252655,
      "learning_rate": 0.0006922287303558006,
      "loss": 0.5118,
      "num_input_tokens_seen": 10888136,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 0.1301669180393219,
      "learning_rate": 0.0006920176007218175,
      "loss": 0.4123,
      "num_input_tokens_seen": 10891080,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 0.16281014680862427,
      "learning_rate": 0.000691806430919943,
      "loss": 0.4515,
      "num_input_tokens_seen": 10894056,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 0.13677366077899933,
      "learning_rate": 0.0006915952209943514,
      "loss": 0.4591,
      "num_input_tokens_seen": 10897416,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.06001295894384384,
      "learning_rate": 0.0006913839709892256,
      "loss": 0.5091,
      "num_input_tokens_seen": 10900712,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 0.09055802971124649,
      "learning_rate": 0.0006911726809487561,
      "loss": 0.5292,
      "num_input_tokens_seen": 10904200,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 0.1594056338071823,
      "learning_rate": 0.0006909613509171431,
      "loss": 0.6292,
      "num_input_tokens_seen": 10907176,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 0.09657332301139832,
      "learning_rate": 0.0006907499809385941,
      "loss": 0.5508,
      "num_input_tokens_seen": 10911400,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 0.14560848474502563,
      "learning_rate": 0.0006905385710573252,
      "loss": 0.3445,
      "num_input_tokens_seen": 10914792,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 0.09502632170915604,
      "learning_rate": 0.000690327121317561,
      "loss": 0.3417,
      "num_input_tokens_seen": 10918536,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 0.1863841712474823,
      "learning_rate": 0.0006901156317635342,
      "loss": 0.4895,
      "num_input_tokens_seen": 10921704,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 0.1603793203830719,
      "learning_rate": 0.0006899041024394864,
      "loss": 0.4132,
      "num_input_tokens_seen": 10924104,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 0.14604714512825012,
      "learning_rate": 0.000689692533389667,
      "loss": 0.3756,
      "num_input_tokens_seen": 10927368,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 0.08919350057840347,
      "learning_rate": 0.0006894809246583334,
      "loss": 0.4422,
      "num_input_tokens_seen": 10930280,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 0.16462723910808563,
      "learning_rate": 0.0006892692762897523,
      "loss": 0.4879,
      "num_input_tokens_seen": 10933256,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 0.24123649299144745,
      "learning_rate": 0.000689057588328198,
      "loss": 0.5821,
      "num_input_tokens_seen": 10935432,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 0.09779203683137894,
      "learning_rate": 0.0006888458608179528,
      "loss": 0.4426,
      "num_input_tokens_seen": 10939400,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 0.21634800732135773,
      "learning_rate": 0.0006886340938033085,
      "loss": 0.5576,
      "num_input_tokens_seen": 10941928,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 0.13475140929222107,
      "learning_rate": 0.0006884222873285632,
      "loss": 0.3986,
      "num_input_tokens_seen": 10945256,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 0.18837782740592957,
      "learning_rate": 0.0006882104414380254,
      "loss": 0.5403,
      "num_input_tokens_seen": 10949448,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 0.22460249066352844,
      "learning_rate": 0.0006879985561760105,
      "loss": 0.5198,
      "num_input_tokens_seen": 10952264,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 0.11860767006874084,
      "learning_rate": 0.0006877866315868422,
      "loss": 0.6528,
      "num_input_tokens_seen": 10954952,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 0.10531772673130035,
      "learning_rate": 0.0006875746677148531,
      "loss": 0.4562,
      "num_input_tokens_seen": 10958472,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 0.11207634210586548,
      "learning_rate": 0.0006873626646043835,
      "loss": 0.4821,
      "num_input_tokens_seen": 10961256,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 0.1043039858341217,
      "learning_rate": 0.000687150622299782,
      "loss": 0.3967,
      "num_input_tokens_seen": 10964488,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 0.11426118016242981,
      "learning_rate": 0.0006869385408454056,
      "loss": 0.4057,
      "num_input_tokens_seen": 10967272,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 0.11067374795675278,
      "learning_rate": 0.0006867264202856188,
      "loss": 0.4274,
      "num_input_tokens_seen": 10970472,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 0.406608909368515,
      "learning_rate": 0.0006865142606647954,
      "loss": 0.4975,
      "num_input_tokens_seen": 10973032,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 0.21606984734535217,
      "learning_rate": 0.0006863020620273166,
      "loss": 0.4302,
      "num_input_tokens_seen": 10976136,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 0.09379236400127411,
      "learning_rate": 0.0006860898244175716,
      "loss": 0.3436,
      "num_input_tokens_seen": 10979176,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 0.056505609303712845,
      "learning_rate": 0.0006858775478799586,
      "loss": 0.3823,
      "num_input_tokens_seen": 10983016,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 0.0846385806798935,
      "learning_rate": 0.0006856652324588831,
      "loss": 0.4389,
      "num_input_tokens_seen": 10985704,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 0.1396835595369339,
      "learning_rate": 0.000685452878198759,
      "loss": 0.4378,
      "num_input_tokens_seen": 10988296,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 0.16614040732383728,
      "learning_rate": 0.0006852404851440088,
      "loss": 0.5459,
      "num_input_tokens_seen": 10992232,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 0.10422507673501968,
      "learning_rate": 0.0006850280533390624,
      "loss": 0.4203,
      "num_input_tokens_seen": 10998600,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 0.07663978636264801,
      "learning_rate": 0.0006848155828283581,
      "loss": 0.4427,
      "num_input_tokens_seen": 11002664,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 0.1180863082408905,
      "learning_rate": 0.0006846030736563422,
      "loss": 0.3531,
      "num_input_tokens_seen": 11005320,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 0.10956967622041702,
      "learning_rate": 0.0006843905258674696,
      "loss": 0.4678,
      "num_input_tokens_seen": 11008488,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 0.11152677983045578,
      "learning_rate": 0.0006841779395062026,
      "loss": 0.4479,
      "num_input_tokens_seen": 11011016,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 0.09536878764629364,
      "learning_rate": 0.0006839653146170116,
      "loss": 0.3472,
      "num_input_tokens_seen": 11015016,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 0.24777306616306305,
      "learning_rate": 0.0006837526512443758,
      "loss": 0.5101,
      "num_input_tokens_seen": 11017800,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 0.13411466777324677,
      "learning_rate": 0.0006835399494327818,
      "loss": 0.5476,
      "num_input_tokens_seen": 11020616,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 0.05546148493885994,
      "learning_rate": 0.0006833272092267241,
      "loss": 0.4343,
      "num_input_tokens_seen": 11024296,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 0.2808511257171631,
      "learning_rate": 0.000683114430670706,
      "loss": 0.4877,
      "num_input_tokens_seen": 11027624,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 0.11363092809915543,
      "learning_rate": 0.0006829016138092378,
      "loss": 0.4943,
      "num_input_tokens_seen": 11030472,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 0.10540614277124405,
      "learning_rate": 0.0006826887586868388,
      "loss": 0.5556,
      "num_input_tokens_seen": 11034024,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 0.0755327045917511,
      "learning_rate": 0.0006824758653480356,
      "loss": 0.377,
      "num_input_tokens_seen": 11038472,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 0.1738768368959427,
      "learning_rate": 0.0006822629338373632,
      "loss": 0.3839,
      "num_input_tokens_seen": 11041896,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 0.2554941475391388,
      "learning_rate": 0.0006820499641993644,
      "loss": 0.5108,
      "num_input_tokens_seen": 11044904,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.08075360953807831,
      "learning_rate": 0.0006818369564785902,
      "loss": 0.5021,
      "num_input_tokens_seen": 11048648,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 0.14994870126247406,
      "learning_rate": 0.0006816239107195989,
      "loss": 0.3844,
      "num_input_tokens_seen": 11052168,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 0.09309469908475876,
      "learning_rate": 0.000681410826966958,
      "loss": 0.4307,
      "num_input_tokens_seen": 11055496,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 0.1087491437792778,
      "learning_rate": 0.0006811977052652414,
      "loss": 0.4728,
      "num_input_tokens_seen": 11059432,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 0.08985304087400436,
      "learning_rate": 0.000680984545659032,
      "loss": 0.4782,
      "num_input_tokens_seen": 11062376,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 0.1320965588092804,
      "learning_rate": 0.0006807713481929207,
      "loss": 0.4679,
      "num_input_tokens_seen": 11066888,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 0.2889709770679474,
      "learning_rate": 0.0006805581129115055,
      "loss": 0.4651,
      "num_input_tokens_seen": 11069576,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 0.05170970782637596,
      "learning_rate": 0.0006803448398593931,
      "loss": 0.4363,
      "num_input_tokens_seen": 11075144,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.07989544421434402,
      "learning_rate": 0.0006801315290811976,
      "loss": 0.4528,
      "num_input_tokens_seen": 11078600,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 0.12443763762712479,
      "learning_rate": 0.0006799181806215413,
      "loss": 0.4264,
      "num_input_tokens_seen": 11081416,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 0.08824996650218964,
      "learning_rate": 0.0006797047945250543,
      "loss": 0.4863,
      "num_input_tokens_seen": 11084520,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 0.4262656271457672,
      "learning_rate": 0.0006794913708363745,
      "loss": 0.6495,
      "num_input_tokens_seen": 11087208,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 0.1635698676109314,
      "learning_rate": 0.0006792779096001475,
      "loss": 0.4883,
      "num_input_tokens_seen": 11090184,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 0.2339438647031784,
      "learning_rate": 0.0006790644108610273,
      "loss": 0.4533,
      "num_input_tokens_seen": 11093416,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 0.1285746693611145,
      "learning_rate": 0.0006788508746636751,
      "loss": 0.3837,
      "num_input_tokens_seen": 11097096,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 0.09944939613342285,
      "learning_rate": 0.0006786373010527605,
      "loss": 0.4101,
      "num_input_tokens_seen": 11100360,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 0.09454169124364853,
      "learning_rate": 0.0006784236900729603,
      "loss": 0.5602,
      "num_input_tokens_seen": 11103272,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.2261234074831009,
      "learning_rate": 0.0006782100417689599,
      "loss": 0.5303,
      "num_input_tokens_seen": 11106568,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 0.14358572661876678,
      "learning_rate": 0.0006779963561854517,
      "loss": 0.3559,
      "num_input_tokens_seen": 11109640,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 0.13961243629455566,
      "learning_rate": 0.0006777826333671367,
      "loss": 0.4314,
      "num_input_tokens_seen": 11112776,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 0.12075915187597275,
      "learning_rate": 0.0006775688733587227,
      "loss": 0.4182,
      "num_input_tokens_seen": 11115912,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 0.10507984459400177,
      "learning_rate": 0.0006773550762049265,
      "loss": 0.4316,
      "num_input_tokens_seen": 11119048,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 0.17059223353862762,
      "learning_rate": 0.0006771412419504716,
      "loss": 0.4838,
      "num_input_tokens_seen": 11122280,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 0.11802691966295242,
      "learning_rate": 0.00067692737064009,
      "loss": 0.5712,
      "num_input_tokens_seen": 11125576,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 0.13288447260856628,
      "learning_rate": 0.0006767134623185208,
      "loss": 0.3907,
      "num_input_tokens_seen": 11128904,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 0.11092767119407654,
      "learning_rate": 0.0006764995170305114,
      "loss": 0.6073,
      "num_input_tokens_seen": 11132008,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 0.14892128109931946,
      "learning_rate": 0.0006762855348208166,
      "loss": 0.5273,
      "num_input_tokens_seen": 11135624,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 0.10243025422096252,
      "learning_rate": 0.0006760715157341991,
      "loss": 0.4961,
      "num_input_tokens_seen": 11138728,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 0.13765102624893188,
      "learning_rate": 0.0006758574598154292,
      "loss": 0.5119,
      "num_input_tokens_seen": 11142216,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 0.11022961884737015,
      "learning_rate": 0.0006756433671092852,
      "loss": 0.4528,
      "num_input_tokens_seen": 11144776,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 0.13766561448574066,
      "learning_rate": 0.0006754292376605524,
      "loss": 0.3754,
      "num_input_tokens_seen": 11147688,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 0.15102478861808777,
      "learning_rate": 0.0006752150715140247,
      "loss": 0.4893,
      "num_input_tokens_seen": 11151304,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 0.08803941309452057,
      "learning_rate": 0.0006750008687145028,
      "loss": 0.4594,
      "num_input_tokens_seen": 11154248,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 0.10965096950531006,
      "learning_rate": 0.0006747866293067958,
      "loss": 0.4788,
      "num_input_tokens_seen": 11157192,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 0.1057671308517456,
      "learning_rate": 0.00067457235333572,
      "loss": 0.3608,
      "num_input_tokens_seen": 11160872,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 0.08695471286773682,
      "learning_rate": 0.0006743580408460994,
      "loss": 0.4624,
      "num_input_tokens_seen": 11163272,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 0.06512736529111862,
      "learning_rate": 0.0006741436918827659,
      "loss": 0.354,
      "num_input_tokens_seen": 11166888,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 0.11920178681612015,
      "learning_rate": 0.0006739293064905589,
      "loss": 0.4148,
      "num_input_tokens_seen": 11171272,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 0.22958821058273315,
      "learning_rate": 0.0006737148847143251,
      "loss": 0.4266,
      "num_input_tokens_seen": 11173608,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 0.10525697469711304,
      "learning_rate": 0.0006735004265989195,
      "loss": 0.5723,
      "num_input_tokens_seen": 11177608,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 0.10932978987693787,
      "learning_rate": 0.0006732859321892038,
      "loss": 0.4179,
      "num_input_tokens_seen": 11182088,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 0.1316346377134323,
      "learning_rate": 0.0006730714015300481,
      "loss": 0.423,
      "num_input_tokens_seen": 11185064,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 0.17921511828899384,
      "learning_rate": 0.0006728568346663299,
      "loss": 0.4626,
      "num_input_tokens_seen": 11187944,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 0.09069656580686569,
      "learning_rate": 0.0006726422316429337,
      "loss": 0.3626,
      "num_input_tokens_seen": 11190632,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 0.07393737882375717,
      "learning_rate": 0.0006724275925047523,
      "loss": 0.6272,
      "num_input_tokens_seen": 11194280,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 0.10698480159044266,
      "learning_rate": 0.0006722129172966858,
      "loss": 0.656,
      "num_input_tokens_seen": 11197416,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 0.18943049013614655,
      "learning_rate": 0.0006719982060636416,
      "loss": 0.5653,
      "num_input_tokens_seen": 11200328,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 0.18503382802009583,
      "learning_rate": 0.000671783458850535,
      "loss": 0.4903,
      "num_input_tokens_seen": 11204200,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 0.11136595904827118,
      "learning_rate": 0.0006715686757022886,
      "loss": 0.5199,
      "num_input_tokens_seen": 11207144,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 0.11365007609128952,
      "learning_rate": 0.0006713538566638326,
      "loss": 0.4077,
      "num_input_tokens_seen": 11210888,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 0.10706069320440292,
      "learning_rate": 0.0006711390017801049,
      "loss": 0.3771,
      "num_input_tokens_seen": 11213512,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.48470157384872437,
      "eval_runtime": 13.6891,
      "eval_samples_per_second": 61.947,
      "eval_steps_per_second": 15.487,
      "num_input_tokens_seen": 11214472,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 0.12300987541675568,
      "learning_rate": 0.0006709241110960502,
      "loss": 0.4193,
      "num_input_tokens_seen": 11217000,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 0.1024356558918953,
      "learning_rate": 0.0006707091846566216,
      "loss": 0.3626,
      "num_input_tokens_seen": 11220424,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 0.11652526259422302,
      "learning_rate": 0.0006704942225067791,
      "loss": 0.4578,
      "num_input_tokens_seen": 11222888,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 0.11158929020166397,
      "learning_rate": 0.0006702792246914902,
      "loss": 0.3387,
      "num_input_tokens_seen": 11225480,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 0.07651310414075851,
      "learning_rate": 0.0006700641912557304,
      "loss": 0.5879,
      "num_input_tokens_seen": 11228712,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 0.12223070859909058,
      "learning_rate": 0.0006698491222444818,
      "loss": 0.4227,
      "num_input_tokens_seen": 11231464,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 0.10597201436758041,
      "learning_rate": 0.0006696340177027346,
      "loss": 0.6398,
      "num_input_tokens_seen": 11234568,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 0.11653821170330048,
      "learning_rate": 0.0006694188776754863,
      "loss": 0.5371,
      "num_input_tokens_seen": 11237096,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 0.14709728956222534,
      "learning_rate": 0.0006692037022077415,
      "loss": 0.3618,
      "num_input_tokens_seen": 11239880,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 0.11246100068092346,
      "learning_rate": 0.0006689884913445126,
      "loss": 0.477,
      "num_input_tokens_seen": 11243656,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 0.08112238347530365,
      "learning_rate": 0.0006687732451308193,
      "loss": 0.4238,
      "num_input_tokens_seen": 11246888,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 0.10734150558710098,
      "learning_rate": 0.0006685579636116886,
      "loss": 0.4477,
      "num_input_tokens_seen": 11249608,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 0.12064169347286224,
      "learning_rate": 0.0006683426468321547,
      "loss": 0.3676,
      "num_input_tokens_seen": 11252360,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 0.20570455491542816,
      "learning_rate": 0.0006681272948372598,
      "loss": 0.4688,
      "num_input_tokens_seen": 11254952,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 0.107622429728508,
      "learning_rate": 0.000667911907672053,
      "loss": 0.4822,
      "num_input_tokens_seen": 11257928,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 0.16565725207328796,
      "learning_rate": 0.0006676964853815906,
      "loss": 0.5538,
      "num_input_tokens_seen": 11261224,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 0.08582580089569092,
      "learning_rate": 0.0006674810280109367,
      "loss": 0.4114,
      "num_input_tokens_seen": 11264104,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 0.13149309158325195,
      "learning_rate": 0.0006672655356051625,
      "loss": 0.444,
      "num_input_tokens_seen": 11267752,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 0.09589628875255585,
      "learning_rate": 0.0006670500082093465,
      "loss": 0.3873,
      "num_input_tokens_seen": 11271304,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 0.055307887494564056,
      "learning_rate": 0.0006668344458685745,
      "loss": 0.4818,
      "num_input_tokens_seen": 11275272,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 0.15472424030303955,
      "learning_rate": 0.00066661884862794,
      "loss": 0.3539,
      "num_input_tokens_seen": 11277544,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.2303035706281662,
      "learning_rate": 0.000666403216532543,
      "loss": 0.5871,
      "num_input_tokens_seen": 11281608,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 0.17261120676994324,
      "learning_rate": 0.0006661875496274916,
      "loss": 0.5495,
      "num_input_tokens_seen": 11284328,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 0.1433330774307251,
      "learning_rate": 0.0006659718479579008,
      "loss": 0.5219,
      "num_input_tokens_seen": 11287464,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 0.19469676911830902,
      "learning_rate": 0.0006657561115688929,
      "loss": 0.4126,
      "num_input_tokens_seen": 11290536,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 0.16044816374778748,
      "learning_rate": 0.0006655403405055977,
      "loss": 0.3299,
      "num_input_tokens_seen": 11296520,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 0.08687323331832886,
      "learning_rate": 0.0006653245348131517,
      "loss": 0.3984,
      "num_input_tokens_seen": 11299144,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 0.0917791873216629,
      "learning_rate": 0.0006651086945366991,
      "loss": 0.5685,
      "num_input_tokens_seen": 11302856,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 0.08662895858287811,
      "learning_rate": 0.0006648928197213914,
      "loss": 0.4558,
      "num_input_tokens_seen": 11306824,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 0.2520500719547272,
      "learning_rate": 0.0006646769104123868,
      "loss": 0.5534,
      "num_input_tokens_seen": 11309672,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 0.12014017254114151,
      "learning_rate": 0.0006644609666548513,
      "loss": 0.3516,
      "num_input_tokens_seen": 11312776,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 0.10752210021018982,
      "learning_rate": 0.000664244988493958,
      "loss": 0.3535,
      "num_input_tokens_seen": 11319304,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 0.15549226105213165,
      "learning_rate": 0.0006640289759748867,
      "loss": 0.6658,
      "num_input_tokens_seen": 11322440,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 0.0965450257062912,
      "learning_rate": 0.0006638129291428252,
      "loss": 0.4236,
      "num_input_tokens_seen": 11325736,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 0.18180936574935913,
      "learning_rate": 0.0006635968480429677,
      "loss": 0.5434,
      "num_input_tokens_seen": 11328616,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 0.08311229199171066,
      "learning_rate": 0.0006633807327205162,
      "loss": 0.4844,
      "num_input_tokens_seen": 11332072,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 0.09029735624790192,
      "learning_rate": 0.0006631645832206789,
      "loss": 0.5462,
      "num_input_tokens_seen": 11335624,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 0.2153969258069992,
      "learning_rate": 0.0006629483995886727,
      "loss": 0.4338,
      "num_input_tokens_seen": 11337800,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 0.12511861324310303,
      "learning_rate": 0.0006627321818697202,
      "loss": 0.4747,
      "num_input_tokens_seen": 11341640,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 0.1184840202331543,
      "learning_rate": 0.0006625159301090518,
      "loss": 0.5508,
      "num_input_tokens_seen": 11344680,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 0.07314246892929077,
      "learning_rate": 0.0006622996443519047,
      "loss": 0.5198,
      "num_input_tokens_seen": 11348136,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 0.11709234863519669,
      "learning_rate": 0.0006620833246435238,
      "loss": 0.4049,
      "num_input_tokens_seen": 11351560,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 0.0832161232829094,
      "learning_rate": 0.0006618669710291606,
      "loss": 0.4735,
      "num_input_tokens_seen": 11355240,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 0.19349870085716248,
      "learning_rate": 0.0006616505835540736,
      "loss": 0.4631,
      "num_input_tokens_seen": 11357832,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 0.1360904574394226,
      "learning_rate": 0.0006614341622635287,
      "loss": 0.5145,
      "num_input_tokens_seen": 11361832,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 0.17125554382801056,
      "learning_rate": 0.0006612177072027989,
      "loss": 0.4187,
      "num_input_tokens_seen": 11365288,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 0.11432459205389023,
      "learning_rate": 0.0006610012184171641,
      "loss": 0.3482,
      "num_input_tokens_seen": 11368584,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 0.11819737404584885,
      "learning_rate": 0.0006607846959519109,
      "loss": 0.3071,
      "num_input_tokens_seen": 11371560,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 0.17429494857788086,
      "learning_rate": 0.0006605681398523339,
      "loss": 0.3273,
      "num_input_tokens_seen": 11374280,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 0.15439234673976898,
      "learning_rate": 0.0006603515501637338,
      "loss": 0.493,
      "num_input_tokens_seen": 11377320,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 0.18340152502059937,
      "learning_rate": 0.0006601349269314187,
      "loss": 0.4885,
      "num_input_tokens_seen": 11380424,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 0.11228033900260925,
      "learning_rate": 0.0006599182702007042,
      "loss": 0.3168,
      "num_input_tokens_seen": 11383560,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 0.2660442888736725,
      "learning_rate": 0.0006597015800169116,
      "loss": 0.4303,
      "num_input_tokens_seen": 11386760,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 0.09475035220384598,
      "learning_rate": 0.0006594848564253705,
      "loss": 0.4555,
      "num_input_tokens_seen": 11389352,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 0.16083113849163055,
      "learning_rate": 0.0006592680994714171,
      "loss": 0.5289,
      "num_input_tokens_seen": 11392584,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 0.10245729237794876,
      "learning_rate": 0.0006590513092003943,
      "loss": 0.4557,
      "num_input_tokens_seen": 11395688,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 0.10071221739053726,
      "learning_rate": 0.0006588344856576521,
      "loss": 0.4043,
      "num_input_tokens_seen": 11398696,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 0.1709580421447754,
      "learning_rate": 0.0006586176288885475,
      "loss": 0.5093,
      "num_input_tokens_seen": 11402536,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 0.12784136831760406,
      "learning_rate": 0.0006584007389384446,
      "loss": 0.4833,
      "num_input_tokens_seen": 11405704,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 0.07640431076288223,
      "learning_rate": 0.0006581838158527144,
      "loss": 0.5133,
      "num_input_tokens_seen": 11408648,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 0.0978701189160347,
      "learning_rate": 0.0006579668596767343,
      "loss": 0.4746,
      "num_input_tokens_seen": 11411816,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 0.1308833807706833,
      "learning_rate": 0.0006577498704558894,
      "loss": 0.3955,
      "num_input_tokens_seen": 11414952,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 0.10133248567581177,
      "learning_rate": 0.0006575328482355715,
      "loss": 0.4437,
      "num_input_tokens_seen": 11418344,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.15104259550571442,
      "learning_rate": 0.0006573157930611788,
      "loss": 0.5206,
      "num_input_tokens_seen": 11421128,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 0.07719854265451431,
      "learning_rate": 0.000657098704978117,
      "loss": 0.3981,
      "num_input_tokens_seen": 11424072,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 0.17287948727607727,
      "learning_rate": 0.0006568815840317984,
      "loss": 0.5026,
      "num_input_tokens_seen": 11426952,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 0.08676675707101822,
      "learning_rate": 0.0006566644302676422,
      "loss": 0.3259,
      "num_input_tokens_seen": 11430984,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 0.11899687349796295,
      "learning_rate": 0.0006564472437310746,
      "loss": 0.2978,
      "num_input_tokens_seen": 11434184,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 0.10874760895967484,
      "learning_rate": 0.0006562300244675283,
      "loss": 0.638,
      "num_input_tokens_seen": 11436680,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 0.14059032499790192,
      "learning_rate": 0.0006560127725224432,
      "loss": 0.4052,
      "num_input_tokens_seen": 11439784,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 0.17050570249557495,
      "learning_rate": 0.0006557954879412662,
      "loss": 0.502,
      "num_input_tokens_seen": 11442440,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 0.1673784852027893,
      "learning_rate": 0.0006555781707694506,
      "loss": 0.4366,
      "num_input_tokens_seen": 11445096,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 0.15166135132312775,
      "learning_rate": 0.0006553608210524564,
      "loss": 0.5466,
      "num_input_tokens_seen": 11448712,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 0.12706761062145233,
      "learning_rate": 0.0006551434388357509,
      "loss": 0.4238,
      "num_input_tokens_seen": 11451080,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 0.10832885652780533,
      "learning_rate": 0.000654926024164808,
      "loss": 0.3315,
      "num_input_tokens_seen": 11454120,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 0.13873504102230072,
      "learning_rate": 0.0006547085770851084,
      "loss": 0.4316,
      "num_input_tokens_seen": 11457544,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 0.20401890575885773,
      "learning_rate": 0.0006544910976421395,
      "loss": 0.3645,
      "num_input_tokens_seen": 11461384,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 0.12434090673923492,
      "learning_rate": 0.0006542735858813953,
      "loss": 0.4506,
      "num_input_tokens_seen": 11464328,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 0.12167320400476456,
      "learning_rate": 0.0006540560418483771,
      "loss": 0.4507,
      "num_input_tokens_seen": 11467176,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 0.17762425541877747,
      "learning_rate": 0.0006538384655885925,
      "loss": 0.4381,
      "num_input_tokens_seen": 11469832,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 0.15011239051818848,
      "learning_rate": 0.000653620857147556,
      "loss": 0.3683,
      "num_input_tokens_seen": 11472680,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 0.10186190903186798,
      "learning_rate": 0.0006534032165707887,
      "loss": 0.4458,
      "num_input_tokens_seen": 11475656,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 0.08663310110569,
      "learning_rate": 0.0006531855439038187,
      "loss": 0.4916,
      "num_input_tokens_seen": 11480488,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 0.10621064901351929,
      "learning_rate": 0.0006529678391921805,
      "loss": 0.3878,
      "num_input_tokens_seen": 11484232,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 0.11034819483757019,
      "learning_rate": 0.0006527501024814155,
      "loss": 0.4571,
      "num_input_tokens_seen": 11486984,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 0.11373165994882584,
      "learning_rate": 0.0006525323338170718,
      "loss": 0.3651,
      "num_input_tokens_seen": 11489992,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 0.17349430918693542,
      "learning_rate": 0.000652314533244704,
      "loss": 0.4767,
      "num_input_tokens_seen": 11493960,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 0.10520753264427185,
      "learning_rate": 0.0006520967008098735,
      "loss": 0.5906,
      "num_input_tokens_seen": 11497416,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 0.14310133457183838,
      "learning_rate": 0.0006518788365581485,
      "loss": 0.3957,
      "num_input_tokens_seen": 11501384,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 0.1285993605852127,
      "learning_rate": 0.0006516609405351034,
      "loss": 0.3521,
      "num_input_tokens_seen": 11504744,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 0.0935884639620781,
      "learning_rate": 0.00065144301278632,
      "loss": 0.4977,
      "num_input_tokens_seen": 11507592,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 0.11842448264360428,
      "learning_rate": 0.0006512250533573862,
      "loss": 0.4413,
      "num_input_tokens_seen": 11510568,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 0.1434660106897354,
      "learning_rate": 0.0006510070622938962,
      "loss": 0.5459,
      "num_input_tokens_seen": 11512968,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.1264456808567047,
      "learning_rate": 0.0006507890396414516,
      "loss": 0.3915,
      "num_input_tokens_seen": 11515560,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 0.09020047634840012,
      "learning_rate": 0.0006505709854456602,
      "loss": 0.3795,
      "num_input_tokens_seen": 11519560,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 0.19135773181915283,
      "learning_rate": 0.0006503528997521365,
      "loss": 0.406,
      "num_input_tokens_seen": 11522568,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 0.0739833265542984,
      "learning_rate": 0.0006501347826065017,
      "loss": 0.4921,
      "num_input_tokens_seen": 11526408,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 0.11783110350370407,
      "learning_rate": 0.000649916634054383,
      "loss": 0.4187,
      "num_input_tokens_seen": 11529672,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 0.18306314945220947,
      "learning_rate": 0.0006496984541414147,
      "loss": 0.4594,
      "num_input_tokens_seen": 11533128,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 0.21095068752765656,
      "learning_rate": 0.0006494802429132378,
      "loss": 0.5512,
      "num_input_tokens_seen": 11535496,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 0.13154861330986023,
      "learning_rate": 0.0006492620004154993,
      "loss": 0.5087,
      "num_input_tokens_seen": 11538280,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 0.09018375724554062,
      "learning_rate": 0.0006490437266938534,
      "loss": 0.4077,
      "num_input_tokens_seen": 11543080,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 0.15641456842422485,
      "learning_rate": 0.0006488254217939601,
      "loss": 0.4955,
      "num_input_tokens_seen": 11546248,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 0.19629408419132233,
      "learning_rate": 0.0006486070857614863,
      "loss": 0.5238,
      "num_input_tokens_seen": 11549352,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 0.18829986453056335,
      "learning_rate": 0.0006483887186421059,
      "loss": 0.4462,
      "num_input_tokens_seen": 11552552,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 0.1695438027381897,
      "learning_rate": 0.0006481703204814982,
      "loss": 0.4579,
      "num_input_tokens_seen": 11554920,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 0.10975233465433121,
      "learning_rate": 0.00064795189132535,
      "loss": 0.4701,
      "num_input_tokens_seen": 11557800,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 0.13402464985847473,
      "learning_rate": 0.000647733431219354,
      "loss": 0.5354,
      "num_input_tokens_seen": 11560616,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 0.11337144672870636,
      "learning_rate": 0.0006475149402092097,
      "loss": 0.4934,
      "num_input_tokens_seen": 11564552,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 0.17232248187065125,
      "learning_rate": 0.0006472964183406229,
      "loss": 0.3841,
      "num_input_tokens_seen": 11567976,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 0.16743066906929016,
      "learning_rate": 0.0006470778656593055,
      "loss": 0.6287,
      "num_input_tokens_seen": 11571144,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 0.0981658473610878,
      "learning_rate": 0.0006468592822109769,
      "loss": 0.4433,
      "num_input_tokens_seen": 11574664,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 0.09206516295671463,
      "learning_rate": 0.0006466406680413619,
      "loss": 0.4228,
      "num_input_tokens_seen": 11577896,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 0.08323750644922256,
      "learning_rate": 0.000646422023196192,
      "loss": 0.4995,
      "num_input_tokens_seen": 11581608,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 0.2513657510280609,
      "learning_rate": 0.0006462033477212054,
      "loss": 0.303,
      "num_input_tokens_seen": 11584936,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 0.09627249091863632,
      "learning_rate": 0.0006459846416621462,
      "loss": 0.4762,
      "num_input_tokens_seen": 11588360,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 0.08978918194770813,
      "learning_rate": 0.0006457659050647657,
      "loss": 0.3546,
      "num_input_tokens_seen": 11590760,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 0.14567498862743378,
      "learning_rate": 0.0006455471379748208,
      "loss": 0.5726,
      "num_input_tokens_seen": 11593736,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 0.1143496036529541,
      "learning_rate": 0.000645328340438075,
      "loss": 0.4373,
      "num_input_tokens_seen": 11597192,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 0.21985788643360138,
      "learning_rate": 0.0006451095125002985,
      "loss": 0.4365,
      "num_input_tokens_seen": 11601000,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 0.15556126832962036,
      "learning_rate": 0.0006448906542072674,
      "loss": 0.5281,
      "num_input_tokens_seen": 11603624,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 0.15319068729877472,
      "learning_rate": 0.0006446717656047645,
      "loss": 0.4558,
      "num_input_tokens_seen": 11607656,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 0.1502566784620285,
      "learning_rate": 0.0006444528467385789,
      "loss": 0.5315,
      "num_input_tokens_seen": 11612168,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 0.17322231829166412,
      "learning_rate": 0.0006442338976545054,
      "loss": 0.4434,
      "num_input_tokens_seen": 11614984,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 0.10553935915231705,
      "learning_rate": 0.0006440149183983461,
      "loss": 0.5199,
      "num_input_tokens_seen": 11618056,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 0.11859353631734848,
      "learning_rate": 0.0006437959090159093,
      "loss": 0.3645,
      "num_input_tokens_seen": 11620616,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 0.10931362211704254,
      "learning_rate": 0.0006435768695530085,
      "loss": 0.3747,
      "num_input_tokens_seen": 11624264,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 0.15304356813430786,
      "learning_rate": 0.0006433578000554645,
      "loss": 0.4476,
      "num_input_tokens_seen": 11626600,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 0.10974249988794327,
      "learning_rate": 0.0006431387005691045,
      "loss": 0.3806,
      "num_input_tokens_seen": 11629416,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 0.11021328717470169,
      "learning_rate": 0.0006429195711397611,
      "loss": 0.3288,
      "num_input_tokens_seen": 11633800,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 0.15123145282268524,
      "learning_rate": 0.0006427004118132742,
      "loss": 0.5756,
      "num_input_tokens_seen": 11637896,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 0.09965746104717255,
      "learning_rate": 0.0006424812226354889,
      "loss": 0.364,
      "num_input_tokens_seen": 11640904,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 0.06595002114772797,
      "learning_rate": 0.0006422620036522574,
      "loss": 0.5777,
      "num_input_tokens_seen": 11644744,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 0.12606364488601685,
      "learning_rate": 0.000642042754909438,
      "loss": 0.3922,
      "num_input_tokens_seen": 11648264,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 0.08213119953870773,
      "learning_rate": 0.0006418234764528945,
      "loss": 0.4071,
      "num_input_tokens_seen": 11651784,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 0.11485958844423294,
      "learning_rate": 0.0006416041683284978,
      "loss": 0.4658,
      "num_input_tokens_seen": 11655400,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 0.07699627429246902,
      "learning_rate": 0.0006413848305821246,
      "loss": 0.33,
      "num_input_tokens_seen": 11658568,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 0.1294824481010437,
      "learning_rate": 0.0006411654632596578,
      "loss": 0.4764,
      "num_input_tokens_seen": 11661448,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 0.09675014019012451,
      "learning_rate": 0.0006409460664069867,
      "loss": 0.4628,
      "num_input_tokens_seen": 11664520,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 0.08706793189048767,
      "learning_rate": 0.0006407266400700064,
      "loss": 0.4264,
      "num_input_tokens_seen": 11667816,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 0.13353019952774048,
      "learning_rate": 0.0006405071842946185,
      "loss": 0.4307,
      "num_input_tokens_seen": 11670632,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 0.10676655173301697,
      "learning_rate": 0.0006402876991267308,
      "loss": 0.3389,
      "num_input_tokens_seen": 11673800,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 0.16351211071014404,
      "learning_rate": 0.0006400681846122568,
      "loss": 0.4842,
      "num_input_tokens_seen": 11677224,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 0.13162314891815186,
      "learning_rate": 0.0006398486407971166,
      "loss": 0.4135,
      "num_input_tokens_seen": 11680648,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 0.10508529096841812,
      "learning_rate": 0.0006396290677272361,
      "loss": 0.3986,
      "num_input_tokens_seen": 11684616,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 0.1262485682964325,
      "learning_rate": 0.0006394094654485477,
      "loss": 0.3688,
      "num_input_tokens_seen": 11688296,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 0.1664770245552063,
      "learning_rate": 0.0006391898340069896,
      "loss": 0.3959,
      "num_input_tokens_seen": 11691240,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 0.16549867391586304,
      "learning_rate": 0.0006389701734485061,
      "loss": 0.3738,
      "num_input_tokens_seen": 11693704,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 0.1034146249294281,
      "learning_rate": 0.0006387504838190479,
      "loss": 0.4102,
      "num_input_tokens_seen": 11696776,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 0.1701921969652176,
      "learning_rate": 0.0006385307651645715,
      "loss": 0.3734,
      "num_input_tokens_seen": 11699496,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 0.11938418447971344,
      "learning_rate": 0.0006383110175310393,
      "loss": 0.4025,
      "num_input_tokens_seen": 11703176,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 0.2796365022659302,
      "learning_rate": 0.0006380912409644202,
      "loss": 0.4878,
      "num_input_tokens_seen": 11706728,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 0.162472665309906,
      "learning_rate": 0.000637871435510689,
      "loss": 0.3987,
      "num_input_tokens_seen": 11709864,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 0.13315190374851227,
      "learning_rate": 0.0006376516012158262,
      "loss": 0.476,
      "num_input_tokens_seen": 11714152,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 0.0700833648443222,
      "learning_rate": 0.000637431738125819,
      "loss": 0.4998,
      "num_input_tokens_seen": 11718504,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 0.16252218186855316,
      "learning_rate": 0.00063721184628666,
      "loss": 0.3785,
      "num_input_tokens_seen": 11721640,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 0.12308686971664429,
      "learning_rate": 0.0006369919257443484,
      "loss": 0.3954,
      "num_input_tokens_seen": 11725352,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 0.12635537981987,
      "learning_rate": 0.0006367719765448886,
      "loss": 0.5032,
      "num_input_tokens_seen": 11728168,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 0.19922678172588348,
      "learning_rate": 0.0006365519987342916,
      "loss": 0.5263,
      "num_input_tokens_seen": 11730920,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 0.15870779752731323,
      "learning_rate": 0.0006363319923585746,
      "loss": 0.4328,
      "num_input_tokens_seen": 11733576,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 0.10138425976037979,
      "learning_rate": 0.0006361119574637599,
      "loss": 0.3616,
      "num_input_tokens_seen": 11737096,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 0.10073952376842499,
      "learning_rate": 0.0006358918940958767,
      "loss": 0.512,
      "num_input_tokens_seen": 11741608,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 0.12682335078716278,
      "learning_rate": 0.0006356718023009597,
      "loss": 0.5684,
      "num_input_tokens_seen": 11744776,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 0.15581947565078735,
      "learning_rate": 0.0006354516821250492,
      "loss": 0.4247,
      "num_input_tokens_seen": 11747432,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 0.14171980321407318,
      "learning_rate": 0.0006352315336141923,
      "loss": 0.4206,
      "num_input_tokens_seen": 11750088,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.16444091498851776,
      "learning_rate": 0.0006350113568144414,
      "loss": 0.2861,
      "num_input_tokens_seen": 11753480,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.08887580782175064,
      "learning_rate": 0.0006347911517718547,
      "loss": 0.3948,
      "num_input_tokens_seen": 11756168,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 0.08034353703260422,
      "learning_rate": 0.0006345709185324972,
      "loss": 0.5451,
      "num_input_tokens_seen": 11761032,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 0.10846971720457077,
      "learning_rate": 0.0006343506571424386,
      "loss": 0.4909,
      "num_input_tokens_seen": 11764392,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.1784958392381668,
      "learning_rate": 0.0006341303676477552,
      "loss": 0.4525,
      "num_input_tokens_seen": 11767272,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 0.18627293407917023,
      "learning_rate": 0.0006339100500945293,
      "loss": 0.6169,
      "num_input_tokens_seen": 11770248,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 0.0798565074801445,
      "learning_rate": 0.0006336897045288486,
      "loss": 0.4218,
      "num_input_tokens_seen": 11773704,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 0.11299910396337509,
      "learning_rate": 0.0006334693309968068,
      "loss": 0.3642,
      "num_input_tokens_seen": 11776328,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 0.14486846327781677,
      "learning_rate": 0.0006332489295445038,
      "loss": 0.4545,
      "num_input_tokens_seen": 11779400,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 0.09679491072893143,
      "learning_rate": 0.0006330285002180448,
      "loss": 0.3624,
      "num_input_tokens_seen": 11781672,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 0.09086473286151886,
      "learning_rate": 0.0006328080430635412,
      "loss": 0.5188,
      "num_input_tokens_seen": 11784552,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 0.12095815688371658,
      "learning_rate": 0.0006325875581271103,
      "loss": 0.5165,
      "num_input_tokens_seen": 11787624,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 0.06724695861339569,
      "learning_rate": 0.0006323670454548749,
      "loss": 0.5417,
      "num_input_tokens_seen": 11791784,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 0.18488620221614838,
      "learning_rate": 0.0006321465050929635,
      "loss": 0.382,
      "num_input_tokens_seen": 11794824,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 0.1412302702665329,
      "learning_rate": 0.000631925937087511,
      "loss": 0.4133,
      "num_input_tokens_seen": 11803304,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 0.12457461655139923,
      "learning_rate": 0.0006317053414846576,
      "loss": 0.4393,
      "num_input_tokens_seen": 11807304,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.14798159897327423,
      "learning_rate": 0.0006314847183305492,
      "loss": 0.4642,
      "num_input_tokens_seen": 11810376,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 0.10991599410772324,
      "learning_rate": 0.0006312640676713378,
      "loss": 0.3905,
      "num_input_tokens_seen": 11813352,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 0.11046745628118515,
      "learning_rate": 0.0006310433895531811,
      "loss": 0.4199,
      "num_input_tokens_seen": 11816392,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 0.17358896136283875,
      "learning_rate": 0.0006308226840222422,
      "loss": 0.4911,
      "num_input_tokens_seen": 11819816,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 0.10008607059717178,
      "learning_rate": 0.0006306019511246905,
      "loss": 0.4251,
      "num_input_tokens_seen": 11823752,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 0.1235399916768074,
      "learning_rate": 0.0006303811909067003,
      "loss": 0.4492,
      "num_input_tokens_seen": 11826248,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 0.06767675280570984,
      "learning_rate": 0.0006301604034144527,
      "loss": 0.3951,
      "num_input_tokens_seen": 11830504,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 0.10699004679918289,
      "learning_rate": 0.0006299395886941336,
      "loss": 0.4382,
      "num_input_tokens_seen": 11832872,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 0.12163907289505005,
      "learning_rate": 0.0006297187467919347,
      "loss": 0.5824,
      "num_input_tokens_seen": 11835688,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 0.10451285541057587,
      "learning_rate": 0.0006294978777540542,
      "loss": 0.4658,
      "num_input_tokens_seen": 11839720,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 0.1089344173669815,
      "learning_rate": 0.000629276981626695,
      "loss": 0.4428,
      "num_input_tokens_seen": 11842856,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 0.12046122550964355,
      "learning_rate": 0.0006290560584560661,
      "loss": 0.4442,
      "num_input_tokens_seen": 11846216,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 0.1483493447303772,
      "learning_rate": 0.0006288351082883822,
      "loss": 0.4021,
      "num_input_tokens_seen": 11848776,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 0.06152665242552757,
      "learning_rate": 0.0006286141311698633,
      "loss": 0.4998,
      "num_input_tokens_seen": 11852424,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 0.06721996515989304,
      "learning_rate": 0.0006283931271467355,
      "loss": 0.691,
      "num_input_tokens_seen": 11857064,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 0.10896816849708557,
      "learning_rate": 0.0006281720962652305,
      "loss": 0.3738,
      "num_input_tokens_seen": 11860040,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 0.06904657930135727,
      "learning_rate": 0.0006279510385715849,
      "loss": 0.3913,
      "num_input_tokens_seen": 11863688,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 0.0855000838637352,
      "learning_rate": 0.0006277299541120419,
      "loss": 0.319,
      "num_input_tokens_seen": 11868136,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 0.11048915982246399,
      "learning_rate": 0.0006275088429328499,
      "loss": 0.5036,
      "num_input_tokens_seen": 11871176,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 0.13024479150772095,
      "learning_rate": 0.0006272877050802623,
      "loss": 0.393,
      "num_input_tokens_seen": 11874536,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 0.10791781544685364,
      "learning_rate": 0.0006270665406005393,
      "loss": 0.4441,
      "num_input_tokens_seen": 11877992,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 0.11048801243305206,
      "learning_rate": 0.0006268453495399455,
      "loss": 0.385,
      "num_input_tokens_seen": 11880584,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 0.16069172322750092,
      "learning_rate": 0.0006266241319447517,
      "loss": 0.7339,
      "num_input_tokens_seen": 11884200,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 0.1680268943309784,
      "learning_rate": 0.0006264028878612343,
      "loss": 0.4273,
      "num_input_tokens_seen": 11886760,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 0.15647703409194946,
      "learning_rate": 0.0006261816173356746,
      "loss": 0.4377,
      "num_input_tokens_seen": 11889512,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.17687848210334778,
      "learning_rate": 0.0006259603204143602,
      "loss": 0.4996,
      "num_input_tokens_seen": 11892776,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 0.15774869918823242,
      "learning_rate": 0.0006257389971435839,
      "loss": 0.3639,
      "num_input_tokens_seen": 11895432,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 0.11926122009754181,
      "learning_rate": 0.0006255176475696437,
      "loss": 0.4117,
      "num_input_tokens_seen": 11899112,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 0.3396330177783966,
      "learning_rate": 0.0006252962717388438,
      "loss": 0.4383,
      "num_input_tokens_seen": 11901928,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 0.10551569610834122,
      "learning_rate": 0.0006250748696974932,
      "loss": 0.6031,
      "num_input_tokens_seen": 11905928,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 0.14922383427619934,
      "learning_rate": 0.0006248534414919067,
      "loss": 0.6252,
      "num_input_tokens_seen": 11909416,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 0.15037840604782104,
      "learning_rate": 0.0006246319871684047,
      "loss": 0.4365,
      "num_input_tokens_seen": 11912840,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 0.12003500014543533,
      "learning_rate": 0.000624410506773313,
      "loss": 0.635,
      "num_input_tokens_seen": 11916648,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.11226695030927658,
      "learning_rate": 0.0006241890003529626,
      "loss": 0.5066,
      "num_input_tokens_seen": 11920040,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 0.25133800506591797,
      "learning_rate": 0.00062396746795369,
      "loss": 0.4709,
      "num_input_tokens_seen": 11923048,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 0.13667508959770203,
      "learning_rate": 0.0006237459096218375,
      "loss": 0.4208,
      "num_input_tokens_seen": 11926792,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 0.10280783474445343,
      "learning_rate": 0.0006235243254037525,
      "loss": 0.4249,
      "num_input_tokens_seen": 11930056,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 0.11084027588367462,
      "learning_rate": 0.0006233027153457878,
      "loss": 0.5018,
      "num_input_tokens_seen": 11933096,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 0.09266644716262817,
      "learning_rate": 0.0006230810794943017,
      "loss": 0.3967,
      "num_input_tokens_seen": 11936072,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 0.14926964044570923,
      "learning_rate": 0.0006228594178956581,
      "loss": 0.3885,
      "num_input_tokens_seen": 11938824,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 0.07806407660245895,
      "learning_rate": 0.0006226377305962259,
      "loss": 0.5573,
      "num_input_tokens_seen": 11941224,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 0.12614426016807556,
      "learning_rate": 0.0006224160176423796,
      "loss": 0.5211,
      "num_input_tokens_seen": 11944264,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 0.13259144127368927,
      "learning_rate": 0.0006221942790804989,
      "loss": 0.4249,
      "num_input_tokens_seen": 11946984,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 0.14586402475833893,
      "learning_rate": 0.0006219725149569691,
      "loss": 0.5168,
      "num_input_tokens_seen": 11949864,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 0.1412881314754486,
      "learning_rate": 0.0006217507253181809,
      "loss": 0.4745,
      "num_input_tokens_seen": 11953096,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 0.17733338475227356,
      "learning_rate": 0.0006215289102105297,
      "loss": 0.5258,
      "num_input_tokens_seen": 11958024,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 0.10389561951160431,
      "learning_rate": 0.0006213070696804171,
      "loss": 0.3679,
      "num_input_tokens_seen": 11961960,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 0.06610537320375443,
      "learning_rate": 0.0006210852037742495,
      "loss": 0.5197,
      "num_input_tokens_seen": 11966120,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 0.19727981090545654,
      "learning_rate": 0.0006208633125384385,
      "loss": 0.6778,
      "num_input_tokens_seen": 11969288,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 0.1408163458108902,
      "learning_rate": 0.0006206413960194017,
      "loss": 0.5076,
      "num_input_tokens_seen": 11971880,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 0.10993533581495285,
      "learning_rate": 0.000620419454263561,
      "loss": 0.4584,
      "num_input_tokens_seen": 11975240,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 0.08358088880777359,
      "learning_rate": 0.0006201974873173443,
      "loss": 0.4546,
      "num_input_tokens_seen": 11979432,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 0.21400859951972961,
      "learning_rate": 0.0006199754952271847,
      "loss": 0.4902,
      "num_input_tokens_seen": 11983496,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 0.11797948181629181,
      "learning_rate": 0.00061975347803952,
      "loss": 0.4562,
      "num_input_tokens_seen": 11987208,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 0.133120596408844,
      "learning_rate": 0.000619531435800794,
      "loss": 0.3811,
      "num_input_tokens_seen": 11990408,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 0.1462554782629013,
      "learning_rate": 0.0006193093685574553,
      "loss": 0.4883,
      "num_input_tokens_seen": 11993192,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.07937740534543991,
      "learning_rate": 0.0006190872763559578,
      "loss": 0.5568,
      "num_input_tokens_seen": 11996872,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.07670843601226807,
      "learning_rate": 0.0006188651592427609,
      "loss": 0.4163,
      "num_input_tokens_seen": 12000776,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 0.1162135899066925,
      "learning_rate": 0.0006186430172643287,
      "loss": 0.3179,
      "num_input_tokens_seen": 12004200,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 0.08468711376190186,
      "learning_rate": 0.0006184208504671307,
      "loss": 0.3222,
      "num_input_tokens_seen": 12006952,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 0.11304430663585663,
      "learning_rate": 0.000618198658897642,
      "loss": 0.3982,
      "num_input_tokens_seen": 12009640,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 0.10068337619304657,
      "learning_rate": 0.0006179764426023422,
      "loss": 0.4846,
      "num_input_tokens_seen": 12012680,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 0.10416316241025925,
      "learning_rate": 0.0006177542016277167,
      "loss": 0.436,
      "num_input_tokens_seen": 12016168,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 0.10971858352422714,
      "learning_rate": 0.0006175319360202556,
      "loss": 0.4551,
      "num_input_tokens_seen": 12018664,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 0.1086028441786766,
      "learning_rate": 0.0006173096458264542,
      "loss": 0.489,
      "num_input_tokens_seen": 12022280,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 0.09583026170730591,
      "learning_rate": 0.0006170873310928133,
      "loss": 0.3749,
      "num_input_tokens_seen": 12025512,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 0.15099811553955078,
      "learning_rate": 0.0006168649918658387,
      "loss": 0.5443,
      "num_input_tokens_seen": 12027976,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 0.12015320360660553,
      "learning_rate": 0.0006166426281920407,
      "loss": 0.5422,
      "num_input_tokens_seen": 12031400,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 0.11067882180213928,
      "learning_rate": 0.000616420240117936,
      "loss": 0.4779,
      "num_input_tokens_seen": 12034344,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 0.16471874713897705,
      "learning_rate": 0.000616197827690045,
      "loss": 0.5029,
      "num_input_tokens_seen": 12039368,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 0.09780199080705643,
      "learning_rate": 0.0006159753909548943,
      "loss": 0.299,
      "num_input_tokens_seen": 12042760,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 0.24917037785053253,
      "learning_rate": 0.0006157529299590148,
      "loss": 0.3736,
      "num_input_tokens_seen": 12045800,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 0.18849970400333405,
      "learning_rate": 0.000615530444748943,
      "loss": 0.3984,
      "num_input_tokens_seen": 12049544,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 0.09222005307674408,
      "learning_rate": 0.0006153079353712201,
      "loss": 0.4704,
      "num_input_tokens_seen": 12052296,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 0.11459451168775558,
      "learning_rate": 0.0006150854018723926,
      "loss": 0.4943,
      "num_input_tokens_seen": 12055208,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 0.05597199499607086,
      "learning_rate": 0.0006148628442990122,
      "loss": 0.5035,
      "num_input_tokens_seen": 12059624,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 0.11559450626373291,
      "learning_rate": 0.0006146402626976351,
      "loss": 0.425,
      "num_input_tokens_seen": 12062504,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 0.15933966636657715,
      "learning_rate": 0.0006144176571148228,
      "loss": 0.3654,
      "num_input_tokens_seen": 12066024,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 0.10798667371273041,
      "learning_rate": 0.0006141950275971422,
      "loss": 0.3102,
      "num_input_tokens_seen": 12069000,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 0.1259399950504303,
      "learning_rate": 0.0006139723741911644,
      "loss": 0.4834,
      "num_input_tokens_seen": 12073160,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 0.18115262687206268,
      "learning_rate": 0.0006137496969434664,
      "loss": 0.5581,
      "num_input_tokens_seen": 12076552,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 0.17752565443515778,
      "learning_rate": 0.0006135269959006294,
      "loss": 0.4406,
      "num_input_tokens_seen": 12079592,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 0.18812905251979828,
      "learning_rate": 0.00061330427110924,
      "loss": 0.4899,
      "num_input_tokens_seen": 12082376,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 0.1980348527431488,
      "learning_rate": 0.0006130815226158899,
      "loss": 0.4785,
      "num_input_tokens_seen": 12085224,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 0.19137410819530487,
      "learning_rate": 0.0006128587504671753,
      "loss": 0.4471,
      "num_input_tokens_seen": 12087976,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 0.1597272753715515,
      "learning_rate": 0.0006126359547096974,
      "loss": 0.4444,
      "num_input_tokens_seen": 12090696,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 0.09106873720884323,
      "learning_rate": 0.0006124131353900632,
      "loss": 0.4268,
      "num_input_tokens_seen": 12094312,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 0.12764985859394073,
      "learning_rate": 0.0006121902925548833,
      "loss": 0.5107,
      "num_input_tokens_seen": 12097928,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 0.08013898879289627,
      "learning_rate": 0.0006119674262507743,
      "loss": 0.5502,
      "num_input_tokens_seen": 12101576,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 0.1555919051170349,
      "learning_rate": 0.0006117445365243571,
      "loss": 0.4241,
      "num_input_tokens_seen": 12104200,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 0.10624925047159195,
      "learning_rate": 0.0006115216234222575,
      "loss": 0.3376,
      "num_input_tokens_seen": 12108008,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 0.11426247656345367,
      "learning_rate": 0.0006112986869911068,
      "loss": 0.4373,
      "num_input_tokens_seen": 12111240,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 0.11681117117404938,
      "learning_rate": 0.0006110757272775405,
      "loss": 0.3933,
      "num_input_tokens_seen": 12114216,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 0.2122548371553421,
      "learning_rate": 0.0006108527443281992,
      "loss": 0.4875,
      "num_input_tokens_seen": 12116520,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.17263706028461456,
      "learning_rate": 0.0006106297381897284,
      "loss": 0.4898,
      "num_input_tokens_seen": 12119208,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 0.18242760002613068,
      "learning_rate": 0.0006104067089087787,
      "loss": 0.4682,
      "num_input_tokens_seen": 12122248,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 0.09593898802995682,
      "learning_rate": 0.0006101836565320049,
      "loss": 0.4062,
      "num_input_tokens_seen": 12125800,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 0.12214227765798569,
      "learning_rate": 0.0006099605811060673,
      "loss": 0.5318,
      "num_input_tokens_seen": 12128328,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 0.1528218686580658,
      "learning_rate": 0.0006097374826776307,
      "loss": 0.4216,
      "num_input_tokens_seen": 12131560,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 0.15902753174304962,
      "learning_rate": 0.0006095143612933646,
      "loss": 0.4385,
      "num_input_tokens_seen": 12134024,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 0.1580304205417633,
      "learning_rate": 0.0006092912169999435,
      "loss": 0.5623,
      "num_input_tokens_seen": 12136808,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 0.1471673846244812,
      "learning_rate": 0.0006090680498440465,
      "loss": 0.4756,
      "num_input_tokens_seen": 12140488,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 0.11993728578090668,
      "learning_rate": 0.0006088448598723581,
      "loss": 0.3293,
      "num_input_tokens_seen": 12144104,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 0.09450720250606537,
      "learning_rate": 0.0006086216471315666,
      "loss": 0.4631,
      "num_input_tokens_seen": 12147464,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 0.10302761942148209,
      "learning_rate": 0.0006083984116683659,
      "loss": 0.3998,
      "num_input_tokens_seen": 12150760,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 0.13130547106266022,
      "learning_rate": 0.0006081751535294539,
      "loss": 0.5643,
      "num_input_tokens_seen": 12153896,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 0.15336863696575165,
      "learning_rate": 0.0006079518727615341,
      "loss": 0.3765,
      "num_input_tokens_seen": 12156584,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 0.10200369358062744,
      "learning_rate": 0.0006077285694113141,
      "loss": 0.3757,
      "num_input_tokens_seen": 12160648,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 0.16882291436195374,
      "learning_rate": 0.0006075052435255062,
      "loss": 0.4625,
      "num_input_tokens_seen": 12163496,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 0.16762471199035645,
      "learning_rate": 0.000607281895150828,
      "loss": 0.6049,
      "num_input_tokens_seen": 12165800,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 0.109357550740242,
      "learning_rate": 0.0006070585243340013,
      "loss": 0.5575,
      "num_input_tokens_seen": 12169064,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 0.19289883971214294,
      "learning_rate": 0.0006068351311217526,
      "loss": 0.4776,
      "num_input_tokens_seen": 12171816,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 0.11086839437484741,
      "learning_rate": 0.0006066117155608135,
      "loss": 0.4154,
      "num_input_tokens_seen": 12175368,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 0.2874324917793274,
      "learning_rate": 0.0006063882776979196,
      "loss": 0.535,
      "num_input_tokens_seen": 12179752,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 0.08470305055379868,
      "learning_rate": 0.0006061648175798117,
      "loss": 0.4757,
      "num_input_tokens_seen": 12182888,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 0.20401541888713837,
      "learning_rate": 0.0006059413352532355,
      "loss": 0.3703,
      "num_input_tokens_seen": 12186024,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 0.16307547688484192,
      "learning_rate": 0.0006057178307649402,
      "loss": 0.3969,
      "num_input_tokens_seen": 12188680,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 0.12588031589984894,
      "learning_rate": 0.000605494304161681,
      "loss": 0.4739,
      "num_input_tokens_seen": 12192040,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 0.17539194226264954,
      "learning_rate": 0.0006052707554902169,
      "loss": 0.5169,
      "num_input_tokens_seen": 12195400,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 0.19466078281402588,
      "learning_rate": 0.0006050471847973117,
      "loss": 0.4593,
      "num_input_tokens_seen": 12198888,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 0.1322670578956604,
      "learning_rate": 0.000604823592129734,
      "loss": 0.5208,
      "num_input_tokens_seen": 12201800,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 0.13935109972953796,
      "learning_rate": 0.0006045999775342566,
      "loss": 0.3877,
      "num_input_tokens_seen": 12204328,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 0.06695950031280518,
      "learning_rate": 0.0006043763410576572,
      "loss": 0.3828,
      "num_input_tokens_seen": 12207432,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 0.2022753804922104,
      "learning_rate": 0.0006041526827467182,
      "loss": 0.3415,
      "num_input_tokens_seen": 12209832,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 0.13681721687316895,
      "learning_rate": 0.000603929002648226,
      "loss": 0.4573,
      "num_input_tokens_seen": 12212648,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 0.14972589910030365,
      "learning_rate": 0.0006037053008089723,
      "loss": 0.2992,
      "num_input_tokens_seen": 12216648,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 0.15276537835597992,
      "learning_rate": 0.0006034815772757528,
      "loss": 0.4698,
      "num_input_tokens_seen": 12219816,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 0.12135303020477295,
      "learning_rate": 0.0006032578320953678,
      "loss": 0.4687,
      "num_input_tokens_seen": 12224040,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.0884019210934639,
      "learning_rate": 0.0006030340653146224,
      "loss": 0.4876,
      "num_input_tokens_seen": 12228392,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 0.1119118258357048,
      "learning_rate": 0.0006028102769803261,
      "loss": 0.5814,
      "num_input_tokens_seen": 12231144,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 0.11404945701360703,
      "learning_rate": 0.0006025864671392925,
      "loss": 0.5176,
      "num_input_tokens_seen": 12235176,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 0.12313396483659744,
      "learning_rate": 0.0006023626358383406,
      "loss": 0.5061,
      "num_input_tokens_seen": 12238312,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 0.20531116425991058,
      "learning_rate": 0.0006021387831242932,
      "loss": 0.7069,
      "num_input_tokens_seen": 12240872,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 0.12111493200063705,
      "learning_rate": 0.0006019149090439774,
      "loss": 0.4121,
      "num_input_tokens_seen": 12244008,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 0.1574656218290329,
      "learning_rate": 0.0006016910136442254,
      "loss": 0.6491,
      "num_input_tokens_seen": 12247048,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 0.15116393566131592,
      "learning_rate": 0.0006014670969718735,
      "loss": 0.36,
      "num_input_tokens_seen": 12250056,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 0.11320869624614716,
      "learning_rate": 0.0006012431590737625,
      "loss": 0.3429,
      "num_input_tokens_seen": 12252744,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 0.11611253023147583,
      "learning_rate": 0.0006010191999967378,
      "loss": 0.3385,
      "num_input_tokens_seen": 12255368,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 0.1371999830007553,
      "learning_rate": 0.0006007952197876487,
      "loss": 0.3716,
      "num_input_tokens_seen": 12259304,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 0.11720919609069824,
      "learning_rate": 0.0006005712184933498,
      "loss": 0.4013,
      "num_input_tokens_seen": 12262376,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 0.12045598775148392,
      "learning_rate": 0.0006003471961606992,
      "loss": 0.3318,
      "num_input_tokens_seen": 12265672,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 0.11959918588399887,
      "learning_rate": 0.0006001231528365601,
      "loss": 0.5076,
      "num_input_tokens_seen": 12268904,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 0.15601424872875214,
      "learning_rate": 0.0005998990885677996,
      "loss": 0.4756,
      "num_input_tokens_seen": 12272264,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.07770662754774094,
      "learning_rate": 0.0005996750034012895,
      "loss": 0.4144,
      "num_input_tokens_seen": 12275720,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 0.11248792707920074,
      "learning_rate": 0.0005994508973839058,
      "loss": 0.4095,
      "num_input_tokens_seen": 12278440,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 0.23448801040649414,
      "learning_rate": 0.0005992267705625289,
      "loss": 0.4427,
      "num_input_tokens_seen": 12280776,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 0.06805692613124847,
      "learning_rate": 0.0005990026229840439,
      "loss": 0.5527,
      "num_input_tokens_seen": 12284104,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 0.2342698872089386,
      "learning_rate": 0.0005987784546953394,
      "loss": 0.3789,
      "num_input_tokens_seen": 12287848,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 0.1196206659078598,
      "learning_rate": 0.0005985542657433091,
      "loss": 0.4727,
      "num_input_tokens_seen": 12290664,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 0.09656766057014465,
      "learning_rate": 0.000598330056174851,
      "loss": 0.4057,
      "num_input_tokens_seen": 12293928,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 0.14599813520908356,
      "learning_rate": 0.0005981058260368668,
      "loss": 0.3537,
      "num_input_tokens_seen": 12296488,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 0.07751766592264175,
      "learning_rate": 0.0005978815753762632,
      "loss": 0.4914,
      "num_input_tokens_seen": 12300008,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 0.0976889505982399,
      "learning_rate": 0.0005976573042399505,
      "loss": 0.3619,
      "num_input_tokens_seen": 12303240,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 0.11438395082950592,
      "learning_rate": 0.000597433012674844,
      "loss": 0.477,
      "num_input_tokens_seen": 12306312,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 0.15003958344459534,
      "learning_rate": 0.0005972087007278628,
      "loss": 0.39,
      "num_input_tokens_seen": 12309000,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 0.08094257861375809,
      "learning_rate": 0.0005969843684459305,
      "loss": 0.4153,
      "num_input_tokens_seen": 12312904,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 0.11087790131568909,
      "learning_rate": 0.0005967600158759746,
      "loss": 0.3903,
      "num_input_tokens_seen": 12315624,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 0.15141744911670685,
      "learning_rate": 0.0005965356430649276,
      "loss": 0.4963,
      "num_input_tokens_seen": 12319336,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 0.12688153982162476,
      "learning_rate": 0.0005963112500597251,
      "loss": 0.6399,
      "num_input_tokens_seen": 12322088,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 0.11798018217086792,
      "learning_rate": 0.0005960868369073079,
      "loss": 0.4581,
      "num_input_tokens_seen": 12324648,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 0.20742297172546387,
      "learning_rate": 0.0005958624036546208,
      "loss": 0.4291,
      "num_input_tokens_seen": 12328392,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 0.10113275796175003,
      "learning_rate": 0.0005956379503486123,
      "loss": 0.3235,
      "num_input_tokens_seen": 12332200,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 0.199302077293396,
      "learning_rate": 0.0005954134770362359,
      "loss": 0.3401,
      "num_input_tokens_seen": 12335304,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 0.09613559395074844,
      "learning_rate": 0.0005951889837644486,
      "loss": 0.3496,
      "num_input_tokens_seen": 12338440,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 0.1737687587738037,
      "learning_rate": 0.0005949644705802116,
      "loss": 0.4681,
      "num_input_tokens_seen": 12345192,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 0.1030992716550827,
      "learning_rate": 0.0005947399375304909,
      "loss": 0.4352,
      "num_input_tokens_seen": 12349000,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 0.11714126914739609,
      "learning_rate": 0.0005945153846622563,
      "loss": 0.4082,
      "num_input_tokens_seen": 12351880,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 0.2889925539493561,
      "learning_rate": 0.0005942908120224815,
      "loss": 0.587,
      "num_input_tokens_seen": 12354600,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 0.20518441498279572,
      "learning_rate": 0.0005940662196581443,
      "loss": 0.3585,
      "num_input_tokens_seen": 12357416,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 0.12549865245819092,
      "learning_rate": 0.0005938416076162273,
      "loss": 0.5251,
      "num_input_tokens_seen": 12360616,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 0.16258326172828674,
      "learning_rate": 0.0005936169759437165,
      "loss": 0.4579,
      "num_input_tokens_seen": 12363144,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 0.10843506455421448,
      "learning_rate": 0.0005933923246876024,
      "loss": 0.3993,
      "num_input_tokens_seen": 12366856,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 0.18723738193511963,
      "learning_rate": 0.0005931676538948795,
      "loss": 0.5501,
      "num_input_tokens_seen": 12370824,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 0.11195605993270874,
      "learning_rate": 0.0005929429636125463,
      "loss": 0.4601,
      "num_input_tokens_seen": 12374568,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 0.10533807426691055,
      "learning_rate": 0.0005927182538876055,
      "loss": 0.6144,
      "num_input_tokens_seen": 12377096,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 0.268115371465683,
      "learning_rate": 0.0005924935247670637,
      "loss": 0.4336,
      "num_input_tokens_seen": 12380072,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 0.1125943660736084,
      "learning_rate": 0.0005922687762979317,
      "loss": 0.3917,
      "num_input_tokens_seen": 12383240,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 0.1740208864212036,
      "learning_rate": 0.0005920440085272245,
      "loss": 0.4668,
      "num_input_tokens_seen": 12387368,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 0.14598214626312256,
      "learning_rate": 0.0005918192215019608,
      "loss": 0.4693,
      "num_input_tokens_seen": 12390280,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 0.10503258556127548,
      "learning_rate": 0.0005915944152691633,
      "loss": 0.3611,
      "num_input_tokens_seen": 12396232,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 0.14084947109222412,
      "learning_rate": 0.0005913695898758593,
      "loss": 0.4405,
      "num_input_tokens_seen": 12398760,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 0.24161356687545776,
      "learning_rate": 0.0005911447453690793,
      "loss": 0.4877,
      "num_input_tokens_seen": 12402184,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 0.17117151618003845,
      "learning_rate": 0.0005909198817958585,
      "loss": 0.5124,
      "num_input_tokens_seen": 12405064,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 0.24191926419734955,
      "learning_rate": 0.0005906949992032359,
      "loss": 0.5467,
      "num_input_tokens_seen": 12407976,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 0.16202181577682495,
      "learning_rate": 0.0005904700976382539,
      "loss": 0.5232,
      "num_input_tokens_seen": 12411976,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 0.08085937052965164,
      "learning_rate": 0.0005902451771479596,
      "loss": 0.5234,
      "num_input_tokens_seen": 12416136,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 0.12399529665708542,
      "learning_rate": 0.000590020237779404,
      "loss": 0.459,
      "num_input_tokens_seen": 12418952,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 0.20124852657318115,
      "learning_rate": 0.0005897952795796415,
      "loss": 0.3963,
      "num_input_tokens_seen": 12421768,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.18419094383716583,
      "learning_rate": 0.0005895703025957311,
      "loss": 0.5123,
      "num_input_tokens_seen": 12425416,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 0.20506899058818817,
      "learning_rate": 0.0005893453068747352,
      "loss": 0.4713,
      "num_input_tokens_seen": 12427848,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 0.16195552051067352,
      "learning_rate": 0.0005891202924637202,
      "loss": 0.3528,
      "num_input_tokens_seen": 12431464,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 0.18053720891475677,
      "learning_rate": 0.0005888952594097572,
      "loss": 0.3947,
      "num_input_tokens_seen": 12434792,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 0.1458410620689392,
      "learning_rate": 0.0005886702077599197,
      "loss": 0.5323,
      "num_input_tokens_seen": 12437928,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 0.15457269549369812,
      "learning_rate": 0.0005884451375612865,
      "loss": 0.4135,
      "num_input_tokens_seen": 12441160,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 0.20233674347400665,
      "learning_rate": 0.0005882200488609396,
      "loss": 0.5346,
      "num_input_tokens_seen": 12444712,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 0.12707574665546417,
      "learning_rate": 0.0005879949417059649,
      "loss": 0.4806,
      "num_input_tokens_seen": 12447688,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 0.10558116436004639,
      "learning_rate": 0.0005877698161434522,
      "loss": 0.4668,
      "num_input_tokens_seen": 12450984,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 0.10654469579458237,
      "learning_rate": 0.0005875446722204955,
      "loss": 0.5517,
      "num_input_tokens_seen": 12455176,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 0.12903942167758942,
      "learning_rate": 0.0005873195099841919,
      "loss": 0.4319,
      "num_input_tokens_seen": 12458376,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 0.12002690136432648,
      "learning_rate": 0.0005870943294816431,
      "loss": 0.4432,
      "num_input_tokens_seen": 12460968,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.37165889143943787,
      "learning_rate": 0.0005868691307599543,
      "loss": 0.3931,
      "num_input_tokens_seen": 12463312,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.4845305383205414,
      "eval_runtime": 13.7241,
      "eval_samples_per_second": 61.789,
      "eval_steps_per_second": 15.447,
      "num_input_tokens_seen": 12463312,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 0.07818173617124557,
      "learning_rate": 0.0005866439138662343,
      "loss": 0.3677,
      "num_input_tokens_seen": 12466928,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 0.1255151629447937,
      "learning_rate": 0.000586418678847596,
      "loss": 0.4061,
      "num_input_tokens_seen": 12469776,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 0.10591090470552444,
      "learning_rate": 0.000586193425751156,
      "loss": 0.4473,
      "num_input_tokens_seen": 12472528,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 0.08624587953090668,
      "learning_rate": 0.0005859681546240348,
      "loss": 0.383,
      "num_input_tokens_seen": 12475760,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 0.09550748765468597,
      "learning_rate": 0.0005857428655133562,
      "loss": 0.4966,
      "num_input_tokens_seen": 12479056,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 0.10062125325202942,
      "learning_rate": 0.0005855175584662485,
      "loss": 0.4916,
      "num_input_tokens_seen": 12482224,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 0.08729492127895355,
      "learning_rate": 0.0005852922335298433,
      "loss": 0.374,
      "num_input_tokens_seen": 12485360,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 0.11415597051382065,
      "learning_rate": 0.0005850668907512757,
      "loss": 0.4176,
      "num_input_tokens_seen": 12487920,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 0.11684195697307587,
      "learning_rate": 0.0005848415301776852,
      "loss": 0.5133,
      "num_input_tokens_seen": 12491024,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 0.10661987215280533,
      "learning_rate": 0.0005846161518562144,
      "loss": 0.4717,
      "num_input_tokens_seen": 12493744,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 0.1336270123720169,
      "learning_rate": 0.00058439075583401,
      "loss": 0.3639,
      "num_input_tokens_seen": 12496592,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 0.12663450837135315,
      "learning_rate": 0.0005841653421582223,
      "loss": 0.6352,
      "num_input_tokens_seen": 12499728,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 0.08926717936992645,
      "learning_rate": 0.0005839399108760052,
      "loss": 0.5339,
      "num_input_tokens_seen": 12503248,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 0.09825651347637177,
      "learning_rate": 0.0005837144620345165,
      "loss": 0.4013,
      "num_input_tokens_seen": 12506544,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 0.09046178311109543,
      "learning_rate": 0.0005834889956809173,
      "loss": 0.4015,
      "num_input_tokens_seen": 12510512,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 0.09444357454776764,
      "learning_rate": 0.0005832635118623726,
      "loss": 0.4528,
      "num_input_tokens_seen": 12514096,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 0.08236242830753326,
      "learning_rate": 0.0005830380106260513,
      "loss": 0.399,
      "num_input_tokens_seen": 12518352,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 0.11172337085008621,
      "learning_rate": 0.0005828124920191255,
      "loss": 0.5255,
      "num_input_tokens_seen": 12522736,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 0.22328118979930878,
      "learning_rate": 0.000582586956088771,
      "loss": 0.5269,
      "num_input_tokens_seen": 12525296,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 0.09477156400680542,
      "learning_rate": 0.0005823614028821677,
      "loss": 0.5192,
      "num_input_tokens_seen": 12529616,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 0.1211472898721695,
      "learning_rate": 0.0005821358324464985,
      "loss": 0.4433,
      "num_input_tokens_seen": 12534000,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 0.16690759360790253,
      "learning_rate": 0.0005819102448289501,
      "loss": 0.3564,
      "num_input_tokens_seen": 12536848,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 0.1454850286245346,
      "learning_rate": 0.000581684640076713,
      "loss": 0.4324,
      "num_input_tokens_seen": 12539632,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 0.10659980028867722,
      "learning_rate": 0.0005814590182369811,
      "loss": 0.4131,
      "num_input_tokens_seen": 12543024,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 0.17794162034988403,
      "learning_rate": 0.000581233379356952,
      "loss": 0.4111,
      "num_input_tokens_seen": 12545264,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 0.0770919993519783,
      "learning_rate": 0.0005810077234838267,
      "loss": 0.5067,
      "num_input_tokens_seen": 12549264,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 0.20591261982917786,
      "learning_rate": 0.0005807820506648097,
      "loss": 0.4603,
      "num_input_tokens_seen": 12554864,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 0.1503773182630539,
      "learning_rate": 0.0005805563609471095,
      "loss": 0.5452,
      "num_input_tokens_seen": 12558032,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 0.18474723398685455,
      "learning_rate": 0.0005803306543779375,
      "loss": 0.4291,
      "num_input_tokens_seen": 12560528,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 0.21841280162334442,
      "learning_rate": 0.000580104931004509,
      "loss": 0.4569,
      "num_input_tokens_seen": 12563056,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 0.11594877392053604,
      "learning_rate": 0.000579879190874043,
      "loss": 0.3162,
      "num_input_tokens_seen": 12566896,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 0.10810394585132599,
      "learning_rate": 0.0005796534340337614,
      "loss": 0.5287,
      "num_input_tokens_seen": 12570480,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 0.11315474659204483,
      "learning_rate": 0.0005794276605308902,
      "loss": 0.4823,
      "num_input_tokens_seen": 12573808,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 0.12978997826576233,
      "learning_rate": 0.0005792018704126586,
      "loss": 0.3554,
      "num_input_tokens_seen": 12577072,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 0.08590590953826904,
      "learning_rate": 0.0005789760637262992,
      "loss": 0.5132,
      "num_input_tokens_seen": 12580176,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 0.2473965734243393,
      "learning_rate": 0.0005787502405190483,
      "loss": 0.4005,
      "num_input_tokens_seen": 12584528,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 0.15296563506126404,
      "learning_rate": 0.0005785244008381456,
      "loss": 0.5541,
      "num_input_tokens_seen": 12587376,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 0.0915563702583313,
      "learning_rate": 0.0005782985447308341,
      "loss": 0.383,
      "num_input_tokens_seen": 12590256,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 0.09697214514017105,
      "learning_rate": 0.0005780726722443602,
      "loss": 0.3983,
      "num_input_tokens_seen": 12592848,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 0.09441816806793213,
      "learning_rate": 0.0005778467834259741,
      "loss": 0.3901,
      "num_input_tokens_seen": 12596272,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 0.08672640472650528,
      "learning_rate": 0.000577620878322929,
      "loss": 0.5293,
      "num_input_tokens_seen": 12600464,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 0.20196697115898132,
      "learning_rate": 0.0005773949569824816,
      "loss": 0.4911,
      "num_input_tokens_seen": 12603952,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 0.09703255444765091,
      "learning_rate": 0.0005771690194518922,
      "loss": 0.3695,
      "num_input_tokens_seen": 12606480,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 0.12416037917137146,
      "learning_rate": 0.0005769430657784245,
      "loss": 0.3208,
      "num_input_tokens_seen": 12608784,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 0.12777391076087952,
      "learning_rate": 0.0005767170960093449,
      "loss": 0.4928,
      "num_input_tokens_seen": 12613232,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 0.22420310974121094,
      "learning_rate": 0.0005764911101919244,
      "loss": 0.5742,
      "num_input_tokens_seen": 12616816,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 0.13972443342208862,
      "learning_rate": 0.0005762651083734363,
      "loss": 0.3693,
      "num_input_tokens_seen": 12620624,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 0.16793414950370789,
      "learning_rate": 0.0005760390906011574,
      "loss": 0.4821,
      "num_input_tokens_seen": 12624176,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 0.11975376307964325,
      "learning_rate": 0.0005758130569223684,
      "loss": 0.3706,
      "num_input_tokens_seen": 12627472,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 0.19481967389583588,
      "learning_rate": 0.0005755870073843527,
      "loss": 0.4268,
      "num_input_tokens_seen": 12631152,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 0.1378716081380844,
      "learning_rate": 0.0005753609420343973,
      "loss": 0.4812,
      "num_input_tokens_seen": 12635216,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 0.1448974609375,
      "learning_rate": 0.0005751348609197929,
      "loss": 0.5045,
      "num_input_tokens_seen": 12639120,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 0.16316713392734528,
      "learning_rate": 0.0005749087640878325,
      "loss": 0.4864,
      "num_input_tokens_seen": 12642064,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 0.10152098536491394,
      "learning_rate": 0.0005746826515858135,
      "loss": 0.3291,
      "num_input_tokens_seen": 12644848,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 0.13121090829372406,
      "learning_rate": 0.0005744565234610356,
      "loss": 0.4273,
      "num_input_tokens_seen": 12648112,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 0.07597003132104874,
      "learning_rate": 0.0005742303797608023,
      "loss": 0.3645,
      "num_input_tokens_seen": 12650992,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 0.11566008627414703,
      "learning_rate": 0.0005740042205324208,
      "loss": 0.5099,
      "num_input_tokens_seen": 12654000,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 0.15935321152210236,
      "learning_rate": 0.0005737780458232004,
      "loss": 0.4003,
      "num_input_tokens_seen": 12657008,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 0.18486540019512177,
      "learning_rate": 0.0005735518556804546,
      "loss": 0.4637,
      "num_input_tokens_seen": 12659696,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 0.17570754885673523,
      "learning_rate": 0.0005733256501514999,
      "loss": 0.381,
      "num_input_tokens_seen": 12663024,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 0.12787359952926636,
      "learning_rate": 0.0005730994292836555,
      "loss": 0.4443,
      "num_input_tokens_seen": 12666160,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 0.11426855623722076,
      "learning_rate": 0.0005728731931242445,
      "loss": 0.4733,
      "num_input_tokens_seen": 12668656,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 0.21413186192512512,
      "learning_rate": 0.0005726469417205928,
      "loss": 0.3991,
      "num_input_tokens_seen": 12671792,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 0.1150677427649498,
      "learning_rate": 0.0005724206751200297,
      "loss": 0.3721,
      "num_input_tokens_seen": 12674192,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 0.1371210664510727,
      "learning_rate": 0.0005721943933698879,
      "loss": 0.3647,
      "num_input_tokens_seen": 12677456,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 0.18283796310424805,
      "learning_rate": 0.0005719680965175023,
      "loss": 0.3516,
      "num_input_tokens_seen": 12681968,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 0.170455664396286,
      "learning_rate": 0.000571741784610212,
      "loss": 0.4107,
      "num_input_tokens_seen": 12684560,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 0.08748598396778107,
      "learning_rate": 0.0005715154576953589,
      "loss": 0.6615,
      "num_input_tokens_seen": 12687824,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 0.11667986214160919,
      "learning_rate": 0.0005712891158202879,
      "loss": 0.4601,
      "num_input_tokens_seen": 12690160,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 0.24489635229110718,
      "learning_rate": 0.0005710627590323473,
      "loss": 0.5673,
      "num_input_tokens_seen": 12693840,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 0.12188644707202911,
      "learning_rate": 0.0005708363873788882,
      "loss": 0.491,
      "num_input_tokens_seen": 12696688,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 0.16292105615139008,
      "learning_rate": 0.0005706100009072649,
      "loss": 0.4675,
      "num_input_tokens_seen": 12699472,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 0.12793394923210144,
      "learning_rate": 0.000570383599664835,
      "loss": 0.4301,
      "num_input_tokens_seen": 12702768,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 0.13218817114830017,
      "learning_rate": 0.0005701571836989591,
      "loss": 0.3295,
      "num_input_tokens_seen": 12706096,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 0.12416829913854599,
      "learning_rate": 0.0005699307530570008,
      "loss": 0.545,
      "num_input_tokens_seen": 12709648,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 0.1086953654885292,
      "learning_rate": 0.0005697043077863267,
      "loss": 0.37,
      "num_input_tokens_seen": 12712784,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 0.082944855093956,
      "learning_rate": 0.0005694778479343065,
      "loss": 0.4111,
      "num_input_tokens_seen": 12715920,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 0.22988851368427277,
      "learning_rate": 0.0005692513735483134,
      "loss": 0.3636,
      "num_input_tokens_seen": 12718832,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 0.12256612628698349,
      "learning_rate": 0.0005690248846757228,
      "loss": 0.5676,
      "num_input_tokens_seen": 12722448,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 0.10080935806035995,
      "learning_rate": 0.0005687983813639139,
      "loss": 0.5224,
      "num_input_tokens_seen": 12726096,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 0.2098236382007599,
      "learning_rate": 0.0005685718636602685,
      "loss": 0.4398,
      "num_input_tokens_seen": 12728976,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 0.1044839397072792,
      "learning_rate": 0.0005683453316121716,
      "loss": 0.3228,
      "num_input_tokens_seen": 12732432,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 0.10526978969573975,
      "learning_rate": 0.0005681187852670111,
      "loss": 0.3986,
      "num_input_tokens_seen": 12738288,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 0.11618170142173767,
      "learning_rate": 0.0005678922246721777,
      "loss": 0.5051,
      "num_input_tokens_seen": 12741168,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 0.18974962830543518,
      "learning_rate": 0.0005676656498750656,
      "loss": 0.3426,
      "num_input_tokens_seen": 12745456,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 0.2666926383972168,
      "learning_rate": 0.0005674390609230715,
      "loss": 0.4275,
      "num_input_tokens_seen": 12748336,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 0.18506866693496704,
      "learning_rate": 0.0005672124578635951,
      "loss": 0.443,
      "num_input_tokens_seen": 12751344,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 0.1418944001197815,
      "learning_rate": 0.0005669858407440394,
      "loss": 0.3581,
      "num_input_tokens_seen": 12754256,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 0.19353501498699188,
      "learning_rate": 0.00056675920961181,
      "loss": 0.3859,
      "num_input_tokens_seen": 12757840,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 0.18636015057563782,
      "learning_rate": 0.0005665325645143155,
      "loss": 0.4938,
      "num_input_tokens_seen": 12760944,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 0.1174573078751564,
      "learning_rate": 0.0005663059054989679,
      "loss": 0.4353,
      "num_input_tokens_seen": 12764880,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 0.21738366782665253,
      "learning_rate": 0.0005660792326131809,
      "loss": 0.4503,
      "num_input_tokens_seen": 12767920,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 0.15232312679290771,
      "learning_rate": 0.0005658525459043725,
      "loss": 0.4877,
      "num_input_tokens_seen": 12770416,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 0.14163275063037872,
      "learning_rate": 0.000565625845419963,
      "loss": 0.4211,
      "num_input_tokens_seen": 12773936,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 0.14547035098075867,
      "learning_rate": 0.000565399131207375,
      "loss": 0.3837,
      "num_input_tokens_seen": 12777776,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 0.3177248537540436,
      "learning_rate": 0.000565172403314035,
      "loss": 0.5813,
      "num_input_tokens_seen": 12780912,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 0.0527123361825943,
      "learning_rate": 0.0005649456617873717,
      "loss": 0.3061,
      "num_input_tokens_seen": 12784272,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 0.13071513175964355,
      "learning_rate": 0.0005647189066748169,
      "loss": 0.5443,
      "num_input_tokens_seen": 12787888,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 0.19224028289318085,
      "learning_rate": 0.0005644921380238052,
      "loss": 0.4403,
      "num_input_tokens_seen": 12791472,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 0.11125262081623077,
      "learning_rate": 0.000564265355881774,
      "loss": 0.4055,
      "num_input_tokens_seen": 12794800,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 0.14979249238967896,
      "learning_rate": 0.0005640385602961634,
      "loss": 0.4333,
      "num_input_tokens_seen": 12799184,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 0.13733643293380737,
      "learning_rate": 0.0005638117513144168,
      "loss": 0.4481,
      "num_input_tokens_seen": 12802576,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 0.2195049524307251,
      "learning_rate": 0.0005635849289839797,
      "loss": 0.4476,
      "num_input_tokens_seen": 12805488,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 0.11887489259243011,
      "learning_rate": 0.000563358093352301,
      "loss": 0.4024,
      "num_input_tokens_seen": 12809328,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 0.08613143861293793,
      "learning_rate": 0.0005631312444668318,
      "loss": 0.4913,
      "num_input_tokens_seen": 12812752,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.11022938787937164,
      "learning_rate": 0.0005629043823750266,
      "loss": 0.4081,
      "num_input_tokens_seen": 12815312,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 0.10879461467266083,
      "learning_rate": 0.0005626775071243423,
      "loss": 0.4293,
      "num_input_tokens_seen": 12817936,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 0.2426423877477646,
      "learning_rate": 0.0005624506187622384,
      "loss": 0.5904,
      "num_input_tokens_seen": 12821200,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 0.15026803314685822,
      "learning_rate": 0.0005622237173361777,
      "loss": 0.4076,
      "num_input_tokens_seen": 12824304,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 0.15672586858272552,
      "learning_rate": 0.0005619968028936252,
      "loss": 0.5153,
      "num_input_tokens_seen": 12827184,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 0.08259159326553345,
      "learning_rate": 0.0005617698754820489,
      "loss": 0.6161,
      "num_input_tokens_seen": 12831088,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 0.1551463007926941,
      "learning_rate": 0.0005615429351489192,
      "loss": 0.5387,
      "num_input_tokens_seen": 12834032,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.15797339379787445,
      "learning_rate": 0.0005613159819417095,
      "loss": 0.5118,
      "num_input_tokens_seen": 12836880,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 0.13632318377494812,
      "learning_rate": 0.0005610890159078961,
      "loss": 0.4559,
      "num_input_tokens_seen": 12839792,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 0.286468505859375,
      "learning_rate": 0.0005608620370949575,
      "loss": 0.3683,
      "num_input_tokens_seen": 12842192,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 0.13545063138008118,
      "learning_rate": 0.000560635045550375,
      "loss": 0.4397,
      "num_input_tokens_seen": 12845488,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 0.1467014104127884,
      "learning_rate": 0.0005604080413216328,
      "loss": 0.5247,
      "num_input_tokens_seen": 12852240,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 0.1325000375509262,
      "learning_rate": 0.0005601810244562175,
      "loss": 0.4092,
      "num_input_tokens_seen": 12854928,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 0.14427244663238525,
      "learning_rate": 0.0005599539950016184,
      "loss": 0.3453,
      "num_input_tokens_seen": 12858256,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 0.2159520536661148,
      "learning_rate": 0.0005597269530053277,
      "loss": 0.4727,
      "num_input_tokens_seen": 12862640,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 0.1313207447528839,
      "learning_rate": 0.0005594998985148398,
      "loss": 0.3958,
      "num_input_tokens_seen": 12865200,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 0.09498949348926544,
      "learning_rate": 0.0005592728315776519,
      "loss": 0.4691,
      "num_input_tokens_seen": 12869040,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 0.13053929805755615,
      "learning_rate": 0.0005590457522412638,
      "loss": 0.5661,
      "num_input_tokens_seen": 12872304,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 0.24607791006565094,
      "learning_rate": 0.000558818660553178,
      "loss": 0.3819,
      "num_input_tokens_seen": 12874896,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 0.08928713202476501,
      "learning_rate": 0.0005585915565608995,
      "loss": 0.6151,
      "num_input_tokens_seen": 12878160,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 0.20747876167297363,
      "learning_rate": 0.0005583644403119357,
      "loss": 0.4732,
      "num_input_tokens_seen": 12881328,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 0.11714654415845871,
      "learning_rate": 0.0005581373118537969,
      "loss": 0.2881,
      "num_input_tokens_seen": 12884144,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 0.14252610504627228,
      "learning_rate": 0.0005579101712339957,
      "loss": 0.5831,
      "num_input_tokens_seen": 12887440,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 0.1358191817998886,
      "learning_rate": 0.0005576830185000474,
      "loss": 0.4991,
      "num_input_tokens_seen": 12891088,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 0.09484845399856567,
      "learning_rate": 0.0005574558536994698,
      "loss": 0.4076,
      "num_input_tokens_seen": 12895632,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 0.18553417921066284,
      "learning_rate": 0.0005572286768797829,
      "loss": 0.3647,
      "num_input_tokens_seen": 12898544,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 0.10820348560810089,
      "learning_rate": 0.0005570014880885098,
      "loss": 0.4443,
      "num_input_tokens_seen": 12901136,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 0.13856109976768494,
      "learning_rate": 0.0005567742873731757,
      "loss": 0.4431,
      "num_input_tokens_seen": 12904656,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 0.13502690196037292,
      "learning_rate": 0.0005565470747813085,
      "loss": 0.4902,
      "num_input_tokens_seen": 12907696,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 0.29387810826301575,
      "learning_rate": 0.0005563198503604382,
      "loss": 0.5156,
      "num_input_tokens_seen": 12910064,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 0.1472327709197998,
      "learning_rate": 0.0005560926141580978,
      "loss": 0.3556,
      "num_input_tokens_seen": 12912944,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 0.22612474858760834,
      "learning_rate": 0.0005558653662218225,
      "loss": 0.3614,
      "num_input_tokens_seen": 12915248,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 0.10016725957393646,
      "learning_rate": 0.00055563810659915,
      "loss": 0.3736,
      "num_input_tokens_seen": 12917808,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 0.09701071679592133,
      "learning_rate": 0.0005554108353376201,
      "loss": 0.4223,
      "num_input_tokens_seen": 12921808,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 0.1157044991850853,
      "learning_rate": 0.0005551835524847757,
      "loss": 0.5793,
      "num_input_tokens_seen": 12925264,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 0.12255945056676865,
      "learning_rate": 0.0005549562580881618,
      "loss": 0.4567,
      "num_input_tokens_seen": 12928848,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 0.10459744185209274,
      "learning_rate": 0.0005547289521953257,
      "loss": 0.3765,
      "num_input_tokens_seen": 12932144,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 0.08713077753782272,
      "learning_rate": 0.000554501634853817,
      "loss": 0.5962,
      "num_input_tokens_seen": 12935536,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 0.1105668842792511,
      "learning_rate": 0.0005542743061111883,
      "loss": 0.5452,
      "num_input_tokens_seen": 12938736,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 0.09592252969741821,
      "learning_rate": 0.0005540469660149938,
      "loss": 0.3442,
      "num_input_tokens_seen": 12942288,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 0.26218298077583313,
      "learning_rate": 0.0005538196146127907,
      "loss": 0.4371,
      "num_input_tokens_seen": 12945520,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 0.13016998767852783,
      "learning_rate": 0.0005535922519521381,
      "loss": 0.3844,
      "num_input_tokens_seen": 12948240,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 0.16351811587810516,
      "learning_rate": 0.0005533648780805979,
      "loss": 0.428,
      "num_input_tokens_seen": 12951184,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 0.1609000563621521,
      "learning_rate": 0.000553137493045734,
      "loss": 0.359,
      "num_input_tokens_seen": 12953968,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 0.1600758135318756,
      "learning_rate": 0.0005529100968951127,
      "loss": 0.5585,
      "num_input_tokens_seen": 12956784,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 0.15759843587875366,
      "learning_rate": 0.0005526826896763029,
      "loss": 0.4218,
      "num_input_tokens_seen": 12959408,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 0.09173223376274109,
      "learning_rate": 0.0005524552714368753,
      "loss": 0.5889,
      "num_input_tokens_seen": 12962544,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 0.14652462303638458,
      "learning_rate": 0.0005522278422244035,
      "loss": 0.5135,
      "num_input_tokens_seen": 12965200,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 0.14383649826049805,
      "learning_rate": 0.0005520004020864631,
      "loss": 0.4568,
      "num_input_tokens_seen": 12968816,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 0.16997882723808289,
      "learning_rate": 0.0005517729510706315,
      "loss": 0.4683,
      "num_input_tokens_seen": 12971632,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 0.11360190808773041,
      "learning_rate": 0.0005515454892244892,
      "loss": 0.4327,
      "num_input_tokens_seen": 12975120,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 0.08689777553081512,
      "learning_rate": 0.0005513180165956189,
      "loss": 0.6483,
      "num_input_tokens_seen": 12979088,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 0.12074737995862961,
      "learning_rate": 0.0005510905332316049,
      "loss": 0.4933,
      "num_input_tokens_seen": 12982224,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 0.11691112816333771,
      "learning_rate": 0.0005508630391800341,
      "loss": 0.4859,
      "num_input_tokens_seen": 12984720,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 0.11495727300643921,
      "learning_rate": 0.0005506355344884959,
      "loss": 0.481,
      "num_input_tokens_seen": 12987472,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 0.2590029835700989,
      "learning_rate": 0.0005504080192045813,
      "loss": 0.3951,
      "num_input_tokens_seen": 12990352,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 0.11305680125951767,
      "learning_rate": 0.0005501804933758846,
      "loss": 0.3891,
      "num_input_tokens_seen": 12993616,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 0.2194765955209732,
      "learning_rate": 0.0005499529570500009,
      "loss": 0.484,
      "num_input_tokens_seen": 12996432,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 0.17166654765605927,
      "learning_rate": 0.0005497254102745285,
      "loss": 0.4182,
      "num_input_tokens_seen": 13000560,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 0.1417667716741562,
      "learning_rate": 0.0005494978530970676,
      "loss": 0.5014,
      "num_input_tokens_seen": 13003664,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 0.08915088325738907,
      "learning_rate": 0.0005492702855652206,
      "loss": 0.4477,
      "num_input_tokens_seen": 13007536,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 0.1691567599773407,
      "learning_rate": 0.000549042707726592,
      "loss": 0.4445,
      "num_input_tokens_seen": 13010960,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 0.13119053840637207,
      "learning_rate": 0.0005488151196287882,
      "loss": 0.3981,
      "num_input_tokens_seen": 13014896,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 0.11653225868940353,
      "learning_rate": 0.0005485875213194184,
      "loss": 0.3344,
      "num_input_tokens_seen": 13017808,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 0.15597003698349,
      "learning_rate": 0.0005483599128460935,
      "loss": 0.3522,
      "num_input_tokens_seen": 13020592,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 0.1591482013463974,
      "learning_rate": 0.0005481322942564266,
      "loss": 0.3941,
      "num_input_tokens_seen": 13024016,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 0.13117465376853943,
      "learning_rate": 0.0005479046655980327,
      "loss": 0.5303,
      "num_input_tokens_seen": 13026416,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 0.1502653956413269,
      "learning_rate": 0.0005476770269185295,
      "loss": 0.409,
      "num_input_tokens_seen": 13029168,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 0.17668256163597107,
      "learning_rate": 0.0005474493782655361,
      "loss": 0.4936,
      "num_input_tokens_seen": 13032240,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 0.11182918399572372,
      "learning_rate": 0.0005472217196866742,
      "loss": 0.4757,
      "num_input_tokens_seen": 13035120,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 0.11751163005828857,
      "learning_rate": 0.0005469940512295671,
      "loss": 0.3282,
      "num_input_tokens_seen": 13039088,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 0.08852065354585648,
      "learning_rate": 0.0005467663729418406,
      "loss": 0.4455,
      "num_input_tokens_seen": 13042320,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 0.1654997169971466,
      "learning_rate": 0.0005465386848711225,
      "loss": 0.4273,
      "num_input_tokens_seen": 13046160,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 0.11831577867269516,
      "learning_rate": 0.0005463109870650426,
      "loss": 0.4169,
      "num_input_tokens_seen": 13048944,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 0.2625177800655365,
      "learning_rate": 0.0005460832795712323,
      "loss": 0.4675,
      "num_input_tokens_seen": 13052816,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 0.0736968070268631,
      "learning_rate": 0.0005458555624373258,
      "loss": 0.4106,
      "num_input_tokens_seen": 13056592,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 0.09229853004217148,
      "learning_rate": 0.0005456278357109589,
      "loss": 0.3207,
      "num_input_tokens_seen": 13059632,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 0.13713251054286957,
      "learning_rate": 0.0005454000994397691,
      "loss": 0.4737,
      "num_input_tokens_seen": 13062864,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.07642743736505508,
      "learning_rate": 0.0005451723536713965,
      "loss": 0.3965,
      "num_input_tokens_seen": 13067152,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 0.1875230222940445,
      "learning_rate": 0.0005449445984534828,
      "loss": 0.5018,
      "num_input_tokens_seen": 13071440,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 0.10872694849967957,
      "learning_rate": 0.0005447168338336721,
      "loss": 0.4489,
      "num_input_tokens_seen": 13075248,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 0.10498341172933578,
      "learning_rate": 0.0005444890598596098,
      "loss": 0.4899,
      "num_input_tokens_seen": 13078128,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 0.12605492770671844,
      "learning_rate": 0.0005442612765789437,
      "loss": 0.4697,
      "num_input_tokens_seen": 13082000,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 0.10108339041471481,
      "learning_rate": 0.0005440334840393235,
      "loss": 0.4132,
      "num_input_tokens_seen": 13085680,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 0.13785690069198608,
      "learning_rate": 0.0005438056822884007,
      "loss": 0.5332,
      "num_input_tokens_seen": 13089520,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 0.19421528279781342,
      "learning_rate": 0.0005435778713738292,
      "loss": 0.4888,
      "num_input_tokens_seen": 13092368,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 0.12127745151519775,
      "learning_rate": 0.0005433500513432639,
      "loss": 0.362,
      "num_input_tokens_seen": 13095216,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 0.24386164546012878,
      "learning_rate": 0.0005431222222443626,
      "loss": 0.4815,
      "num_input_tokens_seen": 13097840,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 0.1493818461894989,
      "learning_rate": 0.0005428943841247843,
      "loss": 0.3794,
      "num_input_tokens_seen": 13100784,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 0.18368655443191528,
      "learning_rate": 0.0005426665370321901,
      "loss": 0.3982,
      "num_input_tokens_seen": 13104720,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 0.09569686651229858,
      "learning_rate": 0.0005424386810142433,
      "loss": 0.4194,
      "num_input_tokens_seen": 13109040,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 0.17005513608455658,
      "learning_rate": 0.0005422108161186086,
      "loss": 0.4345,
      "num_input_tokens_seen": 13111888,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 0.2385224550962448,
      "learning_rate": 0.0005419829423929525,
      "loss": 0.4528,
      "num_input_tokens_seen": 13115504,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 0.17850163578987122,
      "learning_rate": 0.0005417550598849441,
      "loss": 0.4994,
      "num_input_tokens_seen": 13119056,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 0.15318083763122559,
      "learning_rate": 0.0005415271686422534,
      "loss": 0.5323,
      "num_input_tokens_seen": 13121552,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 0.0909104123711586,
      "learning_rate": 0.0005412992687125527,
      "loss": 0.3765,
      "num_input_tokens_seen": 13125872,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 0.5944169163703918,
      "learning_rate": 0.0005410713601435164,
      "loss": 0.4729,
      "num_input_tokens_seen": 13128592,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 0.15962517261505127,
      "learning_rate": 0.0005408434429828199,
      "loss": 0.4978,
      "num_input_tokens_seen": 13131216,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 0.10963349044322968,
      "learning_rate": 0.0005406155172781412,
      "loss": 0.539,
      "num_input_tokens_seen": 13134192,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 0.13101986050605774,
      "learning_rate": 0.0005403875830771596,
      "loss": 0.475,
      "num_input_tokens_seen": 13137232,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 0.1306985318660736,
      "learning_rate": 0.0005401596404275564,
      "loss": 0.4517,
      "num_input_tokens_seen": 13140880,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 0.11062967032194138,
      "learning_rate": 0.0005399316893770147,
      "loss": 0.4932,
      "num_input_tokens_seen": 13144464,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 0.09889605641365051,
      "learning_rate": 0.0005397037299732189,
      "loss": 0.5768,
      "num_input_tokens_seen": 13147504,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 0.3622376024723053,
      "learning_rate": 0.000539475762263856,
      "loss": 0.6034,
      "num_input_tokens_seen": 13150288,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 0.12363304197788239,
      "learning_rate": 0.0005392477862966138,
      "loss": 0.3551,
      "num_input_tokens_seen": 13153616,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 0.11783338338136673,
      "learning_rate": 0.0005390198021191827,
      "loss": 0.3868,
      "num_input_tokens_seen": 13156080,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.16032131016254425,
      "learning_rate": 0.000538791809779254,
      "loss": 0.4499,
      "num_input_tokens_seen": 13159792,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 0.10831063985824585,
      "learning_rate": 0.0005385638093245213,
      "loss": 0.5525,
      "num_input_tokens_seen": 13163120,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 0.08500007539987564,
      "learning_rate": 0.0005383358008026795,
      "loss": 0.3426,
      "num_input_tokens_seen": 13166192,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 0.17210356891155243,
      "learning_rate": 0.0005381077842614257,
      "loss": 0.4488,
      "num_input_tokens_seen": 13168944,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 0.1820390522480011,
      "learning_rate": 0.0005378797597484582,
      "loss": 0.4775,
      "num_input_tokens_seen": 13172080,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 0.11143666505813599,
      "learning_rate": 0.0005376517273114771,
      "loss": 0.3141,
      "num_input_tokens_seen": 13174704,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 0.12657895684242249,
      "learning_rate": 0.0005374236869981841,
      "loss": 0.4197,
      "num_input_tokens_seen": 13177648,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 0.16527371108531952,
      "learning_rate": 0.0005371956388562827,
      "loss": 0.5175,
      "num_input_tokens_seen": 13180592,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 0.1770022213459015,
      "learning_rate": 0.0005369675829334782,
      "loss": 0.5428,
      "num_input_tokens_seen": 13183408,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 0.10077229887247086,
      "learning_rate": 0.0005367395192774769,
      "loss": 0.4368,
      "num_input_tokens_seen": 13187280,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 0.11022341996431351,
      "learning_rate": 0.0005365114479359875,
      "loss": 0.4069,
      "num_input_tokens_seen": 13190288,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 0.0733535885810852,
      "learning_rate": 0.0005362833689567197,
      "loss": 0.3305,
      "num_input_tokens_seen": 13193328,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 0.09926003962755203,
      "learning_rate": 0.0005360552823873849,
      "loss": 0.3844,
      "num_input_tokens_seen": 13195888,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 0.13760334253311157,
      "learning_rate": 0.0005358271882756966,
      "loss": 0.6025,
      "num_input_tokens_seen": 13198512,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 0.30590173602104187,
      "learning_rate": 0.0005355990866693689,
      "loss": 0.3611,
      "num_input_tokens_seen": 13201168,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 0.12780757248401642,
      "learning_rate": 0.0005353709776161187,
      "loss": 0.4117,
      "num_input_tokens_seen": 13204432,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 0.1352105587720871,
      "learning_rate": 0.0005351428611636636,
      "loss": 0.4991,
      "num_input_tokens_seen": 13207920,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 0.14387904107570648,
      "learning_rate": 0.0005349147373597226,
      "loss": 0.4114,
      "num_input_tokens_seen": 13210384,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 0.14968660473823547,
      "learning_rate": 0.000534686606252017,
      "loss": 0.6321,
      "num_input_tokens_seen": 13213360,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 0.22181734442710876,
      "learning_rate": 0.0005344584678882692,
      "loss": 0.5773,
      "num_input_tokens_seen": 13216560,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 0.0993194580078125,
      "learning_rate": 0.0005342303223162027,
      "loss": 0.4279,
      "num_input_tokens_seen": 13219920,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 0.1077074259519577,
      "learning_rate": 0.0005340021695835437,
      "loss": 0.5053,
      "num_input_tokens_seen": 13223536,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 0.0955248475074768,
      "learning_rate": 0.0005337740097380184,
      "loss": 0.392,
      "num_input_tokens_seen": 13226992,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 0.10960470139980316,
      "learning_rate": 0.0005335458428273557,
      "loss": 0.3978,
      "num_input_tokens_seen": 13230000,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 0.09877818822860718,
      "learning_rate": 0.0005333176688992855,
      "loss": 0.491,
      "num_input_tokens_seen": 13232912,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 0.1237856075167656,
      "learning_rate": 0.000533089488001539,
      "loss": 0.5413,
      "num_input_tokens_seen": 13236720,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 0.11413590610027313,
      "learning_rate": 0.0005328613001818492,
      "loss": 0.4457,
      "num_input_tokens_seen": 13239600,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 0.104999840259552,
      "learning_rate": 0.0005326331054879502,
      "loss": 0.5259,
      "num_input_tokens_seen": 13242000,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 0.14053641259670258,
      "learning_rate": 0.0005324049039675778,
      "loss": 0.433,
      "num_input_tokens_seen": 13244880,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 0.07591013610363007,
      "learning_rate": 0.0005321766956684693,
      "loss": 0.5755,
      "num_input_tokens_seen": 13248368,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 0.176559716463089,
      "learning_rate": 0.0005319484806383631,
      "loss": 0.4564,
      "num_input_tokens_seen": 13252720,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 0.11880218982696533,
      "learning_rate": 0.0005317202589249991,
      "loss": 0.3142,
      "num_input_tokens_seen": 13255824,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 0.11753948032855988,
      "learning_rate": 0.0005314920305761191,
      "loss": 0.2988,
      "num_input_tokens_seen": 13258192,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 0.10409464687108994,
      "learning_rate": 0.0005312637956394654,
      "loss": 0.508,
      "num_input_tokens_seen": 13261456,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 0.15283647179603577,
      "learning_rate": 0.0005310355541627824,
      "loss": 0.3395,
      "num_input_tokens_seen": 13264656,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 0.1481093019247055,
      "learning_rate": 0.0005308073061938153,
      "loss": 0.3922,
      "num_input_tokens_seen": 13267184,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 0.0954689234495163,
      "learning_rate": 0.0005305790517803115,
      "loss": 0.5448,
      "num_input_tokens_seen": 13270768,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 0.12948444485664368,
      "learning_rate": 0.0005303507909700189,
      "loss": 0.5752,
      "num_input_tokens_seen": 13273296,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 0.12360751628875732,
      "learning_rate": 0.000530122523810687,
      "loss": 0.4722,
      "num_input_tokens_seen": 13278448,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 0.11263084411621094,
      "learning_rate": 0.0005298942503500668,
      "loss": 0.499,
      "num_input_tokens_seen": 13281520,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 0.1503220647573471,
      "learning_rate": 0.0005296659706359105,
      "loss": 0.4121,
      "num_input_tokens_seen": 13284496,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 0.10033177584409714,
      "learning_rate": 0.0005294376847159716,
      "loss": 0.4349,
      "num_input_tokens_seen": 13287760,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 0.1348394900560379,
      "learning_rate": 0.0005292093926380049,
      "loss": 0.4805,
      "num_input_tokens_seen": 13290384,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 0.12744347751140594,
      "learning_rate": 0.0005289810944497663,
      "loss": 0.4758,
      "num_input_tokens_seen": 13294448,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 0.09060568362474442,
      "learning_rate": 0.0005287527901990137,
      "loss": 0.5801,
      "num_input_tokens_seen": 13298480,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 0.15685386955738068,
      "learning_rate": 0.0005285244799335053,
      "loss": 0.4048,
      "num_input_tokens_seen": 13302096,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 0.15099528431892395,
      "learning_rate": 0.0005282961637010009,
      "loss": 0.4465,
      "num_input_tokens_seen": 13305136,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 0.12875941395759583,
      "learning_rate": 0.0005280678415492621,
      "loss": 0.3954,
      "num_input_tokens_seen": 13308464,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 0.1400390863418579,
      "learning_rate": 0.0005278395135260512,
      "loss": 0.4417,
      "num_input_tokens_seen": 13313392,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 0.1248137354850769,
      "learning_rate": 0.0005276111796791314,
      "loss": 0.3728,
      "num_input_tokens_seen": 13316624,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 0.19065769016742706,
      "learning_rate": 0.000527382840056268,
      "loss": 0.5815,
      "num_input_tokens_seen": 13320176,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 0.16534088551998138,
      "learning_rate": 0.0005271544947052267,
      "loss": 0.3871,
      "num_input_tokens_seen": 13323024,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 0.14750154316425323,
      "learning_rate": 0.000526926143673775,
      "loss": 0.5057,
      "num_input_tokens_seen": 13326544,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 0.12689636647701263,
      "learning_rate": 0.0005266977870096813,
      "loss": 0.3308,
      "num_input_tokens_seen": 13330352,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 0.16673308610916138,
      "learning_rate": 0.000526469424760715,
      "loss": 0.4348,
      "num_input_tokens_seen": 13333264,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 0.10181479901075363,
      "learning_rate": 0.0005262410569746471,
      "loss": 0.4265,
      "num_input_tokens_seen": 13335792,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 0.17304642498493195,
      "learning_rate": 0.0005260126836992495,
      "loss": 0.4946,
      "num_input_tokens_seen": 13338256,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 0.08070303499698639,
      "learning_rate": 0.000525784304982295,
      "loss": 0.4352,
      "num_input_tokens_seen": 13341136,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 0.15978139638900757,
      "learning_rate": 0.0005255559208715585,
      "loss": 0.4512,
      "num_input_tokens_seen": 13343984,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 0.16650764644145966,
      "learning_rate": 0.0005253275314148145,
      "loss": 0.4176,
      "num_input_tokens_seen": 13346544,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 0.1767401099205017,
      "learning_rate": 0.0005250991366598401,
      "loss": 0.5049,
      "num_input_tokens_seen": 13349392,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 0.16938655078411102,
      "learning_rate": 0.0005248707366544126,
      "loss": 0.3509,
      "num_input_tokens_seen": 13353040,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 0.05791356414556503,
      "learning_rate": 0.0005246423314463106,
      "loss": 0.4199,
      "num_input_tokens_seen": 13356912,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 0.10095513612031937,
      "learning_rate": 0.000524413921083314,
      "loss": 0.5075,
      "num_input_tokens_seen": 13361328,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 0.12466385960578918,
      "learning_rate": 0.0005241855056132039,
      "loss": 0.4515,
      "num_input_tokens_seen": 13366032,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 0.22816504538059235,
      "learning_rate": 0.0005239570850837617,
      "loss": 0.4378,
      "num_input_tokens_seen": 13368784,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 0.10596133768558502,
      "learning_rate": 0.0005237286595427709,
      "loss": 0.4065,
      "num_input_tokens_seen": 13372816,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 0.1362793892621994,
      "learning_rate": 0.000523500229038015,
      "loss": 0.6014,
      "num_input_tokens_seen": 13376048,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 0.14824141561985016,
      "learning_rate": 0.0005232717936172794,
      "loss": 0.4125,
      "num_input_tokens_seen": 13378896,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 0.16392381489276886,
      "learning_rate": 0.0005230433533283498,
      "loss": 0.3317,
      "num_input_tokens_seen": 13382512,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 0.14256992936134338,
      "learning_rate": 0.0005228149082190139,
      "loss": 0.4575,
      "num_input_tokens_seen": 13385776,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.13564765453338623,
      "learning_rate": 0.0005225864583370594,
      "loss": 0.4121,
      "num_input_tokens_seen": 13389680,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 0.12287687510251999,
      "learning_rate": 0.0005223580037302754,
      "loss": 0.384,
      "num_input_tokens_seen": 13393776,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 0.1435723453760147,
      "learning_rate": 0.0005221295444464522,
      "loss": 0.4659,
      "num_input_tokens_seen": 13396880,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 0.18541781604290009,
      "learning_rate": 0.0005219010805333807,
      "loss": 0.4577,
      "num_input_tokens_seen": 13400176,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 0.1271742880344391,
      "learning_rate": 0.000521672612038853,
      "loss": 0.5616,
      "num_input_tokens_seen": 13402896,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 0.27184948325157166,
      "learning_rate": 0.0005214441390106624,
      "loss": 0.4511,
      "num_input_tokens_seen": 13406672,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 0.17848879098892212,
      "learning_rate": 0.0005212156614966022,
      "loss": 0.5191,
      "num_input_tokens_seen": 13409456,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 0.1641417145729065,
      "learning_rate": 0.0005209871795444678,
      "loss": 0.396,
      "num_input_tokens_seen": 13413104,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 0.11796014755964279,
      "learning_rate": 0.000520758693202055,
      "loss": 0.3768,
      "num_input_tokens_seen": 13419888,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 0.10333086550235748,
      "learning_rate": 0.0005205302025171601,
      "loss": 0.3902,
      "num_input_tokens_seen": 13423248,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 0.1523464471101761,
      "learning_rate": 0.0005203017075375812,
      "loss": 0.5861,
      "num_input_tokens_seen": 13425968,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 0.18609941005706787,
      "learning_rate": 0.0005200732083111169,
      "loss": 0.4273,
      "num_input_tokens_seen": 13429488,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 0.09010405093431473,
      "learning_rate": 0.000519844704885566,
      "loss": 0.4567,
      "num_input_tokens_seen": 13433424,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 0.27939337491989136,
      "learning_rate": 0.0005196161973087297,
      "loss": 0.4964,
      "num_input_tokens_seen": 13436944,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 0.07813097536563873,
      "learning_rate": 0.0005193876856284084,
      "loss": 0.3348,
      "num_input_tokens_seen": 13440080,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 0.14404092729091644,
      "learning_rate": 0.0005191591698924046,
      "loss": 0.4089,
      "num_input_tokens_seen": 13442896,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 0.10755685716867447,
      "learning_rate": 0.0005189306501485211,
      "loss": 0.4451,
      "num_input_tokens_seen": 13445808,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 0.5881959795951843,
      "learning_rate": 0.0005187021264445613,
      "loss": 0.501,
      "num_input_tokens_seen": 13448720,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 0.1639707088470459,
      "learning_rate": 0.0005184735988283302,
      "loss": 0.4168,
      "num_input_tokens_seen": 13452400,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 0.14229339361190796,
      "learning_rate": 0.000518245067347633,
      "loss": 0.4948,
      "num_input_tokens_seen": 13455888,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 0.16594071686267853,
      "learning_rate": 0.0005180165320502756,
      "loss": 0.4615,
      "num_input_tokens_seen": 13459248,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 0.12472287565469742,
      "learning_rate": 0.0005177879929840655,
      "loss": 0.4415,
      "num_input_tokens_seen": 13462832,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 0.1488727629184723,
      "learning_rate": 0.0005175594501968103,
      "loss": 0.4621,
      "num_input_tokens_seen": 13466704,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 0.0661839172244072,
      "learning_rate": 0.0005173309037363182,
      "loss": 0.4288,
      "num_input_tokens_seen": 13470096,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 0.08386586606502533,
      "learning_rate": 0.0005171023536503991,
      "loss": 0.3798,
      "num_input_tokens_seen": 13473904,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 0.07745962589979172,
      "learning_rate": 0.0005168737999868624,
      "loss": 0.3287,
      "num_input_tokens_seen": 13476752,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 0.07908166199922562,
      "learning_rate": 0.0005166452427935193,
      "loss": 0.451,
      "num_input_tokens_seen": 13480208,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 0.1437820941209793,
      "learning_rate": 0.0005164166821181814,
      "loss": 0.4273,
      "num_input_tokens_seen": 13482832,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 0.07729443162679672,
      "learning_rate": 0.0005161881180086607,
      "loss": 0.3749,
      "num_input_tokens_seen": 13485744,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 0.184439554810524,
      "learning_rate": 0.0005159595505127705,
      "loss": 0.3995,
      "num_input_tokens_seen": 13488656,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.1393522322177887,
      "learning_rate": 0.0005157309796783246,
      "loss": 0.3997,
      "num_input_tokens_seen": 13491408,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 0.15527325868606567,
      "learning_rate": 0.000515502405553137,
      "loss": 0.5128,
      "num_input_tokens_seen": 13495056,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 0.07796885073184967,
      "learning_rate": 0.0005152738281850231,
      "loss": 0.4278,
      "num_input_tokens_seen": 13498512,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 0.21173161268234253,
      "learning_rate": 0.0005150452476217986,
      "loss": 0.5601,
      "num_input_tokens_seen": 13501392,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 0.10287503898143768,
      "learning_rate": 0.0005148166639112799,
      "loss": 0.4794,
      "num_input_tokens_seen": 13504464,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.10682222992181778,
      "learning_rate": 0.0005145880771012842,
      "loss": 0.4653,
      "num_input_tokens_seen": 13507472,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 0.10969806462526321,
      "learning_rate": 0.0005143594872396291,
      "loss": 0.4028,
      "num_input_tokens_seen": 13510224,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 0.2153702974319458,
      "learning_rate": 0.0005141308943741333,
      "loss": 0.4696,
      "num_input_tokens_seen": 13512944,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 0.07855704426765442,
      "learning_rate": 0.0005139022985526154,
      "loss": 0.396,
      "num_input_tokens_seen": 13516976,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 0.11331094801425934,
      "learning_rate": 0.0005136736998228952,
      "loss": 0.4443,
      "num_input_tokens_seen": 13520816,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 0.13241717219352722,
      "learning_rate": 0.0005134450982327933,
      "loss": 0.3919,
      "num_input_tokens_seen": 13523888,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 0.13168896734714508,
      "learning_rate": 0.0005132164938301299,
      "loss": 0.333,
      "num_input_tokens_seen": 13526384,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 0.2671434283256531,
      "learning_rate": 0.0005129878866627271,
      "loss": 0.579,
      "num_input_tokens_seen": 13529008,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 0.12068299204111099,
      "learning_rate": 0.0005127592767784063,
      "loss": 0.4566,
      "num_input_tokens_seen": 13532080,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 0.08389627188444138,
      "learning_rate": 0.0005125306642249905,
      "loss": 0.5129,
      "num_input_tokens_seen": 13535504,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.13456569612026215,
      "learning_rate": 0.0005123020490503027,
      "loss": 0.4218,
      "num_input_tokens_seen": 13538256,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 0.16114699840545654,
      "learning_rate": 0.0005120734313021665,
      "loss": 0.5939,
      "num_input_tokens_seen": 13541456,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 0.10190266370773315,
      "learning_rate": 0.0005118448110284063,
      "loss": 0.4001,
      "num_input_tokens_seen": 13544720,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 0.15401574969291687,
      "learning_rate": 0.0005116161882768469,
      "loss": 0.4976,
      "num_input_tokens_seen": 13547504,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 0.14619846642017365,
      "learning_rate": 0.0005113875630953131,
      "loss": 0.397,
      "num_input_tokens_seen": 13551056,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 0.09215210378170013,
      "learning_rate": 0.0005111589355316314,
      "loss": 0.4481,
      "num_input_tokens_seen": 13554608,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 0.13215139508247375,
      "learning_rate": 0.0005109303056336274,
      "loss": 0.4909,
      "num_input_tokens_seen": 13558736,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 0.12774810194969177,
      "learning_rate": 0.0005107016734491281,
      "loss": 0.4435,
      "num_input_tokens_seen": 13562224,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 0.14957697689533234,
      "learning_rate": 0.0005104730390259611,
      "loss": 0.4733,
      "num_input_tokens_seen": 13565040,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 0.16767798364162445,
      "learning_rate": 0.0005102444024119537,
      "loss": 0.4185,
      "num_input_tokens_seen": 13567568,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 0.17881964147090912,
      "learning_rate": 0.0005100157636549342,
      "loss": 0.4028,
      "num_input_tokens_seen": 13570256,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 0.1406484842300415,
      "learning_rate": 0.0005097871228027312,
      "loss": 0.461,
      "num_input_tokens_seen": 13574000,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 0.08790078014135361,
      "learning_rate": 0.0005095584799031736,
      "loss": 0.459,
      "num_input_tokens_seen": 13578128,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 0.09819120913743973,
      "learning_rate": 0.0005093298350040912,
      "loss": 0.4183,
      "num_input_tokens_seen": 13582192,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 0.1549319624900818,
      "learning_rate": 0.0005091011881533139,
      "loss": 0.4762,
      "num_input_tokens_seen": 13585040,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 0.11069794744253159,
      "learning_rate": 0.0005088725393986716,
      "loss": 0.4029,
      "num_input_tokens_seen": 13587792,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 0.07686703652143478,
      "learning_rate": 0.0005086438887879955,
      "loss": 0.342,
      "num_input_tokens_seen": 13591760,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 0.12150775641202927,
      "learning_rate": 0.0005084152363691164,
      "loss": 0.4672,
      "num_input_tokens_seen": 13594096,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 0.24350762367248535,
      "learning_rate": 0.0005081865821898658,
      "loss": 0.4353,
      "num_input_tokens_seen": 13596944,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 0.22685053944587708,
      "learning_rate": 0.0005079579262980756,
      "loss": 0.4551,
      "num_input_tokens_seen": 13599504,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 0.15134549140930176,
      "learning_rate": 0.0005077292687415781,
      "loss": 0.5768,
      "num_input_tokens_seen": 13603632,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 0.17657043039798737,
      "learning_rate": 0.0005075006095682056,
      "loss": 0.3211,
      "num_input_tokens_seen": 13606480,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 0.13084132969379425,
      "learning_rate": 0.0005072719488257914,
      "loss": 0.4339,
      "num_input_tokens_seen": 13609168,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 0.11262065917253494,
      "learning_rate": 0.0005070432865621683,
      "loss": 0.4505,
      "num_input_tokens_seen": 13612080,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 0.09876495599746704,
      "learning_rate": 0.00050681462282517,
      "loss": 0.5031,
      "num_input_tokens_seen": 13615536,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 0.2229916900396347,
      "learning_rate": 0.0005065859576626304,
      "loss": 0.377,
      "num_input_tokens_seen": 13618672,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 0.16732731461524963,
      "learning_rate": 0.0005063572911223836,
      "loss": 0.5228,
      "num_input_tokens_seen": 13621072,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 0.12289201468229294,
      "learning_rate": 0.0005061286232522639,
      "loss": 0.4719,
      "num_input_tokens_seen": 13624400,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 0.2403615266084671,
      "learning_rate": 0.0005058999541001064,
      "loss": 0.5205,
      "num_input_tokens_seen": 13626544,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 0.10970297455787659,
      "learning_rate": 0.0005056712837137456,
      "loss": 0.3858,
      "num_input_tokens_seen": 13629840,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 0.11267611384391785,
      "learning_rate": 0.000505442612141017,
      "loss": 0.3769,
      "num_input_tokens_seen": 13632784,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 0.1301780492067337,
      "learning_rate": 0.000505213939429756,
      "loss": 0.4221,
      "num_input_tokens_seen": 13635600,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 0.13717228174209595,
      "learning_rate": 0.0005049852656277984,
      "loss": 0.4533,
      "num_input_tokens_seen": 13638992,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 0.11683729290962219,
      "learning_rate": 0.0005047565907829801,
      "loss": 0.4026,
      "num_input_tokens_seen": 13641616,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 0.12453383952379227,
      "learning_rate": 0.0005045279149431374,
      "loss": 0.4339,
      "num_input_tokens_seen": 13644240,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 0.06057761237025261,
      "learning_rate": 0.0005042992381561063,
      "loss": 0.5143,
      "num_input_tokens_seen": 13647408,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 0.1270289272069931,
      "learning_rate": 0.0005040705604697238,
      "loss": 0.4705,
      "num_input_tokens_seen": 13650064,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 0.09706737846136093,
      "learning_rate": 0.0005038418819318265,
      "loss": 0.4631,
      "num_input_tokens_seen": 13654416,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 0.11397510021924973,
      "learning_rate": 0.0005036132025902511,
      "loss": 0.398,
      "num_input_tokens_seen": 13657616,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 0.07375575602054596,
      "learning_rate": 0.0005033845224928352,
      "loss": 0.3827,
      "num_input_tokens_seen": 13661424,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 0.14971932768821716,
      "learning_rate": 0.0005031558416874156,
      "loss": 0.5081,
      "num_input_tokens_seen": 13664112,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 0.23061397671699524,
      "learning_rate": 0.0005029271602218301,
      "loss": 0.5222,
      "num_input_tokens_seen": 13667024,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 0.10160969942808151,
      "learning_rate": 0.0005026984781439162,
      "loss": 0.4127,
      "num_input_tokens_seen": 13670032,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 0.08821065723896027,
      "learning_rate": 0.0005024697955015112,
      "loss": 0.5023,
      "num_input_tokens_seen": 13673584,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 0.0600435696542263,
      "learning_rate": 0.0005022411123424532,
      "loss": 0.3711,
      "num_input_tokens_seen": 13676880,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 0.10604142397642136,
      "learning_rate": 0.00050201242871458,
      "loss": 0.5742,
      "num_input_tokens_seen": 13680400,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 0.17286217212677002,
      "learning_rate": 0.0005017837446657298,
      "loss": 0.5404,
      "num_input_tokens_seen": 13683248,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 0.2479478269815445,
      "learning_rate": 0.0005015550602437405,
      "loss": 0.4543,
      "num_input_tokens_seen": 13686128,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 0.12906241416931152,
      "learning_rate": 0.0005013263754964504,
      "loss": 0.3641,
      "num_input_tokens_seen": 13689360,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 0.09925145655870438,
      "learning_rate": 0.0005010976904716977,
      "loss": 0.3287,
      "num_input_tokens_seen": 13691952,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 0.18647728860378265,
      "learning_rate": 0.0005008690052173206,
      "loss": 0.4792,
      "num_input_tokens_seen": 13694544,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 0.11175622791051865,
      "learning_rate": 0.0005006403197811575,
      "loss": 0.4382,
      "num_input_tokens_seen": 13700912,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 0.2525268793106079,
      "learning_rate": 0.0005004116342110469,
      "loss": 0.4924,
      "num_input_tokens_seen": 13704272,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 0.17610815167427063,
      "learning_rate": 0.0005001829485548272,
      "loss": 0.5228,
      "num_input_tokens_seen": 13707344,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.4845702648162842,
      "eval_runtime": 13.7295,
      "eval_samples_per_second": 61.765,
      "eval_steps_per_second": 15.441,
      "num_input_tokens_seen": 13708664,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 0.15772560238838196,
      "learning_rate": 0.0004999542628603364,
      "loss": 0.3973,
      "num_input_tokens_seen": 13709848,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 0.14910653233528137,
      "learning_rate": 0.0004997255771754136,
      "loss": 0.476,
      "num_input_tokens_seen": 13712536,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.06882455945014954,
      "learning_rate": 0.0004994968915478969,
      "loss": 0.4542,
      "num_input_tokens_seen": 13715960,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 0.0690050795674324,
      "learning_rate": 0.0004992682060256246,
      "loss": 0.4002,
      "num_input_tokens_seen": 13719704,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 0.10159433633089066,
      "learning_rate": 0.0004990395206564353,
      "loss": 0.453,
      "num_input_tokens_seen": 13723096,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 0.17649421095848083,
      "learning_rate": 0.0004988108354881674,
      "loss": 0.3573,
      "num_input_tokens_seen": 13726360,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 0.20627589523792267,
      "learning_rate": 0.0004985821505686591,
      "loss": 0.3389,
      "num_input_tokens_seen": 13731128,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.12474219501018524,
      "learning_rate": 0.0004983534659457489,
      "loss": 0.339,
      "num_input_tokens_seen": 13733688,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 0.14212018251419067,
      "learning_rate": 0.0004981247816672745,
      "loss": 0.4468,
      "num_input_tokens_seen": 13736920,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 0.2117454558610916,
      "learning_rate": 0.0004978960977810746,
      "loss": 0.5387,
      "num_input_tokens_seen": 13739800,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 0.10416652262210846,
      "learning_rate": 0.0004976674143349869,
      "loss": 0.4458,
      "num_input_tokens_seen": 13743192,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 0.1647941619157791,
      "learning_rate": 0.0004974387313768497,
      "loss": 0.4832,
      "num_input_tokens_seen": 13746136,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 0.1836225539445877,
      "learning_rate": 0.0004972100489545007,
      "loss": 0.4386,
      "num_input_tokens_seen": 13749048,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 0.18399865925312042,
      "learning_rate": 0.0004969813671157774,
      "loss": 0.4222,
      "num_input_tokens_seen": 13752536,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 0.07714682817459106,
      "learning_rate": 0.0004967526859085177,
      "loss": 0.5974,
      "num_input_tokens_seen": 13755768,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 0.1667437106370926,
      "learning_rate": 0.0004965240053805591,
      "loss": 0.5323,
      "num_input_tokens_seen": 13759256,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 0.11651379615068436,
      "learning_rate": 0.000496295325579739,
      "loss": 0.3017,
      "num_input_tokens_seen": 13762296,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 0.12581250071525574,
      "learning_rate": 0.0004960666465538945,
      "loss": 0.4728,
      "num_input_tokens_seen": 13765432,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 0.20421980321407318,
      "learning_rate": 0.0004958379683508627,
      "loss": 0.4382,
      "num_input_tokens_seen": 13768088,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.15708313882350922,
      "learning_rate": 0.0004956092910184803,
      "loss": 0.4164,
      "num_input_tokens_seen": 13771160,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 0.11570440232753754,
      "learning_rate": 0.0004953806146045843,
      "loss": 0.3684,
      "num_input_tokens_seen": 13775256,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 0.12686249613761902,
      "learning_rate": 0.0004951519391570108,
      "loss": 0.4292,
      "num_input_tokens_seen": 13778488,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 0.15280450880527496,
      "learning_rate": 0.0004949232647235964,
      "loss": 0.4141,
      "num_input_tokens_seen": 13782328,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 0.17674288153648376,
      "learning_rate": 0.0004946945913521771,
      "loss": 0.3561,
      "num_input_tokens_seen": 13785144,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 0.2100675106048584,
      "learning_rate": 0.0004944659190905889,
      "loss": 0.3966,
      "num_input_tokens_seen": 13788184,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 0.08831305056810379,
      "learning_rate": 0.0004942372479866674,
      "loss": 0.4344,
      "num_input_tokens_seen": 13792600,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 0.19831857085227966,
      "learning_rate": 0.0004940085780882475,
      "loss": 0.5491,
      "num_input_tokens_seen": 13796024,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 0.1679784506559372,
      "learning_rate": 0.0004937799094431649,
      "loss": 0.4484,
      "num_input_tokens_seen": 13799000,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 0.10990475863218307,
      "learning_rate": 0.0004935512420992542,
      "loss": 0.5241,
      "num_input_tokens_seen": 13801688,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 0.1195637509226799,
      "learning_rate": 0.0004933225761043502,
      "loss": 0.3979,
      "num_input_tokens_seen": 13804920,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 0.1198105663061142,
      "learning_rate": 0.0004930939115062872,
      "loss": 0.3786,
      "num_input_tokens_seen": 13808472,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 0.09557375311851501,
      "learning_rate": 0.0004928652483528989,
      "loss": 0.2823,
      "num_input_tokens_seen": 13811192,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 0.14217686653137207,
      "learning_rate": 0.0004926365866920193,
      "loss": 0.4602,
      "num_input_tokens_seen": 13814104,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 0.09055997431278229,
      "learning_rate": 0.0004924079265714817,
      "loss": 0.4445,
      "num_input_tokens_seen": 13817144,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 0.11034900695085526,
      "learning_rate": 0.0004921792680391193,
      "loss": 0.3798,
      "num_input_tokens_seen": 13821464,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 0.23866455256938934,
      "learning_rate": 0.0004919506111427649,
      "loss": 0.4024,
      "num_input_tokens_seen": 13824696,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 0.08555756509304047,
      "learning_rate": 0.0004917219559302507,
      "loss": 0.3864,
      "num_input_tokens_seen": 13828248,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 0.14691504836082458,
      "learning_rate": 0.0004914933024494089,
      "loss": 0.4285,
      "num_input_tokens_seen": 13831448,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 0.1336001306772232,
      "learning_rate": 0.0004912646507480713,
      "loss": 0.4452,
      "num_input_tokens_seen": 13834072,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 0.09750374406576157,
      "learning_rate": 0.0004910360008740689,
      "loss": 0.5282,
      "num_input_tokens_seen": 13837208,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.09469394385814667,
      "learning_rate": 0.0004908073528752329,
      "loss": 0.386,
      "num_input_tokens_seen": 13840248,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 0.0838305801153183,
      "learning_rate": 0.0004905787067993941,
      "loss": 0.4836,
      "num_input_tokens_seen": 13844440,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.11324967443943024,
      "learning_rate": 0.000490350062694382,
      "loss": 0.468,
      "num_input_tokens_seen": 13847608,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 0.1314213126897812,
      "learning_rate": 0.0004901214206080269,
      "loss": 0.4978,
      "num_input_tokens_seen": 13851096,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 0.118181012570858,
      "learning_rate": 0.0004898927805881577,
      "loss": 0.2982,
      "num_input_tokens_seen": 13854392,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 0.17496263980865479,
      "learning_rate": 0.0004896641426826036,
      "loss": 0.4575,
      "num_input_tokens_seen": 13857400,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 0.09642907977104187,
      "learning_rate": 0.0004894355069391931,
      "loss": 0.4752,
      "num_input_tokens_seen": 13860056,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 0.1216716319322586,
      "learning_rate": 0.0004892068734057538,
      "loss": 0.4509,
      "num_input_tokens_seen": 13863448,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 0.2264557033777237,
      "learning_rate": 0.0004889782421301134,
      "loss": 0.3799,
      "num_input_tokens_seen": 13866040,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 0.20176927745342255,
      "learning_rate": 0.0004887496131600991,
      "loss": 0.4804,
      "num_input_tokens_seen": 13869176,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 0.07442789524793625,
      "learning_rate": 0.0004885209865435373,
      "loss": 0.4647,
      "num_input_tokens_seen": 13873688,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 0.12016964703798294,
      "learning_rate": 0.0004882923623282543,
      "loss": 0.3997,
      "num_input_tokens_seen": 13876760,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 0.12951132655143738,
      "learning_rate": 0.0004880637405620756,
      "loss": 0.4816,
      "num_input_tokens_seen": 13879224,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 0.18244677782058716,
      "learning_rate": 0.00048783512129282604,
      "loss": 0.4838,
      "num_input_tokens_seen": 13881432,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 0.11205962300300598,
      "learning_rate": 0.00048760650456833033,
      "loss": 0.3425,
      "num_input_tokens_seen": 13885432,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 0.16533830761909485,
      "learning_rate": 0.0004873778904364125,
      "loss": 0.4788,
      "num_input_tokens_seen": 13888056,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 0.16799086332321167,
      "learning_rate": 0.00048714927894489597,
      "loss": 0.5274,
      "num_input_tokens_seen": 13890776,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 0.11895795911550522,
      "learning_rate": 0.00048692067014160357,
      "loss": 0.5243,
      "num_input_tokens_seen": 13894008,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 0.1368561089038849,
      "learning_rate": 0.00048669206407435806,
      "loss": 0.4205,
      "num_input_tokens_seen": 13897144,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 0.18568553030490875,
      "learning_rate": 0.00048646346079098075,
      "loss": 0.448,
      "num_input_tokens_seen": 13900312,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 0.148000106215477,
      "learning_rate": 0.0004862348603392929,
      "loss": 0.4826,
      "num_input_tokens_seen": 13903448,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 0.10695650428533554,
      "learning_rate": 0.0004860062627671154,
      "loss": 0.3941,
      "num_input_tokens_seen": 13906808,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 0.15451069176197052,
      "learning_rate": 0.0004857776681222679,
      "loss": 0.4684,
      "num_input_tokens_seen": 13909720,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 0.1144069954752922,
      "learning_rate": 0.00048554907645257024,
      "loss": 0.4273,
      "num_input_tokens_seen": 13913688,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 0.13484320044517517,
      "learning_rate": 0.0004853204878058407,
      "loss": 0.5117,
      "num_input_tokens_seen": 13916568,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 0.19826363027095795,
      "learning_rate": 0.0004850919022298977,
      "loss": 0.4505,
      "num_input_tokens_seen": 13919160,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 0.10891585052013397,
      "learning_rate": 0.0004848633197725586,
      "loss": 0.3797,
      "num_input_tokens_seen": 13922072,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 0.10723894834518433,
      "learning_rate": 0.0004846347404816404,
      "loss": 0.5046,
      "num_input_tokens_seen": 13925688,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 0.20159736275672913,
      "learning_rate": 0.0004844061644049591,
      "loss": 0.5043,
      "num_input_tokens_seen": 13928408,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 0.11774048954248428,
      "learning_rate": 0.0004841775915903305,
      "loss": 0.5092,
      "num_input_tokens_seen": 13932152,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 0.1895529180765152,
      "learning_rate": 0.00048394902208556917,
      "loss": 0.3755,
      "num_input_tokens_seen": 13935864,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 0.16221708059310913,
      "learning_rate": 0.0004837204559384892,
      "loss": 0.438,
      "num_input_tokens_seen": 13938200,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 0.3116278648376465,
      "learning_rate": 0.0004834918931969042,
      "loss": 0.4792,
      "num_input_tokens_seen": 13940824,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 0.10924316942691803,
      "learning_rate": 0.0004832633339086267,
      "loss": 0.411,
      "num_input_tokens_seen": 13943704,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 0.1898515820503235,
      "learning_rate": 0.0004830347781214689,
      "loss": 0.3874,
      "num_input_tokens_seen": 13946808,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 0.12250484526157379,
      "learning_rate": 0.0004828062258832421,
      "loss": 0.4592,
      "num_input_tokens_seen": 13949880,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 0.18033334612846375,
      "learning_rate": 0.0004825776772417564,
      "loss": 0.3949,
      "num_input_tokens_seen": 13952728,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 0.6319389939308167,
      "learning_rate": 0.00048234913224482205,
      "loss": 0.5028,
      "num_input_tokens_seen": 13955320,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 0.1620103120803833,
      "learning_rate": 0.00048212059094024774,
      "loss": 0.3765,
      "num_input_tokens_seen": 13958712,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.15005479753017426,
      "learning_rate": 0.0004818920533758418,
      "loss": 0.5194,
      "num_input_tokens_seen": 13961912,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 0.08771645277738571,
      "learning_rate": 0.00048166351959941185,
      "loss": 0.4073,
      "num_input_tokens_seen": 13965784,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 0.1251508891582489,
      "learning_rate": 0.0004814349896587645,
      "loss": 0.4477,
      "num_input_tokens_seen": 13968088,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 0.2048286348581314,
      "learning_rate": 0.0004812064636017053,
      "loss": 0.3502,
      "num_input_tokens_seen": 13971000,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 0.11232218891382217,
      "learning_rate": 0.0004809779414760396,
      "loss": 0.3831,
      "num_input_tokens_seen": 13973944,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 0.07372494041919708,
      "learning_rate": 0.00048074942332957136,
      "loss": 0.5359,
      "num_input_tokens_seen": 13977272,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 0.18074649572372437,
      "learning_rate": 0.00048052090921010435,
      "loss": 0.5765,
      "num_input_tokens_seen": 13980536,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 0.23115409910678864,
      "learning_rate": 0.00048029239916544094,
      "loss": 0.4659,
      "num_input_tokens_seen": 13983416,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 0.09762795269489288,
      "learning_rate": 0.0004800638932433826,
      "loss": 0.4141,
      "num_input_tokens_seen": 13987640,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 0.11983642727136612,
      "learning_rate": 0.00047983539149173043,
      "loss": 0.4276,
      "num_input_tokens_seen": 13989944,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 0.10734406113624573,
      "learning_rate": 0.0004796068939582843,
      "loss": 0.5276,
      "num_input_tokens_seen": 13994552,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 0.15420161187648773,
      "learning_rate": 0.00047937840069084333,
      "loss": 0.4743,
      "num_input_tokens_seen": 13997176,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 0.12727388739585876,
      "learning_rate": 0.0004791499117372056,
      "loss": 0.4341,
      "num_input_tokens_seen": 14000056,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 0.17074847221374512,
      "learning_rate": 0.0004789214271451686,
      "loss": 0.5407,
      "num_input_tokens_seen": 14002840,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 0.0974038690328598,
      "learning_rate": 0.00047869294696252854,
      "loss": 0.354,
      "num_input_tokens_seen": 14005880,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 0.11541248112916946,
      "learning_rate": 0.0004784644712370807,
      "loss": 0.523,
      "num_input_tokens_seen": 14009176,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.12264413386583328,
      "learning_rate": 0.00047823600001661984,
      "loss": 0.3119,
      "num_input_tokens_seen": 14012728,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 0.14798583090305328,
      "learning_rate": 0.0004780075333489396,
      "loss": 0.4855,
      "num_input_tokens_seen": 14017560,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 0.10777588188648224,
      "learning_rate": 0.00047777907128183233,
      "loss": 0.4907,
      "num_input_tokens_seen": 14020824,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 0.10196299850940704,
      "learning_rate": 0.0004775506138630901,
      "loss": 0.4871,
      "num_input_tokens_seen": 14024920,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 0.13971978425979614,
      "learning_rate": 0.0004773221611405032,
      "loss": 0.3786,
      "num_input_tokens_seen": 14027864,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 0.17617835104465485,
      "learning_rate": 0.0004770937131618614,
      "loss": 0.4066,
      "num_input_tokens_seen": 14030456,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 0.22681647539138794,
      "learning_rate": 0.0004768652699749535,
      "loss": 0.4644,
      "num_input_tokens_seen": 14034104,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 0.1382366269826889,
      "learning_rate": 0.00047663683162756717,
      "loss": 0.4191,
      "num_input_tokens_seen": 14039640,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 0.10063354671001434,
      "learning_rate": 0.00047640839816748926,
      "loss": 0.5782,
      "num_input_tokens_seen": 14042520,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 0.24708276987075806,
      "learning_rate": 0.00047617996964250525,
      "loss": 0.4601,
      "num_input_tokens_seen": 14045688,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 0.10118605941534042,
      "learning_rate": 0.0004759515461003997,
      "loss": 0.4375,
      "num_input_tokens_seen": 14048728,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 0.14765490591526031,
      "learning_rate": 0.00047572312758895654,
      "loss": 0.45,
      "num_input_tokens_seen": 14051288,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 0.1302272379398346,
      "learning_rate": 0.0004754947141559579,
      "loss": 0.5557,
      "num_input_tokens_seen": 14054616,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 0.09866221249103546,
      "learning_rate": 0.0004752663058491856,
      "loss": 0.4505,
      "num_input_tokens_seen": 14057496,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 0.10753985494375229,
      "learning_rate": 0.0004750379027164201,
      "loss": 0.3673,
      "num_input_tokens_seen": 14060024,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 0.09607435017824173,
      "learning_rate": 0.00047480950480544037,
      "loss": 0.3441,
      "num_input_tokens_seen": 14063288,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.10792625695466995,
      "learning_rate": 0.00047458111216402483,
      "loss": 0.6174,
      "num_input_tokens_seen": 14066840,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 0.15053382515907288,
      "learning_rate": 0.00047435272483995074,
      "loss": 0.3772,
      "num_input_tokens_seen": 14069912,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 0.12421189993619919,
      "learning_rate": 0.00047412434288099386,
      "loss": 0.4785,
      "num_input_tokens_seen": 14073112,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 0.1642599105834961,
      "learning_rate": 0.0004738959663349294,
      "loss": 0.4032,
      "num_input_tokens_seen": 14076504,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 0.08748695254325867,
      "learning_rate": 0.0004736675952495311,
      "loss": 0.4263,
      "num_input_tokens_seen": 14080344,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 0.17556439340114594,
      "learning_rate": 0.0004734392296725712,
      "loss": 0.4134,
      "num_input_tokens_seen": 14083416,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 0.12518936395645142,
      "learning_rate": 0.00047321086965182153,
      "loss": 0.4711,
      "num_input_tokens_seen": 14087032,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.10742117464542389,
      "learning_rate": 0.0004729825152350522,
      "loss": 0.4078,
      "num_input_tokens_seen": 14090936,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 0.12489108741283417,
      "learning_rate": 0.00047275416647003256,
      "loss": 0.3948,
      "num_input_tokens_seen": 14093688,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 0.08606471121311188,
      "learning_rate": 0.0004725258234045306,
      "loss": 0.4668,
      "num_input_tokens_seen": 14097336,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 0.13699458539485931,
      "learning_rate": 0.0004722974860863127,
      "loss": 0.4025,
      "num_input_tokens_seen": 14100984,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 0.10648544132709503,
      "learning_rate": 0.00047206915456314476,
      "loss": 0.4526,
      "num_input_tokens_seen": 14103672,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 0.14246204495429993,
      "learning_rate": 0.00047184082888279097,
      "loss": 0.4947,
      "num_input_tokens_seen": 14106872,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 0.21785767376422882,
      "learning_rate": 0.00047161250909301454,
      "loss": 0.4002,
      "num_input_tokens_seen": 14109720,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 0.10483436286449432,
      "learning_rate": 0.00047138419524157734,
      "loss": 0.4039,
      "num_input_tokens_seen": 14113112,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 0.10602086782455444,
      "learning_rate": 0.0004711558873762402,
      "loss": 0.4924,
      "num_input_tokens_seen": 14116440,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 0.12623190879821777,
      "learning_rate": 0.0004709275855447621,
      "loss": 0.6045,
      "num_input_tokens_seen": 14119384,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 0.19692781567573547,
      "learning_rate": 0.0004706992897949016,
      "loss": 0.5217,
      "num_input_tokens_seen": 14122008,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 0.14341987669467926,
      "learning_rate": 0.00047047100017441525,
      "loss": 0.4293,
      "num_input_tokens_seen": 14125688,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 0.14586953818798065,
      "learning_rate": 0.0004702427167310589,
      "loss": 0.3212,
      "num_input_tokens_seen": 14129144,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 0.09336761385202408,
      "learning_rate": 0.0004700144395125866,
      "loss": 0.3962,
      "num_input_tokens_seen": 14131960,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 0.08285865187644958,
      "learning_rate": 0.00046978616856675173,
      "loss": 0.5033,
      "num_input_tokens_seen": 14136312,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 0.21683555841445923,
      "learning_rate": 0.0004695579039413055,
      "loss": 0.4153,
      "num_input_tokens_seen": 14139416,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 0.1400396078824997,
      "learning_rate": 0.0004693296456839984,
      "loss": 0.405,
      "num_input_tokens_seen": 14142904,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 0.1500183790922165,
      "learning_rate": 0.0004691013938425796,
      "loss": 0.4336,
      "num_input_tokens_seen": 14146328,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 0.17254775762557983,
      "learning_rate": 0.00046887314846479663,
      "loss": 0.3592,
      "num_input_tokens_seen": 14149368,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 0.19880221784114838,
      "learning_rate": 0.000468644909598396,
      "loss": 0.3373,
      "num_input_tokens_seen": 14153400,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 0.19495338201522827,
      "learning_rate": 0.00046841667729112264,
      "loss": 0.4414,
      "num_input_tokens_seen": 14155704,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 0.15996576845645905,
      "learning_rate": 0.00046818845159071987,
      "loss": 0.3895,
      "num_input_tokens_seen": 14158584,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 0.14125381410121918,
      "learning_rate": 0.0004679602325449302,
      "loss": 0.3074,
      "num_input_tokens_seen": 14162040,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 0.12406895309686661,
      "learning_rate": 0.00046773202020149423,
      "loss": 0.3353,
      "num_input_tokens_seen": 14166008,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 0.13993793725967407,
      "learning_rate": 0.0004675038146081516,
      "loss": 0.4715,
      "num_input_tokens_seen": 14170040,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 0.12619364261627197,
      "learning_rate": 0.0004672756158126403,
      "loss": 0.4233,
      "num_input_tokens_seen": 14173016,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 0.310507595539093,
      "learning_rate": 0.0004670474238626968,
      "loss": 0.2968,
      "num_input_tokens_seen": 14176152,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 0.11117317527532578,
      "learning_rate": 0.00046681923880605614,
      "loss": 0.392,
      "num_input_tokens_seen": 14179736,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 0.2644006609916687,
      "learning_rate": 0.0004665910606904523,
      "loss": 0.3312,
      "num_input_tokens_seen": 14183224,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 0.10765819251537323,
      "learning_rate": 0.0004663628895636173,
      "loss": 0.4685,
      "num_input_tokens_seen": 14186200,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.14004386961460114,
      "learning_rate": 0.0004661347254732822,
      "loss": 0.4192,
      "num_input_tokens_seen": 14189848,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 0.14626839756965637,
      "learning_rate": 0.0004659065684671762,
      "loss": 0.4106,
      "num_input_tokens_seen": 14193400,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 0.29469987750053406,
      "learning_rate": 0.00046567841859302696,
      "loss": 0.3713,
      "num_input_tokens_seen": 14195960,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 0.24419932067394257,
      "learning_rate": 0.0004654502758985611,
      "loss": 0.4262,
      "num_input_tokens_seen": 14198520,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 0.22819724678993225,
      "learning_rate": 0.00046522214043150323,
      "loss": 0.425,
      "num_input_tokens_seen": 14201816,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 0.15504343807697296,
      "learning_rate": 0.00046499401223957696,
      "loss": 0.5202,
      "num_input_tokens_seen": 14205048,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 0.13718363642692566,
      "learning_rate": 0.0004647658913705038,
      "loss": 0.4189,
      "num_input_tokens_seen": 14207544,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 0.10060016065835953,
      "learning_rate": 0.00046453777787200444,
      "loss": 0.472,
      "num_input_tokens_seen": 14211128,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 0.16134962439537048,
      "learning_rate": 0.0004643096717917973,
      "loss": 0.4466,
      "num_input_tokens_seen": 14214328,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 0.08551730215549469,
      "learning_rate": 0.0004640815731775995,
      "loss": 0.354,
      "num_input_tokens_seen": 14218904,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 0.13787566125392914,
      "learning_rate": 0.0004638534820771267,
      "loss": 0.4531,
      "num_input_tokens_seen": 14222392,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 0.2132241576910019,
      "learning_rate": 0.0004636253985380934,
      "loss": 0.4865,
      "num_input_tokens_seen": 14224792,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 0.10204747319221497,
      "learning_rate": 0.00046339732260821173,
      "loss": 0.5699,
      "num_input_tokens_seen": 14228248,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 0.1092388778924942,
      "learning_rate": 0.0004631692543351924,
      "loss": 0.4504,
      "num_input_tokens_seen": 14232088,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 0.12752017378807068,
      "learning_rate": 0.00046294119376674486,
      "loss": 0.4191,
      "num_input_tokens_seen": 14235640,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 0.09570431709289551,
      "learning_rate": 0.0004627131409505768,
      "loss": 0.3893,
      "num_input_tokens_seen": 14238904,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 0.12748727202415466,
      "learning_rate": 0.0004624850959343944,
      "loss": 0.4724,
      "num_input_tokens_seen": 14243288,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 0.19836489856243134,
      "learning_rate": 0.00046225705876590175,
      "loss": 0.4187,
      "num_input_tokens_seen": 14246200,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 0.1554318517446518,
      "learning_rate": 0.00046202902949280206,
      "loss": 0.597,
      "num_input_tokens_seen": 14249656,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 0.17096012830734253,
      "learning_rate": 0.00046180100816279614,
      "loss": 0.4054,
      "num_input_tokens_seen": 14252920,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 0.14876225590705872,
      "learning_rate": 0.0004615729948235833,
      "loss": 0.3543,
      "num_input_tokens_seen": 14256152,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 0.12253880500793457,
      "learning_rate": 0.00046134498952286174,
      "loss": 0.3679,
      "num_input_tokens_seen": 14259224,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 0.2206830382347107,
      "learning_rate": 0.00046111699230832727,
      "loss": 0.4603,
      "num_input_tokens_seen": 14262232,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 0.19829323887825012,
      "learning_rate": 0.00046088900322767434,
      "loss": 0.4674,
      "num_input_tokens_seen": 14265752,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 0.19374501705169678,
      "learning_rate": 0.0004606610223285961,
      "loss": 0.4302,
      "num_input_tokens_seen": 14268888,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 0.09722213447093964,
      "learning_rate": 0.00046043304965878284,
      "loss": 0.3379,
      "num_input_tokens_seen": 14272024,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 0.1090613603591919,
      "learning_rate": 0.00046020508526592425,
      "loss": 0.5228,
      "num_input_tokens_seen": 14275128,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 0.18867741525173187,
      "learning_rate": 0.0004599771291977079,
      "loss": 0.5936,
      "num_input_tokens_seen": 14278936,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 0.1661062389612198,
      "learning_rate": 0.0004597491815018194,
      "loss": 0.5038,
      "num_input_tokens_seen": 14282040,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.1665009707212448,
      "learning_rate": 0.0004595212422259431,
      "loss": 0.2976,
      "num_input_tokens_seen": 14285688,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 0.08397595584392548,
      "learning_rate": 0.00045929331141776104,
      "loss": 0.3702,
      "num_input_tokens_seen": 14289688,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 0.24280188977718353,
      "learning_rate": 0.0004590653891249537,
      "loss": 0.4181,
      "num_input_tokens_seen": 14292440,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 0.09344487637281418,
      "learning_rate": 0.0004588374753952001,
      "loss": 0.503,
      "num_input_tokens_seen": 14296568,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 0.1226547509431839,
      "learning_rate": 0.0004586095702761769,
      "loss": 0.5568,
      "num_input_tokens_seen": 14299512,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.13309796154499054,
      "learning_rate": 0.0004583816738155595,
      "loss": 0.5114,
      "num_input_tokens_seen": 14303512,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 0.1388390064239502,
      "learning_rate": 0.0004581537860610213,
      "loss": 0.3615,
      "num_input_tokens_seen": 14306712,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 0.15758763253688812,
      "learning_rate": 0.0004579259070602333,
      "loss": 0.5254,
      "num_input_tokens_seen": 14309784,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 0.11948682367801666,
      "learning_rate": 0.0004576980368608658,
      "loss": 0.4215,
      "num_input_tokens_seen": 14312696,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 0.18370604515075684,
      "learning_rate": 0.0004574701755105862,
      "loss": 0.3751,
      "num_input_tokens_seen": 14315736,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 0.4017069637775421,
      "learning_rate": 0.00045724232305706066,
      "loss": 0.3787,
      "num_input_tokens_seen": 14318488,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.28205934166908264,
      "learning_rate": 0.00045701447954795343,
      "loss": 0.4854,
      "num_input_tokens_seen": 14321240,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 0.09954674541950226,
      "learning_rate": 0.0004567866450309268,
      "loss": 0.4192,
      "num_input_tokens_seen": 14324024,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 0.10691235214471817,
      "learning_rate": 0.00045655881955364086,
      "loss": 0.4635,
      "num_input_tokens_seen": 14327320,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 0.1593731790781021,
      "learning_rate": 0.0004563310031637543,
      "loss": 0.4547,
      "num_input_tokens_seen": 14330584,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 0.18352386355400085,
      "learning_rate": 0.00045610319590892365,
      "loss": 0.4392,
      "num_input_tokens_seen": 14333592,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 0.06445270776748657,
      "learning_rate": 0.00045587539783680366,
      "loss": 0.3748,
      "num_input_tokens_seen": 14337496,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 0.08968200534582138,
      "learning_rate": 0.00045564760899504715,
      "loss": 0.3561,
      "num_input_tokens_seen": 14341304,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 0.0899999663233757,
      "learning_rate": 0.00045541982943130466,
      "loss": 0.3533,
      "num_input_tokens_seen": 14344312,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.2555168569087982,
      "learning_rate": 0.00045519205919322534,
      "loss": 0.3937,
      "num_input_tokens_seen": 14346648,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 0.1410970836877823,
      "learning_rate": 0.00045496429832845594,
      "loss": 0.3312,
      "num_input_tokens_seen": 14349816,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 0.14550898969173431,
      "learning_rate": 0.0004547365468846416,
      "loss": 0.5453,
      "num_input_tokens_seen": 14353656,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 0.09066812694072723,
      "learning_rate": 0.0004545088049094252,
      "loss": 0.5652,
      "num_input_tokens_seen": 14357624,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 0.1946171522140503,
      "learning_rate": 0.00045428107245044805,
      "loss": 0.4017,
      "num_input_tokens_seen": 14360664,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 0.27734851837158203,
      "learning_rate": 0.0004540533495553488,
      "loss": 0.3997,
      "num_input_tokens_seen": 14363960,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.22636857628822327,
      "learning_rate": 0.00045382563627176454,
      "loss": 0.3898,
      "num_input_tokens_seen": 14366424,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 0.15735460817813873,
      "learning_rate": 0.0004535979326473304,
      "loss": 0.6727,
      "num_input_tokens_seen": 14370232,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 0.29900628328323364,
      "learning_rate": 0.00045337023872967945,
      "loss": 0.5644,
      "num_input_tokens_seen": 14373976,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 0.16691488027572632,
      "learning_rate": 0.0004531425545664425,
      "loss": 0.4398,
      "num_input_tokens_seen": 14376568,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 0.30801263451576233,
      "learning_rate": 0.0004529148802052488,
      "loss": 0.4139,
      "num_input_tokens_seen": 14379256,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 0.10866888612508774,
      "learning_rate": 0.00045268721569372483,
      "loss": 0.3919,
      "num_input_tokens_seen": 14381752,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 0.08178889751434326,
      "learning_rate": 0.00045245956107949547,
      "loss": 0.4745,
      "num_input_tokens_seen": 14384984,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 0.1310143768787384,
      "learning_rate": 0.0004522319164101837,
      "loss": 0.4109,
      "num_input_tokens_seen": 14388504,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 0.17378047108650208,
      "learning_rate": 0.00045200428173340987,
      "loss": 0.4969,
      "num_input_tokens_seen": 14391256,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 0.1565741002559662,
      "learning_rate": 0.00045177665709679286,
      "loss": 0.4739,
      "num_input_tokens_seen": 14393880,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 0.22839409112930298,
      "learning_rate": 0.0004515490425479492,
      "loss": 0.4127,
      "num_input_tokens_seen": 14396248,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 0.25417476892471313,
      "learning_rate": 0.0004513214381344928,
      "loss": 0.4201,
      "num_input_tokens_seen": 14398616,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.1322731375694275,
      "learning_rate": 0.00045109384390403635,
      "loss": 0.5156,
      "num_input_tokens_seen": 14401496,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 0.16300828754901886,
      "learning_rate": 0.0004508662599041896,
      "loss": 0.3782,
      "num_input_tokens_seen": 14404216,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 0.14448219537734985,
      "learning_rate": 0.0004506386861825609,
      "loss": 0.6041,
      "num_input_tokens_seen": 14407864,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 0.09811676293611526,
      "learning_rate": 0.00045041112278675603,
      "loss": 0.4819,
      "num_input_tokens_seen": 14411416,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 0.14976029098033905,
      "learning_rate": 0.00045018356976437836,
      "loss": 0.4028,
      "num_input_tokens_seen": 14414168,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 0.154007688164711,
      "learning_rate": 0.0004499560271630295,
      "loss": 0.4432,
      "num_input_tokens_seen": 14418456,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 0.20888520777225494,
      "learning_rate": 0.00044972849503030894,
      "loss": 0.4738,
      "num_input_tokens_seen": 14421624,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 0.09456796199083328,
      "learning_rate": 0.00044950097341381356,
      "loss": 0.4051,
      "num_input_tokens_seen": 14424120,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 0.30070760846138,
      "learning_rate": 0.0004492734623611385,
      "loss": 0.5315,
      "num_input_tokens_seen": 14428120,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 0.10071755945682526,
      "learning_rate": 0.0004490459619198766,
      "loss": 0.6252,
      "num_input_tokens_seen": 14431288,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 0.1609276831150055,
      "learning_rate": 0.00044881847213761787,
      "loss": 0.3866,
      "num_input_tokens_seen": 14434104,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.1393337845802307,
      "learning_rate": 0.000448590993061951,
      "loss": 0.4567,
      "num_input_tokens_seen": 14436664,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 0.149698868393898,
      "learning_rate": 0.00044836352474046177,
      "loss": 0.4246,
      "num_input_tokens_seen": 14443672,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 0.1277666687965393,
      "learning_rate": 0.00044813606722073415,
      "loss": 0.6585,
      "num_input_tokens_seen": 14446200,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 0.10574603080749512,
      "learning_rate": 0.00044790862055034945,
      "loss": 0.4541,
      "num_input_tokens_seen": 14449240,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 0.08658820390701294,
      "learning_rate": 0.00044768118477688727,
      "loss": 0.4664,
      "num_input_tokens_seen": 14452024,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 0.11474493145942688,
      "learning_rate": 0.00044745375994792425,
      "loss": 0.4385,
      "num_input_tokens_seen": 14455992,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 0.1485312581062317,
      "learning_rate": 0.00044722634611103507,
      "loss": 0.322,
      "num_input_tokens_seen": 14460536,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 0.12149445712566376,
      "learning_rate": 0.0004469989433137923,
      "loss": 0.4472,
      "num_input_tokens_seen": 14463000,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 0.1339751034975052,
      "learning_rate": 0.00044677155160376586,
      "loss": 0.4358,
      "num_input_tokens_seen": 14467032,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 0.15794458985328674,
      "learning_rate": 0.00044654417102852375,
      "loss": 0.4193,
      "num_input_tokens_seen": 14470232,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 0.11345532536506653,
      "learning_rate": 0.0004463168016356309,
      "loss": 0.5201,
      "num_input_tokens_seen": 14473880,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 0.18680527806282043,
      "learning_rate": 0.00044608944347265095,
      "loss": 0.4026,
      "num_input_tokens_seen": 14476792,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 0.25150176882743835,
      "learning_rate": 0.0004458620965871442,
      "loss": 0.4276,
      "num_input_tokens_seen": 14479544,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 0.19843968749046326,
      "learning_rate": 0.00044563476102666935,
      "loss": 0.4831,
      "num_input_tokens_seen": 14482424,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 0.07715383917093277,
      "learning_rate": 0.0004454074368387821,
      "loss": 0.318,
      "num_input_tokens_seen": 14485784,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 0.15643124282360077,
      "learning_rate": 0.0004451801240710365,
      "loss": 0.4052,
      "num_input_tokens_seen": 14488216,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 0.05627750605344772,
      "learning_rate": 0.0004449528227709834,
      "loss": 0.4202,
      "num_input_tokens_seen": 14494264,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 0.0983547642827034,
      "learning_rate": 0.0004447255329861716,
      "loss": 0.3935,
      "num_input_tokens_seen": 14498680,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 0.06319372355937958,
      "learning_rate": 0.0004444982547641479,
      "loss": 0.4109,
      "num_input_tokens_seen": 14502488,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 0.1186370700597763,
      "learning_rate": 0.00044427098815245594,
      "loss": 0.5409,
      "num_input_tokens_seen": 14505624,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 0.16066260635852814,
      "learning_rate": 0.00044404373319863763,
      "loss": 0.3641,
      "num_input_tokens_seen": 14508312,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 0.13632960617542267,
      "learning_rate": 0.000443816489950232,
      "loss": 0.4023,
      "num_input_tokens_seen": 14511320,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 0.12915454804897308,
      "learning_rate": 0.00044358925845477547,
      "loss": 0.3778,
      "num_input_tokens_seen": 14514840,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 0.17071211338043213,
      "learning_rate": 0.0004433620387598027,
      "loss": 0.3957,
      "num_input_tokens_seen": 14517944,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 0.13207630813121796,
      "learning_rate": 0.00044313483091284507,
      "loss": 0.4482,
      "num_input_tokens_seen": 14521432,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 0.13203851878643036,
      "learning_rate": 0.0004429076349614321,
      "loss": 0.3555,
      "num_input_tokens_seen": 14524248,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 0.12656114995479584,
      "learning_rate": 0.0004426804509530907,
      "loss": 0.4552,
      "num_input_tokens_seen": 14527448,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 0.12005181610584259,
      "learning_rate": 0.000442453278935345,
      "loss": 0.5117,
      "num_input_tokens_seen": 14529944,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 0.19013337790966034,
      "learning_rate": 0.0004422261189557167,
      "loss": 0.5132,
      "num_input_tokens_seen": 14533144,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.08593453466892242,
      "learning_rate": 0.00044199897106172526,
      "loss": 0.5217,
      "num_input_tokens_seen": 14536888,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 0.09309424459934235,
      "learning_rate": 0.0004417718353008873,
      "loss": 0.5053,
      "num_input_tokens_seen": 14545912,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 0.22856688499450684,
      "learning_rate": 0.0004415447117207172,
      "loss": 0.3705,
      "num_input_tokens_seen": 14548856,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 0.14289309084415436,
      "learning_rate": 0.00044131760036872657,
      "loss": 0.7268,
      "num_input_tokens_seen": 14551800,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 0.13919396698474884,
      "learning_rate": 0.00044109050129242434,
      "loss": 0.4701,
      "num_input_tokens_seen": 14555256,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 0.128069207072258,
      "learning_rate": 0.00044086341453931715,
      "loss": 0.522,
      "num_input_tokens_seen": 14558840,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 0.21287576854228973,
      "learning_rate": 0.00044063634015690893,
      "loss": 0.3606,
      "num_input_tokens_seen": 14561848,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 0.07387420535087585,
      "learning_rate": 0.0004404092781927012,
      "loss": 0.3501,
      "num_input_tokens_seen": 14565464,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.18159449100494385,
      "learning_rate": 0.00044018222869419255,
      "loss": 0.5275,
      "num_input_tokens_seen": 14568664,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 0.2060248702764511,
      "learning_rate": 0.00043995519170887936,
      "loss": 0.4242,
      "num_input_tokens_seen": 14571384,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 0.0714694932103157,
      "learning_rate": 0.00043972816728425504,
      "loss": 0.4199,
      "num_input_tokens_seen": 14574808,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 0.145622119307518,
      "learning_rate": 0.0004395011554678103,
      "loss": 0.4397,
      "num_input_tokens_seen": 14577624,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 0.1336793452501297,
      "learning_rate": 0.00043927415630703364,
      "loss": 0.3781,
      "num_input_tokens_seen": 14581016,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 0.17729514837265015,
      "learning_rate": 0.0004390471698494108,
      "loss": 0.4212,
      "num_input_tokens_seen": 14583736,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 0.10639865696430206,
      "learning_rate": 0.0004388201961424248,
      "loss": 0.4845,
      "num_input_tokens_seen": 14587352,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 0.13763004541397095,
      "learning_rate": 0.0004385932352335555,
      "loss": 0.4622,
      "num_input_tokens_seen": 14590808,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 0.1719980090856552,
      "learning_rate": 0.000438366287170281,
      "loss": 0.3886,
      "num_input_tokens_seen": 14594296,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 0.16627955436706543,
      "learning_rate": 0.0004381393520000759,
      "loss": 0.3395,
      "num_input_tokens_seen": 14597944,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 0.22830455005168915,
      "learning_rate": 0.0004379124297704127,
      "loss": 0.4036,
      "num_input_tokens_seen": 14600920,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 0.1342162936925888,
      "learning_rate": 0.0004376855205287609,
      "loss": 0.433,
      "num_input_tokens_seen": 14604504,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 0.12527871131896973,
      "learning_rate": 0.0004374586243225874,
      "loss": 0.4654,
      "num_input_tokens_seen": 14607928,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 0.14052926003932953,
      "learning_rate": 0.00043723174119935607,
      "loss": 0.505,
      "num_input_tokens_seen": 14611000,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 0.19624502956867218,
      "learning_rate": 0.00043700487120652826,
      "loss": 0.4542,
      "num_input_tokens_seen": 14614456,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 0.11153988540172577,
      "learning_rate": 0.00043677801439156284,
      "loss": 0.4098,
      "num_input_tokens_seen": 14618648,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 0.14413344860076904,
      "learning_rate": 0.00043655117080191545,
      "loss": 0.4281,
      "num_input_tokens_seen": 14621464,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 0.21224476397037506,
      "learning_rate": 0.00043632434048503934,
      "loss": 0.3414,
      "num_input_tokens_seen": 14624472,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 0.2126932591199875,
      "learning_rate": 0.00043609752348838484,
      "loss": 0.5636,
      "num_input_tokens_seen": 14628824,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 0.1398584544658661,
      "learning_rate": 0.0004358707198593992,
      "loss": 0.3721,
      "num_input_tokens_seen": 14631832,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 0.14184299111366272,
      "learning_rate": 0.0004356439296455273,
      "loss": 0.3471,
      "num_input_tokens_seen": 14635416,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 0.14518587291240692,
      "learning_rate": 0.0004354171528942113,
      "loss": 0.5697,
      "num_input_tokens_seen": 14638584,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 0.1563519984483719,
      "learning_rate": 0.00043519038965288994,
      "loss": 0.4739,
      "num_input_tokens_seen": 14641880,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 0.1832299530506134,
      "learning_rate": 0.0004349636399689998,
      "loss": 0.6024,
      "num_input_tokens_seen": 14645144,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 0.1677011102437973,
      "learning_rate": 0.00043473690388997434,
      "loss": 0.5772,
      "num_input_tokens_seen": 14648312,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 0.12260983884334564,
      "learning_rate": 0.0004345101814632438,
      "loss": 0.5057,
      "num_input_tokens_seen": 14651768,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 0.09753179550170898,
      "learning_rate": 0.0004342834727362362,
      "loss": 0.6211,
      "num_input_tokens_seen": 14655640,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 0.23719169199466705,
      "learning_rate": 0.00043405677775637633,
      "loss": 0.5526,
      "num_input_tokens_seen": 14658424,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 0.2084636688232422,
      "learning_rate": 0.0004338300965710863,
      "loss": 0.3738,
      "num_input_tokens_seen": 14660888,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 0.33953288197517395,
      "learning_rate": 0.0004336034292277853,
      "loss": 0.4356,
      "num_input_tokens_seen": 14664888,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 0.21611551940441132,
      "learning_rate": 0.00043337677577388916,
      "loss": 0.4749,
      "num_input_tokens_seen": 14667864,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 0.1994357407093048,
      "learning_rate": 0.0004331501362568116,
      "loss": 0.497,
      "num_input_tokens_seen": 14670296,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 0.14535683393478394,
      "learning_rate": 0.00043292351072396273,
      "loss": 0.4724,
      "num_input_tokens_seen": 14673592,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 0.11912309378385544,
      "learning_rate": 0.0004326968992227503,
      "loss": 0.4518,
      "num_input_tokens_seen": 14676664,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 0.12782636284828186,
      "learning_rate": 0.0004324703018005786,
      "loss": 0.4472,
      "num_input_tokens_seen": 14679704,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 0.15755915641784668,
      "learning_rate": 0.00043224371850484955,
      "loss": 0.5242,
      "num_input_tokens_seen": 14682360,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 0.10344871878623962,
      "learning_rate": 0.0004320171493829614,
      "loss": 0.378,
      "num_input_tokens_seen": 14685528,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 0.1613067388534546,
      "learning_rate": 0.00043179059448231004,
      "loss": 0.5829,
      "num_input_tokens_seen": 14688600,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 0.1260797083377838,
      "learning_rate": 0.0004315640538502881,
      "loss": 0.5765,
      "num_input_tokens_seen": 14691544,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 0.14700840413570404,
      "learning_rate": 0.00043133752753428547,
      "loss": 0.3459,
      "num_input_tokens_seen": 14695320,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 0.1945607215166092,
      "learning_rate": 0.0004311110155816886,
      "loss": 0.4982,
      "num_input_tokens_seen": 14699192,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 0.1050792783498764,
      "learning_rate": 0.00043088451803988153,
      "loss": 0.4018,
      "num_input_tokens_seen": 14702264,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 0.186896413564682,
      "learning_rate": 0.00043065803495624465,
      "loss": 0.5669,
      "num_input_tokens_seen": 14705912,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 0.10432635992765427,
      "learning_rate": 0.00043043156637815575,
      "loss": 0.394,
      "num_input_tokens_seen": 14708888,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 0.17152579128742218,
      "learning_rate": 0.00043020511235298956,
      "loss": 0.3604,
      "num_input_tokens_seen": 14712440,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 0.09103792160749435,
      "learning_rate": 0.00042997867292811744,
      "loss": 0.5126,
      "num_input_tokens_seen": 14714968,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 0.1825615018606186,
      "learning_rate": 0.00042975224815090834,
      "loss": 0.5558,
      "num_input_tokens_seen": 14717240,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 0.179880291223526,
      "learning_rate": 0.0004295258380687274,
      "loss": 0.4849,
      "num_input_tokens_seen": 14720248,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 0.1349543184041977,
      "learning_rate": 0.000429299442728937,
      "loss": 0.4423,
      "num_input_tokens_seen": 14723544,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.1437530815601349,
      "learning_rate": 0.0004290730621788967,
      "loss": 0.473,
      "num_input_tokens_seen": 14727416,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.1849072426557541,
      "learning_rate": 0.0004288466964659625,
      "loss": 0.4032,
      "num_input_tokens_seen": 14730328,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 0.11382696032524109,
      "learning_rate": 0.0004286203456374877,
      "loss": 0.3732,
      "num_input_tokens_seen": 14732952,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 0.2413271963596344,
      "learning_rate": 0.0004283940097408224,
      "loss": 0.544,
      "num_input_tokens_seen": 14735800,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 0.0706309974193573,
      "learning_rate": 0.00042816768882331324,
      "loss": 0.3865,
      "num_input_tokens_seen": 14738584,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 0.12789475917816162,
      "learning_rate": 0.000427941382932304,
      "loss": 0.4264,
      "num_input_tokens_seen": 14741944,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 0.12061990052461624,
      "learning_rate": 0.0004277150921151354,
      "loss": 0.472,
      "num_input_tokens_seen": 14745496,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 0.1609293669462204,
      "learning_rate": 0.0004274888164191448,
      "loss": 0.3723,
      "num_input_tokens_seen": 14748088,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.16427090764045715,
      "learning_rate": 0.00042726255589166666,
      "loss": 0.4577,
      "num_input_tokens_seen": 14751096,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 0.191057488322258,
      "learning_rate": 0.0004270363105800321,
      "loss": 0.3897,
      "num_input_tokens_seen": 14754488,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 0.18617956340312958,
      "learning_rate": 0.0004268100805315688,
      "loss": 0.35,
      "num_input_tokens_seen": 14757016,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 0.11150764673948288,
      "learning_rate": 0.00042658386579360165,
      "loss": 0.6181,
      "num_input_tokens_seen": 14760280,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 0.1925751268863678,
      "learning_rate": 0.00042635766641345213,
      "loss": 0.4615,
      "num_input_tokens_seen": 14762840,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 0.0911838710308075,
      "learning_rate": 0.0004261314824384388,
      "loss": 0.3186,
      "num_input_tokens_seen": 14766392,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 0.14858368039131165,
      "learning_rate": 0.00042590531391587666,
      "loss": 0.3561,
      "num_input_tokens_seen": 14770264,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 0.18160317838191986,
      "learning_rate": 0.00042567916089307733,
      "loss": 0.3722,
      "num_input_tokens_seen": 14773336,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.06641208380460739,
      "learning_rate": 0.00042545302341734977,
      "loss": 0.3967,
      "num_input_tokens_seen": 14777080,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 0.0838356465101242,
      "learning_rate": 0.0004252269015359991,
      "loss": 0.5736,
      "num_input_tokens_seen": 14780280,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 0.10825871676206589,
      "learning_rate": 0.00042500079529632763,
      "loss": 0.452,
      "num_input_tokens_seen": 14783736,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 0.15018095076084137,
      "learning_rate": 0.0004247747047456342,
      "loss": 0.423,
      "num_input_tokens_seen": 14786872,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 0.12956584990024567,
      "learning_rate": 0.0004245486299312144,
      "loss": 0.3553,
      "num_input_tokens_seen": 14789176,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 0.25765877962112427,
      "learning_rate": 0.0004243225709003602,
      "loss": 0.4265,
      "num_input_tokens_seen": 14792152,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 0.12196949869394302,
      "learning_rate": 0.000424096527700361,
      "loss": 0.5258,
      "num_input_tokens_seen": 14795256,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 0.14709743857383728,
      "learning_rate": 0.0004238705003785021,
      "loss": 0.4848,
      "num_input_tokens_seen": 14798360,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 0.28132548928260803,
      "learning_rate": 0.0004236444889820661,
      "loss": 0.5136,
      "num_input_tokens_seen": 14801592,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 0.13752909004688263,
      "learning_rate": 0.0004234184935583318,
      "loss": 0.4938,
      "num_input_tokens_seen": 14804344,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 0.15892961621284485,
      "learning_rate": 0.00042319251415457517,
      "loss": 0.5306,
      "num_input_tokens_seen": 14807288,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 0.1059977188706398,
      "learning_rate": 0.0004229665508180681,
      "loss": 0.4019,
      "num_input_tokens_seen": 14810680,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 0.12792406976222992,
      "learning_rate": 0.0004227406035960798,
      "loss": 0.4963,
      "num_input_tokens_seen": 14813432,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 0.12316926568746567,
      "learning_rate": 0.0004225146725358758,
      "loss": 0.3628,
      "num_input_tokens_seen": 14816664,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 0.16719341278076172,
      "learning_rate": 0.0004222887576847183,
      "loss": 0.4159,
      "num_input_tokens_seen": 14819864,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 0.14574235677719116,
      "learning_rate": 0.00042206285908986626,
      "loss": 0.5369,
      "num_input_tokens_seen": 14823736,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 0.10873974859714508,
      "learning_rate": 0.00042183697679857484,
      "loss": 0.46,
      "num_input_tokens_seen": 14826904,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 0.1206497773528099,
      "learning_rate": 0.00042161111085809604,
      "loss": 0.5257,
      "num_input_tokens_seen": 14830168,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 0.11597403883934021,
      "learning_rate": 0.00042138526131567855,
      "loss": 0.4859,
      "num_input_tokens_seen": 14834520,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 0.10441970825195312,
      "learning_rate": 0.0004211594282185677,
      "loss": 0.4702,
      "num_input_tokens_seen": 14837944,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 0.06840342283248901,
      "learning_rate": 0.0004209336116140048,
      "loss": 0.4367,
      "num_input_tokens_seen": 14842168,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 0.17564088106155396,
      "learning_rate": 0.00042070781154922857,
      "loss": 0.3349,
      "num_input_tokens_seen": 14845112,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 0.1246572807431221,
      "learning_rate": 0.0004204820280714734,
      "loss": 0.5013,
      "num_input_tokens_seen": 14847608,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 0.12770095467567444,
      "learning_rate": 0.00042025626122797066,
      "loss": 0.5022,
      "num_input_tokens_seen": 14852024,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 0.12105519324541092,
      "learning_rate": 0.0004200305110659484,
      "loss": 0.4037,
      "num_input_tokens_seen": 14855288,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 0.1443774551153183,
      "learning_rate": 0.0004198047776326308,
      "loss": 0.3566,
      "num_input_tokens_seen": 14858776,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 0.20785298943519592,
      "learning_rate": 0.00041957906097523897,
      "loss": 0.5025,
      "num_input_tokens_seen": 14862872,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 0.1648014932870865,
      "learning_rate": 0.00041935336114099013,
      "loss": 0.4249,
      "num_input_tokens_seen": 14865496,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 0.3877384662628174,
      "learning_rate": 0.0004191276781770979,
      "loss": 0.4516,
      "num_input_tokens_seen": 14868408,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 0.09799838811159134,
      "learning_rate": 0.00041890201213077286,
      "loss": 0.4697,
      "num_input_tokens_seen": 14871544,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 0.11552876234054565,
      "learning_rate": 0.00041867636304922153,
      "loss": 0.5042,
      "num_input_tokens_seen": 14874520,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 0.10114146769046783,
      "learning_rate": 0.00041845073097964737,
      "loss": 0.5118,
      "num_input_tokens_seen": 14878552,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 0.20924913883209229,
      "learning_rate": 0.0004182251159692498,
      "loss": 0.4589,
      "num_input_tokens_seen": 14881144,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.1013297587633133,
      "learning_rate": 0.0004179995180652253,
      "loss": 0.3721,
      "num_input_tokens_seen": 14885560,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 0.10085237771272659,
      "learning_rate": 0.00041777393731476587,
      "loss": 0.4534,
      "num_input_tokens_seen": 14888888,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 0.17339816689491272,
      "learning_rate": 0.0004175483737650608,
      "loss": 0.5185,
      "num_input_tokens_seen": 14891672,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 0.1765376478433609,
      "learning_rate": 0.0004173228274632951,
      "loss": 0.4631,
      "num_input_tokens_seen": 14894680,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 0.14923273026943207,
      "learning_rate": 0.00041709729845665084,
      "loss": 0.4689,
      "num_input_tokens_seen": 14899704,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 0.1269654631614685,
      "learning_rate": 0.0004168717867923061,
      "loss": 0.3472,
      "num_input_tokens_seen": 14902776,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 0.21186494827270508,
      "learning_rate": 0.00041664629251743487,
      "loss": 0.4155,
      "num_input_tokens_seen": 14905560,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 0.1952824890613556,
      "learning_rate": 0.0004164208156792084,
      "loss": 0.4577,
      "num_input_tokens_seen": 14908216,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 0.23160190880298615,
      "learning_rate": 0.00041619535632479374,
      "loss": 0.3825,
      "num_input_tokens_seen": 14910552,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 0.10930733382701874,
      "learning_rate": 0.00041596991450135445,
      "loss": 0.4729,
      "num_input_tokens_seen": 14914232,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 0.14660851657390594,
      "learning_rate": 0.0004157444902560504,
      "loss": 0.4162,
      "num_input_tokens_seen": 14917592,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 0.15489429235458374,
      "learning_rate": 0.00041551908363603786,
      "loss": 0.4925,
      "num_input_tokens_seen": 14920088,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 0.07722050696611404,
      "learning_rate": 0.00041529369468846914,
      "loss": 0.4974,
      "num_input_tokens_seen": 14923768,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 0.16815932095050812,
      "learning_rate": 0.00041506832346049303,
      "loss": 0.4402,
      "num_input_tokens_seen": 14926328,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 0.11484646052122116,
      "learning_rate": 0.00041484296999925476,
      "loss": 0.3896,
      "num_input_tokens_seen": 14930488,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 0.13719019293785095,
      "learning_rate": 0.00041461763435189555,
      "loss": 0.5609,
      "num_input_tokens_seen": 14933240,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 0.13282105326652527,
      "learning_rate": 0.00041439231656555315,
      "loss": 0.3776,
      "num_input_tokens_seen": 14936568,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 0.11658226698637009,
      "learning_rate": 0.0004141670166873617,
      "loss": 0.3843,
      "num_input_tokens_seen": 14939608,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 0.12447582930326462,
      "learning_rate": 0.00041394173476445097,
      "loss": 0.4664,
      "num_input_tokens_seen": 14942584,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 0.14567427337169647,
      "learning_rate": 0.00041371647084394744,
      "loss": 0.4301,
      "num_input_tokens_seen": 14945336,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 0.20716717839241028,
      "learning_rate": 0.00041349122497297394,
      "loss": 0.3586,
      "num_input_tokens_seen": 14948120,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 0.11720949411392212,
      "learning_rate": 0.0004132659971986491,
      "loss": 0.4694,
      "num_input_tokens_seen": 14951480,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.48665136098861694,
      "eval_runtime": 13.713,
      "eval_samples_per_second": 61.839,
      "eval_steps_per_second": 15.46,
      "num_input_tokens_seen": 14951560,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 0.12290950864553452,
      "learning_rate": 0.00041304078756808837,
      "loss": 0.3937,
      "num_input_tokens_seen": 14953448,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.14423677325248718,
      "learning_rate": 0.0004128155961284027,
      "loss": 0.4956,
      "num_input_tokens_seen": 14957736,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 0.10939598828554153,
      "learning_rate": 0.0004125904229266996,
      "loss": 0.3935,
      "num_input_tokens_seen": 14960648,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 0.2382964789867401,
      "learning_rate": 0.0004123652680100828,
      "loss": 0.4173,
      "num_input_tokens_seen": 14964072,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 0.16037975251674652,
      "learning_rate": 0.0004121401314256521,
      "loss": 0.3777,
      "num_input_tokens_seen": 14966696,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 0.14390812814235687,
      "learning_rate": 0.00041191501322050363,
      "loss": 0.4445,
      "num_input_tokens_seen": 14969480,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 0.14796149730682373,
      "learning_rate": 0.0004116899134417296,
      "loss": 0.5453,
      "num_input_tokens_seen": 14972008,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.07939841598272324,
      "learning_rate": 0.00041146483213641793,
      "loss": 0.4287,
      "num_input_tokens_seen": 14975176,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 0.2054065465927124,
      "learning_rate": 0.00041123976935165355,
      "loss": 0.3576,
      "num_input_tokens_seen": 14978184,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 0.09227768331766129,
      "learning_rate": 0.0004110147251345165,
      "loss": 0.4852,
      "num_input_tokens_seen": 14981704,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 0.13415367901325226,
      "learning_rate": 0.0004107896995320839,
      "loss": 0.3489,
      "num_input_tokens_seen": 14985192,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 0.3299506604671478,
      "learning_rate": 0.0004105646925914284,
      "loss": 0.4307,
      "num_input_tokens_seen": 14987560,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 0.14965832233428955,
      "learning_rate": 0.00041033970435961904,
      "loss": 0.3625,
      "num_input_tokens_seen": 14991272,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 0.16357146203517914,
      "learning_rate": 0.00041011473488372044,
      "loss": 0.4627,
      "num_input_tokens_seen": 14994792,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 0.4164908528327942,
      "learning_rate": 0.0004098897842107939,
      "loss": 0.3698,
      "num_input_tokens_seen": 14997352,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 0.18251995742321014,
      "learning_rate": 0.00040966485238789634,
      "loss": 0.5398,
      "num_input_tokens_seen": 15000584,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 0.11791124939918518,
      "learning_rate": 0.00040943993946208126,
      "loss": 0.4758,
      "num_input_tokens_seen": 15004488,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 0.2790828049182892,
      "learning_rate": 0.00040921504548039765,
      "loss": 0.4676,
      "num_input_tokens_seen": 15007656,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 0.13048656284809113,
      "learning_rate": 0.00040899017048989073,
      "loss": 0.6438,
      "num_input_tokens_seen": 15011720,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 0.16621333360671997,
      "learning_rate": 0.00040876531453760186,
      "loss": 0.4158,
      "num_input_tokens_seen": 15014376,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 0.12492359429597855,
      "learning_rate": 0.00040854047767056824,
      "loss": 0.5513,
      "num_input_tokens_seen": 15017192,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 0.1809747815132141,
      "learning_rate": 0.00040831565993582335,
      "loss": 0.3834,
      "num_input_tokens_seen": 15020232,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 0.13034017384052277,
      "learning_rate": 0.0004080908613803964,
      "loss": 0.4055,
      "num_input_tokens_seen": 15024104,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 0.13369861245155334,
      "learning_rate": 0.0004078660820513128,
      "loss": 0.4007,
      "num_input_tokens_seen": 15028104,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 0.16435420513153076,
      "learning_rate": 0.0004076413219955937,
      "loss": 0.3415,
      "num_input_tokens_seen": 15030760,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 0.16802653670310974,
      "learning_rate": 0.00040741658126025614,
      "loss": 0.442,
      "num_input_tokens_seen": 15033992,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 0.16392016410827637,
      "learning_rate": 0.0004071918598923136,
      "loss": 0.5042,
      "num_input_tokens_seen": 15036936,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 0.16588369011878967,
      "learning_rate": 0.0004069671579387753,
      "loss": 0.5054,
      "num_input_tokens_seen": 15040104,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 0.16434524953365326,
      "learning_rate": 0.000406742475446646,
      "loss": 0.3401,
      "num_input_tokens_seen": 15042856,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 0.2084660530090332,
      "learning_rate": 0.0004065178124629272,
      "loss": 0.5643,
      "num_input_tokens_seen": 15045736,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 0.14831668138504028,
      "learning_rate": 0.0004062931690346153,
      "loss": 0.4578,
      "num_input_tokens_seen": 15048616,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 0.17457802593708038,
      "learning_rate": 0.0004060685452087034,
      "loss": 0.3702,
      "num_input_tokens_seen": 15051272,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 0.1315428465604782,
      "learning_rate": 0.00040584394103218026,
      "loss": 0.5439,
      "num_input_tokens_seen": 15054504,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 0.12697260081768036,
      "learning_rate": 0.0004056193565520304,
      "loss": 0.4545,
      "num_input_tokens_seen": 15056968,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.09942988306283951,
      "learning_rate": 0.0004053947918152344,
      "loss": 0.5651,
      "num_input_tokens_seen": 15060904,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 0.10207436978816986,
      "learning_rate": 0.0004051702468687688,
      "loss": 0.3495,
      "num_input_tokens_seen": 15064136,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 0.12528425455093384,
      "learning_rate": 0.0004049457217596055,
      "loss": 0.4362,
      "num_input_tokens_seen": 15067240,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 0.16136163473129272,
      "learning_rate": 0.0004047212165347129,
      "loss": 0.4298,
      "num_input_tokens_seen": 15070888,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 0.10924218595027924,
      "learning_rate": 0.00040449673124105464,
      "loss": 0.458,
      "num_input_tokens_seen": 15073608,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 0.12392029911279678,
      "learning_rate": 0.00040427226592559064,
      "loss": 0.4388,
      "num_input_tokens_seen": 15076744,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 0.14883805811405182,
      "learning_rate": 0.00040404782063527684,
      "loss": 0.4061,
      "num_input_tokens_seen": 15079560,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 0.14648477733135223,
      "learning_rate": 0.00040382339541706393,
      "loss": 0.4334,
      "num_input_tokens_seen": 15082216,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.12761308252811432,
      "learning_rate": 0.0004035989903178995,
      "loss": 0.5261,
      "num_input_tokens_seen": 15085800,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 0.2174454629421234,
      "learning_rate": 0.0004033746053847266,
      "loss": 0.4223,
      "num_input_tokens_seen": 15089192,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.14483371376991272,
      "learning_rate": 0.00040315024066448374,
      "loss": 0.6173,
      "num_input_tokens_seen": 15092200,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.14040595293045044,
      "learning_rate": 0.00040292589620410577,
      "loss": 0.4896,
      "num_input_tokens_seen": 15095272,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 0.12471495568752289,
      "learning_rate": 0.0004027015720505229,
      "loss": 0.4982,
      "num_input_tokens_seen": 15098088,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 0.13804839551448822,
      "learning_rate": 0.0004024772682506609,
      "loss": 0.4235,
      "num_input_tokens_seen": 15101064,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 0.10459000617265701,
      "learning_rate": 0.0004022529848514419,
      "loss": 0.4127,
      "num_input_tokens_seen": 15104488,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 0.2878374755382538,
      "learning_rate": 0.00040202872189978324,
      "loss": 0.4615,
      "num_input_tokens_seen": 15107784,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 0.1106368899345398,
      "learning_rate": 0.0004018044794425983,
      "loss": 0.4271,
      "num_input_tokens_seen": 15111912,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 0.16473770141601562,
      "learning_rate": 0.00040158025752679596,
      "loss": 0.3526,
      "num_input_tokens_seen": 15114568,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 0.11354470252990723,
      "learning_rate": 0.0004013560561992811,
      "loss": 0.4931,
      "num_input_tokens_seen": 15117384,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 0.08914033323526382,
      "learning_rate": 0.0004011318755069537,
      "loss": 0.3721,
      "num_input_tokens_seen": 15120936,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.10508333891630173,
      "learning_rate": 0.00040090771549670994,
      "loss": 0.4157,
      "num_input_tokens_seen": 15123912,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 0.1520356684923172,
      "learning_rate": 0.00040068357621544167,
      "loss": 0.3658,
      "num_input_tokens_seen": 15126504,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 0.1694401502609253,
      "learning_rate": 0.00040045945771003625,
      "loss": 0.5105,
      "num_input_tokens_seen": 15129384,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.09734438359737396,
      "learning_rate": 0.0004002353600273767,
      "loss": 0.2965,
      "num_input_tokens_seen": 15132360,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.13428835570812225,
      "learning_rate": 0.0004000112832143415,
      "loss": 0.6678,
      "num_input_tokens_seen": 15135784,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 0.16924512386322021,
      "learning_rate": 0.00039978722731780523,
      "loss": 0.4359,
      "num_input_tokens_seen": 15137896,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.21384769678115845,
      "learning_rate": 0.00039956319238463754,
      "loss": 0.4467,
      "num_input_tokens_seen": 15140168,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 0.17629040777683258,
      "learning_rate": 0.00039933917846170436,
      "loss": 0.4147,
      "num_input_tokens_seen": 15143432,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 0.20254462957382202,
      "learning_rate": 0.0003991151855958665,
      "loss": 0.3771,
      "num_input_tokens_seen": 15146440,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 0.1717827469110489,
      "learning_rate": 0.00039889121383398113,
      "loss": 0.4521,
      "num_input_tokens_seen": 15149064,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 0.21316751837730408,
      "learning_rate": 0.00039866726322290014,
      "loss": 0.3694,
      "num_input_tokens_seen": 15151336,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 0.12281869351863861,
      "learning_rate": 0.0003984433338094715,
      "loss": 0.4842,
      "num_input_tokens_seen": 15153928,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 0.12667688727378845,
      "learning_rate": 0.000398219425640539,
      "loss": 0.5913,
      "num_input_tokens_seen": 15157736,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 0.2231023609638214,
      "learning_rate": 0.0003979955387629413,
      "loss": 0.5305,
      "num_input_tokens_seen": 15160616,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 0.09242764115333557,
      "learning_rate": 0.0003977716732235133,
      "loss": 0.4989,
      "num_input_tokens_seen": 15163784,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 0.16553565859794617,
      "learning_rate": 0.000397547829069085,
      "loss": 0.3593,
      "num_input_tokens_seen": 15166888,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.2383621782064438,
      "learning_rate": 0.000397324006346482,
      "loss": 0.5148,
      "num_input_tokens_seen": 15169768,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 0.1353936642408371,
      "learning_rate": 0.0003971002051025253,
      "loss": 0.5865,
      "num_input_tokens_seen": 15172168,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 0.11248360574245453,
      "learning_rate": 0.000396876425384032,
      "loss": 0.3492,
      "num_input_tokens_seen": 15174984,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 0.14537538588047028,
      "learning_rate": 0.00039665266723781377,
      "loss": 0.413,
      "num_input_tokens_seen": 15178344,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 0.15325064957141876,
      "learning_rate": 0.00039642893071067877,
      "loss": 0.4865,
      "num_input_tokens_seen": 15182408,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 0.10424372553825378,
      "learning_rate": 0.0003962052158494298,
      "loss": 0.2909,
      "num_input_tokens_seen": 15186216,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 0.10549073666334152,
      "learning_rate": 0.00039598152270086534,
      "loss": 0.5098,
      "num_input_tokens_seen": 15189480,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 0.17236551642417908,
      "learning_rate": 0.00039575785131177975,
      "loss": 0.4165,
      "num_input_tokens_seen": 15192168,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 0.19573663175106049,
      "learning_rate": 0.0003955342017289624,
      "loss": 0.3839,
      "num_input_tokens_seen": 15195240,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 0.1272163689136505,
      "learning_rate": 0.0003953105739991982,
      "loss": 0.5791,
      "num_input_tokens_seen": 15198056,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 0.1323499083518982,
      "learning_rate": 0.0003950869681692678,
      "loss": 0.4963,
      "num_input_tokens_seen": 15201320,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 0.11227034032344818,
      "learning_rate": 0.0003948633842859465,
      "loss": 0.4726,
      "num_input_tokens_seen": 15205576,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 0.12911014258861542,
      "learning_rate": 0.00039463982239600575,
      "loss": 0.3783,
      "num_input_tokens_seen": 15208232,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 0.10932837426662445,
      "learning_rate": 0.00039441628254621215,
      "loss": 0.3845,
      "num_input_tokens_seen": 15212264,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.17344306409358978,
      "learning_rate": 0.00039419276478332773,
      "loss": 0.4115,
      "num_input_tokens_seen": 15215144,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 0.21697112917900085,
      "learning_rate": 0.0003939692691541097,
      "loss": 0.4337,
      "num_input_tokens_seen": 15217864,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 0.09572982788085938,
      "learning_rate": 0.00039374579570531114,
      "loss": 0.3972,
      "num_input_tokens_seen": 15221096,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 0.25099653005599976,
      "learning_rate": 0.0003935223444836797,
      "loss": 0.4468,
      "num_input_tokens_seen": 15224584,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 0.12177038192749023,
      "learning_rate": 0.0003932989155359591,
      "loss": 0.3527,
      "num_input_tokens_seen": 15226888,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 0.19264283776283264,
      "learning_rate": 0.000393075508908888,
      "loss": 0.4608,
      "num_input_tokens_seen": 15231016,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 0.12417925149202347,
      "learning_rate": 0.00039285212464920064,
      "loss": 0.4293,
      "num_input_tokens_seen": 15233640,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 0.14931978285312653,
      "learning_rate": 0.0003926287628036265,
      "loss": 0.3966,
      "num_input_tokens_seen": 15237256,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 0.12567099928855896,
      "learning_rate": 0.00039240542341889003,
      "loss": 0.3149,
      "num_input_tokens_seen": 15240136,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 0.13410434126853943,
      "learning_rate": 0.0003921821065417116,
      "loss": 0.5136,
      "num_input_tokens_seen": 15243592,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 0.2108437567949295,
      "learning_rate": 0.0003919588122188063,
      "loss": 0.6334,
      "num_input_tokens_seen": 15246216,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 0.11406576633453369,
      "learning_rate": 0.000391735540496885,
      "loss": 0.5423,
      "num_input_tokens_seen": 15249384,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 0.1910782903432846,
      "learning_rate": 0.00039151229142265337,
      "loss": 0.386,
      "num_input_tokens_seen": 15252648,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 0.2037835568189621,
      "learning_rate": 0.00039128906504281295,
      "loss": 0.3963,
      "num_input_tokens_seen": 15257448,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 0.16953781247138977,
      "learning_rate": 0.00039106586140405964,
      "loss": 0.5658,
      "num_input_tokens_seen": 15260168,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 0.08305374532938004,
      "learning_rate": 0.00039084268055308537,
      "loss": 0.5233,
      "num_input_tokens_seen": 15264744,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 0.13857465982437134,
      "learning_rate": 0.00039061952253657703,
      "loss": 0.2595,
      "num_input_tokens_seen": 15267656,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.167386993765831,
      "learning_rate": 0.0003903963874012166,
      "loss": 0.4194,
      "num_input_tokens_seen": 15271080,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 0.2041766345500946,
      "learning_rate": 0.0003901732751936815,
      "loss": 0.3253,
      "num_input_tokens_seen": 15274920,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 0.1576981246471405,
      "learning_rate": 0.00038995018596064447,
      "loss": 0.516,
      "num_input_tokens_seen": 15277384,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 0.07954645901918411,
      "learning_rate": 0.00038972711974877294,
      "loss": 0.4595,
      "num_input_tokens_seen": 15281352,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 0.15965665876865387,
      "learning_rate": 0.0003895040766047298,
      "loss": 0.3461,
      "num_input_tokens_seen": 15283816,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.14995381236076355,
      "learning_rate": 0.0003892810565751733,
      "loss": 0.3421,
      "num_input_tokens_seen": 15286472,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 0.08683546632528305,
      "learning_rate": 0.0003890580597067566,
      "loss": 0.3562,
      "num_input_tokens_seen": 15290088,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 0.06243761256337166,
      "learning_rate": 0.0003888350860461281,
      "loss": 0.4451,
      "num_input_tokens_seen": 15293192,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 0.11290472745895386,
      "learning_rate": 0.0003886121356399315,
      "loss": 0.3654,
      "num_input_tokens_seen": 15295976,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 0.1971772313117981,
      "learning_rate": 0.0003883892085348052,
      "loss": 0.4452,
      "num_input_tokens_seen": 15298664,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 0.11222080141305923,
      "learning_rate": 0.0003881663047773832,
      "loss": 0.5439,
      "num_input_tokens_seen": 15301864,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 0.12991026043891907,
      "learning_rate": 0.00038794342441429426,
      "loss": 0.3187,
      "num_input_tokens_seen": 15305288,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 0.21753908693790436,
      "learning_rate": 0.00038772056749216267,
      "loss": 0.531,
      "num_input_tokens_seen": 15308456,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 0.0964599996805191,
      "learning_rate": 0.00038749773405760744,
      "loss": 0.4306,
      "num_input_tokens_seen": 15311656,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 0.0829571858048439,
      "learning_rate": 0.00038727492415724265,
      "loss": 0.4054,
      "num_input_tokens_seen": 15314472,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 0.19300027191638947,
      "learning_rate": 0.00038705213783767767,
      "loss": 0.4681,
      "num_input_tokens_seen": 15317256,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 0.12903659045696259,
      "learning_rate": 0.00038682937514551686,
      "loss": 0.3279,
      "num_input_tokens_seen": 15319976,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 0.1498861461877823,
      "learning_rate": 0.0003866066361273596,
      "loss": 0.3945,
      "num_input_tokens_seen": 15323080,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 0.1307489424943924,
      "learning_rate": 0.00038638392082980056,
      "loss": 0.5033,
      "num_input_tokens_seen": 15326280,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 0.15341304242610931,
      "learning_rate": 0.0003861612292994292,
      "loss": 0.383,
      "num_input_tokens_seen": 15328872,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.12408335506916046,
      "learning_rate": 0.0003859385615828297,
      "loss": 0.4495,
      "num_input_tokens_seen": 15332008,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 0.10831256955862045,
      "learning_rate": 0.00038571591772658186,
      "loss": 0.5076,
      "num_input_tokens_seen": 15336712,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 0.1485883742570877,
      "learning_rate": 0.0003854932977772602,
      "loss": 0.5204,
      "num_input_tokens_seen": 15339368,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 0.08675266802310944,
      "learning_rate": 0.00038527070178143435,
      "loss": 0.2908,
      "num_input_tokens_seen": 15343656,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 0.15999028086662292,
      "learning_rate": 0.00038504812978566873,
      "loss": 0.4929,
      "num_input_tokens_seen": 15347368,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 0.12658759951591492,
      "learning_rate": 0.00038482558183652307,
      "loss": 0.3918,
      "num_input_tokens_seen": 15350536,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 0.2839531898498535,
      "learning_rate": 0.00038460305798055164,
      "loss": 0.4208,
      "num_input_tokens_seen": 15353448,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 0.2004721760749817,
      "learning_rate": 0.00038438055826430385,
      "loss": 0.5213,
      "num_input_tokens_seen": 15357288,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 0.11177010089159012,
      "learning_rate": 0.0003841580827343243,
      "loss": 0.3631,
      "num_input_tokens_seen": 15360200,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 0.09889119863510132,
      "learning_rate": 0.00038393563143715215,
      "loss": 0.4629,
      "num_input_tokens_seen": 15363496,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 0.20919877290725708,
      "learning_rate": 0.00038371320441932195,
      "loss": 0.7057,
      "num_input_tokens_seen": 15366120,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 0.21289271116256714,
      "learning_rate": 0.00038349080172736267,
      "loss": 0.4816,
      "num_input_tokens_seen": 15369960,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 0.15084610879421234,
      "learning_rate": 0.00038326842340779833,
      "loss": 0.3905,
      "num_input_tokens_seen": 15372904,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 0.148229718208313,
      "learning_rate": 0.00038304606950714803,
      "loss": 0.4294,
      "num_input_tokens_seen": 15375880,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.4731249511241913,
      "learning_rate": 0.0003828237400719259,
      "loss": 0.4898,
      "num_input_tokens_seen": 15379016,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 0.08787409216165543,
      "learning_rate": 0.00038260143514864037,
      "loss": 0.4491,
      "num_input_tokens_seen": 15381704,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 0.12356802076101303,
      "learning_rate": 0.0003823791547837955,
      "loss": 0.4157,
      "num_input_tokens_seen": 15384488,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 0.11761949956417084,
      "learning_rate": 0.0003821568990238894,
      "loss": 0.5008,
      "num_input_tokens_seen": 15387688,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 0.11176585406064987,
      "learning_rate": 0.0003819346679154155,
      "loss": 0.3809,
      "num_input_tokens_seen": 15391176,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 0.08717233687639236,
      "learning_rate": 0.0003817124615048623,
      "loss": 0.3425,
      "num_input_tokens_seen": 15394600,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 0.2725333273410797,
      "learning_rate": 0.00038149027983871243,
      "loss": 0.4687,
      "num_input_tokens_seen": 15397448,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 0.11408508569002151,
      "learning_rate": 0.00038126812296344415,
      "loss": 0.4004,
      "num_input_tokens_seen": 15401096,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 0.12063401937484741,
      "learning_rate": 0.0003810459909255301,
      "loss": 0.4917,
      "num_input_tokens_seen": 15403912,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 0.1369999349117279,
      "learning_rate": 0.0003808238837714374,
      "loss": 0.5309,
      "num_input_tokens_seen": 15406632,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 0.121792271733284,
      "learning_rate": 0.0003806018015476287,
      "loss": 0.4491,
      "num_input_tokens_seen": 15410888,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 0.2634393572807312,
      "learning_rate": 0.0003803797443005609,
      "loss": 0.5601,
      "num_input_tokens_seen": 15413864,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 0.09769542515277863,
      "learning_rate": 0.0003801577120766859,
      "loss": 0.3132,
      "num_input_tokens_seen": 15416936,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 0.22865056991577148,
      "learning_rate": 0.0003799357049224505,
      "loss": 0.4275,
      "num_input_tokens_seen": 15420648,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 0.10631036013364792,
      "learning_rate": 0.0003797137228842956,
      "loss": 0.3617,
      "num_input_tokens_seen": 15423240,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.15880505740642548,
      "learning_rate": 0.0003794917660086576,
      "loss": 0.4928,
      "num_input_tokens_seen": 15427240,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 0.07749488204717636,
      "learning_rate": 0.0003792698343419674,
      "loss": 0.4828,
      "num_input_tokens_seen": 15430696,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 0.10534035414457321,
      "learning_rate": 0.0003790479279306505,
      "loss": 0.4679,
      "num_input_tokens_seen": 15434696,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 0.133040189743042,
      "learning_rate": 0.0003788260468211271,
      "loss": 0.5475,
      "num_input_tokens_seen": 15438824,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 0.14559166133403778,
      "learning_rate": 0.0003786041910598125,
      "loss": 0.4566,
      "num_input_tokens_seen": 15441736,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 0.20324933528900146,
      "learning_rate": 0.0003783823606931159,
      "loss": 0.5884,
      "num_input_tokens_seen": 15444200,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 0.16929784417152405,
      "learning_rate": 0.0003781605557674421,
      "loss": 0.4351,
      "num_input_tokens_seen": 15447432,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 0.09912587702274323,
      "learning_rate": 0.0003779387763291899,
      "loss": 0.3531,
      "num_input_tokens_seen": 15451240,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 0.15340213477611542,
      "learning_rate": 0.0003777170224247533,
      "loss": 0.5841,
      "num_input_tokens_seen": 15454024,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 0.13746820390224457,
      "learning_rate": 0.0003774952941005204,
      "loss": 0.399,
      "num_input_tokens_seen": 15457160,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 0.09712470322847366,
      "learning_rate": 0.00037727359140287455,
      "loss": 0.5963,
      "num_input_tokens_seen": 15460840,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 0.13902761042118073,
      "learning_rate": 0.00037705191437819316,
      "loss": 0.3715,
      "num_input_tokens_seen": 15463752,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.14176276326179504,
      "learning_rate": 0.00037683026307284853,
      "loss": 0.3637,
      "num_input_tokens_seen": 15466408,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 0.1141437515616417,
      "learning_rate": 0.0003766086375332077,
      "loss": 0.4374,
      "num_input_tokens_seen": 15469864,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 0.13401439785957336,
      "learning_rate": 0.0003763870378056321,
      "loss": 0.4101,
      "num_input_tokens_seen": 15474504,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 0.11861176043748856,
      "learning_rate": 0.00037616546393647824,
      "loss": 0.3931,
      "num_input_tokens_seen": 15478408,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 0.1045609712600708,
      "learning_rate": 0.0003759439159720962,
      "loss": 0.5243,
      "num_input_tokens_seen": 15481640,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 0.25294631719589233,
      "learning_rate": 0.0003757223939588318,
      "loss": 0.4379,
      "num_input_tokens_seen": 15484648,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 0.16838707029819489,
      "learning_rate": 0.00037550089794302464,
      "loss": 0.4271,
      "num_input_tokens_seen": 15487560,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 0.0972713828086853,
      "learning_rate": 0.0003752794279710094,
      "loss": 0.4045,
      "num_input_tokens_seen": 15491784,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 0.08908140659332275,
      "learning_rate": 0.0003750579840891148,
      "loss": 0.4926,
      "num_input_tokens_seen": 15495080,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 0.21701037883758545,
      "learning_rate": 0.0003748365663436647,
      "loss": 0.4608,
      "num_input_tokens_seen": 15498472,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 0.11678215861320496,
      "learning_rate": 0.0003746151747809769,
      "loss": 0.4011,
      "num_input_tokens_seen": 15501032,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 0.14909520745277405,
      "learning_rate": 0.000374393809447364,
      "loss": 0.4272,
      "num_input_tokens_seen": 15503944,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 0.4047136604785919,
      "learning_rate": 0.0003741724703891333,
      "loss": 0.427,
      "num_input_tokens_seen": 15507272,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 0.11916977167129517,
      "learning_rate": 0.00037395115765258616,
      "loss": 0.4214,
      "num_input_tokens_seen": 15510248,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 0.14327140152454376,
      "learning_rate": 0.0003737298712840188,
      "loss": 0.3605,
      "num_input_tokens_seen": 15512968,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 0.12063303589820862,
      "learning_rate": 0.000373508611329722,
      "loss": 0.3302,
      "num_input_tokens_seen": 15517160,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 0.10639140009880066,
      "learning_rate": 0.00037328737783598036,
      "loss": 0.4027,
      "num_input_tokens_seen": 15520520,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 0.2848038375377655,
      "learning_rate": 0.0003730661708490738,
      "loss": 0.4036,
      "num_input_tokens_seen": 15524488,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 0.2641674876213074,
      "learning_rate": 0.0003728449904152761,
      "loss": 0.4045,
      "num_input_tokens_seen": 15527400,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 0.1184362843632698,
      "learning_rate": 0.00037262383658085563,
      "loss": 0.4954,
      "num_input_tokens_seen": 15530472,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 0.2305123507976532,
      "learning_rate": 0.00037240270939207555,
      "loss": 0.2726,
      "num_input_tokens_seen": 15535912,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 0.19837556779384613,
      "learning_rate": 0.000372181608895193,
      "loss": 0.3398,
      "num_input_tokens_seen": 15539336,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 0.337339848279953,
      "learning_rate": 0.00037196053513645957,
      "loss": 0.3063,
      "num_input_tokens_seen": 15542472,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 0.10039318352937698,
      "learning_rate": 0.00037173948816212146,
      "loss": 0.3766,
      "num_input_tokens_seen": 15545640,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 0.1321684569120407,
      "learning_rate": 0.00037151846801841904,
      "loss": 0.5186,
      "num_input_tokens_seen": 15550024,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 0.18519118428230286,
      "learning_rate": 0.00037129747475158736,
      "loss": 0.5506,
      "num_input_tokens_seen": 15552904,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 0.37006810307502747,
      "learning_rate": 0.0003710765084078558,
      "loss": 0.3315,
      "num_input_tokens_seen": 15555528,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 0.09296007454395294,
      "learning_rate": 0.0003708555690334477,
      "loss": 0.4432,
      "num_input_tokens_seen": 15559240,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.20368339121341705,
      "learning_rate": 0.00037063465667458125,
      "loss": 0.5439,
      "num_input_tokens_seen": 15562184,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 0.3550611138343811,
      "learning_rate": 0.0003704137713774686,
      "loss": 0.4724,
      "num_input_tokens_seen": 15565384,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 0.08654165267944336,
      "learning_rate": 0.0003701929131883167,
      "loss": 0.5023,
      "num_input_tokens_seen": 15569960,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 0.37417611479759216,
      "learning_rate": 0.0003699720821533264,
      "loss": 0.5079,
      "num_input_tokens_seen": 15575272,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 0.13261952996253967,
      "learning_rate": 0.00036975127831869326,
      "loss": 0.4735,
      "num_input_tokens_seen": 15578504,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.32919782400131226,
      "learning_rate": 0.0003695305017306066,
      "loss": 0.424,
      "num_input_tokens_seen": 15581224,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 0.1305406391620636,
      "learning_rate": 0.00036930975243525046,
      "loss": 0.4434,
      "num_input_tokens_seen": 15584744,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 0.12987542152404785,
      "learning_rate": 0.00036908903047880304,
      "loss": 0.3904,
      "num_input_tokens_seen": 15587624,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.13439176976680756,
      "learning_rate": 0.00036886833590743707,
      "loss": 0.4795,
      "num_input_tokens_seen": 15590792,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 0.19111232459545135,
      "learning_rate": 0.00036864766876731913,
      "loss": 0.4364,
      "num_input_tokens_seen": 15593320,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 0.12148039788007736,
      "learning_rate": 0.00036842702910461054,
      "loss": 0.4961,
      "num_input_tokens_seen": 15597192,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 0.2117842733860016,
      "learning_rate": 0.0003682064169654663,
      "loss": 0.4201,
      "num_input_tokens_seen": 15601288,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 0.2763199806213379,
      "learning_rate": 0.00036798583239603587,
      "loss": 0.4508,
      "num_input_tokens_seen": 15604872,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.17410725355148315,
      "learning_rate": 0.0003677652754424634,
      "loss": 0.5361,
      "num_input_tokens_seen": 15607880,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.1305951327085495,
      "learning_rate": 0.0003675447461508865,
      "loss": 0.4259,
      "num_input_tokens_seen": 15610408,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 0.17562425136566162,
      "learning_rate": 0.00036732424456743784,
      "loss": 0.4493,
      "num_input_tokens_seen": 15613416,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 0.09772121161222458,
      "learning_rate": 0.0003671037707382435,
      "loss": 0.4746,
      "num_input_tokens_seen": 15617288,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 0.13464853167533875,
      "learning_rate": 0.000366883324709424,
      "loss": 0.4014,
      "num_input_tokens_seen": 15620008,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 0.23478716611862183,
      "learning_rate": 0.00036666290652709446,
      "loss": 0.5529,
      "num_input_tokens_seen": 15623272,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 0.5255036354064941,
      "learning_rate": 0.0003664425162373635,
      "loss": 0.4215,
      "num_input_tokens_seen": 15625224,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 0.16708169877529144,
      "learning_rate": 0.0003662221538863346,
      "loss": 0.5847,
      "num_input_tokens_seen": 15628072,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 0.09825912863016129,
      "learning_rate": 0.0003660018195201049,
      "loss": 0.4951,
      "num_input_tokens_seen": 15631304,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 0.12982460856437683,
      "learning_rate": 0.0003657815131847657,
      "loss": 0.4268,
      "num_input_tokens_seen": 15634152,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 0.1640521138906479,
      "learning_rate": 0.0003655612349264027,
      "loss": 0.6297,
      "num_input_tokens_seen": 15637160,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.1991739273071289,
      "learning_rate": 0.0003653409847910957,
      "loss": 0.3665,
      "num_input_tokens_seen": 15639944,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 0.14418871700763702,
      "learning_rate": 0.0003651207628249182,
      "loss": 0.5287,
      "num_input_tokens_seen": 15647592,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 0.14264142513275146,
      "learning_rate": 0.0003649005690739386,
      "loss": 0.4001,
      "num_input_tokens_seen": 15650344,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 0.18377667665481567,
      "learning_rate": 0.0003646804035842187,
      "loss": 0.3852,
      "num_input_tokens_seen": 15653672,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 0.22818416357040405,
      "learning_rate": 0.0003644602664018143,
      "loss": 0.6509,
      "num_input_tokens_seen": 15656008,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 0.1535719782114029,
      "learning_rate": 0.000364240157572776,
      "loss": 0.3971,
      "num_input_tokens_seen": 15659656,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.18388718366622925,
      "learning_rate": 0.0003640200771431478,
      "loss": 0.3362,
      "num_input_tokens_seen": 15661736,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 0.1017843633890152,
      "learning_rate": 0.0003638000251589683,
      "loss": 0.3233,
      "num_input_tokens_seen": 15665288,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 0.169859379529953,
      "learning_rate": 0.0003635800016662696,
      "loss": 0.425,
      "num_input_tokens_seen": 15667912,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 0.20675161480903625,
      "learning_rate": 0.00036336000671107816,
      "loss": 0.5754,
      "num_input_tokens_seen": 15671176,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 0.1808212697505951,
      "learning_rate": 0.00036314004033941445,
      "loss": 0.4249,
      "num_input_tokens_seen": 15673864,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.12945283949375153,
      "learning_rate": 0.00036292010259729283,
      "loss": 0.3581,
      "num_input_tokens_seen": 15676744,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 0.10133857280015945,
      "learning_rate": 0.000362700193530722,
      "loss": 0.4418,
      "num_input_tokens_seen": 15680168,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 0.09676481038331985,
      "learning_rate": 0.0003624803131857042,
      "loss": 0.3254,
      "num_input_tokens_seen": 15683336,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 0.13444049656391144,
      "learning_rate": 0.0003622604616082361,
      "loss": 0.6099,
      "num_input_tokens_seen": 15686216,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 0.08643554151058197,
      "learning_rate": 0.0003620406388443078,
      "loss": 0.5125,
      "num_input_tokens_seen": 15689544,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 0.1931578814983368,
      "learning_rate": 0.00036182084493990407,
      "loss": 0.3858,
      "num_input_tokens_seen": 15692424,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 0.11732988059520721,
      "learning_rate": 0.0003616010799410031,
      "loss": 0.4599,
      "num_input_tokens_seen": 15695368,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 0.11490893363952637,
      "learning_rate": 0.0003613813438935773,
      "loss": 0.5699,
      "num_input_tokens_seen": 15698696,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 0.18497011065483093,
      "learning_rate": 0.0003611616368435928,
      "loss": 0.2837,
      "num_input_tokens_seen": 15701640,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 0.1663033813238144,
      "learning_rate": 0.0003609419588370102,
      "loss": 0.5585,
      "num_input_tokens_seen": 15704776,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.19863922894001007,
      "learning_rate": 0.00036072230991978326,
      "loss": 0.4135,
      "num_input_tokens_seen": 15707496,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 0.22860437631607056,
      "learning_rate": 0.00036050269013785996,
      "loss": 0.4067,
      "num_input_tokens_seen": 15710440,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 0.07869651168584824,
      "learning_rate": 0.0003602830995371825,
      "loss": 0.4864,
      "num_input_tokens_seen": 15715336,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 0.14495941996574402,
      "learning_rate": 0.0003600635381636866,
      "loss": 0.4508,
      "num_input_tokens_seen": 15718216,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.16013412177562714,
      "learning_rate": 0.0003598440060633022,
      "loss": 0.3701,
      "num_input_tokens_seen": 15722280,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 0.24317212402820587,
      "learning_rate": 0.00035962450328195264,
      "loss": 0.3867,
      "num_input_tokens_seen": 15725608,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 0.15014897286891937,
      "learning_rate": 0.00035940502986555543,
      "loss": 0.5318,
      "num_input_tokens_seen": 15728072,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 0.16425682604312897,
      "learning_rate": 0.00035918558586002205,
      "loss": 0.4282,
      "num_input_tokens_seen": 15731048,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 0.16442929208278656,
      "learning_rate": 0.0003589661713112575,
      "loss": 0.31,
      "num_input_tokens_seen": 15734184,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.15091265738010406,
      "learning_rate": 0.000358746786265161,
      "loss": 0.5139,
      "num_input_tokens_seen": 15737896,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 0.16170759499073029,
      "learning_rate": 0.0003585274307676254,
      "loss": 0.4505,
      "num_input_tokens_seen": 15741800,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.13030286133289337,
      "learning_rate": 0.00035830810486453725,
      "loss": 0.4294,
      "num_input_tokens_seen": 15744808,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.11862701922655106,
      "learning_rate": 0.00035808880860177696,
      "loss": 0.4496,
      "num_input_tokens_seen": 15747976,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 0.1110406443476677,
      "learning_rate": 0.00035786954202521895,
      "loss": 0.3529,
      "num_input_tokens_seen": 15751336,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 0.15400120615959167,
      "learning_rate": 0.00035765030518073116,
      "loss": 0.4358,
      "num_input_tokens_seen": 15753896,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 0.13611112534999847,
      "learning_rate": 0.0003574310981141756,
      "loss": 0.4763,
      "num_input_tokens_seen": 15757192,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 0.13878610730171204,
      "learning_rate": 0.00035721192087140796,
      "loss": 0.4765,
      "num_input_tokens_seen": 15760296,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 0.11015725880861282,
      "learning_rate": 0.00035699277349827724,
      "loss": 0.5255,
      "num_input_tokens_seen": 15763176,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 0.15005232393741608,
      "learning_rate": 0.0003567736560406269,
      "loss": 0.5249,
      "num_input_tokens_seen": 15766600,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 0.11497718840837479,
      "learning_rate": 0.0003565545685442936,
      "loss": 0.4758,
      "num_input_tokens_seen": 15769640,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 0.15476736426353455,
      "learning_rate": 0.00035633551105510806,
      "loss": 0.3555,
      "num_input_tokens_seen": 15772296,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 0.22290123999118805,
      "learning_rate": 0.00035611648361889457,
      "loss": 0.3915,
      "num_input_tokens_seen": 15774952,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 0.21411152184009552,
      "learning_rate": 0.00035589748628147145,
      "loss": 0.4639,
      "num_input_tokens_seen": 15777800,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 0.11815599352121353,
      "learning_rate": 0.00035567851908865,
      "loss": 0.2571,
      "num_input_tokens_seen": 15780424,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 0.11518323421478271,
      "learning_rate": 0.0003554595820862358,
      "loss": 0.4413,
      "num_input_tokens_seen": 15783304,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 0.21124477684497833,
      "learning_rate": 0.00035524067532002803,
      "loss": 0.4019,
      "num_input_tokens_seen": 15785928,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 0.17497166991233826,
      "learning_rate": 0.00035502179883581956,
      "loss": 0.3855,
      "num_input_tokens_seen": 15788712,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 0.1418498158454895,
      "learning_rate": 0.0003548029526793969,
      "loss": 0.3641,
      "num_input_tokens_seen": 15792296,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 0.15983617305755615,
      "learning_rate": 0.0003545841368965398,
      "loss": 0.3962,
      "num_input_tokens_seen": 15795272,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.21710337698459625,
      "learning_rate": 0.00035436535153302235,
      "loss": 0.4551,
      "num_input_tokens_seen": 15798248,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 0.15111088752746582,
      "learning_rate": 0.0003541465966346118,
      "loss": 0.4094,
      "num_input_tokens_seen": 15802760,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 0.14987999200820923,
      "learning_rate": 0.00035392787224706936,
      "loss": 0.4692,
      "num_input_tokens_seen": 15805608,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 0.1075805053114891,
      "learning_rate": 0.0003537091784161495,
      "loss": 0.4225,
      "num_input_tokens_seen": 15808264,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 0.176035076379776,
      "learning_rate": 0.0003534905151876007,
      "loss": 0.3983,
      "num_input_tokens_seen": 15811240,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 0.255801796913147,
      "learning_rate": 0.0003532718826071646,
      "loss": 0.367,
      "num_input_tokens_seen": 15814504,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 0.11561009287834167,
      "learning_rate": 0.00035305328072057654,
      "loss": 0.3509,
      "num_input_tokens_seen": 15818056,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 0.12588539719581604,
      "learning_rate": 0.00035283470957356576,
      "loss": 0.4641,
      "num_input_tokens_seen": 15822024,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 0.1297672986984253,
      "learning_rate": 0.0003526161692118548,
      "loss": 0.4916,
      "num_input_tokens_seen": 15825768,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 0.2798996865749359,
      "learning_rate": 0.0003523976596811597,
      "loss": 0.4372,
      "num_input_tokens_seen": 15829576,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.07350795716047287,
      "learning_rate": 0.0003521791810271906,
      "loss": 0.272,
      "num_input_tokens_seen": 15832744,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 0.17833366990089417,
      "learning_rate": 0.0003519607332956501,
      "loss": 0.5343,
      "num_input_tokens_seen": 15836328,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 0.151450514793396,
      "learning_rate": 0.0003517423165322353,
      "loss": 0.3503,
      "num_input_tokens_seen": 15839816,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 0.20526538789272308,
      "learning_rate": 0.0003515239307826366,
      "loss": 0.5683,
      "num_input_tokens_seen": 15843528,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 0.16460974514484406,
      "learning_rate": 0.00035130557609253744,
      "loss": 0.4605,
      "num_input_tokens_seen": 15847272,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 0.14055147767066956,
      "learning_rate": 0.00035108725250761566,
      "loss": 0.3439,
      "num_input_tokens_seen": 15850600,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 0.14492888748645782,
      "learning_rate": 0.0003508689600735416,
      "loss": 0.4745,
      "num_input_tokens_seen": 15853416,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 0.09410499036312103,
      "learning_rate": 0.0003506506988359797,
      "loss": 0.375,
      "num_input_tokens_seen": 15856648,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 0.09032679349184036,
      "learning_rate": 0.0003504324688405878,
      "loss": 0.3837,
      "num_input_tokens_seen": 15860232,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 0.13268545269966125,
      "learning_rate": 0.000350214270133017,
      "loss": 0.4454,
      "num_input_tokens_seen": 15865032,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 0.14695844054222107,
      "learning_rate": 0.00034999610275891204,
      "loss": 0.3995,
      "num_input_tokens_seen": 15868008,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 0.12454329431056976,
      "learning_rate": 0.0003497779667639113,
      "loss": 0.4502,
      "num_input_tokens_seen": 15870824,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 0.10929283499717712,
      "learning_rate": 0.0003495598621936458,
      "loss": 0.4802,
      "num_input_tokens_seen": 15874184,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 0.1628391444683075,
      "learning_rate": 0.00034934178909374084,
      "loss": 0.4618,
      "num_input_tokens_seen": 15877032,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 0.19822989404201508,
      "learning_rate": 0.0003491237475098148,
      "loss": 0.5044,
      "num_input_tokens_seen": 15881160,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.12807273864746094,
      "learning_rate": 0.00034890573748747945,
      "loss": 0.4781,
      "num_input_tokens_seen": 15885768,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 0.19862230122089386,
      "learning_rate": 0.00034868775907234017,
      "loss": 0.3829,
      "num_input_tokens_seen": 15888872,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 0.11818253248929977,
      "learning_rate": 0.0003484698123099956,
      "loss": 0.5179,
      "num_input_tokens_seen": 15892936,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 0.18030017614364624,
      "learning_rate": 0.00034825189724603723,
      "loss": 0.4909,
      "num_input_tokens_seen": 15896328,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 0.1751364916563034,
      "learning_rate": 0.0003480340139260509,
      "loss": 0.4607,
      "num_input_tokens_seen": 15898856,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 0.20364804565906525,
      "learning_rate": 0.000347816162395615,
      "loss": 0.4291,
      "num_input_tokens_seen": 15902632,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 0.15923795104026794,
      "learning_rate": 0.00034759834270030185,
      "loss": 0.4132,
      "num_input_tokens_seen": 15905448,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.2105800062417984,
      "learning_rate": 0.0003473805548856768,
      "loss": 0.3579,
      "num_input_tokens_seen": 15912168,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 0.1334500014781952,
      "learning_rate": 0.00034716279899729826,
      "loss": 0.486,
      "num_input_tokens_seen": 15915880,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 0.12234565615653992,
      "learning_rate": 0.00034694507508071864,
      "loss": 0.4769,
      "num_input_tokens_seen": 15919432,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 0.24737216532230377,
      "learning_rate": 0.00034672738318148303,
      "loss": 0.3687,
      "num_input_tokens_seen": 15923144,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.20480699837207794,
      "learning_rate": 0.00034650972334513043,
      "loss": 0.4032,
      "num_input_tokens_seen": 15925320,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 0.20773926377296448,
      "learning_rate": 0.00034629209561719244,
      "loss": 0.5678,
      "num_input_tokens_seen": 15928488,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 0.23621487617492676,
      "learning_rate": 0.0003460745000431946,
      "loss": 0.318,
      "num_input_tokens_seen": 15931816,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 0.11618348956108093,
      "learning_rate": 0.00034585693666865525,
      "loss": 0.3812,
      "num_input_tokens_seen": 15934856,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 0.16266097128391266,
      "learning_rate": 0.00034563940553908605,
      "loss": 0.4252,
      "num_input_tokens_seen": 15938024,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 0.2510982155799866,
      "learning_rate": 0.00034542190669999216,
      "loss": 0.5131,
      "num_input_tokens_seen": 15941224,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.1499318778514862,
      "learning_rate": 0.000345204440196872,
      "loss": 0.5,
      "num_input_tokens_seen": 15944424,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 0.20437026023864746,
      "learning_rate": 0.00034498700607521693,
      "loss": 0.3668,
      "num_input_tokens_seen": 15946568,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 0.14548547565937042,
      "learning_rate": 0.00034476960438051187,
      "loss": 0.3718,
      "num_input_tokens_seen": 15950408,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 0.16686101257801056,
      "learning_rate": 0.00034455223515823444,
      "loss": 0.4119,
      "num_input_tokens_seen": 15953384,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 0.12728454172611237,
      "learning_rate": 0.0003443348984538559,
      "loss": 0.3829,
      "num_input_tokens_seen": 15957032,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 0.07660499960184097,
      "learning_rate": 0.0003441175943128407,
      "loss": 0.4145,
      "num_input_tokens_seen": 15961480,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 0.15766234695911407,
      "learning_rate": 0.0003439003227806464,
      "loss": 0.5607,
      "num_input_tokens_seen": 15965000,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 0.16992440819740295,
      "learning_rate": 0.00034368308390272386,
      "loss": 0.4223,
      "num_input_tokens_seen": 15967624,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 0.14666178822517395,
      "learning_rate": 0.0003434658777245167,
      "loss": 0.5209,
      "num_input_tokens_seen": 15970056,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 0.27174806594848633,
      "learning_rate": 0.000343248704291462,
      "loss": 0.4566,
      "num_input_tokens_seen": 15972680,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.12644340097904205,
      "learning_rate": 0.00034303156364899016,
      "loss": 0.5034,
      "num_input_tokens_seen": 15975592,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 0.2328181266784668,
      "learning_rate": 0.00034281445584252425,
      "loss": 0.4114,
      "num_input_tokens_seen": 15978664,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 0.20510150492191315,
      "learning_rate": 0.00034259738091748103,
      "loss": 0.4316,
      "num_input_tokens_seen": 15983048,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 0.07869666814804077,
      "learning_rate": 0.00034238033891927013,
      "loss": 0.4461,
      "num_input_tokens_seen": 15986344,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 0.18747058510780334,
      "learning_rate": 0.00034216332989329387,
      "loss": 0.6127,
      "num_input_tokens_seen": 15989864,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 0.17444279789924622,
      "learning_rate": 0.00034194635388494835,
      "loss": 0.458,
      "num_input_tokens_seen": 15993224,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 0.08943110704421997,
      "learning_rate": 0.00034172941093962237,
      "loss": 0.4417,
      "num_input_tokens_seen": 15996296,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 0.0679711401462555,
      "learning_rate": 0.00034151250110269805,
      "loss": 0.4369,
      "num_input_tokens_seen": 15999944,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 0.15725304186344147,
      "learning_rate": 0.00034129562441955044,
      "loss": 0.3849,
      "num_input_tokens_seen": 16003112,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.19359587132930756,
      "learning_rate": 0.00034107878093554765,
      "loss": 0.503,
      "num_input_tokens_seen": 16005832,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 0.14291587471961975,
      "learning_rate": 0.00034086197069605064,
      "loss": 0.4698,
      "num_input_tokens_seen": 16009096,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 0.20303936302661896,
      "learning_rate": 0.0003406451937464139,
      "loss": 0.4131,
      "num_input_tokens_seen": 16011592,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.08586468547582626,
      "learning_rate": 0.0003404284501319845,
      "loss": 0.5143,
      "num_input_tokens_seen": 16015784,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 0.16179026663303375,
      "learning_rate": 0.00034021173989810306,
      "loss": 0.552,
      "num_input_tokens_seen": 16019208,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 0.11024821549654007,
      "learning_rate": 0.0003399950630901025,
      "loss": 0.4809,
      "num_input_tokens_seen": 16022856,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 0.11883845925331116,
      "learning_rate": 0.0003397784197533095,
      "loss": 0.4736,
      "num_input_tokens_seen": 16025480,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 0.20727834105491638,
      "learning_rate": 0.0003395618099330431,
      "loss": 0.4912,
      "num_input_tokens_seen": 16028072,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 0.1322183609008789,
      "learning_rate": 0.0003393452336746156,
      "loss": 0.5371,
      "num_input_tokens_seen": 16031592,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 0.18513406813144684,
      "learning_rate": 0.00033912869102333237,
      "loss": 0.3815,
      "num_input_tokens_seen": 16035048,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.10563770681619644,
      "learning_rate": 0.00033891218202449164,
      "loss": 0.5073,
      "num_input_tokens_seen": 16039720,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 0.10618118196725845,
      "learning_rate": 0.00033869570672338486,
      "loss": 0.391,
      "num_input_tokens_seen": 16043016,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 0.14926785230636597,
      "learning_rate": 0.00033847926516529574,
      "loss": 0.4342,
      "num_input_tokens_seen": 16045896,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 0.15739049017429352,
      "learning_rate": 0.0003382628573955018,
      "loss": 0.4211,
      "num_input_tokens_seen": 16048552,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 0.13773146271705627,
      "learning_rate": 0.00033804648345927267,
      "loss": 0.5317,
      "num_input_tokens_seen": 16051752,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 0.16476017236709595,
      "learning_rate": 0.00033783014340187166,
      "loss": 0.416,
      "num_input_tokens_seen": 16054760,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 0.12605588138103485,
      "learning_rate": 0.0003376138372685545,
      "loss": 0.3307,
      "num_input_tokens_seen": 16057672,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 0.15125305950641632,
      "learning_rate": 0.00033739756510457015,
      "loss": 0.4199,
      "num_input_tokens_seen": 16060744,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 0.29392996430397034,
      "learning_rate": 0.00033718132695516003,
      "loss": 0.4709,
      "num_input_tokens_seen": 16063720,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 0.07872213423252106,
      "learning_rate": 0.00033696512286555865,
      "loss": 0.4297,
      "num_input_tokens_seen": 16069448,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 0.20956820249557495,
      "learning_rate": 0.0003367489528809936,
      "loss": 0.2168,
      "num_input_tokens_seen": 16074280,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 0.1249823272228241,
      "learning_rate": 0.00033653281704668515,
      "loss": 0.4193,
      "num_input_tokens_seen": 16077896,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 0.1019059494137764,
      "learning_rate": 0.0003363167154078465,
      "loss": 0.5499,
      "num_input_tokens_seen": 16080680,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 0.1739431917667389,
      "learning_rate": 0.0003361006480096837,
      "loss": 0.5363,
      "num_input_tokens_seen": 16083656,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 0.1975255012512207,
      "learning_rate": 0.0003358846148973953,
      "loss": 0.3585,
      "num_input_tokens_seen": 16087144,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 0.09858639538288116,
      "learning_rate": 0.0003356686161161732,
      "loss": 0.5328,
      "num_input_tokens_seen": 16091400,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 0.16442914307117462,
      "learning_rate": 0.00033545265171120163,
      "loss": 0.3496,
      "num_input_tokens_seen": 16093864,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 0.3296131491661072,
      "learning_rate": 0.0003352367217276582,
      "loss": 0.3885,
      "num_input_tokens_seen": 16096520,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 0.18685203790664673,
      "learning_rate": 0.0003350208262107131,
      "loss": 0.4159,
      "num_input_tokens_seen": 16099688,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 0.16561810672283173,
      "learning_rate": 0.00033480496520552874,
      "loss": 0.5062,
      "num_input_tokens_seen": 16103528,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 0.1927213817834854,
      "learning_rate": 0.00033458913875726093,
      "loss": 0.6932,
      "num_input_tokens_seen": 16107080,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.12937644124031067,
      "learning_rate": 0.00033437334691105837,
      "loss": 0.4403,
      "num_input_tokens_seen": 16110120,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 0.10513673722743988,
      "learning_rate": 0.0003341575897120619,
      "loss": 0.3986,
      "num_input_tokens_seen": 16113640,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 0.1313045769929886,
      "learning_rate": 0.00033394186720540577,
      "loss": 0.4241,
      "num_input_tokens_seen": 16116520,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 0.09051655977964401,
      "learning_rate": 0.00033372617943621665,
      "loss": 0.4602,
      "num_input_tokens_seen": 16119208,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 0.15762783586978912,
      "learning_rate": 0.00033351052644961353,
      "loss": 0.5048,
      "num_input_tokens_seen": 16121960,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 0.07560233771800995,
      "learning_rate": 0.00033329490829070907,
      "loss": 0.3496,
      "num_input_tokens_seen": 16125960,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 0.15263602137565613,
      "learning_rate": 0.0003330793250046078,
      "loss": 0.4144,
      "num_input_tokens_seen": 16128744,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 0.1481407731771469,
      "learning_rate": 0.0003328637766364075,
      "loss": 0.6096,
      "num_input_tokens_seen": 16131944,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 0.1531498283147812,
      "learning_rate": 0.00033264826323119824,
      "loss": 0.5553,
      "num_input_tokens_seen": 16135624,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 0.19474299252033234,
      "learning_rate": 0.00033243278483406326,
      "loss": 0.3843,
      "num_input_tokens_seen": 16139304,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 0.15382270514965057,
      "learning_rate": 0.00033221734149007785,
      "loss": 0.4211,
      "num_input_tokens_seen": 16142312,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 0.10413149744272232,
      "learning_rate": 0.0003320019332443104,
      "loss": 0.4737,
      "num_input_tokens_seen": 16146376,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 0.12987864017486572,
      "learning_rate": 0.00033178656014182185,
      "loss": 0.3968,
      "num_input_tokens_seen": 16149384,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 0.1446792632341385,
      "learning_rate": 0.00033157122222766597,
      "loss": 0.3665,
      "num_input_tokens_seen": 16153768,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 0.09971868246793747,
      "learning_rate": 0.0003313559195468889,
      "loss": 0.3444,
      "num_input_tokens_seen": 16158536,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 0.1635800153017044,
      "learning_rate": 0.0003311406521445293,
      "loss": 0.4374,
      "num_input_tokens_seen": 16162312,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.16831274330615997,
      "learning_rate": 0.000330925420065619,
      "loss": 0.5132,
      "num_input_tokens_seen": 16165064,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 0.31987228989601135,
      "learning_rate": 0.0003307102233551817,
      "loss": 0.4186,
      "num_input_tokens_seen": 16167304,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 0.15234187245368958,
      "learning_rate": 0.0003304950620582344,
      "loss": 0.4842,
      "num_input_tokens_seen": 16171016,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 0.15800565481185913,
      "learning_rate": 0.0003302799362197864,
      "loss": 0.4199,
      "num_input_tokens_seen": 16174824,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 0.09298521280288696,
      "learning_rate": 0.00033006484588483947,
      "loss": 0.5065,
      "num_input_tokens_seen": 16179880,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 0.2087797224521637,
      "learning_rate": 0.00032984979109838803,
      "loss": 0.5328,
      "num_input_tokens_seen": 16182856,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 0.10598179697990417,
      "learning_rate": 0.000329634771905419,
      "loss": 0.5173,
      "num_input_tokens_seen": 16185672,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 0.11872106790542603,
      "learning_rate": 0.0003294197883509122,
      "loss": 0.3515,
      "num_input_tokens_seen": 16188392,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 0.13637390732765198,
      "learning_rate": 0.00032920484047983947,
      "loss": 0.471,
      "num_input_tokens_seen": 16191560,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.48564380407333374,
      "eval_runtime": 13.6776,
      "eval_samples_per_second": 61.999,
      "eval_steps_per_second": 15.5,
      "num_input_tokens_seen": 16193456,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.07902836799621582,
      "learning_rate": 0.0003289899283371657,
      "loss": 0.3282,
      "num_input_tokens_seen": 16194544,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 0.07736322283744812,
      "learning_rate": 0.00032877505196784795,
      "loss": 0.4105,
      "num_input_tokens_seen": 16198000,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.12862908840179443,
      "learning_rate": 0.0003285602114168358,
      "loss": 0.3907,
      "num_input_tokens_seen": 16200816,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.08328378200531006,
      "learning_rate": 0.00032834540672907144,
      "loss": 0.4899,
      "num_input_tokens_seen": 16206096,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 0.14412735402584076,
      "learning_rate": 0.00032813063794948983,
      "loss": 0.4096,
      "num_input_tokens_seen": 16209552,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 0.10211469233036041,
      "learning_rate": 0.00032791590512301784,
      "loss": 0.4326,
      "num_input_tokens_seen": 16211984,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 0.10399968922138214,
      "learning_rate": 0.00032770120829457534,
      "loss": 0.4477,
      "num_input_tokens_seen": 16215920,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.14546921849250793,
      "learning_rate": 0.00032748654750907447,
      "loss": 0.4736,
      "num_input_tokens_seen": 16218864,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 0.18565411865711212,
      "learning_rate": 0.0003272719228114196,
      "loss": 0.3878,
      "num_input_tokens_seen": 16225936,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.17991934716701508,
      "learning_rate": 0.0003270573342465079,
      "loss": 0.5126,
      "num_input_tokens_seen": 16228688,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 0.09378987550735474,
      "learning_rate": 0.00032684278185922877,
      "loss": 0.4544,
      "num_input_tokens_seen": 16231504,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 0.13753606379032135,
      "learning_rate": 0.0003266282656944643,
      "loss": 0.3571,
      "num_input_tokens_seen": 16234736,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 0.13851051032543182,
      "learning_rate": 0.0003264137857970888,
      "loss": 0.4022,
      "num_input_tokens_seen": 16239152,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.15994340181350708,
      "learning_rate": 0.0003261993422119688,
      "loss": 0.4134,
      "num_input_tokens_seen": 16242352,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 0.10349547863006592,
      "learning_rate": 0.00032598493498396365,
      "loss": 0.367,
      "num_input_tokens_seen": 16246416,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 0.17318935692310333,
      "learning_rate": 0.0003257705641579247,
      "loss": 0.3716,
      "num_input_tokens_seen": 16250192,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 0.20793558657169342,
      "learning_rate": 0.00032555622977869614,
      "loss": 0.406,
      "num_input_tokens_seen": 16254192,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 0.1930878460407257,
      "learning_rate": 0.00032534193189111407,
      "loss": 0.4259,
      "num_input_tokens_seen": 16257104,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 0.20336338877677917,
      "learning_rate": 0.0003251276705400074,
      "loss": 0.4329,
      "num_input_tokens_seen": 16259888,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 0.13133449852466583,
      "learning_rate": 0.0003249134457701968,
      "loss": 0.4168,
      "num_input_tokens_seen": 16263728,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.13576044142246246,
      "learning_rate": 0.000324699257626496,
      "loss": 0.3518,
      "num_input_tokens_seen": 16266576,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.164794921875,
      "learning_rate": 0.00032448510615371045,
      "loss": 0.3578,
      "num_input_tokens_seen": 16269104,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 0.0931117832660675,
      "learning_rate": 0.0003242709913966384,
      "loss": 0.5384,
      "num_input_tokens_seen": 16271984,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 0.10569535940885544,
      "learning_rate": 0.00032405691340007,
      "loss": 0.5167,
      "num_input_tokens_seen": 16276272,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 0.18209457397460938,
      "learning_rate": 0.0003238428722087884,
      "loss": 0.4383,
      "num_input_tokens_seen": 16279280,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 0.26709261536598206,
      "learning_rate": 0.00032362886786756797,
      "loss": 0.4051,
      "num_input_tokens_seen": 16283440,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.11140014231204987,
      "learning_rate": 0.00032341490042117615,
      "loss": 0.4695,
      "num_input_tokens_seen": 16286448,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 0.1466987282037735,
      "learning_rate": 0.00032320096991437277,
      "loss": 0.3277,
      "num_input_tokens_seen": 16289008,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 0.1404561847448349,
      "learning_rate": 0.0003229870763919093,
      "loss": 0.5603,
      "num_input_tokens_seen": 16292592,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 0.2509412467479706,
      "learning_rate": 0.00032277321989853015,
      "loss": 0.3735,
      "num_input_tokens_seen": 16295664,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 0.15899017453193665,
      "learning_rate": 0.00032255940047897137,
      "loss": 0.4121,
      "num_input_tokens_seen": 16298064,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 0.20508478581905365,
      "learning_rate": 0.0003223456181779616,
      "loss": 0.3194,
      "num_input_tokens_seen": 16300400,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 0.220086008310318,
      "learning_rate": 0.0003221318730402218,
      "loss": 0.3916,
      "num_input_tokens_seen": 16303664,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 0.15846028923988342,
      "learning_rate": 0.00032191816511046483,
      "loss": 0.3964,
      "num_input_tokens_seen": 16306288,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 0.08734763413667679,
      "learning_rate": 0.0003217044944333961,
      "loss": 0.4815,
      "num_input_tokens_seen": 16309808,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 0.1452879160642624,
      "learning_rate": 0.00032149086105371315,
      "loss": 0.3674,
      "num_input_tokens_seen": 16313136,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.14702294766902924,
      "learning_rate": 0.00032127726501610554,
      "loss": 0.4199,
      "num_input_tokens_seen": 16316464,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.14305594563484192,
      "learning_rate": 0.00032106370636525493,
      "loss": 0.4312,
      "num_input_tokens_seen": 16319376,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.16495570540428162,
      "learning_rate": 0.00032085018514583564,
      "loss": 0.5762,
      "num_input_tokens_seen": 16323280,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.17873519659042358,
      "learning_rate": 0.00032063670140251366,
      "loss": 0.4152,
      "num_input_tokens_seen": 16326288,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 0.1273943930864334,
      "learning_rate": 0.0003204232551799476,
      "loss": 0.4534,
      "num_input_tokens_seen": 16329008,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.27870234847068787,
      "learning_rate": 0.00032020984652278797,
      "loss": 0.36,
      "num_input_tokens_seen": 16331408,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.14129067957401276,
      "learning_rate": 0.000319996475475677,
      "loss": 0.4766,
      "num_input_tokens_seen": 16335024,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 0.25641128420829773,
      "learning_rate": 0.0003197831420832499,
      "loss": 0.4278,
      "num_input_tokens_seen": 16337744,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 0.23820006847381592,
      "learning_rate": 0.0003195698463901334,
      "loss": 0.4648,
      "num_input_tokens_seen": 16340656,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 0.12329072505235672,
      "learning_rate": 0.0003193565884409466,
      "loss": 0.4759,
      "num_input_tokens_seen": 16343440,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 0.07067862153053284,
      "learning_rate": 0.0003191433682803008,
      "loss": 0.3356,
      "num_input_tokens_seen": 16349008,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.10575675219297409,
      "learning_rate": 0.00031893018595279877,
      "loss": 0.3522,
      "num_input_tokens_seen": 16351728,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 0.25131329894065857,
      "learning_rate": 0.00031871704150303624,
      "loss": 0.5028,
      "num_input_tokens_seen": 16354384,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 0.18218287825584412,
      "learning_rate": 0.00031850393497560034,
      "loss": 0.6667,
      "num_input_tokens_seen": 16357616,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 0.20950117707252502,
      "learning_rate": 0.00031829086641507054,
      "loss": 0.3956,
      "num_input_tokens_seen": 16360592,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 0.13206495344638824,
      "learning_rate": 0.00031807783586601865,
      "loss": 0.408,
      "num_input_tokens_seen": 16363600,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 0.14756625890731812,
      "learning_rate": 0.00031786484337300805,
      "loss": 0.3349,
      "num_input_tokens_seen": 16366352,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 0.1433517336845398,
      "learning_rate": 0.0003176518889805941,
      "loss": 0.3896,
      "num_input_tokens_seen": 16368944,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 0.23877421021461487,
      "learning_rate": 0.0003174389727333248,
      "loss": 0.3994,
      "num_input_tokens_seen": 16371952,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 0.17216995358467102,
      "learning_rate": 0.00031722609467573946,
      "loss": 0.4102,
      "num_input_tokens_seen": 16374704,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.13940468430519104,
      "learning_rate": 0.0003170132548523701,
      "loss": 0.4014,
      "num_input_tokens_seen": 16377584,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 0.3173772394657135,
      "learning_rate": 0.00031680045330774003,
      "loss": 0.531,
      "num_input_tokens_seen": 16380560,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 0.13785749673843384,
      "learning_rate": 0.00031658769008636534,
      "loss": 0.5134,
      "num_input_tokens_seen": 16384016,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 0.1227414682507515,
      "learning_rate": 0.0003163749652327532,
      "loss": 0.4428,
      "num_input_tokens_seen": 16387376,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 0.1288779228925705,
      "learning_rate": 0.00031616227879140347,
      "loss": 0.5026,
      "num_input_tokens_seen": 16390704,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.09464778006076813,
      "learning_rate": 0.0003159496308068077,
      "loss": 0.4012,
      "num_input_tokens_seen": 16394192,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.2886532247066498,
      "learning_rate": 0.00031573702132344926,
      "loss": 0.4016,
      "num_input_tokens_seen": 16397552,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 0.18636758625507355,
      "learning_rate": 0.00031552445038580414,
      "loss": 0.3799,
      "num_input_tokens_seen": 16400400,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 0.12721367180347443,
      "learning_rate": 0.0003153119180383391,
      "loss": 0.3673,
      "num_input_tokens_seen": 16403632,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 0.13013534247875214,
      "learning_rate": 0.0003150994243255138,
      "loss": 0.4368,
      "num_input_tokens_seen": 16406832,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 0.10465541481971741,
      "learning_rate": 0.00031488696929177944,
      "loss": 0.5263,
      "num_input_tokens_seen": 16411760,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 0.2215556800365448,
      "learning_rate": 0.00031467455298157934,
      "loss": 0.4141,
      "num_input_tokens_seen": 16414800,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 0.2325436919927597,
      "learning_rate": 0.0003144621754393483,
      "loss": 0.4173,
      "num_input_tokens_seen": 16418096,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 0.174685999751091,
      "learning_rate": 0.0003142498367095137,
      "loss": 0.4796,
      "num_input_tokens_seen": 16421552,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.15079893171787262,
      "learning_rate": 0.00031403753683649396,
      "loss": 0.3749,
      "num_input_tokens_seen": 16424432,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 0.25451022386550903,
      "learning_rate": 0.0003138252758646999,
      "loss": 0.4403,
      "num_input_tokens_seen": 16426960,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.11792421340942383,
      "learning_rate": 0.00031361305383853413,
      "loss": 0.3188,
      "num_input_tokens_seen": 16430320,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 0.14817187190055847,
      "learning_rate": 0.00031340087080239105,
      "loss": 0.4474,
      "num_input_tokens_seen": 16433328,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 0.11362159252166748,
      "learning_rate": 0.0003131887268006571,
      "loss": 0.427,
      "num_input_tokens_seen": 16436688,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 0.19369575381278992,
      "learning_rate": 0.00031297662187771026,
      "loss": 0.3687,
      "num_input_tokens_seen": 16439856,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 0.19993677735328674,
      "learning_rate": 0.0003127645560779203,
      "loss": 0.3806,
      "num_input_tokens_seen": 16442480,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.11667437851428986,
      "learning_rate": 0.0003125525294456492,
      "loss": 0.3897,
      "num_input_tokens_seen": 16446192,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 0.10571152716875076,
      "learning_rate": 0.0003123405420252503,
      "loss": 0.3813,
      "num_input_tokens_seen": 16449200,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 0.22757935523986816,
      "learning_rate": 0.00031212859386106927,
      "loss": 0.5637,
      "num_input_tokens_seen": 16452112,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 0.11827582865953445,
      "learning_rate": 0.00031191668499744286,
      "loss": 0.5165,
      "num_input_tokens_seen": 16455760,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 0.2886931002140045,
      "learning_rate": 0.00031170481547870046,
      "loss": 0.4279,
      "num_input_tokens_seen": 16458704,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 0.19885802268981934,
      "learning_rate": 0.00031149298534916215,
      "loss": 0.4444,
      "num_input_tokens_seen": 16461680,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.1503104269504547,
      "learning_rate": 0.00031128119465314094,
      "loss": 0.4712,
      "num_input_tokens_seen": 16464496,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 0.3086051046848297,
      "learning_rate": 0.0003110694434349406,
      "loss": 0.3037,
      "num_input_tokens_seen": 16466992,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 0.2250315546989441,
      "learning_rate": 0.0003108577317388574,
      "loss": 0.4106,
      "num_input_tokens_seen": 16469392,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 0.17793522775173187,
      "learning_rate": 0.00031064605960917896,
      "loss": 0.539,
      "num_input_tokens_seen": 16472848,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 0.08657737076282501,
      "learning_rate": 0.00031043442709018446,
      "loss": 0.4651,
      "num_input_tokens_seen": 16477040,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.41229256987571716,
      "learning_rate": 0.0003102228342261451,
      "loss": 0.4954,
      "num_input_tokens_seen": 16480432,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 0.16975224018096924,
      "learning_rate": 0.0003100112810613237,
      "loss": 0.4277,
      "num_input_tokens_seen": 16483088,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 0.22321480512619019,
      "learning_rate": 0.00030979976763997484,
      "loss": 0.3725,
      "num_input_tokens_seen": 16486640,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 0.15677523612976074,
      "learning_rate": 0.0003095882940063447,
      "loss": 0.3411,
      "num_input_tokens_seen": 16489168,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.1196863055229187,
      "learning_rate": 0.0003093768602046712,
      "loss": 0.4978,
      "num_input_tokens_seen": 16491824,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 0.14753854274749756,
      "learning_rate": 0.00030916546627918375,
      "loss": 0.5388,
      "num_input_tokens_seen": 16495408,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 0.12593871355056763,
      "learning_rate": 0.00030895411227410354,
      "loss": 0.4264,
      "num_input_tokens_seen": 16499280,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 0.16460734605789185,
      "learning_rate": 0.0003087427982336435,
      "loss": 0.3821,
      "num_input_tokens_seen": 16502352,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.21481236815452576,
      "learning_rate": 0.00030853152420200804,
      "loss": 0.5734,
      "num_input_tokens_seen": 16505456,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 0.09169020503759384,
      "learning_rate": 0.0003083202902233932,
      "loss": 0.394,
      "num_input_tokens_seen": 16508464,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.24791806936264038,
      "learning_rate": 0.00030810909634198727,
      "loss": 0.3944,
      "num_input_tokens_seen": 16511440,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 0.14941829442977905,
      "learning_rate": 0.00030789794260196893,
      "loss": 0.3933,
      "num_input_tokens_seen": 16514800,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.13302259147167206,
      "learning_rate": 0.00030768682904750935,
      "loss": 0.3844,
      "num_input_tokens_seen": 16517488,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 0.197451651096344,
      "learning_rate": 0.00030747575572277107,
      "loss": 0.4775,
      "num_input_tokens_seen": 16520528,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 0.14301209151744843,
      "learning_rate": 0.0003072647226719083,
      "loss": 0.3996,
      "num_input_tokens_seen": 16524144,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 0.1546056717634201,
      "learning_rate": 0.0003070537299390669,
      "loss": 0.3744,
      "num_input_tokens_seen": 16526768,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 0.15949633717536926,
      "learning_rate": 0.0003068427775683838,
      "loss": 0.5481,
      "num_input_tokens_seen": 16531088,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 0.2350752055644989,
      "learning_rate": 0.00030663186560398793,
      "loss": 0.4553,
      "num_input_tokens_seen": 16533904,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.21559609472751617,
      "learning_rate": 0.0003064209940899998,
      "loss": 0.5308,
      "num_input_tokens_seen": 16537104,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.14989881217479706,
      "learning_rate": 0.00030621016307053105,
      "loss": 0.3238,
      "num_input_tokens_seen": 16540240,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 0.14719635248184204,
      "learning_rate": 0.0003059993725896855,
      "loss": 0.4443,
      "num_input_tokens_seen": 16543248,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.21844428777694702,
      "learning_rate": 0.0003057886226915581,
      "loss": 0.4429,
      "num_input_tokens_seen": 16546544,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 0.10136774182319641,
      "learning_rate": 0.00030557791342023476,
      "loss": 0.4728,
      "num_input_tokens_seen": 16549744,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.11440835148096085,
      "learning_rate": 0.0003053672448197941,
      "loss": 0.5453,
      "num_input_tokens_seen": 16552656,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 0.18262259662151337,
      "learning_rate": 0.0003051566169343052,
      "loss": 0.7457,
      "num_input_tokens_seen": 16555824,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 0.15652212500572205,
      "learning_rate": 0.00030494602980782917,
      "loss": 0.4338,
      "num_input_tokens_seen": 16559056,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 0.19024793803691864,
      "learning_rate": 0.00030473548348441845,
      "loss": 0.3174,
      "num_input_tokens_seen": 16561744,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 0.13167202472686768,
      "learning_rate": 0.000304524978008117,
      "loss": 0.4161,
      "num_input_tokens_seen": 16564784,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 0.24666932225227356,
      "learning_rate": 0.0003043145134229599,
      "loss": 0.4984,
      "num_input_tokens_seen": 16567280,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.16636382043361664,
      "learning_rate": 0.0003041040897729742,
      "loss": 0.4461,
      "num_input_tokens_seen": 16569840,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 0.3140530288219452,
      "learning_rate": 0.0003038937071021778,
      "loss": 0.4929,
      "num_input_tokens_seen": 16572976,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 0.12847799062728882,
      "learning_rate": 0.0003036833654545808,
      "loss": 0.4637,
      "num_input_tokens_seen": 16575792,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 0.177087664604187,
      "learning_rate": 0.0003034730648741841,
      "loss": 0.4486,
      "num_input_tokens_seen": 16578800,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 0.1459130197763443,
      "learning_rate": 0.00030326280540497996,
      "loss": 0.4268,
      "num_input_tokens_seen": 16582864,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 0.28460729122161865,
      "learning_rate": 0.0003030525870909525,
      "loss": 0.606,
      "num_input_tokens_seen": 16585968,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 0.2450941801071167,
      "learning_rate": 0.0003028424099760768,
      "loss": 0.4532,
      "num_input_tokens_seen": 16592752,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 0.1459309458732605,
      "learning_rate": 0.0003026322741043198,
      "loss": 0.5094,
      "num_input_tokens_seen": 16595632,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 0.1498739868402481,
      "learning_rate": 0.0003024221795196393,
      "loss": 0.4875,
      "num_input_tokens_seen": 16598096,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 0.15628334879875183,
      "learning_rate": 0.0003022121262659849,
      "loss": 0.413,
      "num_input_tokens_seen": 16601040,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.1257224678993225,
      "learning_rate": 0.00030200211438729705,
      "loss": 0.5357,
      "num_input_tokens_seen": 16603984,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 0.19174733757972717,
      "learning_rate": 0.000301792143927508,
      "loss": 0.4883,
      "num_input_tokens_seen": 16607280,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 0.13769282400608063,
      "learning_rate": 0.0003015822149305411,
      "loss": 0.4161,
      "num_input_tokens_seen": 16610576,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 0.11156196147203445,
      "learning_rate": 0.0003013723274403113,
      "loss": 0.4775,
      "num_input_tokens_seen": 16615728,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.17496469616889954,
      "learning_rate": 0.00030116248150072444,
      "loss": 0.47,
      "num_input_tokens_seen": 16618576,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.14507021009922028,
      "learning_rate": 0.00030095267715567816,
      "loss": 0.4333,
      "num_input_tokens_seen": 16622832,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.18067997694015503,
      "learning_rate": 0.00030074291444906085,
      "loss": 0.3891,
      "num_input_tokens_seen": 16625168,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 0.1472092568874359,
      "learning_rate": 0.00030053319342475247,
      "loss": 0.4146,
      "num_input_tokens_seen": 16628944,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 0.1661386638879776,
      "learning_rate": 0.00030032351412662465,
      "loss": 0.4132,
      "num_input_tokens_seen": 16632208,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 0.10340750962495804,
      "learning_rate": 0.0003001138765985394,
      "loss": 0.4333,
      "num_input_tokens_seen": 16634960,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 0.11354885995388031,
      "learning_rate": 0.00029990428088435097,
      "loss": 0.4132,
      "num_input_tokens_seen": 16637616,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 0.2378402203321457,
      "learning_rate": 0.000299694727027904,
      "loss": 0.4864,
      "num_input_tokens_seen": 16640880,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 0.2077239751815796,
      "learning_rate": 0.00029948521507303495,
      "loss": 0.4037,
      "num_input_tokens_seen": 16643504,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 0.25895121693611145,
      "learning_rate": 0.0002992757450635714,
      "loss": 0.4191,
      "num_input_tokens_seen": 16647120,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 0.16501112282276154,
      "learning_rate": 0.00029906631704333187,
      "loss": 0.446,
      "num_input_tokens_seen": 16651248,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 0.16359935700893402,
      "learning_rate": 0.0002988569310561264,
      "loss": 0.3254,
      "num_input_tokens_seen": 16653744,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 0.10871928930282593,
      "learning_rate": 0.0002986475871457566,
      "loss": 0.3403,
      "num_input_tokens_seen": 16656752,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 0.15228454768657684,
      "learning_rate": 0.000298438285356014,
      "loss": 0.3861,
      "num_input_tokens_seen": 16659984,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.14464521408081055,
      "learning_rate": 0.00029822902573068257,
      "loss": 0.4718,
      "num_input_tokens_seen": 16663440,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 0.1072569191455841,
      "learning_rate": 0.0002980198083135372,
      "loss": 0.4315,
      "num_input_tokens_seen": 16667440,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 0.20399510860443115,
      "learning_rate": 0.0002978106331483435,
      "loss": 0.4739,
      "num_input_tokens_seen": 16670256,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 0.14155690371990204,
      "learning_rate": 0.00029760150027885874,
      "loss": 0.4063,
      "num_input_tokens_seen": 16673040,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.24286428093910217,
      "learning_rate": 0.0002973924097488311,
      "loss": 0.4029,
      "num_input_tokens_seen": 16676016,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 0.09894047677516937,
      "learning_rate": 0.00029718336160199977,
      "loss": 0.4456,
      "num_input_tokens_seen": 16679536,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 0.18425840139389038,
      "learning_rate": 0.00029697435588209537,
      "loss": 0.4485,
      "num_input_tokens_seen": 16682672,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 0.2914249300956726,
      "learning_rate": 0.00029676539263283943,
      "loss": 0.3726,
      "num_input_tokens_seen": 16685424,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 0.15899285674095154,
      "learning_rate": 0.0002965564718979448,
      "loss": 0.5266,
      "num_input_tokens_seen": 16689328,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 0.11517851799726486,
      "learning_rate": 0.0002963475937211151,
      "loss": 0.3734,
      "num_input_tokens_seen": 16692656,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.1246858537197113,
      "learning_rate": 0.0002961387581460456,
      "loss": 0.4899,
      "num_input_tokens_seen": 16696912,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 0.15576158463954926,
      "learning_rate": 0.0002959299652164221,
      "loss": 0.4921,
      "num_input_tokens_seen": 16699600,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 0.1899450123310089,
      "learning_rate": 0.00029572121497592154,
      "loss": 0.5146,
      "num_input_tokens_seen": 16702608,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 0.14884600043296814,
      "learning_rate": 0.00029551250746821236,
      "loss": 0.3582,
      "num_input_tokens_seen": 16705968,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 0.16611459851264954,
      "learning_rate": 0.0002953038427369537,
      "loss": 0.4707,
      "num_input_tokens_seen": 16710320,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 0.27233394980430603,
      "learning_rate": 0.000295095220825796,
      "loss": 0.437,
      "num_input_tokens_seen": 16713200,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 0.17341557145118713,
      "learning_rate": 0.00029488664177838023,
      "loss": 0.4364,
      "num_input_tokens_seen": 16716176,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 0.19571469724178314,
      "learning_rate": 0.0002946781056383391,
      "loss": 0.3963,
      "num_input_tokens_seen": 16719504,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.08105491101741791,
      "learning_rate": 0.0002944696124492956,
      "loss": 0.419,
      "num_input_tokens_seen": 16724176,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 0.0975620225071907,
      "learning_rate": 0.0002942611622548645,
      "loss": 0.2662,
      "num_input_tokens_seen": 16727344,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 0.1677730828523636,
      "learning_rate": 0.00029405275509865103,
      "loss": 0.4267,
      "num_input_tokens_seen": 16730256,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 0.09448865801095963,
      "learning_rate": 0.00029384439102425174,
      "loss": 0.5198,
      "num_input_tokens_seen": 16733168,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 0.22177474200725555,
      "learning_rate": 0.0002936360700752539,
      "loss": 0.5226,
      "num_input_tokens_seen": 16736816,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 0.15006448328495026,
      "learning_rate": 0.00029342779229523573,
      "loss": 0.3794,
      "num_input_tokens_seen": 16739632,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 0.21216490864753723,
      "learning_rate": 0.00029321955772776675,
      "loss": 0.383,
      "num_input_tokens_seen": 16743792,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 0.13284042477607727,
      "learning_rate": 0.00029301136641640723,
      "loss": 0.3305,
      "num_input_tokens_seen": 16747504,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 0.10513447970151901,
      "learning_rate": 0.0002928032184047085,
      "loss": 0.3724,
      "num_input_tokens_seen": 16751312,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 0.5575754642486572,
      "learning_rate": 0.0002925951137362126,
      "loss": 0.4017,
      "num_input_tokens_seen": 16754768,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 0.1775345504283905,
      "learning_rate": 0.00029238705245445264,
      "loss": 0.4237,
      "num_input_tokens_seen": 16757488,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 0.16405139863491058,
      "learning_rate": 0.0002921790346029527,
      "loss": 0.4815,
      "num_input_tokens_seen": 16760048,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.12116178125143051,
      "learning_rate": 0.0002919710602252279,
      "loss": 0.5181,
      "num_input_tokens_seen": 16763152,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 0.17773117125034332,
      "learning_rate": 0.0002917631293647838,
      "loss": 0.4193,
      "num_input_tokens_seen": 16766608,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.1657658964395523,
      "learning_rate": 0.0002915552420651176,
      "loss": 0.7365,
      "num_input_tokens_seen": 16773584,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 0.10450134426355362,
      "learning_rate": 0.00029134739836971657,
      "loss": 0.3426,
      "num_input_tokens_seen": 16776976,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.11061562597751617,
      "learning_rate": 0.0002911395983220595,
      "loss": 0.3537,
      "num_input_tokens_seen": 16779568,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.2070806920528412,
      "learning_rate": 0.0002909318419656154,
      "loss": 0.5102,
      "num_input_tokens_seen": 16782576,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 0.11059892177581787,
      "learning_rate": 0.00029072412934384484,
      "loss": 0.4478,
      "num_input_tokens_seen": 16785968,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 0.16218729317188263,
      "learning_rate": 0.00029051646050019884,
      "loss": 0.52,
      "num_input_tokens_seen": 16789872,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.11933229118585587,
      "learning_rate": 0.00029030883547811966,
      "loss": 0.3665,
      "num_input_tokens_seen": 16792944,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 0.16767573356628418,
      "learning_rate": 0.0002901012543210397,
      "loss": 0.3392,
      "num_input_tokens_seen": 16796176,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 0.12569692730903625,
      "learning_rate": 0.00028989371707238264,
      "loss": 0.4537,
      "num_input_tokens_seen": 16799504,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.15532924234867096,
      "learning_rate": 0.00028968622377556285,
      "loss": 0.4152,
      "num_input_tokens_seen": 16802384,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 0.14378753304481506,
      "learning_rate": 0.00028947877447398567,
      "loss": 0.3576,
      "num_input_tokens_seen": 16806064,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 0.24663496017456055,
      "learning_rate": 0.00028927136921104733,
      "loss": 0.511,
      "num_input_tokens_seen": 16808560,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.1495647132396698,
      "learning_rate": 0.00028906400803013446,
      "loss": 0.606,
      "num_input_tokens_seen": 16811696,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 0.16115739941596985,
      "learning_rate": 0.00028885669097462435,
      "loss": 0.369,
      "num_input_tokens_seen": 16815600,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 0.15082082152366638,
      "learning_rate": 0.0002886494180878857,
      "loss": 0.3211,
      "num_input_tokens_seen": 16818064,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 0.21012140810489655,
      "learning_rate": 0.00028844218941327757,
      "loss": 0.4336,
      "num_input_tokens_seen": 16821104,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 0.13391070067882538,
      "learning_rate": 0.0002882350049941498,
      "loss": 0.3712,
      "num_input_tokens_seen": 16824848,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 0.27869749069213867,
      "learning_rate": 0.0002880278648738432,
      "loss": 0.3964,
      "num_input_tokens_seen": 16828144,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 0.10081039369106293,
      "learning_rate": 0.0002878207690956891,
      "loss": 0.4377,
      "num_input_tokens_seen": 16831408,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 0.11773241311311722,
      "learning_rate": 0.00028761371770300915,
      "loss": 0.5307,
      "num_input_tokens_seen": 16834544,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 0.17642664909362793,
      "learning_rate": 0.0002874067107391164,
      "loss": 0.5539,
      "num_input_tokens_seen": 16837360,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 0.10966332256793976,
      "learning_rate": 0.00028719974824731445,
      "loss": 0.3911,
      "num_input_tokens_seen": 16840176,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 0.21954962611198425,
      "learning_rate": 0.0002869928302708975,
      "loss": 0.4586,
      "num_input_tokens_seen": 16844080,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 0.1208355650305748,
      "learning_rate": 0.00028678595685315044,
      "loss": 0.3891,
      "num_input_tokens_seen": 16846928,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.11875946074724197,
      "learning_rate": 0.00028657912803734854,
      "loss": 0.4514,
      "num_input_tokens_seen": 16849840,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 0.22157111763954163,
      "learning_rate": 0.00028637234386675816,
      "loss": 0.3618,
      "num_input_tokens_seen": 16853328,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 0.07830411195755005,
      "learning_rate": 0.00028616560438463626,
      "loss": 0.4805,
      "num_input_tokens_seen": 16856080,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 0.17876946926116943,
      "learning_rate": 0.00028595890963423057,
      "loss": 0.3617,
      "num_input_tokens_seen": 16858992,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 0.18706713616847992,
      "learning_rate": 0.0002857522596587789,
      "loss": 0.4013,
      "num_input_tokens_seen": 16861968,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.15381695330142975,
      "learning_rate": 0.00028554565450151045,
      "loss": 0.3946,
      "num_input_tokens_seen": 16865232,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 0.24293403327465057,
      "learning_rate": 0.00028533909420564417,
      "loss": 0.3882,
      "num_input_tokens_seen": 16868432,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 0.16064989566802979,
      "learning_rate": 0.00028513257881439047,
      "loss": 0.4201,
      "num_input_tokens_seen": 16871408,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 0.11530432850122452,
      "learning_rate": 0.00028492610837095003,
      "loss": 0.4228,
      "num_input_tokens_seen": 16875184,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 0.17562921345233917,
      "learning_rate": 0.00028471968291851407,
      "loss": 0.3974,
      "num_input_tokens_seen": 16878064,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 0.12837223708629608,
      "learning_rate": 0.0002845133025002645,
      "loss": 0.5388,
      "num_input_tokens_seen": 16881104,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 0.19984827935695648,
      "learning_rate": 0.0002843069671593734,
      "loss": 0.5791,
      "num_input_tokens_seen": 16884272,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 0.1390974223613739,
      "learning_rate": 0.000284100676939004,
      "loss": 0.4704,
      "num_input_tokens_seen": 16887856,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.1510249227285385,
      "learning_rate": 0.0002838944318823099,
      "loss": 0.4171,
      "num_input_tokens_seen": 16890928,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 0.11489678174257278,
      "learning_rate": 0.0002836882320324354,
      "loss": 0.4286,
      "num_input_tokens_seen": 16893680,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 0.11625460535287857,
      "learning_rate": 0.00028348207743251477,
      "loss": 0.4829,
      "num_input_tokens_seen": 16897392,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 0.10311868786811829,
      "learning_rate": 0.0002832759681256735,
      "loss": 0.5186,
      "num_input_tokens_seen": 16900016,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 0.11004000902175903,
      "learning_rate": 0.000283069904155027,
      "loss": 0.3666,
      "num_input_tokens_seen": 16904400,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 0.22042988240718842,
      "learning_rate": 0.0002828638855636817,
      "loss": 0.544,
      "num_input_tokens_seen": 16908336,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 0.0756988450884819,
      "learning_rate": 0.0002826579123947345,
      "loss": 0.4132,
      "num_input_tokens_seen": 16911920,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 0.1996413618326187,
      "learning_rate": 0.0002824519846912723,
      "loss": 0.4442,
      "num_input_tokens_seen": 16914896,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 0.17176322638988495,
      "learning_rate": 0.00028224610249637293,
      "loss": 0.4446,
      "num_input_tokens_seen": 16917744,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 0.1680779606103897,
      "learning_rate": 0.00028204026585310483,
      "loss": 0.3888,
      "num_input_tokens_seen": 16920464,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 0.17860601842403412,
      "learning_rate": 0.0002818344748045264,
      "loss": 0.5066,
      "num_input_tokens_seen": 16923920,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 0.09086955338716507,
      "learning_rate": 0.0002816287293936868,
      "loss": 0.39,
      "num_input_tokens_seen": 16927856,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 0.17618489265441895,
      "learning_rate": 0.0002814230296636259,
      "loss": 0.6391,
      "num_input_tokens_seen": 16930672,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 0.08690611273050308,
      "learning_rate": 0.0002812173756573734,
      "loss": 0.4862,
      "num_input_tokens_seen": 16933584,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 0.2011360079050064,
      "learning_rate": 0.00028101176741794984,
      "loss": 0.4368,
      "num_input_tokens_seen": 16935952,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 0.1569489985704422,
      "learning_rate": 0.0002808062049883664,
      "loss": 0.3378,
      "num_input_tokens_seen": 16939024,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 0.12694284319877625,
      "learning_rate": 0.00028060068841162403,
      "loss": 0.4192,
      "num_input_tokens_seen": 16942576,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.11767596006393433,
      "learning_rate": 0.00028039521773071477,
      "loss": 0.4955,
      "num_input_tokens_seen": 16945808,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.31806308031082153,
      "learning_rate": 0.0002801897929886204,
      "loss": 0.4461,
      "num_input_tokens_seen": 16948752,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.17362767457962036,
      "learning_rate": 0.0002799844142283135,
      "loss": 0.4804,
      "num_input_tokens_seen": 16951856,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.13476113975048065,
      "learning_rate": 0.00027977908149275733,
      "loss": 0.3922,
      "num_input_tokens_seen": 16955088,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 0.21236652135849,
      "learning_rate": 0.00027957379482490476,
      "loss": 0.3727,
      "num_input_tokens_seen": 16958576,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 0.15874479711055756,
      "learning_rate": 0.0002793685542676997,
      "loss": 0.4497,
      "num_input_tokens_seen": 16961904,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 0.2294979989528656,
      "learning_rate": 0.0002791633598640758,
      "loss": 0.5589,
      "num_input_tokens_seen": 16965008,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 0.10235747694969177,
      "learning_rate": 0.0002789582116569576,
      "loss": 0.5298,
      "num_input_tokens_seen": 16968592,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.16386812925338745,
      "learning_rate": 0.0002787531096892597,
      "loss": 0.3892,
      "num_input_tokens_seen": 16971280,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 0.15166881680488586,
      "learning_rate": 0.0002785480540038874,
      "loss": 0.4893,
      "num_input_tokens_seen": 16975696,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.16596129536628723,
      "learning_rate": 0.0002783430446437355,
      "loss": 0.476,
      "num_input_tokens_seen": 16979088,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 0.16175182163715363,
      "learning_rate": 0.00027813808165169006,
      "loss": 0.458,
      "num_input_tokens_seen": 16981552,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 0.12541380524635315,
      "learning_rate": 0.00027793316507062667,
      "loss": 0.2852,
      "num_input_tokens_seen": 16984208,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 0.18618732690811157,
      "learning_rate": 0.00027772829494341163,
      "loss": 0.468,
      "num_input_tokens_seen": 16987824,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 0.10243634879589081,
      "learning_rate": 0.00027752347131290147,
      "loss": 0.4537,
      "num_input_tokens_seen": 16991568,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.19529369473457336,
      "learning_rate": 0.0002773186942219431,
      "loss": 0.3679,
      "num_input_tokens_seen": 16994256,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 0.1815524697303772,
      "learning_rate": 0.0002771139637133735,
      "loss": 0.4545,
      "num_input_tokens_seen": 16996976,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 0.12224006652832031,
      "learning_rate": 0.00027690927983001966,
      "loss": 0.4453,
      "num_input_tokens_seen": 16999824,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 0.14770887792110443,
      "learning_rate": 0.00027670464261469937,
      "loss": 0.4558,
      "num_input_tokens_seen": 17004112,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 0.16384050250053406,
      "learning_rate": 0.0002765000521102203,
      "loss": 0.4527,
      "num_input_tokens_seen": 17006672,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 0.26492246985435486,
      "learning_rate": 0.0002762955083593807,
      "loss": 0.4398,
      "num_input_tokens_seen": 17009712,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 0.1042708158493042,
      "learning_rate": 0.0002760910114049686,
      "loss": 0.4832,
      "num_input_tokens_seen": 17013680,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.1253710836172104,
      "learning_rate": 0.0002758865612897623,
      "loss": 0.531,
      "num_input_tokens_seen": 17016880,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 0.21092846989631653,
      "learning_rate": 0.00027568215805653045,
      "loss": 0.5601,
      "num_input_tokens_seen": 17019312,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 0.16565605998039246,
      "learning_rate": 0.00027547780174803205,
      "loss": 0.5105,
      "num_input_tokens_seen": 17022032,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 0.0891982764005661,
      "learning_rate": 0.0002752734924070161,
      "loss": 0.4301,
      "num_input_tokens_seen": 17026384,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.301059752702713,
      "learning_rate": 0.00027506923007622177,
      "loss": 0.5461,
      "num_input_tokens_seen": 17028688,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.15479440987110138,
      "learning_rate": 0.00027486501479837846,
      "loss": 0.3949,
      "num_input_tokens_seen": 17031792,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 0.14683972299098969,
      "learning_rate": 0.0002746608466162053,
      "loss": 0.5074,
      "num_input_tokens_seen": 17035088,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 0.1573181301355362,
      "learning_rate": 0.0002744567255724123,
      "loss": 0.3223,
      "num_input_tokens_seen": 17038128,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 0.3055976629257202,
      "learning_rate": 0.0002742526517096992,
      "loss": 0.4912,
      "num_input_tokens_seen": 17041744,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 0.15572169423103333,
      "learning_rate": 0.000274048625070756,
      "loss": 0.4433,
      "num_input_tokens_seen": 17044880,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 0.09570127725601196,
      "learning_rate": 0.0002738446456982627,
      "loss": 0.5657,
      "num_input_tokens_seen": 17049040,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.1327255368232727,
      "learning_rate": 0.0002736407136348892,
      "loss": 0.5425,
      "num_input_tokens_seen": 17052368,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.17561456561088562,
      "learning_rate": 0.0002734368289232959,
      "loss": 0.3682,
      "num_input_tokens_seen": 17056272,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 0.18728196620941162,
      "learning_rate": 0.00027323299160613335,
      "loss": 0.3802,
      "num_input_tokens_seen": 17059440,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 0.19719132781028748,
      "learning_rate": 0.00027302920172604173,
      "loss": 0.4128,
      "num_input_tokens_seen": 17062832,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 0.21281325817108154,
      "learning_rate": 0.00027282545932565214,
      "loss": 0.4744,
      "num_input_tokens_seen": 17065520,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 0.09769036620855331,
      "learning_rate": 0.00027262176444758435,
      "loss": 0.3971,
      "num_input_tokens_seen": 17068848,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 0.14273661375045776,
      "learning_rate": 0.00027241811713444935,
      "loss": 0.3293,
      "num_input_tokens_seen": 17071984,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 0.1657363772392273,
      "learning_rate": 0.0002722145174288478,
      "loss": 0.4758,
      "num_input_tokens_seen": 17075248,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.15411770343780518,
      "learning_rate": 0.0002720109653733706,
      "loss": 0.5049,
      "num_input_tokens_seen": 17079120,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.19610090553760529,
      "learning_rate": 0.00027180746101059856,
      "loss": 0.466,
      "num_input_tokens_seen": 17083856,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.10428302735090256,
      "learning_rate": 0.0002716040043831023,
      "loss": 0.4966,
      "num_input_tokens_seen": 17087056,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.10683199763298035,
      "learning_rate": 0.0002714005955334424,
      "loss": 0.4799,
      "num_input_tokens_seen": 17090800,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 0.20535746216773987,
      "learning_rate": 0.00027119723450417,
      "loss": 0.3818,
      "num_input_tokens_seen": 17093168,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 0.18980900943279266,
      "learning_rate": 0.0002709939213378258,
      "loss": 0.3703,
      "num_input_tokens_seen": 17096304,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 0.15270036458969116,
      "learning_rate": 0.0002707906560769408,
      "loss": 0.3505,
      "num_input_tokens_seen": 17099568,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 0.12021928280591965,
      "learning_rate": 0.0002705874387640354,
      "loss": 0.3803,
      "num_input_tokens_seen": 17103696,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.1423855870962143,
      "learning_rate": 0.00027038426944162074,
      "loss": 0.4141,
      "num_input_tokens_seen": 17106384,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 0.2617117464542389,
      "learning_rate": 0.0002701811481521971,
      "loss": 0.4473,
      "num_input_tokens_seen": 17109040,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.08770669251680374,
      "learning_rate": 0.0002699780749382554,
      "loss": 0.3732,
      "num_input_tokens_seen": 17112496,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 0.1964341551065445,
      "learning_rate": 0.0002697750498422761,
      "loss": 0.4831,
      "num_input_tokens_seen": 17115472,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.13078905642032623,
      "learning_rate": 0.0002695720729067301,
      "loss": 0.4105,
      "num_input_tokens_seen": 17118224,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.21608027815818787,
      "learning_rate": 0.00026936914417407756,
      "loss": 0.5263,
      "num_input_tokens_seen": 17122384,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.08966416865587234,
      "learning_rate": 0.0002691662636867687,
      "loss": 0.3596,
      "num_input_tokens_seen": 17125936,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 0.13964630663394928,
      "learning_rate": 0.000268963431487244,
      "loss": 0.5523,
      "num_input_tokens_seen": 17128976,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 0.0805930346250534,
      "learning_rate": 0.0002687606476179336,
      "loss": 0.443,
      "num_input_tokens_seen": 17132912,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.13144482672214508,
      "learning_rate": 0.0002685579121212578,
      "loss": 0.5324,
      "num_input_tokens_seen": 17136016,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 0.14566931128501892,
      "learning_rate": 0.00026835522503962616,
      "loss": 0.4102,
      "num_input_tokens_seen": 17139184,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.10982950776815414,
      "learning_rate": 0.00026815258641543896,
      "loss": 0.4259,
      "num_input_tokens_seen": 17142352,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.18736587464809418,
      "learning_rate": 0.0002679499962910853,
      "loss": 0.4399,
      "num_input_tokens_seen": 17145808,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 0.12029816955327988,
      "learning_rate": 0.0002677474547089451,
      "loss": 0.324,
      "num_input_tokens_seen": 17148720,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.353693425655365,
      "learning_rate": 0.000267544961711388,
      "loss": 0.5694,
      "num_input_tokens_seen": 17151792,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 0.153315469622612,
      "learning_rate": 0.0002673425173407726,
      "loss": 0.516,
      "num_input_tokens_seen": 17154896,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 0.14784283936023712,
      "learning_rate": 0.00026714012163944826,
      "loss": 0.4093,
      "num_input_tokens_seen": 17158352,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.161689892411232,
      "learning_rate": 0.0002669377746497541,
      "loss": 0.4066,
      "num_input_tokens_seen": 17161648,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 0.11562293022871017,
      "learning_rate": 0.0002667354764140184,
      "loss": 0.3856,
      "num_input_tokens_seen": 17164784,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 0.17383678257465363,
      "learning_rate": 0.00026653322697455987,
      "loss": 0.3866,
      "num_input_tokens_seen": 17167408,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.12049202620983124,
      "learning_rate": 0.00026633102637368684,
      "loss": 0.3582,
      "num_input_tokens_seen": 17170704,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 0.12888957560062408,
      "learning_rate": 0.00026612887465369704,
      "loss": 0.3416,
      "num_input_tokens_seen": 17173328,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 0.16710804402828217,
      "learning_rate": 0.0002659267718568786,
      "loss": 0.4875,
      "num_input_tokens_seen": 17176752,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 0.09010455012321472,
      "learning_rate": 0.0002657247180255091,
      "loss": 0.4549,
      "num_input_tokens_seen": 17179824,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.09091602265834808,
      "learning_rate": 0.0002655227132018558,
      "loss": 0.3731,
      "num_input_tokens_seen": 17185200,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.1482154279947281,
      "learning_rate": 0.00026532075742817594,
      "loss": 0.4021,
      "num_input_tokens_seen": 17189360,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.1580861359834671,
      "learning_rate": 0.0002651188507467161,
      "loss": 0.4144,
      "num_input_tokens_seen": 17192208,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 0.13489830493927002,
      "learning_rate": 0.00026491699319971303,
      "loss": 0.5534,
      "num_input_tokens_seen": 17196336,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 0.15058794617652893,
      "learning_rate": 0.00026471518482939317,
      "loss": 0.4093,
      "num_input_tokens_seen": 17199536,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.16746294498443604,
      "learning_rate": 0.0002645134256779722,
      "loss": 0.4169,
      "num_input_tokens_seen": 17202064,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 0.19053691625595093,
      "learning_rate": 0.00026431171578765624,
      "loss": 0.4786,
      "num_input_tokens_seen": 17205072,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 0.07907552272081375,
      "learning_rate": 0.0002641100552006402,
      "loss": 0.4481,
      "num_input_tokens_seen": 17208976,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 0.21316371858119965,
      "learning_rate": 0.0002639084439591095,
      "loss": 0.4244,
      "num_input_tokens_seen": 17212272,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.13881827890872955,
      "learning_rate": 0.00026370688210523873,
      "loss": 0.5671,
      "num_input_tokens_seen": 17215632,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.205917626619339,
      "learning_rate": 0.00026350536968119275,
      "loss": 0.4832,
      "num_input_tokens_seen": 17218992,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 0.18507814407348633,
      "learning_rate": 0.0002633039067291252,
      "loss": 0.5282,
      "num_input_tokens_seen": 17222480,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 0.1567348837852478,
      "learning_rate": 0.00026310249329118007,
      "loss": 0.4221,
      "num_input_tokens_seen": 17225424,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 0.13986627757549286,
      "learning_rate": 0.0002629011294094905,
      "loss": 0.5111,
      "num_input_tokens_seen": 17228784,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.10581115633249283,
      "learning_rate": 0.0002626998151261798,
      "loss": 0.4351,
      "num_input_tokens_seen": 17232144,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 0.13221071660518646,
      "learning_rate": 0.0002624985504833604,
      "loss": 0.3624,
      "num_input_tokens_seen": 17236560,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 0.14614836871623993,
      "learning_rate": 0.0002622973355231349,
      "loss": 0.4361,
      "num_input_tokens_seen": 17240784,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 0.13664628565311432,
      "learning_rate": 0.00026209617028759497,
      "loss": 0.5118,
      "num_input_tokens_seen": 17243856,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 0.18944944441318512,
      "learning_rate": 0.00026189505481882184,
      "loss": 0.5478,
      "num_input_tokens_seen": 17247536,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 0.14172767102718353,
      "learning_rate": 0.00026169398915888687,
      "loss": 0.4479,
      "num_input_tokens_seen": 17250832,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 0.09708503633737564,
      "learning_rate": 0.0002614929733498506,
      "loss": 0.3302,
      "num_input_tokens_seen": 17254256,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 0.13893654942512512,
      "learning_rate": 0.0002612920074337634,
      "loss": 0.3105,
      "num_input_tokens_seen": 17256976,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 0.14620596170425415,
      "learning_rate": 0.00026109109145266496,
      "loss": 0.4523,
      "num_input_tokens_seen": 17259664,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 0.18587306141853333,
      "learning_rate": 0.00026089022544858445,
      "loss": 0.5195,
      "num_input_tokens_seen": 17262736,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.16245903074741364,
      "learning_rate": 0.00026068940946354075,
      "loss": 0.492,
      "num_input_tokens_seen": 17265040,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 0.10312163084745407,
      "learning_rate": 0.0002604886435395425,
      "loss": 0.3667,
      "num_input_tokens_seen": 17268656,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 0.19737181067466736,
      "learning_rate": 0.00026028792771858744,
      "loss": 0.3571,
      "num_input_tokens_seen": 17272272,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 0.14059191942214966,
      "learning_rate": 0.00026008726204266333,
      "loss": 0.4686,
      "num_input_tokens_seen": 17275408,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 0.11925626546144485,
      "learning_rate": 0.00025988664655374693,
      "loss": 0.3434,
      "num_input_tokens_seen": 17278224,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 0.09976308047771454,
      "learning_rate": 0.00025968608129380455,
      "loss": 0.4071,
      "num_input_tokens_seen": 17281808,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.18677811324596405,
      "learning_rate": 0.00025948556630479234,
      "loss": 0.4825,
      "num_input_tokens_seen": 17285072,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.12111429870128632,
      "learning_rate": 0.0002592851016286557,
      "loss": 0.4175,
      "num_input_tokens_seen": 17287952,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 0.2505093514919281,
      "learning_rate": 0.0002590846873073298,
      "loss": 0.4012,
      "num_input_tokens_seen": 17290768,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 0.20942668616771698,
      "learning_rate": 0.0002588843233827387,
      "loss": 0.4462,
      "num_input_tokens_seen": 17294000,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 0.1084689125418663,
      "learning_rate": 0.0002586840098967963,
      "loss": 0.3816,
      "num_input_tokens_seen": 17297520,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.14337550103664398,
      "learning_rate": 0.00025848374689140587,
      "loss": 0.359,
      "num_input_tokens_seen": 17300816,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 0.08478417247533798,
      "learning_rate": 0.0002582835344084602,
      "loss": 0.5321,
      "num_input_tokens_seen": 17305232,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.16262172162532806,
      "learning_rate": 0.00025808337248984175,
      "loss": 0.344,
      "num_input_tokens_seen": 17308592,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.09085209667682648,
      "learning_rate": 0.00025788326117742185,
      "loss": 0.462,
      "num_input_tokens_seen": 17312144,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 0.1167442798614502,
      "learning_rate": 0.00025768320051306127,
      "loss": 0.4276,
      "num_input_tokens_seen": 17314960,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 0.1725235879421234,
      "learning_rate": 0.00025748319053861063,
      "loss": 0.4831,
      "num_input_tokens_seen": 17318512,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 0.09864541888237,
      "learning_rate": 0.0002572832312959098,
      "loss": 0.4286,
      "num_input_tokens_seen": 17320912,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 0.15613950788974762,
      "learning_rate": 0.0002570833228267879,
      "loss": 0.3355,
      "num_input_tokens_seen": 17323920,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.28002408146858215,
      "learning_rate": 0.00025688346517306366,
      "loss": 0.3765,
      "num_input_tokens_seen": 17327344,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 0.2524358630180359,
      "learning_rate": 0.00025668365837654497,
      "loss": 0.4709,
      "num_input_tokens_seen": 17330096,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 0.13347090780735016,
      "learning_rate": 0.0002564839024790288,
      "loss": 0.5458,
      "num_input_tokens_seen": 17332816,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 0.1070876345038414,
      "learning_rate": 0.000256284197522302,
      "loss": 0.4799,
      "num_input_tokens_seen": 17336432,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 0.11537092924118042,
      "learning_rate": 0.00025608454354814075,
      "loss": 0.5024,
      "num_input_tokens_seen": 17340432,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 0.10592243075370789,
      "learning_rate": 0.0002558849405983104,
      "loss": 0.3809,
      "num_input_tokens_seen": 17345232,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 0.10743961483240128,
      "learning_rate": 0.00025568538871456527,
      "loss": 0.3945,
      "num_input_tokens_seen": 17349520,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 0.13206630945205688,
      "learning_rate": 0.00025548588793864976,
      "loss": 0.3912,
      "num_input_tokens_seen": 17353040,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.1723301261663437,
      "learning_rate": 0.0002552864383122967,
      "loss": 0.6175,
      "num_input_tokens_seen": 17356048,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 0.1570001244544983,
      "learning_rate": 0.0002550870398772289,
      "loss": 0.4119,
      "num_input_tokens_seen": 17359344,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.14863133430480957,
      "learning_rate": 0.0002548876926751584,
      "loss": 0.5819,
      "num_input_tokens_seen": 17363408,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 0.11569096893072128,
      "learning_rate": 0.0002546883967477861,
      "loss": 0.3801,
      "num_input_tokens_seen": 17366672,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 0.33118224143981934,
      "learning_rate": 0.00025448915213680245,
      "loss": 0.308,
      "num_input_tokens_seen": 17368592,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.2011452317237854,
      "learning_rate": 0.0002542899588838875,
      "loss": 0.4188,
      "num_input_tokens_seen": 17371152,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 0.31822869181632996,
      "learning_rate": 0.0002540908170307097,
      "loss": 0.4327,
      "num_input_tokens_seen": 17373872,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 0.11219184100627899,
      "learning_rate": 0.00025389172661892753,
      "loss": 0.392,
      "num_input_tokens_seen": 17377200,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 0.17270103096961975,
      "learning_rate": 0.00025369268769018856,
      "loss": 0.4872,
      "num_input_tokens_seen": 17380400,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 0.07763908803462982,
      "learning_rate": 0.00025349370028612914,
      "loss": 0.4194,
      "num_input_tokens_seen": 17383792,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 0.1691770851612091,
      "learning_rate": 0.0002532947644483755,
      "loss": 0.4433,
      "num_input_tokens_seen": 17386352,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 0.38750511407852173,
      "learning_rate": 0.00025309588021854237,
      "loss": 0.5065,
      "num_input_tokens_seen": 17388976,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 0.1974402517080307,
      "learning_rate": 0.0002528970476382343,
      "loss": 0.5132,
      "num_input_tokens_seen": 17393104,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.22038167715072632,
      "learning_rate": 0.00025269826674904493,
      "loss": 0.382,
      "num_input_tokens_seen": 17395728,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.10166282206773758,
      "learning_rate": 0.0002524995375925566,
      "loss": 0.4187,
      "num_input_tokens_seen": 17399024,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 0.10407013446092606,
      "learning_rate": 0.0002523008602103414,
      "loss": 0.4833,
      "num_input_tokens_seen": 17403312,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 0.14735619723796844,
      "learning_rate": 0.00025210223464396055,
      "loss": 0.3836,
      "num_input_tokens_seen": 17406320,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.16186101734638214,
      "learning_rate": 0.00025190366093496376,
      "loss": 0.4067,
      "num_input_tokens_seen": 17409040,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.17884381115436554,
      "learning_rate": 0.0002517051391248909,
      "loss": 0.5066,
      "num_input_tokens_seen": 17411824,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 0.18150396645069122,
      "learning_rate": 0.0002515066692552701,
      "loss": 0.4644,
      "num_input_tokens_seen": 17415472,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 0.17621463537216187,
      "learning_rate": 0.00025130825136761916,
      "loss": 0.4866,
      "num_input_tokens_seen": 17418416,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.28957706689834595,
      "learning_rate": 0.0002511098855034447,
      "loss": 0.4551,
      "num_input_tokens_seen": 17421904,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 0.2201811820268631,
      "learning_rate": 0.00025091157170424296,
      "loss": 0.4754,
      "num_input_tokens_seen": 17424912,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 0.24049495160579681,
      "learning_rate": 0.0002507133100114984,
      "loss": 0.4721,
      "num_input_tokens_seen": 17428496,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.2779186964035034,
      "learning_rate": 0.0002505151004666857,
      "loss": 0.5442,
      "num_input_tokens_seen": 17431920,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 0.14651858806610107,
      "learning_rate": 0.0002503169431112674,
      "loss": 0.3706,
      "num_input_tokens_seen": 17435152,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 0.15808098018169403,
      "learning_rate": 0.0002501188379866961,
      "loss": 0.4008,
      "num_input_tokens_seen": 17437680,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.488889217376709,
      "eval_runtime": 13.7263,
      "eval_samples_per_second": 61.779,
      "eval_steps_per_second": 15.445,
      "num_input_tokens_seen": 17439088,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.1564636528491974,
      "learning_rate": 0.0002499207851344133,
      "loss": 0.3368,
      "num_input_tokens_seen": 17441328,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 0.10542064160108566,
      "learning_rate": 0.00024972278459584903,
      "loss": 0.4754,
      "num_input_tokens_seen": 17444016,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.191155344247818,
      "learning_rate": 0.00024952483641242304,
      "loss": 0.3599,
      "num_input_tokens_seen": 17447088,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.10253865271806717,
      "learning_rate": 0.0002493269406255435,
      "loss": 0.5022,
      "num_input_tokens_seen": 17451280,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.15954901278018951,
      "learning_rate": 0.0002491290972766082,
      "loss": 0.5289,
      "num_input_tokens_seen": 17455792,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 0.11711404472589493,
      "learning_rate": 0.0002489313064070037,
      "loss": 0.3316,
      "num_input_tokens_seen": 17458896,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.16093645989894867,
      "learning_rate": 0.00024873356805810566,
      "loss": 0.4803,
      "num_input_tokens_seen": 17461904,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.18125955760478973,
      "learning_rate": 0.00024853588227127864,
      "loss": 0.3655,
      "num_input_tokens_seen": 17465424,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 0.10234969109296799,
      "learning_rate": 0.000248338249087876,
      "loss": 0.4328,
      "num_input_tokens_seen": 17468784,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.15121294558048248,
      "learning_rate": 0.0002481406685492405,
      "loss": 0.3793,
      "num_input_tokens_seen": 17471632,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 0.13198646903038025,
      "learning_rate": 0.00024794314069670383,
      "loss": 0.3581,
      "num_input_tokens_seen": 17475024,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.13788388669490814,
      "learning_rate": 0.0002477456655715865,
      "loss": 0.4833,
      "num_input_tokens_seen": 17477904,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 0.12315964698791504,
      "learning_rate": 0.0002475482432151982,
      "loss": 0.3585,
      "num_input_tokens_seen": 17480464,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 0.1271524727344513,
      "learning_rate": 0.00024735087366883733,
      "loss": 0.4091,
      "num_input_tokens_seen": 17483440,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 0.11901547014713287,
      "learning_rate": 0.00024715355697379115,
      "loss": 0.4513,
      "num_input_tokens_seen": 17485968,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 0.1393316686153412,
      "learning_rate": 0.0002469562931713362,
      "loss": 0.5101,
      "num_input_tokens_seen": 17489104,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.11792337149381638,
      "learning_rate": 0.00024675908230273785,
      "loss": 0.4733,
      "num_input_tokens_seen": 17493264,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.12920229136943817,
      "learning_rate": 0.00024656192440925055,
      "loss": 0.4732,
      "num_input_tokens_seen": 17496880,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 0.17376717925071716,
      "learning_rate": 0.0002463648195321173,
      "loss": 0.3343,
      "num_input_tokens_seen": 17498992,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 0.11475850641727448,
      "learning_rate": 0.00024616776771257,
      "loss": 0.314,
      "num_input_tokens_seen": 17505136,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 0.09057436138391495,
      "learning_rate": 0.00024597076899182977,
      "loss": 0.3297,
      "num_input_tokens_seen": 17509200,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.15323248505592346,
      "learning_rate": 0.0002457738234111066,
      "loss": 0.5492,
      "num_input_tokens_seen": 17512080,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.16008801758289337,
      "learning_rate": 0.00024557693101159937,
      "loss": 0.386,
      "num_input_tokens_seen": 17515152,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 0.13003186881542206,
      "learning_rate": 0.00024538009183449553,
      "loss": 0.494,
      "num_input_tokens_seen": 17518576,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.1364624947309494,
      "learning_rate": 0.0002451833059209715,
      "loss": 0.3923,
      "num_input_tokens_seen": 17521264,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 0.11017634719610214,
      "learning_rate": 0.00024498657331219274,
      "loss": 0.3794,
      "num_input_tokens_seen": 17524176,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 0.15935225784778595,
      "learning_rate": 0.0002447898940493135,
      "loss": 0.4116,
      "num_input_tokens_seen": 17527600,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.19248850643634796,
      "learning_rate": 0.00024459326817347684,
      "loss": 0.4074,
      "num_input_tokens_seen": 17529904,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 0.22912484407424927,
      "learning_rate": 0.0002443966957258148,
      "loss": 0.5548,
      "num_input_tokens_seen": 17532624,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 0.1515667587518692,
      "learning_rate": 0.000244200176747448,
      "loss": 0.3473,
      "num_input_tokens_seen": 17536432,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 0.12080492824316025,
      "learning_rate": 0.00024400371127948568,
      "loss": 0.3697,
      "num_input_tokens_seen": 17539184,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 0.09333871304988861,
      "learning_rate": 0.00024380729936302636,
      "loss": 0.4115,
      "num_input_tokens_seen": 17542288,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 0.1498965471982956,
      "learning_rate": 0.00024361094103915725,
      "loss": 0.3676,
      "num_input_tokens_seen": 17545200,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 0.18739116191864014,
      "learning_rate": 0.00024341463634895444,
      "loss": 0.4023,
      "num_input_tokens_seen": 17549328,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.14451797306537628,
      "learning_rate": 0.0002432183853334824,
      "loss": 0.4667,
      "num_input_tokens_seen": 17552496,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 0.11547960340976715,
      "learning_rate": 0.00024302218803379445,
      "loss": 0.4917,
      "num_input_tokens_seen": 17555600,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.11331246793270111,
      "learning_rate": 0.000242826044490933,
      "loss": 0.4314,
      "num_input_tokens_seen": 17558928,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 0.2094225138425827,
      "learning_rate": 0.00024262995474592903,
      "loss": 0.4234,
      "num_input_tokens_seen": 17561456,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 0.10780687630176544,
      "learning_rate": 0.0002424339188398025,
      "loss": 0.361,
      "num_input_tokens_seen": 17565072,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 0.12328751385211945,
      "learning_rate": 0.00024223793681356148,
      "loss": 0.4404,
      "num_input_tokens_seen": 17567920,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 0.17985819280147552,
      "learning_rate": 0.00024204200870820358,
      "loss": 0.4768,
      "num_input_tokens_seen": 17570384,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.2127649188041687,
      "learning_rate": 0.0002418461345647143,
      "loss": 0.4258,
      "num_input_tokens_seen": 17573616,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 0.14864397048950195,
      "learning_rate": 0.00024165031442406853,
      "loss": 0.3882,
      "num_input_tokens_seen": 17578128,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 0.1454264521598816,
      "learning_rate": 0.00024145454832722956,
      "loss": 0.5445,
      "num_input_tokens_seen": 17580944,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 0.11020299792289734,
      "learning_rate": 0.00024125883631514967,
      "loss": 0.3367,
      "num_input_tokens_seen": 17584208,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.13457120954990387,
      "learning_rate": 0.00024106317842876907,
      "loss": 0.4987,
      "num_input_tokens_seen": 17588112,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.07485706359148026,
      "learning_rate": 0.0002408675747090177,
      "loss": 0.3558,
      "num_input_tokens_seen": 17591216,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 0.09998713433742523,
      "learning_rate": 0.00024067202519681313,
      "loss": 0.4596,
      "num_input_tokens_seen": 17594352,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 0.15229125320911407,
      "learning_rate": 0.00024047652993306235,
      "loss": 0.3296,
      "num_input_tokens_seen": 17596976,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 0.13969501852989197,
      "learning_rate": 0.00024028108895866084,
      "loss": 0.4031,
      "num_input_tokens_seen": 17599824,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.12184711545705795,
      "learning_rate": 0.00024008570231449239,
      "loss": 0.3583,
      "num_input_tokens_seen": 17603312,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.3458669185638428,
      "learning_rate": 0.00023989037004142966,
      "loss": 0.4526,
      "num_input_tokens_seen": 17605712,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 0.1183592826128006,
      "learning_rate": 0.0002396950921803343,
      "loss": 0.6139,
      "num_input_tokens_seen": 17609424,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.14461903274059296,
      "learning_rate": 0.0002394998687720557,
      "loss": 0.523,
      "num_input_tokens_seen": 17611952,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 0.18524861335754395,
      "learning_rate": 0.0002393046998574328,
      "loss": 0.4089,
      "num_input_tokens_seen": 17614960,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 0.16333763301372528,
      "learning_rate": 0.0002391095854772924,
      "loss": 0.3611,
      "num_input_tokens_seen": 17619152,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 0.0932527706027031,
      "learning_rate": 0.00023891452567245027,
      "loss": 0.4591,
      "num_input_tokens_seen": 17623024,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.11673516780138016,
      "learning_rate": 0.00023871952048371093,
      "loss": 0.3109,
      "num_input_tokens_seen": 17626736,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 0.2036648839712143,
      "learning_rate": 0.00023852456995186684,
      "loss": 0.4492,
      "num_input_tokens_seen": 17629808,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 0.11180226504802704,
      "learning_rate": 0.0002383296741176997,
      "loss": 0.5346,
      "num_input_tokens_seen": 17633168,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.11408074200153351,
      "learning_rate": 0.00023813483302197964,
      "loss": 0.5687,
      "num_input_tokens_seen": 17636560,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 0.13203641772270203,
      "learning_rate": 0.00023794004670546476,
      "loss": 0.4554,
      "num_input_tokens_seen": 17640784,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 0.11694186180830002,
      "learning_rate": 0.00023774531520890242,
      "loss": 0.3503,
      "num_input_tokens_seen": 17643984,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.12614783644676208,
      "learning_rate": 0.00023755063857302833,
      "loss": 0.4586,
      "num_input_tokens_seen": 17647184,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.17920273542404175,
      "learning_rate": 0.00023735601683856628,
      "loss": 0.3703,
      "num_input_tokens_seen": 17649808,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 0.21273887157440186,
      "learning_rate": 0.0002371614500462293,
      "loss": 0.4269,
      "num_input_tokens_seen": 17652112,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.3443225622177124,
      "learning_rate": 0.00023696693823671816,
      "loss": 0.4319,
      "num_input_tokens_seen": 17655312,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.13490329682826996,
      "learning_rate": 0.00023677248145072272,
      "loss": 0.4777,
      "num_input_tokens_seen": 17659792,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 0.12579795718193054,
      "learning_rate": 0.0002365780797289211,
      "loss": 0.3542,
      "num_input_tokens_seen": 17663056,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.39929625391960144,
      "learning_rate": 0.00023638373311198014,
      "loss": 0.5029,
      "num_input_tokens_seen": 17665904,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.10747040063142776,
      "learning_rate": 0.00023618944164055468,
      "loss": 0.3732,
      "num_input_tokens_seen": 17668912,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 0.22884757816791534,
      "learning_rate": 0.00023599520535528813,
      "loss": 0.3582,
      "num_input_tokens_seen": 17672080,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.13235804438591003,
      "learning_rate": 0.00023580102429681266,
      "loss": 0.482,
      "num_input_tokens_seen": 17674736,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.2006685882806778,
      "learning_rate": 0.00023560689850574885,
      "loss": 0.3552,
      "num_input_tokens_seen": 17678384,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 0.1380346715450287,
      "learning_rate": 0.00023541282802270563,
      "loss": 0.4382,
      "num_input_tokens_seen": 17682128,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 0.19595816731452942,
      "learning_rate": 0.00023521881288828005,
      "loss": 0.5291,
      "num_input_tokens_seen": 17684944,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 0.0891830325126648,
      "learning_rate": 0.00023502485314305828,
      "loss": 0.4457,
      "num_input_tokens_seen": 17687728,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.14359009265899658,
      "learning_rate": 0.00023483094882761397,
      "loss": 0.3825,
      "num_input_tokens_seen": 17690608,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 0.11676640063524246,
      "learning_rate": 0.00023463709998250998,
      "loss": 0.5177,
      "num_input_tokens_seen": 17694512,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.3716685473918915,
      "learning_rate": 0.00023444330664829728,
      "loss": 0.3662,
      "num_input_tokens_seen": 17697424,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 0.20817455649375916,
      "learning_rate": 0.00023424956886551536,
      "loss": 0.3149,
      "num_input_tokens_seen": 17700688,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 0.09431692212820053,
      "learning_rate": 0.0002340558866746918,
      "loss": 0.3947,
      "num_input_tokens_seen": 17704336,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 0.24387235939502716,
      "learning_rate": 0.0002338622601163425,
      "loss": 0.3388,
      "num_input_tokens_seen": 17706896,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 0.13346652686595917,
      "learning_rate": 0.00023366868923097213,
      "loss": 0.3415,
      "num_input_tokens_seen": 17710320,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 0.14550332725048065,
      "learning_rate": 0.0002334751740590735,
      "loss": 0.3883,
      "num_input_tokens_seen": 17713296,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 0.2365768998861313,
      "learning_rate": 0.00023328171464112796,
      "loss": 0.3164,
      "num_input_tokens_seen": 17716912,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 0.20141363143920898,
      "learning_rate": 0.00023308831101760485,
      "loss": 0.481,
      "num_input_tokens_seen": 17719632,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 0.12601575255393982,
      "learning_rate": 0.00023289496322896174,
      "loss": 0.4385,
      "num_input_tokens_seen": 17722960,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.20402869582176208,
      "learning_rate": 0.00023270167131564512,
      "loss": 0.4879,
      "num_input_tokens_seen": 17726288,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 0.16082903742790222,
      "learning_rate": 0.0002325084353180893,
      "loss": 0.3801,
      "num_input_tokens_seen": 17729488,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 0.14020103216171265,
      "learning_rate": 0.00023231525527671716,
      "loss": 0.4758,
      "num_input_tokens_seen": 17732496,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 0.05015021562576294,
      "learning_rate": 0.00023212213123193986,
      "loss": 0.333,
      "num_input_tokens_seen": 17739088,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 0.17222824692726135,
      "learning_rate": 0.0002319290632241566,
      "loss": 0.4062,
      "num_input_tokens_seen": 17742224,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 0.1910252422094345,
      "learning_rate": 0.0002317360512937548,
      "loss": 0.4419,
      "num_input_tokens_seen": 17746800,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.1940694898366928,
      "learning_rate": 0.00023154309548111058,
      "loss": 0.4865,
      "num_input_tokens_seen": 17749936,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.10339370369911194,
      "learning_rate": 0.000231350195826588,
      "loss": 0.4761,
      "num_input_tokens_seen": 17752848,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.11830491572618484,
      "learning_rate": 0.00023115735237053982,
      "loss": 0.3616,
      "num_input_tokens_seen": 17756592,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 0.1849275827407837,
      "learning_rate": 0.00023096456515330632,
      "loss": 0.5391,
      "num_input_tokens_seen": 17759728,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 0.307342529296875,
      "learning_rate": 0.0002307718342152163,
      "loss": 0.4588,
      "num_input_tokens_seen": 17762320,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 0.18434587121009827,
      "learning_rate": 0.0002305791595965871,
      "loss": 0.3975,
      "num_input_tokens_seen": 17765424,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.23326030373573303,
      "learning_rate": 0.00023038654133772395,
      "loss": 0.4465,
      "num_input_tokens_seen": 17768528,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 0.15557296574115753,
      "learning_rate": 0.00023019397947892062,
      "loss": 0.3198,
      "num_input_tokens_seen": 17771184,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 0.12787146866321564,
      "learning_rate": 0.00023000147406045853,
      "loss": 0.306,
      "num_input_tokens_seen": 17776368,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.21954697370529175,
      "learning_rate": 0.00022980902512260786,
      "loss": 0.4554,
      "num_input_tokens_seen": 17779024,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 0.20192478597164154,
      "learning_rate": 0.00022961663270562656,
      "loss": 0.4261,
      "num_input_tokens_seen": 17782544,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 0.2476818561553955,
      "learning_rate": 0.0002294242968497609,
      "loss": 0.3274,
      "num_input_tokens_seen": 17785328,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 0.1221131980419159,
      "learning_rate": 0.00022923201759524552,
      "loss": 0.3651,
      "num_input_tokens_seen": 17788176,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 0.1694859266281128,
      "learning_rate": 0.00022903979498230316,
      "loss": 0.4101,
      "num_input_tokens_seen": 17791504,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 0.05697573721408844,
      "learning_rate": 0.00022884762905114436,
      "loss": 0.4288,
      "num_input_tokens_seen": 17796304,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 0.13949672877788544,
      "learning_rate": 0.00022865551984196792,
      "loss": 0.3499,
      "num_input_tokens_seen": 17798928,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 0.0936320498585701,
      "learning_rate": 0.0002284634673949611,
      "loss": 0.4843,
      "num_input_tokens_seen": 17802480,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 0.07902836054563522,
      "learning_rate": 0.00022827147175029906,
      "loss": 0.4198,
      "num_input_tokens_seen": 17805680,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 0.18390299379825592,
      "learning_rate": 0.0002280795329481452,
      "loss": 0.4185,
      "num_input_tokens_seen": 17808336,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 0.20007550716400146,
      "learning_rate": 0.00022788765102865066,
      "loss": 0.3922,
      "num_input_tokens_seen": 17811280,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 0.08136424422264099,
      "learning_rate": 0.00022769582603195533,
      "loss": 0.4046,
      "num_input_tokens_seen": 17815120,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 0.19863125681877136,
      "learning_rate": 0.00022750405799818634,
      "loss": 0.3822,
      "num_input_tokens_seen": 17818608,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 0.11398672312498093,
      "learning_rate": 0.00022731234696745967,
      "loss": 0.4243,
      "num_input_tokens_seen": 17822640,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.3765931725502014,
      "learning_rate": 0.0002271206929798792,
      "loss": 0.5681,
      "num_input_tokens_seen": 17825712,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 0.22924403846263885,
      "learning_rate": 0.00022692909607553642,
      "loss": 0.5637,
      "num_input_tokens_seen": 17828624,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 0.1393180936574936,
      "learning_rate": 0.00022673755629451132,
      "loss": 0.5243,
      "num_input_tokens_seen": 17831696,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 0.1513340175151825,
      "learning_rate": 0.00022654607367687213,
      "loss": 0.2973,
      "num_input_tokens_seen": 17835280,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 0.17796440422534943,
      "learning_rate": 0.00022635464826267442,
      "loss": 0.4992,
      "num_input_tokens_seen": 17839472,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 0.10307279229164124,
      "learning_rate": 0.00022616328009196236,
      "loss": 0.3817,
      "num_input_tokens_seen": 17843152,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 0.1319958120584488,
      "learning_rate": 0.0002259719692047682,
      "loss": 0.5328,
      "num_input_tokens_seen": 17846640,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 0.1577751636505127,
      "learning_rate": 0.00022578071564111163,
      "loss": 0.3739,
      "num_input_tokens_seen": 17849776,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.14624004065990448,
      "learning_rate": 0.00022558951944100087,
      "loss": 0.4772,
      "num_input_tokens_seen": 17853488,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 0.270740807056427,
      "learning_rate": 0.00022539838064443223,
      "loss": 0.3493,
      "num_input_tokens_seen": 17855984,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 0.11709238588809967,
      "learning_rate": 0.00022520729929138933,
      "loss": 0.4171,
      "num_input_tokens_seen": 17858992,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.13249555230140686,
      "learning_rate": 0.0002250162754218446,
      "loss": 0.3724,
      "num_input_tokens_seen": 17862352,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 0.10370559990406036,
      "learning_rate": 0.00022482530907575767,
      "loss": 0.3881,
      "num_input_tokens_seen": 17865872,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 0.09453070163726807,
      "learning_rate": 0.00022463440029307674,
      "loss": 0.3507,
      "num_input_tokens_seen": 17868880,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 0.13627010583877563,
      "learning_rate": 0.00022444354911373794,
      "loss": 0.4915,
      "num_input_tokens_seen": 17872272,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 0.12430219352245331,
      "learning_rate": 0.00022425275557766473,
      "loss": 0.4061,
      "num_input_tokens_seen": 17875376,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 0.14417964220046997,
      "learning_rate": 0.00022406201972476937,
      "loss": 0.4094,
      "num_input_tokens_seen": 17879408,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.1492076814174652,
      "learning_rate": 0.00022387134159495122,
      "loss": 0.4041,
      "num_input_tokens_seen": 17883120,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 0.09617063403129578,
      "learning_rate": 0.00022368072122809813,
      "loss": 0.4827,
      "num_input_tokens_seen": 17886256,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 0.21188020706176758,
      "learning_rate": 0.0002234901586640858,
      "loss": 0.4087,
      "num_input_tokens_seen": 17890160,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.2769501507282257,
      "learning_rate": 0.00022329965394277785,
      "loss": 0.5357,
      "num_input_tokens_seen": 17893904,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.20169416069984436,
      "learning_rate": 0.00022310920710402532,
      "loss": 0.4048,
      "num_input_tokens_seen": 17896400,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 0.1811653971672058,
      "learning_rate": 0.00022291881818766796,
      "loss": 0.329,
      "num_input_tokens_seen": 17903056,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 0.1653462052345276,
      "learning_rate": 0.00022272848723353252,
      "loss": 0.583,
      "num_input_tokens_seen": 17906032,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.1789606809616089,
      "learning_rate": 0.00022253821428143422,
      "loss": 0.4427,
      "num_input_tokens_seen": 17909040,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 0.19533109664916992,
      "learning_rate": 0.0002223479993711761,
      "loss": 0.4625,
      "num_input_tokens_seen": 17911920,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 0.16902145743370056,
      "learning_rate": 0.00022215784254254906,
      "loss": 0.4982,
      "num_input_tokens_seen": 17914608,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 0.2382948249578476,
      "learning_rate": 0.00022196774383533157,
      "loss": 0.3412,
      "num_input_tokens_seen": 17918352,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 0.17278815805912018,
      "learning_rate": 0.0002217777032892899,
      "loss": 0.3611,
      "num_input_tokens_seen": 17922512,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.14746877551078796,
      "learning_rate": 0.00022158772094417863,
      "loss": 0.5707,
      "num_input_tokens_seen": 17925904,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.38156646490097046,
      "learning_rate": 0.00022139779683973983,
      "loss": 0.5187,
      "num_input_tokens_seen": 17928400,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 0.1525062769651413,
      "learning_rate": 0.00022120793101570364,
      "loss": 0.3943,
      "num_input_tokens_seen": 17932080,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 0.1600160151720047,
      "learning_rate": 0.00022101812351178762,
      "loss": 0.6396,
      "num_input_tokens_seen": 17936016,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 0.1313450187444687,
      "learning_rate": 0.00022082837436769725,
      "loss": 0.5971,
      "num_input_tokens_seen": 17939664,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 0.2165328413248062,
      "learning_rate": 0.00022063868362312596,
      "loss": 0.3858,
      "num_input_tokens_seen": 17942448,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 0.20564262568950653,
      "learning_rate": 0.00022044905131775495,
      "loss": 0.5047,
      "num_input_tokens_seen": 17945456,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.11599201709032059,
      "learning_rate": 0.00022025947749125313,
      "loss": 0.3964,
      "num_input_tokens_seen": 17949488,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.16602084040641785,
      "learning_rate": 0.0002200699621832773,
      "loss": 0.3688,
      "num_input_tokens_seen": 17952592,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 0.11058059334754944,
      "learning_rate": 0.0002198805054334718,
      "loss": 0.5905,
      "num_input_tokens_seen": 17955696,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 0.08624603599309921,
      "learning_rate": 0.00021969110728146862,
      "loss": 0.5963,
      "num_input_tokens_seen": 17959440,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 0.15664665400981903,
      "learning_rate": 0.00021950176776688784,
      "loss": 0.3618,
      "num_input_tokens_seen": 17962320,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 0.4980505406856537,
      "learning_rate": 0.0002193124869293372,
      "loss": 0.4749,
      "num_input_tokens_seen": 17965360,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 0.2653607726097107,
      "learning_rate": 0.00021912326480841223,
      "loss": 0.3914,
      "num_input_tokens_seen": 17968112,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.12968586385250092,
      "learning_rate": 0.0002189341014436958,
      "loss": 0.502,
      "num_input_tokens_seen": 17971344,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.15143609046936035,
      "learning_rate": 0.00021874499687475857,
      "loss": 0.3485,
      "num_input_tokens_seen": 17974480,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 0.10920349508523941,
      "learning_rate": 0.00021855595114115935,
      "loss": 0.5049,
      "num_input_tokens_seen": 17978128,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 0.13654837012290955,
      "learning_rate": 0.00021836696428244418,
      "loss": 0.3497,
      "num_input_tokens_seen": 17981808,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 0.21310020983219147,
      "learning_rate": 0.0002181780363381473,
      "loss": 0.4084,
      "num_input_tokens_seen": 17984624,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.1250733733177185,
      "learning_rate": 0.00021798916734779,
      "loss": 0.3961,
      "num_input_tokens_seen": 17987344,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.10711005330085754,
      "learning_rate": 0.00021780035735088132,
      "loss": 0.6958,
      "num_input_tokens_seen": 17990928,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 0.1110001653432846,
      "learning_rate": 0.0002176116063869184,
      "loss": 0.4041,
      "num_input_tokens_seen": 17994384,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.18912705779075623,
      "learning_rate": 0.00021742291449538582,
      "loss": 0.4656,
      "num_input_tokens_seen": 17996944,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.2776673436164856,
      "learning_rate": 0.00021723428171575566,
      "loss": 0.4641,
      "num_input_tokens_seen": 18000464,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 0.10408354550600052,
      "learning_rate": 0.00021704570808748802,
      "loss": 0.3922,
      "num_input_tokens_seen": 18003664,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 0.24925945699214935,
      "learning_rate": 0.00021685719365003008,
      "loss": 0.4248,
      "num_input_tokens_seen": 18006480,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 0.21046343445777893,
      "learning_rate": 0.00021666873844281682,
      "loss": 0.5848,
      "num_input_tokens_seen": 18009424,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 0.08263245970010757,
      "learning_rate": 0.0002164803425052711,
      "loss": 0.3376,
      "num_input_tokens_seen": 18012816,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 0.09733432531356812,
      "learning_rate": 0.0002162920058768031,
      "loss": 0.6486,
      "num_input_tokens_seen": 18017808,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 0.25160452723503113,
      "learning_rate": 0.000216103728596811,
      "loss": 0.3762,
      "num_input_tokens_seen": 18021168,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 0.188520610332489,
      "learning_rate": 0.00021591551070467975,
      "loss": 0.5142,
      "num_input_tokens_seen": 18023920,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.27962249517440796,
      "learning_rate": 0.0002157273522397829,
      "loss": 0.3676,
      "num_input_tokens_seen": 18027248,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 0.2957867980003357,
      "learning_rate": 0.0002155392532414806,
      "loss": 0.46,
      "num_input_tokens_seen": 18030096,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 0.11264652013778687,
      "learning_rate": 0.00021535121374912126,
      "loss": 0.3913,
      "num_input_tokens_seen": 18033584,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 0.14433377981185913,
      "learning_rate": 0.00021516323380204073,
      "loss": 0.4215,
      "num_input_tokens_seen": 18037936,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 0.23771370947360992,
      "learning_rate": 0.00021497531343956188,
      "loss": 0.5958,
      "num_input_tokens_seen": 18041136,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 0.23261646926403046,
      "learning_rate": 0.00021478745270099591,
      "loss": 0.3933,
      "num_input_tokens_seen": 18044368,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.1464540660381317,
      "learning_rate": 0.0002145996516256408,
      "loss": 0.5094,
      "num_input_tokens_seen": 18046832,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 0.0909845158457756,
      "learning_rate": 0.00021441191025278256,
      "loss": 0.4683,
      "num_input_tokens_seen": 18050096,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 0.13754813373088837,
      "learning_rate": 0.00021422422862169455,
      "loss": 0.4459,
      "num_input_tokens_seen": 18052528,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 0.1408250778913498,
      "learning_rate": 0.00021403660677163773,
      "loss": 0.3308,
      "num_input_tokens_seen": 18055536,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.10626823455095291,
      "learning_rate": 0.00021384904474186018,
      "loss": 0.4092,
      "num_input_tokens_seen": 18059152,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 0.16991300880908966,
      "learning_rate": 0.00021366154257159808,
      "loss": 0.569,
      "num_input_tokens_seen": 18061936,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 0.18192531168460846,
      "learning_rate": 0.00021347410030007435,
      "loss": 0.4286,
      "num_input_tokens_seen": 18065264,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 0.14497049152851105,
      "learning_rate": 0.0002132867179665,
      "loss": 0.4508,
      "num_input_tokens_seen": 18068336,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.24666301906108856,
      "learning_rate": 0.00021309939561007341,
      "loss": 0.4374,
      "num_input_tokens_seen": 18071600,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 0.19118370115756989,
      "learning_rate": 0.00021291213326997998,
      "loss": 0.4274,
      "num_input_tokens_seen": 18075472,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 0.2325969785451889,
      "learning_rate": 0.00021272493098539296,
      "loss": 0.3908,
      "num_input_tokens_seen": 18078992,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 0.11397197097539902,
      "learning_rate": 0.0002125377887954732,
      "loss": 0.514,
      "num_input_tokens_seen": 18081872,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.14545838534832,
      "learning_rate": 0.00021235070673936824,
      "loss": 0.3714,
      "num_input_tokens_seen": 18084944,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.2076072245836258,
      "learning_rate": 0.00021216368485621394,
      "loss": 0.5238,
      "num_input_tokens_seen": 18088016,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 0.2379673272371292,
      "learning_rate": 0.00021197672318513282,
      "loss": 0.5574,
      "num_input_tokens_seen": 18091216,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 0.20488405227661133,
      "learning_rate": 0.00021178982176523525,
      "loss": 0.4645,
      "num_input_tokens_seen": 18094736,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 0.24736326932907104,
      "learning_rate": 0.0002116029806356189,
      "loss": 0.4659,
      "num_input_tokens_seen": 18097392,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 0.22711828351020813,
      "learning_rate": 0.00021141619983536893,
      "loss": 0.4246,
      "num_input_tokens_seen": 18099888,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.24760641157627106,
      "learning_rate": 0.00021122947940355747,
      "loss": 0.3949,
      "num_input_tokens_seen": 18103248,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 0.22834165394306183,
      "learning_rate": 0.00021104281937924462,
      "loss": 0.4464,
      "num_input_tokens_seen": 18105808,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 0.15913733839988708,
      "learning_rate": 0.00021085621980147716,
      "loss": 0.4186,
      "num_input_tokens_seen": 18108880,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 0.16295987367630005,
      "learning_rate": 0.00021066968070928982,
      "loss": 0.4191,
      "num_input_tokens_seen": 18112496,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 0.2197885662317276,
      "learning_rate": 0.00021048320214170463,
      "loss": 0.3526,
      "num_input_tokens_seen": 18116272,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 0.18454858660697937,
      "learning_rate": 0.00021029678413773034,
      "loss": 0.3931,
      "num_input_tokens_seen": 18119696,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 0.169926717877388,
      "learning_rate": 0.0002101104267363639,
      "loss": 0.4052,
      "num_input_tokens_seen": 18124464,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 0.1670735627412796,
      "learning_rate": 0.00020992412997658877,
      "loss": 0.355,
      "num_input_tokens_seen": 18127472,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 0.18545563519001007,
      "learning_rate": 0.0002097378938973763,
      "loss": 0.4528,
      "num_input_tokens_seen": 18130224,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.21481943130493164,
      "learning_rate": 0.0002095517185376849,
      "loss": 0.3677,
      "num_input_tokens_seen": 18132976,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.21011567115783691,
      "learning_rate": 0.0002093656039364606,
      "loss": 0.4321,
      "num_input_tokens_seen": 18136336,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.13547366857528687,
      "learning_rate": 0.00020917955013263618,
      "loss": 0.3491,
      "num_input_tokens_seen": 18139056,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 0.15864308178424835,
      "learning_rate": 0.00020899355716513186,
      "loss": 0.4062,
      "num_input_tokens_seen": 18141552,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.24023281037807465,
      "learning_rate": 0.00020880762507285544,
      "loss": 0.5064,
      "num_input_tokens_seen": 18144496,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.26287850737571716,
      "learning_rate": 0.00020862175389470172,
      "loss": 0.3658,
      "num_input_tokens_seen": 18147792,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.15072685480117798,
      "learning_rate": 0.00020843594366955288,
      "loss": 0.5015,
      "num_input_tokens_seen": 18150864,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 0.1869698017835617,
      "learning_rate": 0.0002082501944362784,
      "loss": 0.3773,
      "num_input_tokens_seen": 18154192,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 0.19021156430244446,
      "learning_rate": 0.00020806450623373486,
      "loss": 0.4562,
      "num_input_tokens_seen": 18157616,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.15905673801898956,
      "learning_rate": 0.00020787887910076586,
      "loss": 0.3914,
      "num_input_tokens_seen": 18160624,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.09920423477888107,
      "learning_rate": 0.00020769331307620258,
      "loss": 0.5241,
      "num_input_tokens_seen": 18164048,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 0.12872445583343506,
      "learning_rate": 0.0002075078081988635,
      "loss": 0.5825,
      "num_input_tokens_seen": 18167408,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 0.13566003739833832,
      "learning_rate": 0.00020732236450755415,
      "loss": 0.3209,
      "num_input_tokens_seen": 18170608,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 0.31402266025543213,
      "learning_rate": 0.00020713698204106708,
      "loss": 0.472,
      "num_input_tokens_seen": 18173136,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.10535918921232224,
      "learning_rate": 0.00020695166083818206,
      "loss": 0.5252,
      "num_input_tokens_seen": 18176048,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.1061309203505516,
      "learning_rate": 0.0002067664009376663,
      "loss": 0.5091,
      "num_input_tokens_seen": 18179440,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 0.18923068046569824,
      "learning_rate": 0.0002065812023782741,
      "loss": 0.4915,
      "num_input_tokens_seen": 18182032,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 0.20624293386936188,
      "learning_rate": 0.00020639606519874703,
      "loss": 0.4245,
      "num_input_tokens_seen": 18184656,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.16533468663692474,
      "learning_rate": 0.00020621098943781352,
      "loss": 0.4473,
      "num_input_tokens_seen": 18187088,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.32823312282562256,
      "learning_rate": 0.0002060259751341891,
      "loss": 0.3805,
      "num_input_tokens_seen": 18189936,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 0.1573360711336136,
      "learning_rate": 0.00020584102232657688,
      "loss": 0.3587,
      "num_input_tokens_seen": 18193360,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.23335348069667816,
      "learning_rate": 0.0002056561310536668,
      "loss": 0.2909,
      "num_input_tokens_seen": 18196080,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 0.10554984956979752,
      "learning_rate": 0.0002054713013541361,
      "loss": 0.3938,
      "num_input_tokens_seen": 18199248,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 0.12508682906627655,
      "learning_rate": 0.00020528653326664915,
      "loss": 0.3668,
      "num_input_tokens_seen": 18203088,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.18262483179569244,
      "learning_rate": 0.00020510182682985717,
      "loss": 0.4116,
      "num_input_tokens_seen": 18205936,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 0.15172336995601654,
      "learning_rate": 0.00020491718208239847,
      "loss": 0.4469,
      "num_input_tokens_seen": 18209712,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.12367741018533707,
      "learning_rate": 0.00020473259906289877,
      "loss": 0.45,
      "num_input_tokens_seen": 18214864,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 0.21971182525157928,
      "learning_rate": 0.00020454807780997077,
      "loss": 0.3866,
      "num_input_tokens_seen": 18218032,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.15149186551570892,
      "learning_rate": 0.0002043636183622144,
      "loss": 0.3935,
      "num_input_tokens_seen": 18222096,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 0.11221767216920853,
      "learning_rate": 0.00020417922075821626,
      "loss": 0.4527,
      "num_input_tokens_seen": 18225360,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 0.26211097836494446,
      "learning_rate": 0.00020399488503655005,
      "loss": 0.5327,
      "num_input_tokens_seen": 18228304,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 0.19079484045505524,
      "learning_rate": 0.00020381061123577688,
      "loss": 0.5401,
      "num_input_tokens_seen": 18231504,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.15140356123447418,
      "learning_rate": 0.00020362639939444473,
      "loss": 0.4637,
      "num_input_tokens_seen": 18234320,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 0.1370520144701004,
      "learning_rate": 0.0002034422495510888,
      "loss": 0.4121,
      "num_input_tokens_seen": 18237296,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.1386454850435257,
      "learning_rate": 0.00020325816174423077,
      "loss": 0.4625,
      "num_input_tokens_seen": 18239696,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 0.15912896394729614,
      "learning_rate": 0.00020307413601238012,
      "loss": 0.5882,
      "num_input_tokens_seen": 18243088,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 0.11468283832073212,
      "learning_rate": 0.00020289017239403246,
      "loss": 0.5,
      "num_input_tokens_seen": 18246096,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 0.09112390130758286,
      "learning_rate": 0.0002027062709276712,
      "loss": 0.3387,
      "num_input_tokens_seen": 18249424,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 0.12080901861190796,
      "learning_rate": 0.00020252243165176632,
      "loss": 0.3408,
      "num_input_tokens_seen": 18252976,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 0.24936844408512115,
      "learning_rate": 0.00020233865460477508,
      "loss": 0.4808,
      "num_input_tokens_seen": 18255568,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 0.14952600002288818,
      "learning_rate": 0.00020215493982514128,
      "loss": 0.3687,
      "num_input_tokens_seen": 18259664,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 0.11976466327905655,
      "learning_rate": 0.00020197128735129622,
      "loss": 0.3184,
      "num_input_tokens_seen": 18262896,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 0.2758728861808777,
      "learning_rate": 0.0002017876972216575,
      "loss": 0.4143,
      "num_input_tokens_seen": 18265776,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 0.19948063790798187,
      "learning_rate": 0.00020160416947463046,
      "loss": 0.4659,
      "num_input_tokens_seen": 18268336,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 0.14379452168941498,
      "learning_rate": 0.00020142070414860702,
      "loss": 0.4876,
      "num_input_tokens_seen": 18271312,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 0.14449837803840637,
      "learning_rate": 0.0002012373012819657,
      "loss": 0.3474,
      "num_input_tokens_seen": 18274224,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 0.23689493536949158,
      "learning_rate": 0.00020105396091307247,
      "loss": 0.4803,
      "num_input_tokens_seen": 18277200,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.14304102957248688,
      "learning_rate": 0.0002008706830802803,
      "loss": 0.4094,
      "num_input_tokens_seen": 18280016,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 0.1974773108959198,
      "learning_rate": 0.00020068746782192837,
      "loss": 0.4543,
      "num_input_tokens_seen": 18283312,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 0.14047037065029144,
      "learning_rate": 0.00020050431517634366,
      "loss": 0.4791,
      "num_input_tokens_seen": 18286064,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 0.2267046719789505,
      "learning_rate": 0.00020032122518183915,
      "loss": 0.6439,
      "num_input_tokens_seen": 18288496,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 0.13530445098876953,
      "learning_rate": 0.00020013819787671538,
      "loss": 0.4623,
      "num_input_tokens_seen": 18290768,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 0.2612515985965729,
      "learning_rate": 0.00019995523329925985,
      "loss": 0.4255,
      "num_input_tokens_seen": 18294096,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.2030300796031952,
      "learning_rate": 0.00019977233148774627,
      "loss": 0.4384,
      "num_input_tokens_seen": 18297296,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.1463489830493927,
      "learning_rate": 0.00019958949248043573,
      "loss": 0.3535,
      "num_input_tokens_seen": 18300176,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 0.15647000074386597,
      "learning_rate": 0.00019940671631557628,
      "loss": 0.5187,
      "num_input_tokens_seen": 18303280,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 0.1559503674507141,
      "learning_rate": 0.0001992240030314022,
      "loss": 0.4306,
      "num_input_tokens_seen": 18306320,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 0.28628507256507874,
      "learning_rate": 0.00019904135266613532,
      "loss": 0.4601,
      "num_input_tokens_seen": 18309264,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.16101700067520142,
      "learning_rate": 0.00019885876525798407,
      "loss": 0.389,
      "num_input_tokens_seen": 18312144,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.260356605052948,
      "learning_rate": 0.0001986762408451434,
      "loss": 0.3739,
      "num_input_tokens_seen": 18314448,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 0.2630551755428314,
      "learning_rate": 0.00019849377946579562,
      "loss": 0.4019,
      "num_input_tokens_seen": 18317200,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.16570505499839783,
      "learning_rate": 0.00019831138115810926,
      "loss": 0.4763,
      "num_input_tokens_seen": 18320464,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 0.16096854209899902,
      "learning_rate": 0.00019812904596024005,
      "loss": 0.4822,
      "num_input_tokens_seen": 18323568,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 0.09289422631263733,
      "learning_rate": 0.00019794677391033055,
      "loss": 0.5554,
      "num_input_tokens_seen": 18326928,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 0.14531657099723816,
      "learning_rate": 0.00019776456504651014,
      "loss": 0.5122,
      "num_input_tokens_seen": 18329808,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.09819141775369644,
      "learning_rate": 0.0001975824194068946,
      "loss": 0.5033,
      "num_input_tokens_seen": 18333424,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 0.13762040436267853,
      "learning_rate": 0.00019740033702958653,
      "loss": 0.3356,
      "num_input_tokens_seen": 18336176,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 0.15430229902267456,
      "learning_rate": 0.0001972183179526758,
      "loss": 0.3632,
      "num_input_tokens_seen": 18339568,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 0.14472688734531403,
      "learning_rate": 0.0001970363622142386,
      "loss": 0.3456,
      "num_input_tokens_seen": 18342640,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 0.08277973532676697,
      "learning_rate": 0.00019685446985233824,
      "loss": 0.3701,
      "num_input_tokens_seen": 18345712,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.15648598968982697,
      "learning_rate": 0.00019667264090502412,
      "loss": 0.2969,
      "num_input_tokens_seen": 18348528,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.13622917234897614,
      "learning_rate": 0.0001964908754103331,
      "loss": 0.4981,
      "num_input_tokens_seen": 18352720,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 0.08660513162612915,
      "learning_rate": 0.00019630917340628822,
      "loss": 0.3792,
      "num_input_tokens_seen": 18356752,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 0.1381758749485016,
      "learning_rate": 0.00019612753493089959,
      "loss": 0.3193,
      "num_input_tokens_seen": 18360080,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.28705745935440063,
      "learning_rate": 0.00019594596002216387,
      "loss": 0.5289,
      "num_input_tokens_seen": 18363056,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.1410190910100937,
      "learning_rate": 0.00019576444871806465,
      "loss": 0.4409,
      "num_input_tokens_seen": 18365488,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.10429570823907852,
      "learning_rate": 0.0001955830010565719,
      "loss": 0.4176,
      "num_input_tokens_seen": 18368368,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 0.27028846740722656,
      "learning_rate": 0.00019540161707564213,
      "loss": 0.483,
      "num_input_tokens_seen": 18371696,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 0.20803174376487732,
      "learning_rate": 0.000195220296813219,
      "loss": 0.4323,
      "num_input_tokens_seen": 18374576,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 0.21602414548397064,
      "learning_rate": 0.00019503904030723274,
      "loss": 0.403,
      "num_input_tokens_seen": 18378512,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.2013111710548401,
      "learning_rate": 0.00019485784759560004,
      "loss": 0.5063,
      "num_input_tokens_seen": 18382032,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 0.15444250404834747,
      "learning_rate": 0.0001946767187162247,
      "loss": 0.3538,
      "num_input_tokens_seen": 18385520,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 0.3262253701686859,
      "learning_rate": 0.00019449565370699618,
      "loss": 0.4559,
      "num_input_tokens_seen": 18388016,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 0.19165116548538208,
      "learning_rate": 0.00019431465260579152,
      "loss": 0.3027,
      "num_input_tokens_seen": 18390928,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 0.21114088594913483,
      "learning_rate": 0.00019413371545047404,
      "loss": 0.2985,
      "num_input_tokens_seen": 18393072,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 0.19640108942985535,
      "learning_rate": 0.00019395284227889382,
      "loss": 0.4906,
      "num_input_tokens_seen": 18395728,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 0.22010253369808197,
      "learning_rate": 0.00019377203312888764,
      "loss": 0.4946,
      "num_input_tokens_seen": 18398704,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 0.16936415433883667,
      "learning_rate": 0.0001935912880382784,
      "loss": 0.4787,
      "num_input_tokens_seen": 18402096,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.11812127381563187,
      "learning_rate": 0.00019341060704487596,
      "loss": 0.4358,
      "num_input_tokens_seen": 18406256,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.11648322641849518,
      "learning_rate": 0.00019322999018647668,
      "loss": 0.5,
      "num_input_tokens_seen": 18409968,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.12841761112213135,
      "learning_rate": 0.00019304943750086362,
      "loss": 0.372,
      "num_input_tokens_seen": 18416272,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.13524097204208374,
      "learning_rate": 0.0001928689490258066,
      "loss": 0.5812,
      "num_input_tokens_seen": 18419248,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 0.26682767271995544,
      "learning_rate": 0.00019268852479906146,
      "loss": 0.5099,
      "num_input_tokens_seen": 18421968,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.1299811154603958,
      "learning_rate": 0.0001925081648583708,
      "loss": 0.361,
      "num_input_tokens_seen": 18424816,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 0.20134377479553223,
      "learning_rate": 0.00019232786924146394,
      "loss": 0.3564,
      "num_input_tokens_seen": 18428272,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.1275637149810791,
      "learning_rate": 0.00019214763798605677,
      "loss": 0.5927,
      "num_input_tokens_seen": 18431024,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 0.11430669575929642,
      "learning_rate": 0.00019196747112985158,
      "loss": 0.4123,
      "num_input_tokens_seen": 18434416,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 0.1134219616651535,
      "learning_rate": 0.00019178736871053737,
      "loss": 0.4507,
      "num_input_tokens_seen": 18437680,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 0.09492295980453491,
      "learning_rate": 0.00019160733076578935,
      "loss": 0.4178,
      "num_input_tokens_seen": 18441296,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 0.1729457825422287,
      "learning_rate": 0.0001914273573332692,
      "loss": 0.5899,
      "num_input_tokens_seen": 18444336,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.15250442922115326,
      "learning_rate": 0.00019124744845062553,
      "loss": 0.4666,
      "num_input_tokens_seen": 18447216,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 0.23445509374141693,
      "learning_rate": 0.00019106760415549324,
      "loss": 0.4863,
      "num_input_tokens_seen": 18449840,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 0.16367250680923462,
      "learning_rate": 0.0001908878244854938,
      "loss": 0.3588,
      "num_input_tokens_seen": 18452848,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 0.2019394338130951,
      "learning_rate": 0.00019070810947823498,
      "loss": 0.4351,
      "num_input_tokens_seen": 18456336,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 0.1184241995215416,
      "learning_rate": 0.0001905284591713109,
      "loss": 0.4117,
      "num_input_tokens_seen": 18458896,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 0.15816554427146912,
      "learning_rate": 0.00019034887360230246,
      "loss": 0.463,
      "num_input_tokens_seen": 18462224,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 0.13743658363819122,
      "learning_rate": 0.000190169352808777,
      "loss": 0.423,
      "num_input_tokens_seen": 18465424,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.08708614110946655,
      "learning_rate": 0.00018998989682828844,
      "loss": 0.3837,
      "num_input_tokens_seen": 18469328,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 0.16109302639961243,
      "learning_rate": 0.00018981050569837643,
      "loss": 0.3493,
      "num_input_tokens_seen": 18472016,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 0.20603087544441223,
      "learning_rate": 0.000189631179456568,
      "loss": 0.4786,
      "num_input_tokens_seen": 18474928,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 0.15462522208690643,
      "learning_rate": 0.00018945191814037578,
      "loss": 0.5456,
      "num_input_tokens_seen": 18479088,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.3067294955253601,
      "learning_rate": 0.00018927272178729938,
      "loss": 0.3567,
      "num_input_tokens_seen": 18481840,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 0.18518328666687012,
      "learning_rate": 0.00018909359043482477,
      "loss": 0.5811,
      "num_input_tokens_seen": 18485104,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.11796843260526657,
      "learning_rate": 0.00018891452412042386,
      "loss": 0.5143,
      "num_input_tokens_seen": 18489680,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.17014352977275848,
      "learning_rate": 0.00018873552288155542,
      "loss": 0.6819,
      "num_input_tokens_seen": 18492688,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 0.1475915163755417,
      "learning_rate": 0.00018855658675566471,
      "loss": 0.5184,
      "num_input_tokens_seen": 18499696,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 0.23823152482509613,
      "learning_rate": 0.00018837771578018275,
      "loss": 0.548,
      "num_input_tokens_seen": 18502192,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 0.10396343469619751,
      "learning_rate": 0.00018819890999252747,
      "loss": 0.4848,
      "num_input_tokens_seen": 18506128,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.11154931038618088,
      "learning_rate": 0.0001880201694301032,
      "loss": 0.5111,
      "num_input_tokens_seen": 18509680,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 0.24734191596508026,
      "learning_rate": 0.00018784149413030005,
      "loss": 0.4115,
      "num_input_tokens_seen": 18513264,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 0.1742577850818634,
      "learning_rate": 0.00018766288413049503,
      "loss": 0.3784,
      "num_input_tokens_seen": 18516464,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 0.13063062727451324,
      "learning_rate": 0.00018748433946805148,
      "loss": 0.5533,
      "num_input_tokens_seen": 18519920,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 0.14011496305465698,
      "learning_rate": 0.00018730586018031854,
      "loss": 0.3499,
      "num_input_tokens_seen": 18523888,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 0.23310589790344238,
      "learning_rate": 0.00018712744630463248,
      "loss": 0.3951,
      "num_input_tokens_seen": 18528560,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 0.11982885748147964,
      "learning_rate": 0.000186949097878315,
      "loss": 0.4068,
      "num_input_tokens_seen": 18532272,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 0.11001361161470413,
      "learning_rate": 0.0001867708149386748,
      "loss": 0.4779,
      "num_input_tokens_seen": 18535216,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 0.14096486568450928,
      "learning_rate": 0.0001865925975230068,
      "loss": 0.3157,
      "num_input_tokens_seen": 18538512,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.23633113503456116,
      "learning_rate": 0.0001864144456685916,
      "loss": 0.4064,
      "num_input_tokens_seen": 18541776,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 0.1484234780073166,
      "learning_rate": 0.00018623635941269706,
      "loss": 0.4717,
      "num_input_tokens_seen": 18545776,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.15232740342617035,
      "learning_rate": 0.00018605833879257628,
      "loss": 0.4383,
      "num_input_tokens_seen": 18549552,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.32378873229026794,
      "learning_rate": 0.0001858803838454694,
      "loss": 0.4376,
      "num_input_tokens_seen": 18553328,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 0.14829294383525848,
      "learning_rate": 0.0001857024946086026,
      "loss": 0.3934,
      "num_input_tokens_seen": 18557072,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 0.13982459902763367,
      "learning_rate": 0.00018552467111918836,
      "loss": 0.3635,
      "num_input_tokens_seen": 18559824,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 0.11998125165700912,
      "learning_rate": 0.00018534691341442507,
      "loss": 0.3774,
      "num_input_tokens_seen": 18562928,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 0.20984169840812683,
      "learning_rate": 0.0001851692215314979,
      "loss": 0.4597,
      "num_input_tokens_seen": 18565744,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 0.1596655398607254,
      "learning_rate": 0.0001849915955075776,
      "loss": 0.4863,
      "num_input_tokens_seen": 18568880,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.1334826499223709,
      "learning_rate": 0.0001848140353798217,
      "loss": 0.4629,
      "num_input_tokens_seen": 18571600,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 0.09992161393165588,
      "learning_rate": 0.00018463654118537382,
      "loss": 0.3894,
      "num_input_tokens_seen": 18575056,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.1422446370124817,
      "learning_rate": 0.00018445911296136386,
      "loss": 0.3638,
      "num_input_tokens_seen": 18579568,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 0.1498279571533203,
      "learning_rate": 0.00018428175074490754,
      "loss": 0.3952,
      "num_input_tokens_seen": 18582640,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 0.19164709746837616,
      "learning_rate": 0.00018410445457310687,
      "loss": 0.4886,
      "num_input_tokens_seen": 18585584,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.28855085372924805,
      "learning_rate": 0.00018392722448305044,
      "loss": 0.5693,
      "num_input_tokens_seen": 18589488,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 0.1456354856491089,
      "learning_rate": 0.00018375006051181265,
      "loss": 0.5592,
      "num_input_tokens_seen": 18593424,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 0.2658443748950958,
      "learning_rate": 0.0001835729626964544,
      "loss": 0.3375,
      "num_input_tokens_seen": 18596368,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 0.2301715612411499,
      "learning_rate": 0.00018339593107402242,
      "loss": 0.4387,
      "num_input_tokens_seen": 18599952,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 0.1322786509990692,
      "learning_rate": 0.00018321896568154945,
      "loss": 0.4033,
      "num_input_tokens_seen": 18603248,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 0.10753120481967926,
      "learning_rate": 0.00018304206655605476,
      "loss": 0.3808,
      "num_input_tokens_seen": 18606768,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 0.12107590585947037,
      "learning_rate": 0.00018286523373454378,
      "loss": 0.397,
      "num_input_tokens_seen": 18609648,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.1380418986082077,
      "learning_rate": 0.00018268846725400783,
      "loss": 0.4019,
      "num_input_tokens_seen": 18612624,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 0.17695246636867523,
      "learning_rate": 0.00018251176715142458,
      "loss": 0.4914,
      "num_input_tokens_seen": 18615728,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 0.1343560814857483,
      "learning_rate": 0.00018233513346375756,
      "loss": 0.2881,
      "num_input_tokens_seen": 18618832,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 0.15229009091854095,
      "learning_rate": 0.00018215856622795634,
      "loss": 0.4759,
      "num_input_tokens_seen": 18623056,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 0.08137120306491852,
      "learning_rate": 0.00018198206548095697,
      "loss": 0.3835,
      "num_input_tokens_seen": 18626032,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.18422569334506989,
      "learning_rate": 0.00018180563125968135,
      "loss": 0.383,
      "num_input_tokens_seen": 18628784,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.16192996501922607,
      "learning_rate": 0.00018162926360103766,
      "loss": 0.4505,
      "num_input_tokens_seen": 18633648,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 0.11385789513587952,
      "learning_rate": 0.0001814529625419199,
      "loss": 0.4821,
      "num_input_tokens_seen": 18637520,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 0.13646605610847473,
      "learning_rate": 0.00018127672811920798,
      "loss": 0.4906,
      "num_input_tokens_seen": 18641488,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.1483057588338852,
      "learning_rate": 0.00018110056036976834,
      "loss": 0.369,
      "num_input_tokens_seen": 18644336,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 0.1383325159549713,
      "learning_rate": 0.00018092445933045332,
      "loss": 0.3521,
      "num_input_tokens_seen": 18647984,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.12338674068450928,
      "learning_rate": 0.00018074842503810118,
      "loss": 0.4816,
      "num_input_tokens_seen": 18650704,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 0.14990223944187164,
      "learning_rate": 0.00018057245752953666,
      "loss": 0.4727,
      "num_input_tokens_seen": 18654640,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.16734787821769714,
      "learning_rate": 0.00018039655684156948,
      "loss": 0.5566,
      "num_input_tokens_seen": 18657392,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 0.11715219914913177,
      "learning_rate": 0.00018022072301099635,
      "loss": 0.371,
      "num_input_tokens_seen": 18659696,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.1566895693540573,
      "learning_rate": 0.00018004495607459975,
      "loss": 0.4108,
      "num_input_tokens_seen": 18662608,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 0.17643524706363678,
      "learning_rate": 0.00017986925606914805,
      "loss": 0.5954,
      "num_input_tokens_seen": 18665328,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 0.20301760733127594,
      "learning_rate": 0.00017969362303139598,
      "loss": 0.3526,
      "num_input_tokens_seen": 18668560,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.19565284252166748,
      "learning_rate": 0.00017951805699808365,
      "loss": 0.7004,
      "num_input_tokens_seen": 18672784,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.15442758798599243,
      "learning_rate": 0.00017934255800593747,
      "loss": 0.4813,
      "num_input_tokens_seen": 18675344,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 0.20131488144397736,
      "learning_rate": 0.0001791671260916698,
      "loss": 0.5485,
      "num_input_tokens_seen": 18679312,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 0.5580204129219055,
      "learning_rate": 0.00017899176129197914,
      "loss": 0.4369,
      "num_input_tokens_seen": 18682160,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.1404315084218979,
      "learning_rate": 0.00017881646364354998,
      "loss": 0.4837,
      "num_input_tokens_seen": 18685264,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.0668374300003052,
      "learning_rate": 0.00017864123318305213,
      "loss": 0.4933,
      "num_input_tokens_seen": 18687760,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.4885338842868805,
      "eval_runtime": 13.7008,
      "eval_samples_per_second": 61.894,
      "eval_steps_per_second": 15.474,
      "num_input_tokens_seen": 18687760,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.11856909841299057,
      "learning_rate": 0.00017846606994714222,
      "loss": 0.5488,
      "num_input_tokens_seen": 18691216,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.12434202432632446,
      "learning_rate": 0.0001782909739724621,
      "loss": 0.4591,
      "num_input_tokens_seen": 18694256,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.368596613407135,
      "learning_rate": 0.0001781159452956399,
      "loss": 0.4116,
      "num_input_tokens_seen": 18701296,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.09235739707946777,
      "learning_rate": 0.00017794098395328977,
      "loss": 0.5121,
      "num_input_tokens_seen": 18704592,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.1041882261633873,
      "learning_rate": 0.00017776608998201172,
      "loss": 0.3881,
      "num_input_tokens_seen": 18707984,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.1368275135755539,
      "learning_rate": 0.00017759126341839132,
      "loss": 0.5712,
      "num_input_tokens_seen": 18710608,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 0.18921688199043274,
      "learning_rate": 0.0001774165042990003,
      "loss": 0.4356,
      "num_input_tokens_seen": 18713904,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.16774733364582062,
      "learning_rate": 0.00017724181266039635,
      "loss": 0.369,
      "num_input_tokens_seen": 18716816,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 0.1361616998910904,
      "learning_rate": 0.00017706718853912292,
      "loss": 0.6404,
      "num_input_tokens_seen": 18719696,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 0.1931544840335846,
      "learning_rate": 0.00017689263197170964,
      "loss": 0.3505,
      "num_input_tokens_seen": 18723600,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 0.14186608791351318,
      "learning_rate": 0.00017671814299467136,
      "loss": 0.4887,
      "num_input_tokens_seen": 18727088,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.0975758284330368,
      "learning_rate": 0.00017654372164450954,
      "loss": 0.5053,
      "num_input_tokens_seen": 18730576,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 0.15157358348369598,
      "learning_rate": 0.0001763693679577108,
      "loss": 0.4437,
      "num_input_tokens_seen": 18733840,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 0.14712665975093842,
      "learning_rate": 0.00017619508197074812,
      "loss": 0.415,
      "num_input_tokens_seen": 18736592,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.1900644600391388,
      "learning_rate": 0.0001760208637200803,
      "loss": 0.4597,
      "num_input_tokens_seen": 18740208,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.17554354667663574,
      "learning_rate": 0.00017584671324215152,
      "loss": 0.4325,
      "num_input_tokens_seen": 18743472,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.17181457579135895,
      "learning_rate": 0.00017567263057339212,
      "loss": 0.4106,
      "num_input_tokens_seen": 18745968,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.12295497953891754,
      "learning_rate": 0.00017549861575021848,
      "loss": 0.4217,
      "num_input_tokens_seen": 18749744,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 0.23317456245422363,
      "learning_rate": 0.00017532466880903214,
      "loss": 0.375,
      "num_input_tokens_seen": 18752752,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.18140274286270142,
      "learning_rate": 0.00017515078978622102,
      "loss": 0.347,
      "num_input_tokens_seen": 18755248,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 0.1867455095052719,
      "learning_rate": 0.0001749769787181587,
      "loss": 0.4537,
      "num_input_tokens_seen": 18758704,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.16028344631195068,
      "learning_rate": 0.00017480323564120426,
      "loss": 0.5111,
      "num_input_tokens_seen": 18761872,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.1902151256799698,
      "learning_rate": 0.00017462956059170287,
      "loss": 0.3957,
      "num_input_tokens_seen": 18764144,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.14474371075630188,
      "learning_rate": 0.00017445595360598553,
      "loss": 0.2815,
      "num_input_tokens_seen": 18767728,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.12309259921312332,
      "learning_rate": 0.0001742824147203686,
      "loss": 0.4478,
      "num_input_tokens_seen": 18770736,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 0.15778136253356934,
      "learning_rate": 0.00017410894397115463,
      "loss": 0.3451,
      "num_input_tokens_seen": 18773872,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 0.11850537359714508,
      "learning_rate": 0.00017393554139463147,
      "loss": 0.3551,
      "num_input_tokens_seen": 18777040,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.18016399443149567,
      "learning_rate": 0.00017376220702707308,
      "loss": 0.3662,
      "num_input_tokens_seen": 18781232,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.10908918082714081,
      "learning_rate": 0.00017358894090473926,
      "loss": 0.4136,
      "num_input_tokens_seen": 18784656,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.11793628334999084,
      "learning_rate": 0.00017341574306387492,
      "loss": 0.427,
      "num_input_tokens_seen": 18788848,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 0.17130136489868164,
      "learning_rate": 0.00017324261354071147,
      "loss": 0.3869,
      "num_input_tokens_seen": 18792368,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.19687999784946442,
      "learning_rate": 0.00017306955237146522,
      "loss": 0.2757,
      "num_input_tokens_seen": 18794480,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 0.21571923792362213,
      "learning_rate": 0.0001728965595923388,
      "loss": 0.5335,
      "num_input_tokens_seen": 18797680,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 0.18714182078838348,
      "learning_rate": 0.00017272363523952033,
      "loss": 0.3696,
      "num_input_tokens_seen": 18800656,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.10668183118104935,
      "learning_rate": 0.0001725507793491838,
      "loss": 0.4511,
      "num_input_tokens_seen": 18803792,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.1605851799249649,
      "learning_rate": 0.00017237799195748827,
      "loss": 0.4445,
      "num_input_tokens_seen": 18806768,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 0.19219519197940826,
      "learning_rate": 0.00017220527310057927,
      "loss": 0.3638,
      "num_input_tokens_seen": 18809680,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 0.21897931396961212,
      "learning_rate": 0.0001720326228145873,
      "loss": 0.3876,
      "num_input_tokens_seen": 18813296,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.187015101313591,
      "learning_rate": 0.00017186004113562903,
      "loss": 0.3442,
      "num_input_tokens_seen": 18817040,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 0.14635755121707916,
      "learning_rate": 0.00017168752809980655,
      "loss": 0.446,
      "num_input_tokens_seen": 18820400,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.13318780064582825,
      "learning_rate": 0.00017151508374320767,
      "loss": 0.2906,
      "num_input_tokens_seen": 18823152,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 0.25627952814102173,
      "learning_rate": 0.00017134270810190583,
      "loss": 0.5056,
      "num_input_tokens_seen": 18825776,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 0.17443841695785522,
      "learning_rate": 0.0001711704012119597,
      "loss": 0.4419,
      "num_input_tokens_seen": 18828816,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 0.08205059915781021,
      "learning_rate": 0.0001709981631094142,
      "loss": 0.4313,
      "num_input_tokens_seen": 18832624,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 0.11324058473110199,
      "learning_rate": 0.00017082599383029946,
      "loss": 0.4224,
      "num_input_tokens_seen": 18836848,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.14819177985191345,
      "learning_rate": 0.0001706538934106316,
      "loss": 0.535,
      "num_input_tokens_seen": 18839824,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.16995921730995178,
      "learning_rate": 0.00017048186188641185,
      "loss": 0.5364,
      "num_input_tokens_seen": 18842416,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 0.19619989395141602,
      "learning_rate": 0.0001703098992936271,
      "loss": 0.363,
      "num_input_tokens_seen": 18845424,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 0.14682930707931519,
      "learning_rate": 0.00017013800566825012,
      "loss": 0.4947,
      "num_input_tokens_seen": 18848592,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.12123067677021027,
      "learning_rate": 0.00016996618104623905,
      "loss": 0.467,
      "num_input_tokens_seen": 18851408,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.1339324414730072,
      "learning_rate": 0.00016979442546353774,
      "loss": 0.4333,
      "num_input_tokens_seen": 18854512,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.09292460978031158,
      "learning_rate": 0.00016962273895607555,
      "loss": 0.4992,
      "num_input_tokens_seen": 18858704,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 0.17353937029838562,
      "learning_rate": 0.00016945112155976723,
      "loss": 0.4566,
      "num_input_tokens_seen": 18862256,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 0.1293090432882309,
      "learning_rate": 0.000169279573310513,
      "loss": 0.4819,
      "num_input_tokens_seen": 18866640,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 0.1589345782995224,
      "learning_rate": 0.00016910809424419904,
      "loss": 0.2971,
      "num_input_tokens_seen": 18869392,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.09522192180156708,
      "learning_rate": 0.00016893668439669674,
      "loss": 0.4439,
      "num_input_tokens_seen": 18872336,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.3231721818447113,
      "learning_rate": 0.0001687653438038632,
      "loss": 0.4975,
      "num_input_tokens_seen": 18876016,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.18447108566761017,
      "learning_rate": 0.00016859407250154084,
      "loss": 0.4004,
      "num_input_tokens_seen": 18878896,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.17625156044960022,
      "learning_rate": 0.0001684228705255575,
      "loss": 0.4212,
      "num_input_tokens_seen": 18881680,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 0.1722463220357895,
      "learning_rate": 0.00016825173791172677,
      "loss": 0.3506,
      "num_input_tokens_seen": 18884912,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.2735006809234619,
      "learning_rate": 0.0001680806746958477,
      "loss": 0.4028,
      "num_input_tokens_seen": 18888048,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.1317298263311386,
      "learning_rate": 0.00016790968091370494,
      "loss": 0.5457,
      "num_input_tokens_seen": 18892656,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.2049083560705185,
      "learning_rate": 0.00016773875660106824,
      "loss": 0.4306,
      "num_input_tokens_seen": 18895312,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.13056950271129608,
      "learning_rate": 0.0001675679017936928,
      "loss": 0.4367,
      "num_input_tokens_seen": 18897712,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 0.16516441106796265,
      "learning_rate": 0.00016739711652731976,
      "loss": 0.4676,
      "num_input_tokens_seen": 18900112,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.13707402348518372,
      "learning_rate": 0.00016722640083767542,
      "loss": 0.3495,
      "num_input_tokens_seen": 18903920,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 0.17717613279819489,
      "learning_rate": 0.00016705575476047153,
      "loss": 0.5532,
      "num_input_tokens_seen": 18911056,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.2510211169719696,
      "learning_rate": 0.0001668851783314055,
      "loss": 0.4506,
      "num_input_tokens_seen": 18913360,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.1092391163110733,
      "learning_rate": 0.00016671467158615977,
      "loss": 0.5377,
      "num_input_tokens_seen": 18917648,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 0.12030302733182907,
      "learning_rate": 0.00016654423456040224,
      "loss": 0.6281,
      "num_input_tokens_seen": 18921008,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 0.10332666337490082,
      "learning_rate": 0.0001663738672897866,
      "loss": 0.3589,
      "num_input_tokens_seen": 18924816,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 0.15943361818790436,
      "learning_rate": 0.0001662035698099517,
      "loss": 0.4208,
      "num_input_tokens_seen": 18927664,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.20697705447673798,
      "learning_rate": 0.00016603334215652206,
      "loss": 0.3598,
      "num_input_tokens_seen": 18930384,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 0.24501225352287292,
      "learning_rate": 0.0001658631843651069,
      "loss": 0.4989,
      "num_input_tokens_seen": 18933488,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.20466084778308868,
      "learning_rate": 0.00016569309647130177,
      "loss": 0.4271,
      "num_input_tokens_seen": 18936080,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.12275506556034088,
      "learning_rate": 0.00016552307851068666,
      "loss": 0.4355,
      "num_input_tokens_seen": 18940112,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 0.16108573973178864,
      "learning_rate": 0.00016535313051882765,
      "loss": 0.3741,
      "num_input_tokens_seen": 18944144,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 0.10953127592802048,
      "learning_rate": 0.00016518325253127604,
      "loss": 0.3702,
      "num_input_tokens_seen": 18948112,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 0.11687792092561722,
      "learning_rate": 0.00016501344458356803,
      "loss": 0.3239,
      "num_input_tokens_seen": 18951792,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.20572203397750854,
      "learning_rate": 0.00016484370671122588,
      "loss": 0.4435,
      "num_input_tokens_seen": 18955248,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.1864776909351349,
      "learning_rate": 0.00016467403894975641,
      "loss": 0.4672,
      "num_input_tokens_seen": 18958384,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 0.20828233659267426,
      "learning_rate": 0.0001645044413346524,
      "loss": 0.351,
      "num_input_tokens_seen": 18961584,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 0.17182905972003937,
      "learning_rate": 0.00016433491390139177,
      "loss": 0.3327,
      "num_input_tokens_seen": 18964720,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 0.13774870336055756,
      "learning_rate": 0.00016416545668543786,
      "loss": 0.5405,
      "num_input_tokens_seen": 18967888,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.16692489385604858,
      "learning_rate": 0.0001639960697222388,
      "loss": 0.3949,
      "num_input_tokens_seen": 18970384,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.19016438722610474,
      "learning_rate": 0.00016382675304722877,
      "loss": 0.3285,
      "num_input_tokens_seen": 18973072,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 0.29814931750297546,
      "learning_rate": 0.00016365750669582662,
      "loss": 0.3735,
      "num_input_tokens_seen": 18976016,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.12998522818088531,
      "learning_rate": 0.00016348833070343682,
      "loss": 0.4216,
      "num_input_tokens_seen": 18978992,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 0.15796875953674316,
      "learning_rate": 0.00016331922510544932,
      "loss": 0.4699,
      "num_input_tokens_seen": 18982480,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.1294303983449936,
      "learning_rate": 0.0001631501899372387,
      "loss": 0.4248,
      "num_input_tokens_seen": 18985936,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.11588280647993088,
      "learning_rate": 0.0001629812252341653,
      "loss": 0.4733,
      "num_input_tokens_seen": 18989072,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.13109257817268372,
      "learning_rate": 0.0001628123310315749,
      "loss": 0.3176,
      "num_input_tokens_seen": 18992080,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 0.14320909976959229,
      "learning_rate": 0.00016264350736479777,
      "loss": 0.5097,
      "num_input_tokens_seen": 18995120,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.14789672195911407,
      "learning_rate": 0.00016247475426915038,
      "loss": 0.5032,
      "num_input_tokens_seen": 18997712,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.13173694908618927,
      "learning_rate": 0.0001623060717799335,
      "loss": 0.401,
      "num_input_tokens_seen": 19000368,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 0.16280098259449005,
      "learning_rate": 0.0001621374599324338,
      "loss": 0.3558,
      "num_input_tokens_seen": 19003856,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.19962844252586365,
      "learning_rate": 0.00016196891876192292,
      "loss": 0.3611,
      "num_input_tokens_seen": 19006448,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 0.15979765355587006,
      "learning_rate": 0.00016180044830365798,
      "loss": 0.42,
      "num_input_tokens_seen": 19009712,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.11037486046552658,
      "learning_rate": 0.00016163204859288066,
      "loss": 0.4396,
      "num_input_tokens_seen": 19012688,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 0.07676874101161957,
      "learning_rate": 0.00016146371966481866,
      "loss": 0.3575,
      "num_input_tokens_seen": 19016112,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 0.18376480042934418,
      "learning_rate": 0.00016129546155468411,
      "loss": 0.4792,
      "num_input_tokens_seen": 19020304,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.21023988723754883,
      "learning_rate": 0.00016112727429767483,
      "loss": 0.4235,
      "num_input_tokens_seen": 19022512,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.1463756561279297,
      "learning_rate": 0.00016095915792897392,
      "loss": 0.4834,
      "num_input_tokens_seen": 19025552,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.16568025946617126,
      "learning_rate": 0.000160791112483749,
      "loss": 0.3912,
      "num_input_tokens_seen": 19028496,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.14607949554920197,
      "learning_rate": 0.00016062313799715362,
      "loss": 0.4226,
      "num_input_tokens_seen": 19031280,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.197066068649292,
      "learning_rate": 0.00016045523450432576,
      "loss": 0.4944,
      "num_input_tokens_seen": 19033872,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 0.1329343318939209,
      "learning_rate": 0.0001602874020403891,
      "loss": 0.4383,
      "num_input_tokens_seen": 19036688,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.2814933955669403,
      "learning_rate": 0.00016011964064045232,
      "loss": 0.4051,
      "num_input_tokens_seen": 19039792,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 0.1650296449661255,
      "learning_rate": 0.00015995195033960936,
      "loss": 0.3988,
      "num_input_tokens_seen": 19043696,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.129103422164917,
      "learning_rate": 0.00015978433117293884,
      "loss": 0.3418,
      "num_input_tokens_seen": 19046512,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 0.13778789341449738,
      "learning_rate": 0.0001596167831755047,
      "loss": 0.4569,
      "num_input_tokens_seen": 19050128,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 0.27614259719848633,
      "learning_rate": 0.00015944930638235626,
      "loss": 0.3238,
      "num_input_tokens_seen": 19052848,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.2020554095506668,
      "learning_rate": 0.00015928190082852773,
      "loss": 0.3002,
      "num_input_tokens_seen": 19055792,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.15351906418800354,
      "learning_rate": 0.0001591145665490384,
      "loss": 0.4481,
      "num_input_tokens_seen": 19058992,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 0.15682345628738403,
      "learning_rate": 0.00015894730357889287,
      "loss": 0.4776,
      "num_input_tokens_seen": 19063280,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 0.11607090383768082,
      "learning_rate": 0.0001587801119530805,
      "loss": 0.4965,
      "num_input_tokens_seen": 19067312,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.2503510117530823,
      "learning_rate": 0.00015861299170657572,
      "loss": 0.4676,
      "num_input_tokens_seen": 19070992,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.1019723191857338,
      "learning_rate": 0.00015844594287433834,
      "loss": 0.4726,
      "num_input_tokens_seen": 19073520,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.2232237160205841,
      "learning_rate": 0.00015827896549131304,
      "loss": 0.3916,
      "num_input_tokens_seen": 19075664,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 0.2655409872531891,
      "learning_rate": 0.0001581120595924298,
      "loss": 0.5275,
      "num_input_tokens_seen": 19078704,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 0.1858530193567276,
      "learning_rate": 0.00015794522521260317,
      "loss": 0.4648,
      "num_input_tokens_seen": 19082416,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 0.1492135375738144,
      "learning_rate": 0.00015777846238673293,
      "loss": 0.4344,
      "num_input_tokens_seen": 19085040,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.17619556188583374,
      "learning_rate": 0.00015761177114970404,
      "loss": 0.4983,
      "num_input_tokens_seen": 19089008,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 0.09393715858459473,
      "learning_rate": 0.0001574451515363865,
      "loss": 0.3898,
      "num_input_tokens_seen": 19092560,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 0.12019946426153183,
      "learning_rate": 0.00015727860358163536,
      "loss": 0.3615,
      "num_input_tokens_seen": 19094832,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 0.20721180737018585,
      "learning_rate": 0.00015711212732029035,
      "loss": 0.4101,
      "num_input_tokens_seen": 19098192,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 0.15652930736541748,
      "learning_rate": 0.00015694572278717627,
      "loss": 0.3535,
      "num_input_tokens_seen": 19101808,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 0.16729183495044708,
      "learning_rate": 0.00015677939001710323,
      "loss": 0.3131,
      "num_input_tokens_seen": 19104336,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.17604057490825653,
      "learning_rate": 0.00015661312904486613,
      "loss": 0.4022,
      "num_input_tokens_seen": 19107600,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.12809789180755615,
      "learning_rate": 0.00015644693990524483,
      "loss": 0.4827,
      "num_input_tokens_seen": 19110640,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 0.11877525597810745,
      "learning_rate": 0.00015628082263300448,
      "loss": 0.2954,
      "num_input_tokens_seen": 19113552,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.23403632640838623,
      "learning_rate": 0.00015611477726289465,
      "loss": 0.4814,
      "num_input_tokens_seen": 19117104,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.11287970840930939,
      "learning_rate": 0.00015594880382964994,
      "loss": 0.4881,
      "num_input_tokens_seen": 19119856,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 0.17107626795768738,
      "learning_rate": 0.00015578290236799037,
      "loss": 0.4139,
      "num_input_tokens_seen": 19122864,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 0.14026787877082825,
      "learning_rate": 0.00015561707291262061,
      "loss": 0.4468,
      "num_input_tokens_seen": 19125872,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.40097615122795105,
      "learning_rate": 0.00015545131549823043,
      "loss": 0.5134,
      "num_input_tokens_seen": 19128400,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 0.27967458963394165,
      "learning_rate": 0.0001552856301594942,
      "loss": 0.4517,
      "num_input_tokens_seen": 19131280,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 0.24241624772548676,
      "learning_rate": 0.00015512001693107126,
      "loss": 0.3303,
      "num_input_tokens_seen": 19134032,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.13828252255916595,
      "learning_rate": 0.0001549544758476062,
      "loss": 0.3757,
      "num_input_tokens_seen": 19137904,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 0.1889721155166626,
      "learning_rate": 0.00015478900694372832,
      "loss": 0.4704,
      "num_input_tokens_seen": 19142544,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.13032881915569305,
      "learning_rate": 0.00015462361025405196,
      "loss": 0.4689,
      "num_input_tokens_seen": 19145712,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.11327105015516281,
      "learning_rate": 0.0001544582858131759,
      "loss": 0.3993,
      "num_input_tokens_seen": 19149328,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 0.12389823794364929,
      "learning_rate": 0.00015429303365568448,
      "loss": 0.3401,
      "num_input_tokens_seen": 19152912,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.11702805757522583,
      "learning_rate": 0.00015412785381614625,
      "loss": 0.5934,
      "num_input_tokens_seen": 19156496,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.20444448292255402,
      "learning_rate": 0.000153962746329115,
      "loss": 0.4191,
      "num_input_tokens_seen": 19162224,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.23808816075325012,
      "learning_rate": 0.0001537977112291295,
      "loss": 0.5305,
      "num_input_tokens_seen": 19166000,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 0.15788820385932922,
      "learning_rate": 0.00015363274855071323,
      "loss": 0.4798,
      "num_input_tokens_seen": 19168720,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 0.15308663249015808,
      "learning_rate": 0.0001534678583283743,
      "loss": 0.3575,
      "num_input_tokens_seen": 19171824,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.12076956778764725,
      "learning_rate": 0.00015330304059660605,
      "loss": 0.4041,
      "num_input_tokens_seen": 19175536,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.09599313139915466,
      "learning_rate": 0.00015313829538988628,
      "loss": 0.4355,
      "num_input_tokens_seen": 19179120,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 0.12925055623054504,
      "learning_rate": 0.00015297362274267783,
      "loss": 0.4362,
      "num_input_tokens_seen": 19183344,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.181119903922081,
      "learning_rate": 0.00015280902268942866,
      "loss": 0.4215,
      "num_input_tokens_seen": 19187600,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 0.2248019576072693,
      "learning_rate": 0.00015264449526457074,
      "loss": 0.4462,
      "num_input_tokens_seen": 19191440,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.14074450731277466,
      "learning_rate": 0.00015248004050252178,
      "loss": 0.4672,
      "num_input_tokens_seen": 19194352,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 0.11228227615356445,
      "learning_rate": 0.00015231565843768348,
      "loss": 0.4159,
      "num_input_tokens_seen": 19197040,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.14786799252033234,
      "learning_rate": 0.00015215134910444284,
      "loss": 0.3992,
      "num_input_tokens_seen": 19200624,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 0.16897748410701752,
      "learning_rate": 0.00015198711253717163,
      "loss": 0.5715,
      "num_input_tokens_seen": 19203600,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.16641613841056824,
      "learning_rate": 0.00015182294877022596,
      "loss": 0.3287,
      "num_input_tokens_seen": 19206032,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.32337501645088196,
      "learning_rate": 0.00015165885783794726,
      "loss": 0.3791,
      "num_input_tokens_seen": 19208624,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 0.11463494598865509,
      "learning_rate": 0.0001514948397746615,
      "loss": 0.3376,
      "num_input_tokens_seen": 19210896,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.16409043967723846,
      "learning_rate": 0.0001513308946146792,
      "loss": 0.4489,
      "num_input_tokens_seen": 19213680,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 0.22242961823940277,
      "learning_rate": 0.0001511670223922959,
      "loss": 0.4185,
      "num_input_tokens_seen": 19216464,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.15114040672779083,
      "learning_rate": 0.000151003223141792,
      "loss": 0.4732,
      "num_input_tokens_seen": 19219184,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 0.13596990704536438,
      "learning_rate": 0.00015083949689743202,
      "loss": 0.4356,
      "num_input_tokens_seen": 19222288,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 0.13137656450271606,
      "learning_rate": 0.00015067584369346594,
      "loss": 0.4084,
      "num_input_tokens_seen": 19225776,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.23970380425453186,
      "learning_rate": 0.00015051226356412818,
      "loss": 0.3511,
      "num_input_tokens_seen": 19228496,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.12829908728599548,
      "learning_rate": 0.00015034875654363754,
      "loss": 0.4635,
      "num_input_tokens_seen": 19232272,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.16824211180210114,
      "learning_rate": 0.00015018532266619818,
      "loss": 0.4897,
      "num_input_tokens_seen": 19235568,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.2649550437927246,
      "learning_rate": 0.00015002196196599828,
      "loss": 0.4255,
      "num_input_tokens_seen": 19238352,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 0.1486699879169464,
      "learning_rate": 0.00014985867447721114,
      "loss": 0.5226,
      "num_input_tokens_seen": 19241616,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.11100004613399506,
      "learning_rate": 0.00014969546023399467,
      "loss": 0.3538,
      "num_input_tokens_seen": 19246512,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 0.20785324275493622,
      "learning_rate": 0.00014953231927049165,
      "loss": 0.3973,
      "num_input_tokens_seen": 19250128,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 0.14388951659202576,
      "learning_rate": 0.00014936925162082904,
      "loss": 0.4912,
      "num_input_tokens_seen": 19252976,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 0.20198820531368256,
      "learning_rate": 0.00014920625731911863,
      "loss": 0.4989,
      "num_input_tokens_seen": 19255664,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.10203548520803452,
      "learning_rate": 0.00014904333639945716,
      "loss": 0.3421,
      "num_input_tokens_seen": 19259088,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.4323825538158417,
      "learning_rate": 0.00014888048889592575,
      "loss": 0.4446,
      "num_input_tokens_seen": 19263792,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.2568484842777252,
      "learning_rate": 0.00014871771484259046,
      "loss": 0.3507,
      "num_input_tokens_seen": 19267312,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 0.20441214740276337,
      "learning_rate": 0.00014855501427350144,
      "loss": 0.5406,
      "num_input_tokens_seen": 19270256,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.2920196056365967,
      "learning_rate": 0.00014839238722269405,
      "loss": 0.5137,
      "num_input_tokens_seen": 19273200,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 0.0994417816400528,
      "learning_rate": 0.00014822983372418774,
      "loss": 0.4439,
      "num_input_tokens_seen": 19276240,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 0.15425819158554077,
      "learning_rate": 0.00014806735381198705,
      "loss": 0.4049,
      "num_input_tokens_seen": 19280656,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 0.08005275577306747,
      "learning_rate": 0.00014790494752008087,
      "loss": 0.3863,
      "num_input_tokens_seen": 19284656,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 0.1483469307422638,
      "learning_rate": 0.00014774261488244295,
      "loss": 0.4347,
      "num_input_tokens_seen": 19287888,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.18657593429088593,
      "learning_rate": 0.00014758035593303114,
      "loss": 0.4399,
      "num_input_tokens_seen": 19290704,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.10843473672866821,
      "learning_rate": 0.00014741817070578817,
      "loss": 0.4091,
      "num_input_tokens_seen": 19293968,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 0.23702208697795868,
      "learning_rate": 0.0001472560592346414,
      "loss": 0.5227,
      "num_input_tokens_seen": 19297360,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.16575096547603607,
      "learning_rate": 0.0001470940215535027,
      "loss": 0.4502,
      "num_input_tokens_seen": 19300688,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.17365041375160217,
      "learning_rate": 0.00014693205769626878,
      "loss": 0.3507,
      "num_input_tokens_seen": 19303888,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 0.18066807091236115,
      "learning_rate": 0.00014677016769682028,
      "loss": 0.4811,
      "num_input_tokens_seen": 19307344,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 0.1387360692024231,
      "learning_rate": 0.00014660835158902268,
      "loss": 0.4277,
      "num_input_tokens_seen": 19311312,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 0.12586846947669983,
      "learning_rate": 0.00014644660940672628,
      "loss": 0.4081,
      "num_input_tokens_seen": 19314768,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 0.41918981075286865,
      "learning_rate": 0.0001462849411837656,
      "loss": 0.474,
      "num_input_tokens_seen": 19317744,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 0.12730123102664948,
      "learning_rate": 0.00014612334695395985,
      "loss": 0.471,
      "num_input_tokens_seen": 19321840,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.18923930823802948,
      "learning_rate": 0.0001459618267511128,
      "loss": 0.3769,
      "num_input_tokens_seen": 19325232,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 0.19123181700706482,
      "learning_rate": 0.00014580038060901256,
      "loss": 0.4163,
      "num_input_tokens_seen": 19328240,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 0.1690431386232376,
      "learning_rate": 0.00014563900856143157,
      "loss": 0.3352,
      "num_input_tokens_seen": 19331216,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 0.1338927149772644,
      "learning_rate": 0.0001454777106421273,
      "loss": 0.4891,
      "num_input_tokens_seen": 19335760,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 0.21250607073307037,
      "learning_rate": 0.00014531648688484127,
      "loss": 0.4058,
      "num_input_tokens_seen": 19338672,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.1638348549604416,
      "learning_rate": 0.0001451553373233,
      "loss": 0.4931,
      "num_input_tokens_seen": 19341424,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 0.15092617273330688,
      "learning_rate": 0.00014499426199121379,
      "loss": 0.3959,
      "num_input_tokens_seen": 19345104,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.1276407092809677,
      "learning_rate": 0.0001448332609222777,
      "loss": 0.4159,
      "num_input_tokens_seen": 19348688,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.17285509407520294,
      "learning_rate": 0.0001446723341501715,
      "loss": 0.4304,
      "num_input_tokens_seen": 19352560,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.1445787400007248,
      "learning_rate": 0.00014451148170855922,
      "loss": 0.3851,
      "num_input_tokens_seen": 19355600,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.21298456192016602,
      "learning_rate": 0.0001443507036310895,
      "loss": 0.4061,
      "num_input_tokens_seen": 19359280,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.13533680140972137,
      "learning_rate": 0.00014418999995139493,
      "loss": 0.389,
      "num_input_tokens_seen": 19363600,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 0.18513964116573334,
      "learning_rate": 0.00014402937070309325,
      "loss": 0.4415,
      "num_input_tokens_seen": 19366608,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.12321288138628006,
      "learning_rate": 0.0001438688159197859,
      "loss": 0.5023,
      "num_input_tokens_seen": 19369936,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 0.11640258878469467,
      "learning_rate": 0.0001437083356350593,
      "loss": 0.3541,
      "num_input_tokens_seen": 19373968,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 0.2551698684692383,
      "learning_rate": 0.0001435479298824841,
      "loss": 0.6039,
      "num_input_tokens_seen": 19376528,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 0.13818277418613434,
      "learning_rate": 0.00014338759869561557,
      "loss": 0.4405,
      "num_input_tokens_seen": 19380752,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 0.10946255177259445,
      "learning_rate": 0.00014322734210799287,
      "loss": 0.4055,
      "num_input_tokens_seen": 19383824,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.2207973599433899,
      "learning_rate": 0.0001430671601531398,
      "loss": 0.5299,
      "num_input_tokens_seen": 19387184,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.22737984359264374,
      "learning_rate": 0.00014290705286456473,
      "loss": 0.414,
      "num_input_tokens_seen": 19390704,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 0.14507828652858734,
      "learning_rate": 0.00014274702027576024,
      "loss": 0.3137,
      "num_input_tokens_seen": 19393456,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.14740411937236786,
      "learning_rate": 0.00014258706242020354,
      "loss": 0.3485,
      "num_input_tokens_seen": 19395920,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.215681791305542,
      "learning_rate": 0.00014242717933135558,
      "loss": 0.3749,
      "num_input_tokens_seen": 19398288,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 0.1284986436367035,
      "learning_rate": 0.0001422673710426625,
      "loss": 0.3887,
      "num_input_tokens_seen": 19401488,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.10895128548145294,
      "learning_rate": 0.00014210763758755403,
      "loss": 0.5705,
      "num_input_tokens_seen": 19405584,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.1289391964673996,
      "learning_rate": 0.00014194797899944468,
      "loss": 0.519,
      "num_input_tokens_seen": 19409328,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 0.14646272361278534,
      "learning_rate": 0.00014178839531173344,
      "loss": 0.3863,
      "num_input_tokens_seen": 19412304,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 0.12974123656749725,
      "learning_rate": 0.00014162888655780303,
      "loss": 0.323,
      "num_input_tokens_seen": 19415792,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.11837605386972427,
      "learning_rate": 0.00014146945277102108,
      "loss": 0.4624,
      "num_input_tokens_seen": 19419440,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.8699066042900085,
      "learning_rate": 0.00014131009398473938,
      "loss": 0.4577,
      "num_input_tokens_seen": 19422256,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 0.14258494973182678,
      "learning_rate": 0.00014115081023229376,
      "loss": 0.485,
      "num_input_tokens_seen": 19425680,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 0.13469596207141876,
      "learning_rate": 0.00014099160154700462,
      "loss": 0.3569,
      "num_input_tokens_seen": 19429200,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 0.1830468475818634,
      "learning_rate": 0.00014083246796217684,
      "loss": 0.4202,
      "num_input_tokens_seen": 19432528,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.15092766284942627,
      "learning_rate": 0.00014067340951109898,
      "loss": 0.3941,
      "num_input_tokens_seen": 19436112,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.18450838327407837,
      "learning_rate": 0.00014051442622704447,
      "loss": 0.4531,
      "num_input_tokens_seen": 19439120,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 0.14453858137130737,
      "learning_rate": 0.0001403555181432709,
      "loss": 0.5349,
      "num_input_tokens_seen": 19442256,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.17033730447292328,
      "learning_rate": 0.00014019668529301977,
      "loss": 0.3953,
      "num_input_tokens_seen": 19445040,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.18497593700885773,
      "learning_rate": 0.00014003792770951733,
      "loss": 0.5115,
      "num_input_tokens_seen": 19447472,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.19051723182201385,
      "learning_rate": 0.00013987924542597363,
      "loss": 0.5159,
      "num_input_tokens_seen": 19449840,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 0.11675233393907547,
      "learning_rate": 0.0001397206384755833,
      "loss": 0.4579,
      "num_input_tokens_seen": 19453040,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 0.16089212894439697,
      "learning_rate": 0.00013956210689152526,
      "loss": 0.5498,
      "num_input_tokens_seen": 19455792,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 0.20295843482017517,
      "learning_rate": 0.0001394036507069622,
      "loss": 0.4811,
      "num_input_tokens_seen": 19458928,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 0.1513179987668991,
      "learning_rate": 0.00013924526995504172,
      "loss": 0.3913,
      "num_input_tokens_seen": 19462128,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.396107941865921,
      "learning_rate": 0.00013908696466889486,
      "loss": 0.3915,
      "num_input_tokens_seen": 19464368,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 0.14489254355430603,
      "learning_rate": 0.0001389287348816375,
      "loss": 0.4034,
      "num_input_tokens_seen": 19467792,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.20648010075092316,
      "learning_rate": 0.00013877058062636944,
      "loss": 0.3856,
      "num_input_tokens_seen": 19473584,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.22427858412265778,
      "learning_rate": 0.00013861250193617498,
      "loss": 0.4936,
      "num_input_tokens_seen": 19477936,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.09402719140052795,
      "learning_rate": 0.00013845449884412203,
      "loss": 0.4599,
      "num_input_tokens_seen": 19481488,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 0.14754585921764374,
      "learning_rate": 0.00013829657138326334,
      "loss": 0.5806,
      "num_input_tokens_seen": 19484432,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.3559110462665558,
      "learning_rate": 0.00013813871958663526,
      "loss": 0.4234,
      "num_input_tokens_seen": 19487376,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 0.10311909019947052,
      "learning_rate": 0.0001379809434872587,
      "loss": 0.4108,
      "num_input_tokens_seen": 19490224,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.1362171173095703,
      "learning_rate": 0.0001378232431181386,
      "loss": 0.3899,
      "num_input_tokens_seen": 19493072,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.16232894361019135,
      "learning_rate": 0.00013766561851226428,
      "loss": 0.3395,
      "num_input_tokens_seen": 19495952,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 0.21679016947746277,
      "learning_rate": 0.00013750806970260886,
      "loss": 0.5679,
      "num_input_tokens_seen": 19499440,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.15188443660736084,
      "learning_rate": 0.00013735059672212959,
      "loss": 0.4042,
      "num_input_tokens_seen": 19501904,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 0.14722195267677307,
      "learning_rate": 0.00013719319960376808,
      "loss": 0.5306,
      "num_input_tokens_seen": 19505200,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 0.24934175610542297,
      "learning_rate": 0.00013703587838045017,
      "loss": 0.3832,
      "num_input_tokens_seen": 19508784,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.24433234333992004,
      "learning_rate": 0.00013687863308508568,
      "loss": 0.3627,
      "num_input_tokens_seen": 19511312,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.20618213713169098,
      "learning_rate": 0.00013672146375056849,
      "loss": 0.3632,
      "num_input_tokens_seen": 19514064,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.20938435196876526,
      "learning_rate": 0.00013656437040977643,
      "loss": 0.4446,
      "num_input_tokens_seen": 19516816,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 0.23097188770771027,
      "learning_rate": 0.00013640735309557178,
      "loss": 0.4131,
      "num_input_tokens_seen": 19519888,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 0.1643245667219162,
      "learning_rate": 0.00013625041184080078,
      "loss": 0.4133,
      "num_input_tokens_seen": 19523472,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.12565761804580688,
      "learning_rate": 0.00013609354667829378,
      "loss": 0.308,
      "num_input_tokens_seen": 19527184,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 0.09192240983247757,
      "learning_rate": 0.00013593675764086521,
      "loss": 0.3918,
      "num_input_tokens_seen": 19531056,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.12012733519077301,
      "learning_rate": 0.00013578004476131356,
      "loss": 0.5211,
      "num_input_tokens_seen": 19534832,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 0.16507592797279358,
      "learning_rate": 0.00013562340807242112,
      "loss": 0.4078,
      "num_input_tokens_seen": 19538064,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.1423185020685196,
      "learning_rate": 0.00013546684760695472,
      "loss": 0.3275,
      "num_input_tokens_seen": 19541584,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 0.1302088499069214,
      "learning_rate": 0.00013531036339766494,
      "loss": 0.5082,
      "num_input_tokens_seen": 19544816,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 0.11944817006587982,
      "learning_rate": 0.00013515395547728675,
      "loss": 0.4756,
      "num_input_tokens_seen": 19547440,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 0.1598578542470932,
      "learning_rate": 0.00013499762387853866,
      "loss": 0.3949,
      "num_input_tokens_seen": 19550480,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 0.14598670601844788,
      "learning_rate": 0.00013484136863412333,
      "loss": 0.3888,
      "num_input_tokens_seen": 19554416,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 0.1623748391866684,
      "learning_rate": 0.00013468518977672773,
      "loss": 0.4122,
      "num_input_tokens_seen": 19557200,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.1976444274187088,
      "learning_rate": 0.0001345290873390227,
      "loss": 0.3794,
      "num_input_tokens_seen": 19560368,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.14321528375148773,
      "learning_rate": 0.00013437306135366322,
      "loss": 0.3317,
      "num_input_tokens_seen": 19563536,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.12351281195878983,
      "learning_rate": 0.00013421711185328806,
      "loss": 0.3918,
      "num_input_tokens_seen": 19566448,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 0.15858592092990875,
      "learning_rate": 0.00013406123887051984,
      "loss": 0.4233,
      "num_input_tokens_seen": 19569104,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.13020366430282593,
      "learning_rate": 0.0001339054424379656,
      "loss": 0.316,
      "num_input_tokens_seen": 19575952,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.17023757100105286,
      "learning_rate": 0.00013374972258821621,
      "loss": 0.5069,
      "num_input_tokens_seen": 19578864,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.15205606818199158,
      "learning_rate": 0.00013359407935384642,
      "loss": 0.43,
      "num_input_tokens_seen": 19582160,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.16460859775543213,
      "learning_rate": 0.0001334385127674152,
      "loss": 0.502,
      "num_input_tokens_seen": 19585392,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.17831768095493317,
      "learning_rate": 0.00013328302286146515,
      "loss": 0.5646,
      "num_input_tokens_seen": 19588976,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.7223708629608154,
      "learning_rate": 0.00013312760966852283,
      "loss": 0.4608,
      "num_input_tokens_seen": 19592496,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.1358892172574997,
      "learning_rate": 0.00013297227322109912,
      "loss": 0.5499,
      "num_input_tokens_seen": 19595888,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 0.12010820209980011,
      "learning_rate": 0.00013281701355168852,
      "loss": 0.3561,
      "num_input_tokens_seen": 19600272,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 0.1373126357793808,
      "learning_rate": 0.00013266183069276982,
      "loss": 0.3745,
      "num_input_tokens_seen": 19604272,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.1183888167142868,
      "learning_rate": 0.00013250672467680514,
      "loss": 0.4434,
      "num_input_tokens_seen": 19608176,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 0.1619797945022583,
      "learning_rate": 0.00013235169553624127,
      "loss": 0.3698,
      "num_input_tokens_seen": 19611728,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 0.12814755737781525,
      "learning_rate": 0.00013219674330350816,
      "loss": 0.3517,
      "num_input_tokens_seen": 19614896,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 0.24106663465499878,
      "learning_rate": 0.00013204186801102024,
      "loss": 0.4442,
      "num_input_tokens_seen": 19619024,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 0.2113795131444931,
      "learning_rate": 0.0001318870696911758,
      "loss": 0.4936,
      "num_input_tokens_seen": 19622448,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 0.10214617848396301,
      "learning_rate": 0.00013173234837635656,
      "loss": 0.4844,
      "num_input_tokens_seen": 19626096,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 0.09769435971975327,
      "learning_rate": 0.00013157770409892878,
      "loss": 0.4055,
      "num_input_tokens_seen": 19629168,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 0.17327779531478882,
      "learning_rate": 0.00013142313689124197,
      "loss": 0.3401,
      "num_input_tokens_seen": 19632464,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.16940078139305115,
      "learning_rate": 0.00013126864678562994,
      "loss": 0.5991,
      "num_input_tokens_seen": 19636016,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.1605699062347412,
      "learning_rate": 0.00013111423381441034,
      "loss": 0.4501,
      "num_input_tokens_seen": 19639184,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 0.12753728032112122,
      "learning_rate": 0.00013095989800988468,
      "loss": 0.3918,
      "num_input_tokens_seen": 19641648,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.33637526631355286,
      "learning_rate": 0.000130805639404338,
      "loss": 0.4776,
      "num_input_tokens_seen": 19644816,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.10924304276704788,
      "learning_rate": 0.0001306514580300397,
      "loss": 0.4493,
      "num_input_tokens_seen": 19647504,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.14238883554935455,
      "learning_rate": 0.00013049735391924244,
      "loss": 0.3845,
      "num_input_tokens_seen": 19650480,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 0.10928884893655777,
      "learning_rate": 0.00013034332710418328,
      "loss": 0.4496,
      "num_input_tokens_seen": 19653712,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 0.20466233789920807,
      "learning_rate": 0.00013018937761708298,
      "loss": 0.4409,
      "num_input_tokens_seen": 19656144,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 0.12794509530067444,
      "learning_rate": 0.0001300355054901457,
      "loss": 0.5713,
      "num_input_tokens_seen": 19658640,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 0.13444434106349945,
      "learning_rate": 0.0001298817107555599,
      "loss": 0.5243,
      "num_input_tokens_seen": 19661872,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 0.17632342875003815,
      "learning_rate": 0.0001297279934454978,
      "loss": 0.4038,
      "num_input_tokens_seen": 19665136,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 0.1677790880203247,
      "learning_rate": 0.000129574353592115,
      "loss": 0.4387,
      "num_input_tokens_seen": 19668112,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 0.19840872287750244,
      "learning_rate": 0.00012942079122755162,
      "loss": 0.3652,
      "num_input_tokens_seen": 19671312,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.1386357992887497,
      "learning_rate": 0.00012926730638393075,
      "loss": 0.4187,
      "num_input_tokens_seen": 19674064,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 0.16661888360977173,
      "learning_rate": 0.0001291138990933598,
      "loss": 0.3944,
      "num_input_tokens_seen": 19676752,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.20742107927799225,
      "learning_rate": 0.00012896056938792994,
      "loss": 0.4309,
      "num_input_tokens_seen": 19679376,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.23359373211860657,
      "learning_rate": 0.000128807317299716,
      "loss": 0.4614,
      "num_input_tokens_seen": 19681872,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.13625018298625946,
      "learning_rate": 0.00012865414286077637,
      "loss": 0.5544,
      "num_input_tokens_seen": 19684816,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 0.12043025344610214,
      "learning_rate": 0.00012850104610315365,
      "loss": 0.4126,
      "num_input_tokens_seen": 19690224,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.20402997732162476,
      "learning_rate": 0.00012834802705887372,
      "loss": 0.4515,
      "num_input_tokens_seen": 19693040,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 0.1367642730474472,
      "learning_rate": 0.00012819508575994643,
      "loss": 0.3173,
      "num_input_tokens_seen": 19695952,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 0.09711230546236038,
      "learning_rate": 0.0001280422222383656,
      "loss": 0.3793,
      "num_input_tokens_seen": 19699216,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.18558301031589508,
      "learning_rate": 0.00012788943652610818,
      "loss": 0.4826,
      "num_input_tokens_seen": 19702832,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 0.17896690964698792,
      "learning_rate": 0.0001277367286551356,
      "loss": 0.4683,
      "num_input_tokens_seen": 19705552,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.1430564671754837,
      "learning_rate": 0.00012758409865739213,
      "loss": 0.4527,
      "num_input_tokens_seen": 19709104,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.18970254063606262,
      "learning_rate": 0.00012743154656480655,
      "loss": 0.3915,
      "num_input_tokens_seen": 19713424,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 0.08011998236179352,
      "learning_rate": 0.00012727907240929094,
      "loss": 0.5219,
      "num_input_tokens_seen": 19716976,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 0.1494903266429901,
      "learning_rate": 0.00012712667622274127,
      "loss": 0.4243,
      "num_input_tokens_seen": 19720048,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.16788344085216522,
      "learning_rate": 0.000126974358037037,
      "loss": 0.4477,
      "num_input_tokens_seen": 19722704,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.14438490569591522,
      "learning_rate": 0.00012682211788404114,
      "loss": 0.4198,
      "num_input_tokens_seen": 19725872,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.18479523062705994,
      "learning_rate": 0.0001266699557956008,
      "loss": 0.5915,
      "num_input_tokens_seen": 19728336,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.17196929454803467,
      "learning_rate": 0.00012651787180354657,
      "loss": 0.3167,
      "num_input_tokens_seen": 19732304,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 0.11135955154895782,
      "learning_rate": 0.00012636586593969262,
      "loss": 0.3588,
      "num_input_tokens_seen": 19735952,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.25317874550819397,
      "learning_rate": 0.000126213938235837,
      "loss": 0.4124,
      "num_input_tokens_seen": 19739792,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 0.11223344504833221,
      "learning_rate": 0.0001260620887237611,
      "loss": 0.3513,
      "num_input_tokens_seen": 19742640,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 0.22421795129776,
      "learning_rate": 0.00012591031743523006,
      "loss": 0.489,
      "num_input_tokens_seen": 19745200,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.15522299706935883,
      "learning_rate": 0.00012575862440199277,
      "loss": 0.3807,
      "num_input_tokens_seen": 19748080,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 0.15853485465049744,
      "learning_rate": 0.00012560700965578165,
      "loss": 0.3993,
      "num_input_tokens_seen": 19750960,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.1676369309425354,
      "learning_rate": 0.0001254554732283129,
      "loss": 0.3764,
      "num_input_tokens_seen": 19755408,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.29448366165161133,
      "learning_rate": 0.0001253040151512862,
      "loss": 0.4874,
      "num_input_tokens_seen": 19758608,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 0.3180912435054779,
      "learning_rate": 0.0001251526354563846,
      "loss": 0.4399,
      "num_input_tokens_seen": 19762064,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 0.17462337017059326,
      "learning_rate": 0.0001250013341752752,
      "loss": 0.4557,
      "num_input_tokens_seen": 19765072,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.13476505875587463,
      "learning_rate": 0.00012485011133960843,
      "loss": 0.3975,
      "num_input_tokens_seen": 19768624,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.1388636827468872,
      "learning_rate": 0.00012469896698101863,
      "loss": 0.4004,
      "num_input_tokens_seen": 19771984,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 0.16610302031040192,
      "learning_rate": 0.00012454790113112325,
      "loss": 0.5666,
      "num_input_tokens_seen": 19776720,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.14852876961231232,
      "learning_rate": 0.00012439691382152345,
      "loss": 0.5333,
      "num_input_tokens_seen": 19779984,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.12519484758377075,
      "learning_rate": 0.00012424600508380412,
      "loss": 0.4412,
      "num_input_tokens_seen": 19782544,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 0.19608205556869507,
      "learning_rate": 0.00012409517494953377,
      "loss": 0.4035,
      "num_input_tokens_seen": 19785456,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.2282322496175766,
      "learning_rate": 0.00012394442345026418,
      "loss": 0.448,
      "num_input_tokens_seen": 19788624,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 0.13103806972503662,
      "learning_rate": 0.00012379375061753118,
      "loss": 0.3957,
      "num_input_tokens_seen": 19791088,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 0.14004874229431152,
      "learning_rate": 0.00012364315648285353,
      "loss": 0.3259,
      "num_input_tokens_seen": 19794576,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 0.12495126575231552,
      "learning_rate": 0.00012349264107773363,
      "loss": 0.3468,
      "num_input_tokens_seen": 19798160,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.12422256916761398,
      "learning_rate": 0.00012334220443365785,
      "loss": 0.3858,
      "num_input_tokens_seen": 19801040,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.15125302970409393,
      "learning_rate": 0.00012319184658209575,
      "loss": 0.3763,
      "num_input_tokens_seen": 19804016,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 0.13033097982406616,
      "learning_rate": 0.00012304156755450063,
      "loss": 0.4152,
      "num_input_tokens_seen": 19807504,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.2581271231174469,
      "learning_rate": 0.00012289136738230906,
      "loss": 0.5289,
      "num_input_tokens_seen": 19809712,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.1477566808462143,
      "learning_rate": 0.00012274124609694092,
      "loss": 0.3882,
      "num_input_tokens_seen": 19812496,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.14876441657543182,
      "learning_rate": 0.00012259120372980014,
      "loss": 0.4432,
      "num_input_tokens_seen": 19815568,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.15876781940460205,
      "learning_rate": 0.0001224412403122739,
      "loss": 0.3441,
      "num_input_tokens_seen": 19817808,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 0.14586445689201355,
      "learning_rate": 0.0001222913558757328,
      "loss": 0.3779,
      "num_input_tokens_seen": 19820432,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 0.16397494077682495,
      "learning_rate": 0.00012214155045153085,
      "loss": 0.527,
      "num_input_tokens_seen": 19823696,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.17409919202327728,
      "learning_rate": 0.00012199182407100584,
      "loss": 0.3824,
      "num_input_tokens_seen": 19827408,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 0.16453079879283905,
      "learning_rate": 0.00012184217676547855,
      "loss": 0.3972,
      "num_input_tokens_seen": 19830384,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.12977094948291779,
      "learning_rate": 0.00012169260856625358,
      "loss": 0.4529,
      "num_input_tokens_seen": 19833520,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 0.15708774328231812,
      "learning_rate": 0.000121543119504619,
      "loss": 0.4195,
      "num_input_tokens_seen": 19837360,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 0.14912401139736176,
      "learning_rate": 0.00012139370961184626,
      "loss": 0.4524,
      "num_input_tokens_seen": 19841232,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.33444616198539734,
      "learning_rate": 0.00012124437891918994,
      "loss": 0.4941,
      "num_input_tokens_seen": 19843728,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.1770254671573639,
      "learning_rate": 0.0001210951274578887,
      "loss": 0.4516,
      "num_input_tokens_seen": 19846512,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.11745202541351318,
      "learning_rate": 0.00012094595525916379,
      "loss": 0.4062,
      "num_input_tokens_seen": 19849360,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.1295435130596161,
      "learning_rate": 0.0001207968623542206,
      "loss": 0.4661,
      "num_input_tokens_seen": 19852720,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.1515352874994278,
      "learning_rate": 0.00012064784877424778,
      "loss": 0.4615,
      "num_input_tokens_seen": 19856592,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.13330088555812836,
      "learning_rate": 0.00012049891455041695,
      "loss": 0.4508,
      "num_input_tokens_seen": 19859728,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 0.14162112772464752,
      "learning_rate": 0.00012035005971388379,
      "loss": 0.5123,
      "num_input_tokens_seen": 19863120,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.3032267689704895,
      "learning_rate": 0.0001202012842957867,
      "loss": 0.4417,
      "num_input_tokens_seen": 19866256,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 0.19100402295589447,
      "learning_rate": 0.00012005258832724797,
      "loss": 0.4893,
      "num_input_tokens_seen": 19869008,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.27546560764312744,
      "learning_rate": 0.00011990397183937324,
      "loss": 0.4502,
      "num_input_tokens_seen": 19871600,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.3110720217227936,
      "learning_rate": 0.00011975543486325108,
      "loss": 0.4739,
      "num_input_tokens_seen": 19874224,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 0.22299887239933014,
      "learning_rate": 0.00011960697742995392,
      "loss": 0.3962,
      "num_input_tokens_seen": 19876848,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 0.0812748596072197,
      "learning_rate": 0.00011945859957053746,
      "loss": 0.3827,
      "num_input_tokens_seen": 19880432,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 0.10460903495550156,
      "learning_rate": 0.00011931030131604036,
      "loss": 0.3652,
      "num_input_tokens_seen": 19883184,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.1157841756939888,
      "learning_rate": 0.00011916208269748507,
      "loss": 0.316,
      "num_input_tokens_seen": 19886416,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 0.38839107751846313,
      "learning_rate": 0.00011901394374587743,
      "loss": 0.4162,
      "num_input_tokens_seen": 19889776,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.1467401534318924,
      "learning_rate": 0.00011886588449220609,
      "loss": 0.4118,
      "num_input_tokens_seen": 19893488,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.17859184741973877,
      "learning_rate": 0.00011871790496744351,
      "loss": 0.3591,
      "num_input_tokens_seen": 19896848,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.11966340988874435,
      "learning_rate": 0.00011857000520254546,
      "loss": 0.3762,
      "num_input_tokens_seen": 19901808,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 0.12207722663879395,
      "learning_rate": 0.0001184221852284506,
      "loss": 0.4045,
      "num_input_tokens_seen": 19905040,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.2544942796230316,
      "learning_rate": 0.00011827444507608143,
      "loss": 0.5211,
      "num_input_tokens_seen": 19908368,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.19212616980075836,
      "learning_rate": 0.00011812678477634325,
      "loss": 0.4026,
      "num_input_tokens_seen": 19911440,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.14714573323726654,
      "learning_rate": 0.0001179792043601251,
      "loss": 0.3329,
      "num_input_tokens_seen": 19915184,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 0.09123262017965317,
      "learning_rate": 0.00011783170385829905,
      "loss": 0.3127,
      "num_input_tokens_seen": 19918896,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.14201787114143372,
      "learning_rate": 0.00011768428330172071,
      "loss": 0.4243,
      "num_input_tokens_seen": 19922128,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 0.2668967545032501,
      "learning_rate": 0.00011753694272122856,
      "loss": 0.4916,
      "num_input_tokens_seen": 19924688,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 0.14085382223129272,
      "learning_rate": 0.00011738968214764456,
      "loss": 0.3634,
      "num_input_tokens_seen": 19928816,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.24447830021381378,
      "learning_rate": 0.00011724250161177391,
      "loss": 0.4268,
      "num_input_tokens_seen": 19931792,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.2650306522846222,
      "learning_rate": 0.00011709540114440525,
      "loss": 0.3565,
      "num_input_tokens_seen": 19934288,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.4923257529735565,
      "eval_runtime": 13.7583,
      "eval_samples_per_second": 61.635,
      "eval_steps_per_second": 15.409,
      "num_input_tokens_seen": 19936088,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.2703168988227844,
      "learning_rate": 0.00011694838077631043,
      "loss": 0.415,
      "num_input_tokens_seen": 19936984,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.28767409920692444,
      "learning_rate": 0.0001168014405382441,
      "loss": 0.4216,
      "num_input_tokens_seen": 19939704,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.09184684604406357,
      "learning_rate": 0.0001166545804609448,
      "loss": 0.4092,
      "num_input_tokens_seen": 19943928,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.2628633975982666,
      "learning_rate": 0.00011650780057513367,
      "loss": 0.5021,
      "num_input_tokens_seen": 19948088,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 0.15570934116840363,
      "learning_rate": 0.00011636110091151553,
      "loss": 0.514,
      "num_input_tokens_seen": 19950968,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.17734597623348236,
      "learning_rate": 0.00011621448150077834,
      "loss": 0.3633,
      "num_input_tokens_seen": 19953048,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.2000391036272049,
      "learning_rate": 0.0001160679423735933,
      "loss": 0.3872,
      "num_input_tokens_seen": 19956888,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.22025536000728607,
      "learning_rate": 0.00011592148356061455,
      "loss": 0.4271,
      "num_input_tokens_seen": 19960152,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.08329905569553375,
      "learning_rate": 0.00011577510509247951,
      "loss": 0.2645,
      "num_input_tokens_seen": 19962584,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.1490245759487152,
      "learning_rate": 0.00011562880699980904,
      "loss": 0.4472,
      "num_input_tokens_seen": 19965560,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.22418220341205597,
      "learning_rate": 0.00011548258931320704,
      "loss": 0.3832,
      "num_input_tokens_seen": 19968408,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.10587762296199799,
      "learning_rate": 0.00011533645206326049,
      "loss": 0.4308,
      "num_input_tokens_seen": 19972504,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.203156977891922,
      "learning_rate": 0.00011519039528053999,
      "loss": 0.4075,
      "num_input_tokens_seen": 19975224,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 0.3055334687232971,
      "learning_rate": 0.00011504441899559837,
      "loss": 0.5058,
      "num_input_tokens_seen": 19979096,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 0.15727882087230682,
      "learning_rate": 0.00011489852323897249,
      "loss": 0.4247,
      "num_input_tokens_seen": 19982232,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 0.15295204520225525,
      "learning_rate": 0.0001147527080411821,
      "loss": 0.6223,
      "num_input_tokens_seen": 19985816,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 0.18305476009845734,
      "learning_rate": 0.00011460697343273002,
      "loss": 0.4858,
      "num_input_tokens_seen": 19987960,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.178559809923172,
      "learning_rate": 0.00011446131944410249,
      "loss": 0.4379,
      "num_input_tokens_seen": 19991096,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 0.1742219179868698,
      "learning_rate": 0.00011431574610576844,
      "loss": 0.486,
      "num_input_tokens_seen": 19993656,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.15710526704788208,
      "learning_rate": 0.00011417025344818005,
      "loss": 0.5278,
      "num_input_tokens_seen": 19997624,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 0.12497246265411377,
      "learning_rate": 0.00011402484150177289,
      "loss": 0.5246,
      "num_input_tokens_seen": 20000248,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.19412140548229218,
      "learning_rate": 0.00011387951029696542,
      "loss": 0.3825,
      "num_input_tokens_seen": 20003352,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 0.1343112587928772,
      "learning_rate": 0.00011373425986415941,
      "loss": 0.3341,
      "num_input_tokens_seen": 20006968,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 0.20963598787784576,
      "learning_rate": 0.00011358909023373953,
      "loss": 0.3459,
      "num_input_tokens_seen": 20010296,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.07452138513326645,
      "learning_rate": 0.00011344400143607342,
      "loss": 0.4384,
      "num_input_tokens_seen": 20014136,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.14047099649906158,
      "learning_rate": 0.00011329899350151212,
      "loss": 0.3388,
      "num_input_tokens_seen": 20017848,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 0.17347553372383118,
      "learning_rate": 0.00011315406646038973,
      "loss": 0.4232,
      "num_input_tokens_seen": 20020984,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.11461217701435089,
      "learning_rate": 0.0001130092203430232,
      "loss": 0.5275,
      "num_input_tokens_seen": 20024792,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.1384919136762619,
      "learning_rate": 0.00011286445517971289,
      "loss": 0.4638,
      "num_input_tokens_seen": 20029208,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 0.16030719876289368,
      "learning_rate": 0.00011271977100074188,
      "loss": 0.4941,
      "num_input_tokens_seen": 20032120,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 0.16562725603580475,
      "learning_rate": 0.00011257516783637633,
      "loss": 0.332,
      "num_input_tokens_seen": 20035896,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.21347734332084656,
      "learning_rate": 0.00011243064571686573,
      "loss": 0.381,
      "num_input_tokens_seen": 20038232,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.16472628712654114,
      "learning_rate": 0.00011228620467244238,
      "loss": 0.5784,
      "num_input_tokens_seen": 20041016,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.17639528214931488,
      "learning_rate": 0.00011214184473332183,
      "loss": 0.3291,
      "num_input_tokens_seen": 20043832,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 0.17560306191444397,
      "learning_rate": 0.0001119975659297025,
      "loss": 0.4055,
      "num_input_tokens_seen": 20047576,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.09522248804569244,
      "learning_rate": 0.00011185336829176568,
      "loss": 0.4921,
      "num_input_tokens_seen": 20051288,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.12465832382440567,
      "learning_rate": 0.00011170925184967601,
      "loss": 0.3525,
      "num_input_tokens_seen": 20054264,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 0.17161913216114044,
      "learning_rate": 0.00011156521663358094,
      "loss": 0.5081,
      "num_input_tokens_seen": 20057112,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.17584539949893951,
      "learning_rate": 0.00011142126267361124,
      "loss": 0.4813,
      "num_input_tokens_seen": 20059608,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 0.09872372448444366,
      "learning_rate": 0.00011127738999988007,
      "loss": 0.3561,
      "num_input_tokens_seen": 20063064,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 0.24156554043293,
      "learning_rate": 0.00011113359864248429,
      "loss": 0.418,
      "num_input_tokens_seen": 20066616,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.21822257339954376,
      "learning_rate": 0.00011098988863150317,
      "loss": 0.3658,
      "num_input_tokens_seen": 20069848,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.28694960474967957,
      "learning_rate": 0.0001108462599969992,
      "loss": 0.4763,
      "num_input_tokens_seen": 20073688,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.10399723798036575,
      "learning_rate": 0.000110702712769018,
      "loss": 0.392,
      "num_input_tokens_seen": 20077016,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.18386702239513397,
      "learning_rate": 0.00011055924697758801,
      "loss": 0.4363,
      "num_input_tokens_seen": 20080408,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.14598849415779114,
      "learning_rate": 0.00011041586265272052,
      "loss": 0.3639,
      "num_input_tokens_seen": 20083640,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 0.2957543432712555,
      "learning_rate": 0.00011027255982441003,
      "loss": 0.3944,
      "num_input_tokens_seen": 20086616,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 0.19090533256530762,
      "learning_rate": 0.00011012933852263369,
      "loss": 0.3927,
      "num_input_tokens_seen": 20089592,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.1630672961473465,
      "learning_rate": 0.00010998619877735184,
      "loss": 0.3446,
      "num_input_tokens_seen": 20092952,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.12501776218414307,
      "learning_rate": 0.0001098431406185078,
      "loss": 0.4131,
      "num_input_tokens_seen": 20096120,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.16720746457576752,
      "learning_rate": 0.00010970016407602751,
      "loss": 0.467,
      "num_input_tokens_seen": 20100376,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.2430938333272934,
      "learning_rate": 0.00010955726917982023,
      "loss": 0.5038,
      "num_input_tokens_seen": 20104184,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.15038958191871643,
      "learning_rate": 0.00010941445595977767,
      "loss": 0.3962,
      "num_input_tokens_seen": 20107224,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 0.1604689359664917,
      "learning_rate": 0.00010927172444577494,
      "loss": 0.436,
      "num_input_tokens_seen": 20110008,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.11250144988298416,
      "learning_rate": 0.00010912907466766985,
      "loss": 0.4086,
      "num_input_tokens_seen": 20113560,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.16916070878505707,
      "learning_rate": 0.00010898650665530302,
      "loss": 0.5696,
      "num_input_tokens_seen": 20117720,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.1608601212501526,
      "learning_rate": 0.000108844020438498,
      "loss": 0.449,
      "num_input_tokens_seen": 20121496,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 0.1931924819946289,
      "learning_rate": 0.00010870161604706152,
      "loss": 0.3677,
      "num_input_tokens_seen": 20124312,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 0.13259267807006836,
      "learning_rate": 0.00010855929351078264,
      "loss": 0.5221,
      "num_input_tokens_seen": 20127608,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.26632434129714966,
      "learning_rate": 0.00010841705285943382,
      "loss": 0.4725,
      "num_input_tokens_seen": 20130168,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.0759875476360321,
      "learning_rate": 0.0001082748941227702,
      "loss": 0.5515,
      "num_input_tokens_seen": 20133368,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.256203293800354,
      "learning_rate": 0.00010813281733052959,
      "loss": 0.4323,
      "num_input_tokens_seen": 20136504,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 0.12806366384029388,
      "learning_rate": 0.00010799082251243292,
      "loss": 0.3363,
      "num_input_tokens_seen": 20139672,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.19217991828918457,
      "learning_rate": 0.00010784890969818407,
      "loss": 0.4943,
      "num_input_tokens_seen": 20142840,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 0.13286222517490387,
      "learning_rate": 0.00010770707891746928,
      "loss": 0.3931,
      "num_input_tokens_seen": 20145688,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 0.146938294172287,
      "learning_rate": 0.00010756533019995817,
      "loss": 0.4667,
      "num_input_tokens_seen": 20148216,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.14609839022159576,
      "learning_rate": 0.0001074236635753027,
      "loss": 0.5211,
      "num_input_tokens_seen": 20152280,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 0.1988513320684433,
      "learning_rate": 0.00010728207907313809,
      "loss": 0.435,
      "num_input_tokens_seen": 20154968,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 0.2678910791873932,
      "learning_rate": 0.0001071405767230822,
      "loss": 0.3953,
      "num_input_tokens_seen": 20158680,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 0.16310979425907135,
      "learning_rate": 0.0001069991565547358,
      "loss": 0.4381,
      "num_input_tokens_seen": 20161528,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.06573208421468735,
      "learning_rate": 0.00010685781859768223,
      "loss": 0.3361,
      "num_input_tokens_seen": 20167000,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.24836575984954834,
      "learning_rate": 0.00010671656288148768,
      "loss": 0.4142,
      "num_input_tokens_seen": 20170168,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 0.16930066049098969,
      "learning_rate": 0.00010657538943570138,
      "loss": 0.3966,
      "num_input_tokens_seen": 20175608,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.12675082683563232,
      "learning_rate": 0.00010643429828985518,
      "loss": 0.5315,
      "num_input_tokens_seen": 20178520,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 0.23994030058383942,
      "learning_rate": 0.0001062932894734639,
      "loss": 0.4163,
      "num_input_tokens_seen": 20181624,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.16837024688720703,
      "learning_rate": 0.00010615236301602476,
      "loss": 0.4684,
      "num_input_tokens_seen": 20184728,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 0.19278816878795624,
      "learning_rate": 0.00010601151894701794,
      "loss": 0.3471,
      "num_input_tokens_seen": 20188120,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 0.3460373282432556,
      "learning_rate": 0.0001058707572959065,
      "loss": 0.421,
      "num_input_tokens_seen": 20190456,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.11380910128355026,
      "learning_rate": 0.00010573007809213614,
      "loss": 0.4898,
      "num_input_tokens_seen": 20193944,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.14607098698616028,
      "learning_rate": 0.00010558948136513536,
      "loss": 0.4091,
      "num_input_tokens_seen": 20196568,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.116322822868824,
      "learning_rate": 0.00010544896714431557,
      "loss": 0.3422,
      "num_input_tokens_seen": 20199576,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.11078549921512604,
      "learning_rate": 0.0001053085354590706,
      "loss": 0.4593,
      "num_input_tokens_seen": 20202712,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.18580669164657593,
      "learning_rate": 0.00010516818633877695,
      "loss": 0.4985,
      "num_input_tokens_seen": 20205368,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.18481597304344177,
      "learning_rate": 0.00010502791981279425,
      "loss": 0.4723,
      "num_input_tokens_seen": 20209464,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.29111233353614807,
      "learning_rate": 0.00010488773591046469,
      "loss": 0.3963,
      "num_input_tokens_seen": 20213016,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.20352894067764282,
      "learning_rate": 0.0001047476346611132,
      "loss": 0.404,
      "num_input_tokens_seen": 20215640,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.19389234483242035,
      "learning_rate": 0.00010460761609404724,
      "loss": 0.3817,
      "num_input_tokens_seen": 20218648,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 0.11622629314661026,
      "learning_rate": 0.00010446768023855701,
      "loss": 0.326,
      "num_input_tokens_seen": 20223576,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.1795436590909958,
      "learning_rate": 0.00010432782712391559,
      "loss": 0.3575,
      "num_input_tokens_seen": 20226808,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.1724606603384018,
      "learning_rate": 0.00010418805677937871,
      "loss": 0.3254,
      "num_input_tokens_seen": 20229944,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.20447030663490295,
      "learning_rate": 0.00010404836923418465,
      "loss": 0.5754,
      "num_input_tokens_seen": 20233304,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 0.14697463810443878,
      "learning_rate": 0.00010390876451755477,
      "loss": 0.4635,
      "num_input_tokens_seen": 20237112,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 0.1463802605867386,
      "learning_rate": 0.0001037692426586922,
      "loss": 0.4145,
      "num_input_tokens_seen": 20239992,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 0.12903131544589996,
      "learning_rate": 0.0001036298036867837,
      "loss": 0.3588,
      "num_input_tokens_seen": 20243320,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 0.10111534595489502,
      "learning_rate": 0.00010349044763099819,
      "loss": 0.3417,
      "num_input_tokens_seen": 20247512,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.028684360906481743,
      "learning_rate": 0.00010335117452048742,
      "loss": 0.2584,
      "num_input_tokens_seen": 20254168,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 0.1491490751504898,
      "learning_rate": 0.00010321198438438589,
      "loss": 0.4317,
      "num_input_tokens_seen": 20258872,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 0.1828986257314682,
      "learning_rate": 0.00010307287725181036,
      "loss": 0.411,
      "num_input_tokens_seen": 20261976,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.24199825525283813,
      "learning_rate": 0.00010293385315186049,
      "loss": 0.3702,
      "num_input_tokens_seen": 20264696,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.21356624364852905,
      "learning_rate": 0.00010279491211361853,
      "loss": 0.426,
      "num_input_tokens_seen": 20267992,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.19211889803409576,
      "learning_rate": 0.00010265605416614938,
      "loss": 0.5465,
      "num_input_tokens_seen": 20271736,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.10591129958629608,
      "learning_rate": 0.00010251727933850069,
      "loss": 0.3315,
      "num_input_tokens_seen": 20274968,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 0.07908423244953156,
      "learning_rate": 0.00010237858765970231,
      "loss": 0.3938,
      "num_input_tokens_seen": 20278936,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 0.11848459392786026,
      "learning_rate": 0.00010223997915876726,
      "loss": 0.348,
      "num_input_tokens_seen": 20282296,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.17069055140018463,
      "learning_rate": 0.00010210145386469049,
      "loss": 0.3748,
      "num_input_tokens_seen": 20285464,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 0.15929768979549408,
      "learning_rate": 0.00010196301180645012,
      "loss": 0.4635,
      "num_input_tokens_seen": 20289176,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.1244526281952858,
      "learning_rate": 0.0001018246530130067,
      "loss": 0.4278,
      "num_input_tokens_seen": 20292504,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 0.11180970817804337,
      "learning_rate": 0.00010168637751330328,
      "loss": 0.5305,
      "num_input_tokens_seen": 20295992,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.25448712706565857,
      "learning_rate": 0.00010154818533626552,
      "loss": 0.4638,
      "num_input_tokens_seen": 20299672,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 0.2048693597316742,
      "learning_rate": 0.00010141007651080153,
      "loss": 0.3934,
      "num_input_tokens_seen": 20302520,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 0.14556510746479034,
      "learning_rate": 0.00010127205106580212,
      "loss": 0.4915,
      "num_input_tokens_seen": 20305272,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.21053878962993622,
      "learning_rate": 0.0001011341090301407,
      "loss": 0.428,
      "num_input_tokens_seen": 20307992,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 0.13532894849777222,
      "learning_rate": 0.00010099625043267336,
      "loss": 0.3305,
      "num_input_tokens_seen": 20310520,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.1536724716424942,
      "learning_rate": 0.00010085847530223812,
      "loss": 0.4228,
      "num_input_tokens_seen": 20312824,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.29362666606903076,
      "learning_rate": 0.00010072078366765641,
      "loss": 0.5431,
      "num_input_tokens_seen": 20315032,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.1622372716665268,
      "learning_rate": 0.00010058317555773139,
      "loss": 0.5125,
      "num_input_tokens_seen": 20318456,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 0.14018777012825012,
      "learning_rate": 0.00010044565100124925,
      "loss": 0.5032,
      "num_input_tokens_seen": 20321464,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.23009242117404938,
      "learning_rate": 0.00010030821002697871,
      "loss": 0.5014,
      "num_input_tokens_seen": 20324984,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.16910941898822784,
      "learning_rate": 0.00010017085266367054,
      "loss": 0.3957,
      "num_input_tokens_seen": 20327512,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.17676161229610443,
      "learning_rate": 0.00010003357894005854,
      "loss": 0.4305,
      "num_input_tokens_seen": 20330520,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 0.17726679146289825,
      "learning_rate": 9.989638888485885e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 20333368,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.12301202863454819,
      "learning_rate": 9.975928252676991e-05,
      "loss": 0.3693,
      "num_input_tokens_seen": 20336440,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 0.19090503454208374,
      "learning_rate": 9.962225989447288e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 20339160,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.1457090526819229,
      "learning_rate": 9.948532101663144e-05,
      "loss": 0.3151,
      "num_input_tokens_seen": 20342232,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.171113520860672,
      "learning_rate": 9.934846592189134e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 20345304,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 0.44284898042678833,
      "learning_rate": 9.921169463888153e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 20347736,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.17070218920707703,
      "learning_rate": 9.907500719621253e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 20350552,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.2536681294441223,
      "learning_rate": 9.893840362247809e-05,
      "loss": 0.2715,
      "num_input_tokens_seen": 20353432,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.15813402831554413,
      "learning_rate": 9.880188394625417e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 20356664,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.1408439576625824,
      "learning_rate": 9.86654481960989e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 20360152,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.16689401865005493,
      "learning_rate": 9.852909640055325e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 20363128,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 0.2542721927165985,
      "learning_rate": 9.839282858814047e-05,
      "loss": 0.3667,
      "num_input_tokens_seen": 20366168,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.13231442868709564,
      "learning_rate": 9.825664478736607e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 20368664,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 0.3399045169353485,
      "learning_rate": 9.812054502671835e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 20371864,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.21671287715435028,
      "learning_rate": 9.798452933466761e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 20374136,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.17304673790931702,
      "learning_rate": 9.784859773966693e-05,
      "loss": 0.4062,
      "num_input_tokens_seen": 20377304,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 0.13501958549022675,
      "learning_rate": 9.771275027015159e-05,
      "loss": 0.425,
      "num_input_tokens_seen": 20380504,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.13719670474529266,
      "learning_rate": 9.757698695453954e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 20384536,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.3397917151451111,
      "learning_rate": 9.74413078212305e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 20387640,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.2674510180950165,
      "learning_rate": 9.730571289860746e-05,
      "loss": 0.5446,
      "num_input_tokens_seen": 20390456,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.12806479632854462,
      "learning_rate": 9.717020221503493e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 20393592,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.1716957837343216,
      "learning_rate": 9.703477579886038e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 20396760,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.12432948499917984,
      "learning_rate": 9.689943367841347e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 20400504,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 0.15210233628749847,
      "learning_rate": 9.676417588200632e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 20405176,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.13576015830039978,
      "learning_rate": 9.662900243793321e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 20408888,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.1566624939441681,
      "learning_rate": 9.649391337447084e-05,
      "loss": 0.6191,
      "num_input_tokens_seen": 20412088,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 0.22023528814315796,
      "learning_rate": 9.635890871987829e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 20416120,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.13763085007667542,
      "learning_rate": 9.622398850239705e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 20420696,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.10678231716156006,
      "learning_rate": 9.608915275025104e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 20424248,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 0.2524702250957489,
      "learning_rate": 9.595440149164619e-05,
      "loss": 0.3968,
      "num_input_tokens_seen": 20427768,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 0.15434673428535461,
      "learning_rate": 9.581973475477085e-05,
      "loss": 0.3874,
      "num_input_tokens_seen": 20431224,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 0.17933832108974457,
      "learning_rate": 9.568515256779587e-05,
      "loss": 0.3779,
      "num_input_tokens_seen": 20434232,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 0.10631255805492401,
      "learning_rate": 9.555065495887433e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 20437720,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.242062047123909,
      "learning_rate": 9.541624195614152e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 20440696,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 0.19653873145580292,
      "learning_rate": 9.528191358771532e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 20443544,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 0.19900986552238464,
      "learning_rate": 9.514766988169549e-05,
      "loss": 0.5983,
      "num_input_tokens_seen": 20448280,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.15171265602111816,
      "learning_rate": 9.501351086616422e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 20452344,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.20965078473091125,
      "learning_rate": 9.487943656918613e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 20455480,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.1157369390130043,
      "learning_rate": 9.474544701880805e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 20458168,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.14451543986797333,
      "learning_rate": 9.461154224305923e-05,
      "loss": 0.2963,
      "num_input_tokens_seen": 20460984,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.1498684138059616,
      "learning_rate": 9.447772226995082e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 20464056,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 0.11832431703805923,
      "learning_rate": 9.434398712747639e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 20467096,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 0.3838707506656647,
      "learning_rate": 9.421033684361185e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 20469720,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 0.18887197971343994,
      "learning_rate": 9.407677144631533e-05,
      "loss": 0.6117,
      "num_input_tokens_seen": 20473240,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.16413237154483795,
      "learning_rate": 9.394329096352732e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 20476600,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 0.16760826110839844,
      "learning_rate": 9.380989542317037e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 20479448,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.21084661781787872,
      "learning_rate": 9.367658485314907e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 20483128,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.12310316413640976,
      "learning_rate": 9.354335928135066e-05,
      "loss": 0.4798,
      "num_input_tokens_seen": 20487896,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.26206061244010925,
      "learning_rate": 9.341021873564432e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 20491352,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.3529069423675537,
      "learning_rate": 9.327716324388164e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 20493912,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 0.1485096514225006,
      "learning_rate": 9.314419283389641e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 20499128,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 0.27621111273765564,
      "learning_rate": 9.30113075335044e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 20502648,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.1896938532590866,
      "learning_rate": 9.287850737050352e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 20506040,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 0.19303616881370544,
      "learning_rate": 9.274579237267422e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 20508984,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.1520790159702301,
      "learning_rate": 9.261316256777897e-05,
      "loss": 0.4624,
      "num_input_tokens_seen": 20512184,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.1618368774652481,
      "learning_rate": 9.24806179835625e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 20515128,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.26306745409965515,
      "learning_rate": 9.234815864775137e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 20518392,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.20160844922065735,
      "learning_rate": 9.221578458805485e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 20520952,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 0.11386403441429138,
      "learning_rate": 9.208349583216385e-05,
      "loss": 0.4541,
      "num_input_tokens_seen": 20523576,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 0.12457655370235443,
      "learning_rate": 9.195129240775174e-05,
      "loss": 0.3765,
      "num_input_tokens_seen": 20526872,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 0.10336605459451675,
      "learning_rate": 9.181917434247417e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 20531544,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.1880929172039032,
      "learning_rate": 9.168714166396835e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 20534392,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.13436934351921082,
      "learning_rate": 9.155519439985438e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 20537272,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 0.2045949399471283,
      "learning_rate": 9.142333257773383e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 20540568,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.19134557247161865,
      "learning_rate": 9.12915562251908e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 20543832,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 0.15001843869686127,
      "learning_rate": 9.115986536979149e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 20546808,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 0.2673313021659851,
      "learning_rate": 9.10282600390841e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 20549880,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.11389413475990295,
      "learning_rate": 9.08967402605988e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 20553816,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 0.26360809803009033,
      "learning_rate": 9.07653060618483e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 20558072,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.5112423896789551,
      "learning_rate": 9.063395747032676e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 20560632,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.26767048239707947,
      "learning_rate": 9.050269451351112e-05,
      "loss": 0.438,
      "num_input_tokens_seen": 20563960,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 0.15844188630580902,
      "learning_rate": 9.037151721886006e-05,
      "loss": 0.3332,
      "num_input_tokens_seen": 20566776,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.2180923968553543,
      "learning_rate": 9.024042561381424e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 20569400,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.1944580078125,
      "learning_rate": 9.010941972579656e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 20571480,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.17619135975837708,
      "learning_rate": 8.99784995822121e-05,
      "loss": 0.5062,
      "num_input_tokens_seen": 20574008,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 0.1551586538553238,
      "learning_rate": 8.984766521044769e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 20576696,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 0.18248192965984344,
      "learning_rate": 8.971691663787252e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 20579960,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.10993718355894089,
      "learning_rate": 8.958625389183756e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 20582936,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.11578256636857986,
      "learning_rate": 8.9455676999676e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 20586168,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.13046078383922577,
      "learning_rate": 8.932518598870309e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 20589112,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.139658585190773,
      "learning_rate": 8.919478088621614e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 20593240,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.10987629741430283,
      "learning_rate": 8.906446171949422e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 20596152,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 0.13221782445907593,
      "learning_rate": 8.893422851579885e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 20600152,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.1756429374217987,
      "learning_rate": 8.8804081302373e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 20603672,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 0.2859979271888733,
      "learning_rate": 8.867402010644221e-05,
      "loss": 0.3497,
      "num_input_tokens_seen": 20606104,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 0.09731169790029526,
      "learning_rate": 8.854404495521389e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 20609688,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.12679660320281982,
      "learning_rate": 8.841415587587709e-05,
      "loss": 0.2892,
      "num_input_tokens_seen": 20613336,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 0.27138829231262207,
      "learning_rate": 8.828435289560344e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 20615992,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.1547587513923645,
      "learning_rate": 8.815463604154588e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 20618520,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.1462714821100235,
      "learning_rate": 8.80250053408399e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 20622008,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 0.17479318380355835,
      "learning_rate": 8.789546082060273e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 20625240,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.24290338158607483,
      "learning_rate": 8.776600250793371e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 20628056,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 0.2883767783641815,
      "learning_rate": 8.763663042991399e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 20630776,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.1198028177022934,
      "learning_rate": 8.75073446136066e-05,
      "loss": 0.4038,
      "num_input_tokens_seen": 20634776,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 0.16558896005153656,
      "learning_rate": 8.737814508605674e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 20637400,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 0.11057262122631073,
      "learning_rate": 8.724903187429145e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 20641432,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 0.1179882362484932,
      "learning_rate": 8.71200050053198e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 20644696,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.124358169734478,
      "learning_rate": 8.699106450613287e-05,
      "loss": 0.3515,
      "num_input_tokens_seen": 20648088,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 0.12315940111875534,
      "learning_rate": 8.686221040370334e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 20652312,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 0.11147724837064743,
      "learning_rate": 8.673344272498596e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 20655288,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.12434571236371994,
      "learning_rate": 8.660476149691759e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 20658200,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.1891758143901825,
      "learning_rate": 8.647616674641684e-05,
      "loss": 0.3402,
      "num_input_tokens_seen": 20660696,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 0.14304126799106598,
      "learning_rate": 8.63476585003844e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 20663416,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 0.1599562168121338,
      "learning_rate": 8.621923678570259e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 20667352,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 0.20044976472854614,
      "learning_rate": 8.609090162923567e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 20669752,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.1568746715784073,
      "learning_rate": 8.596265305783002e-05,
      "loss": 0.3907,
      "num_input_tokens_seen": 20672664,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 0.2162720263004303,
      "learning_rate": 8.583449109831375e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 20675672,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 0.15915590524673462,
      "learning_rate": 8.570641577749705e-05,
      "loss": 0.4167,
      "num_input_tokens_seen": 20678520,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.10893993824720383,
      "learning_rate": 8.557842712217162e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 20680984,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.21064861118793488,
      "learning_rate": 8.545052515911112e-05,
      "loss": 0.5806,
      "num_input_tokens_seen": 20684184,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.1248994767665863,
      "learning_rate": 8.532270991507136e-05,
      "loss": 0.3641,
      "num_input_tokens_seen": 20687640,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.1531379073858261,
      "learning_rate": 8.519498141678983e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 20690392,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 0.1105019599199295,
      "learning_rate": 8.506733969098579e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 20694264,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.14648759365081787,
      "learning_rate": 8.49397847643606e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 20696888,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.17634683847427368,
      "learning_rate": 8.481231666359723e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 20700056,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.13347283005714417,
      "learning_rate": 8.468493541536031e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 20703000,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.1887809932231903,
      "learning_rate": 8.455764104629681e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 20706360,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 0.21165981888771057,
      "learning_rate": 8.443043358303515e-05,
      "loss": 0.3218,
      "num_input_tokens_seen": 20710328,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 0.20212428271770477,
      "learning_rate": 8.430331305218585e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 20713496,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.1255805641412735,
      "learning_rate": 8.417627948034096e-05,
      "loss": 0.2957,
      "num_input_tokens_seen": 20717240,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.17191922664642334,
      "learning_rate": 8.404933289407424e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 20720696,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.1550610065460205,
      "learning_rate": 8.392247331994174e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 20727736,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.14662057161331177,
      "learning_rate": 8.37957007844809e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 20730584,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 0.19528356194496155,
      "learning_rate": 8.366901531421134e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 20732984,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.25425052642822266,
      "learning_rate": 8.354241693563385e-05,
      "loss": 0.3157,
      "num_input_tokens_seen": 20736440,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 0.1851949840784073,
      "learning_rate": 8.341590567523166e-05,
      "loss": 0.34,
      "num_input_tokens_seen": 20743480,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 0.2665478587150574,
      "learning_rate": 8.328948155946924e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 20746360,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.11987520009279251,
      "learning_rate": 8.316314461479318e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 20750424,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.33107632398605347,
      "learning_rate": 8.303689486763177e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 20753112,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.2319696992635727,
      "learning_rate": 8.291073234439512e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 20756216,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.09579677134752274,
      "learning_rate": 8.27846570714747e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 20760792,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 0.1750054955482483,
      "learning_rate": 8.265866907524427e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 20764120,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.14126400649547577,
      "learning_rate": 8.253276838205892e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 20767384,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.23725228011608124,
      "learning_rate": 8.240695501825568e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 20770680,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.26810163259506226,
      "learning_rate": 8.228122901015345e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 20773080,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 0.1627776026725769,
      "learning_rate": 8.21555903840524e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 20775416,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.1278105229139328,
      "learning_rate": 8.203003916623491e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 20778840,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.17389671504497528,
      "learning_rate": 8.190457538296464e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 20781592,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.17180390655994415,
      "learning_rate": 8.177919906048736e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 20784856,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.1803843379020691,
      "learning_rate": 8.165391022503044e-05,
      "loss": 0.3617,
      "num_input_tokens_seen": 20787352,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.18825916945934296,
      "learning_rate": 8.152870890280261e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 20791416,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.2076369971036911,
      "learning_rate": 8.140359511999473e-05,
      "loss": 0.3537,
      "num_input_tokens_seen": 20794200,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.11613041907548904,
      "learning_rate": 8.127856890277923e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 20797720,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.1472306251525879,
      "learning_rate": 8.115363027730998e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 20801784,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 0.18440665304660797,
      "learning_rate": 8.102877926972286e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 20804984,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 0.31398189067840576,
      "learning_rate": 8.090401590613533e-05,
      "loss": 0.3478,
      "num_input_tokens_seen": 20808280,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.1533323973417282,
      "learning_rate": 8.077934021264627e-05,
      "loss": 0.3834,
      "num_input_tokens_seen": 20810584,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.1501612812280655,
      "learning_rate": 8.065475221533652e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 20813624,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.1781538724899292,
      "learning_rate": 8.053025194026858e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 20816504,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.19095417857170105,
      "learning_rate": 8.040583941348623e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 20820952,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.15138903260231018,
      "learning_rate": 8.028151466101541e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 20825240,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 0.14659005403518677,
      "learning_rate": 8.015727770886321e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 20828856,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.12443071603775024,
      "learning_rate": 8.00331285830187e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 20831928,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.2029862403869629,
      "learning_rate": 7.990906730945247e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 20836184,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.18721474707126617,
      "learning_rate": 7.978509391411681e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 20839544,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.21035094559192657,
      "learning_rate": 7.966120842294544e-05,
      "loss": 0.3382,
      "num_input_tokens_seen": 20842360,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 0.13315027952194214,
      "learning_rate": 7.953741086185368e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 20846040,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.12625901401042938,
      "learning_rate": 7.941370125673864e-05,
      "loss": 0.3578,
      "num_input_tokens_seen": 20848952,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.17613187432289124,
      "learning_rate": 7.9290079633479e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 20852184,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 0.13967029750347137,
      "learning_rate": 7.916654601793516e-05,
      "loss": 0.3335,
      "num_input_tokens_seen": 20855224,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.22954581677913666,
      "learning_rate": 7.904310043594859e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 20858808,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.24119774997234344,
      "learning_rate": 7.891974291334303e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 20861784,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.1508529782295227,
      "learning_rate": 7.879647347592322e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 20864888,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 0.13493044674396515,
      "learning_rate": 7.867329214947578e-05,
      "loss": 0.3183,
      "num_input_tokens_seen": 20867768,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.17051935195922852,
      "learning_rate": 7.855019895976889e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 20871096,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 0.188468337059021,
      "learning_rate": 7.842719393255232e-05,
      "loss": 0.5398,
      "num_input_tokens_seen": 20874424,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.1609705090522766,
      "learning_rate": 7.830427709355725e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 20877496,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 0.15303847193717957,
      "learning_rate": 7.818144846849634e-05,
      "loss": 0.56,
      "num_input_tokens_seen": 20880120,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.14176379144191742,
      "learning_rate": 7.805870808306403e-05,
      "loss": 0.3455,
      "num_input_tokens_seen": 20883128,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.4296056628227234,
      "learning_rate": 7.793605596293618e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 20885816,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 0.1724298596382141,
      "learning_rate": 7.781349213377048e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 20889016,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.17276161909103394,
      "learning_rate": 7.769101662120559e-05,
      "loss": 0.4925,
      "num_input_tokens_seen": 20891832,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 0.1234215572476387,
      "learning_rate": 7.756862945086196e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 20895128,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 0.23601923882961273,
      "learning_rate": 7.744633064834172e-05,
      "loss": 0.6022,
      "num_input_tokens_seen": 20897944,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.1207423284649849,
      "learning_rate": 7.732412023922836e-05,
      "loss": 0.5209,
      "num_input_tokens_seen": 20901112,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.09622332453727722,
      "learning_rate": 7.720199824908692e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 20904568,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 0.08774115145206451,
      "learning_rate": 7.707996470346402e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 20908184,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 0.17213574051856995,
      "learning_rate": 7.695801962788756e-05,
      "loss": 0.4328,
      "num_input_tokens_seen": 20911800,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.23233887553215027,
      "learning_rate": 7.683616304786695e-05,
      "loss": 0.4367,
      "num_input_tokens_seen": 20914040,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 0.12531282007694244,
      "learning_rate": 7.671439498889332e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 20917304,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 0.24161307513713837,
      "learning_rate": 7.65927154764392e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 20920344,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.13130532205104828,
      "learning_rate": 7.647112453595862e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 20923928,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 0.135581836104393,
      "learning_rate": 7.634962219288688e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 20926584,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.1810344159603119,
      "learning_rate": 7.622820847264083e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 20929656,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.5015066862106323,
      "learning_rate": 7.610688340061894e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 20932344,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.18638116121292114,
      "learning_rate": 7.598564700220101e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 20935864,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.1618979126214981,
      "learning_rate": 7.586449930274842e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 20938936,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.1310802400112152,
      "learning_rate": 7.574344032760367e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 20942520,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.23334935307502747,
      "learning_rate": 7.562247010209111e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 20945208,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.15911421179771423,
      "learning_rate": 7.550158865151618e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 20948344,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.17664559185504913,
      "learning_rate": 7.538079600116593e-05,
      "loss": 0.3563,
      "num_input_tokens_seen": 20951352,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.22787541151046753,
      "learning_rate": 7.526009217630886e-05,
      "loss": 0.62,
      "num_input_tokens_seen": 20953848,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 0.19712182879447937,
      "learning_rate": 7.513947720219494e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 20957080,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 0.1461261659860611,
      "learning_rate": 7.501895110405533e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 20961336,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 0.20099440217018127,
      "learning_rate": 7.489851390710262e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 20963864,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 0.13963761925697327,
      "learning_rate": 7.477816563653095e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 20966872,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 0.2680201530456543,
      "learning_rate": 7.465790631751584e-05,
      "loss": 0.425,
      "num_input_tokens_seen": 20969432,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 0.24379949271678925,
      "learning_rate": 7.453773597521429e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 20972312,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.10496032983064651,
      "learning_rate": 7.44176546347643e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 20976344,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 0.12550866603851318,
      "learning_rate": 7.429766232128583e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 20980408,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 0.1746000498533249,
      "learning_rate": 7.417775905987956e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 20983928,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.15318281948566437,
      "learning_rate": 7.405794487562811e-05,
      "loss": 0.3455,
      "num_input_tokens_seen": 20986936,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 0.1158071756362915,
      "learning_rate": 7.393821979359528e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 20990200,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.20027920603752136,
      "learning_rate": 7.3818583838826e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 20993368,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.1374317854642868,
      "learning_rate": 7.369903703634684e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 20996888,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.13745619356632233,
      "learning_rate": 7.357957941116572e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 21000152,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.207760289311409,
      "learning_rate": 7.346021098827166e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 21004056,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.1291755884885788,
      "learning_rate": 7.334093179263518e-05,
      "loss": 0.359,
      "num_input_tokens_seen": 21006680,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 0.18136779963970184,
      "learning_rate": 7.322174184920837e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 21010104,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.19746649265289307,
      "learning_rate": 7.310264118292404e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 21014072,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 0.13521088659763336,
      "learning_rate": 7.298362981869705e-05,
      "loss": 0.3868,
      "num_input_tokens_seen": 21017496,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.19665198028087616,
      "learning_rate": 7.286470778142284e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 21020472,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 0.14540134370326996,
      "learning_rate": 7.274587509597886e-05,
      "loss": 0.3835,
      "num_input_tokens_seen": 21023896,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.19092980027198792,
      "learning_rate": 7.262713178722346e-05,
      "loss": 0.5965,
      "num_input_tokens_seen": 21026904,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 0.24550484120845795,
      "learning_rate": 7.250847787999625e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 21029944,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 0.14159396290779114,
      "learning_rate": 7.238991339911844e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 21032504,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.09542216360569,
      "learning_rate": 7.227143836939237e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 21036600,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.262131005525589,
      "learning_rate": 7.215305281560153e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 21039768,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.11544971913099289,
      "learning_rate": 7.203475676251104e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 21043256,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 0.22842833399772644,
      "learning_rate": 7.191655023486682e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 21046072,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 0.23470817506313324,
      "learning_rate": 7.179843325739644e-05,
      "loss": 0.3845,
      "num_input_tokens_seen": 21050424,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.17413011193275452,
      "learning_rate": 7.168040585480861e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 21053144,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.192337766289711,
      "learning_rate": 7.156246805179351e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 21056856,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 0.1676790714263916,
      "learning_rate": 7.144461987302208e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 21060792,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 0.1590968370437622,
      "learning_rate": 7.132686134314714e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 21063768,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.29484760761260986,
      "learning_rate": 7.120919248680208e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 21066680,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.22196249663829803,
      "learning_rate": 7.109161332860203e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 21070968,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.2175784409046173,
      "learning_rate": 7.097412389314322e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 21075032,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 0.25405246019363403,
      "learning_rate": 7.085672420500322e-05,
      "loss": 0.3584,
      "num_input_tokens_seen": 21078712,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 0.22480635344982147,
      "learning_rate": 7.073941428874064e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 21082168,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.1153365895152092,
      "learning_rate": 7.062219416889514e-05,
      "loss": 0.2289,
      "num_input_tokens_seen": 21085304,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.09552455693483353,
      "learning_rate": 7.050506386998806e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 21089144,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 0.1008165180683136,
      "learning_rate": 7.038802341652172e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 21092504,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.26591670513153076,
      "learning_rate": 7.027107283297967e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 21095352,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 0.17387069761753082,
      "learning_rate": 7.015421214382661e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 21098104,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.2641295790672302,
      "learning_rate": 7.003744137350827e-05,
      "loss": 0.2813,
      "num_input_tokens_seen": 21100696,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.15300080180168152,
      "learning_rate": 6.992076054645197e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 21103736,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.15373465418815613,
      "learning_rate": 6.980416968706594e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 21107384,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 0.376687616109848,
      "learning_rate": 6.968766881973965e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 21111096,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.11716236174106598,
      "learning_rate": 6.957125796884395e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 21114808,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.17259754240512848,
      "learning_rate": 6.945493715873046e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 21117560,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.11556055396795273,
      "learning_rate": 6.933870641373208e-05,
      "loss": 0.2995,
      "num_input_tokens_seen": 21120472,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.31648749113082886,
      "learning_rate": 6.92225657581631e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 21122840,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.17083615064620972,
      "learning_rate": 6.910651521631877e-05,
      "loss": 0.5876,
      "num_input_tokens_seen": 21126392,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 0.12559901177883148,
      "learning_rate": 6.899055481247568e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 21129336,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.08119206130504608,
      "learning_rate": 6.88746845708913e-05,
      "loss": 0.3611,
      "num_input_tokens_seen": 21133560,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 0.15897248685359955,
      "learning_rate": 6.87589045158043e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 21136408,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 0.19362296164035797,
      "learning_rate": 6.864321467143459e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 21139224,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.15686815977096558,
      "learning_rate": 6.852761506198319e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 21141848,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.18943090736865997,
      "learning_rate": 6.841210571163231e-05,
      "loss": 0.3553,
      "num_input_tokens_seen": 21146104,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 0.23523856699466705,
      "learning_rate": 6.829668664454513e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 21149816,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.14118804037570953,
      "learning_rate": 6.818135788486584e-05,
      "loss": 0.3513,
      "num_input_tokens_seen": 21152824,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.18942227959632874,
      "learning_rate": 6.806611945672004e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 21156472,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.14316639304161072,
      "learning_rate": 6.79509713842143e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 21159864,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 0.46442198753356934,
      "learning_rate": 6.783591369143626e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 21162264,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.20947222411632538,
      "learning_rate": 6.77209464024548e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 21165816,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 0.17754262685775757,
      "learning_rate": 6.760606954131965e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 21168536,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.2137916088104248,
      "learning_rate": 6.749128313206165e-05,
      "loss": 0.4959,
      "num_input_tokens_seen": 21171448,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 0.23518958687782288,
      "learning_rate": 6.737658719869288e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 21174552,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 0.18252937495708466,
      "learning_rate": 6.726198176520642e-05,
      "loss": 0.3938,
      "num_input_tokens_seen": 21177688,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.11951375752687454,
      "learning_rate": 6.71474668555766e-05,
      "loss": 0.6919,
      "num_input_tokens_seen": 21181112,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 0.13726690411567688,
      "learning_rate": 6.703304249375836e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 21184856,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.4925762712955475,
      "eval_runtime": 13.8503,
      "eval_samples_per_second": 61.226,
      "eval_steps_per_second": 15.306,
      "num_input_tokens_seen": 21184936,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.12452757358551025,
      "learning_rate": 6.691870870368815e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 21187656,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.16947707533836365,
      "learning_rate": 6.680446550928316e-05,
      "loss": 0.3194,
      "num_input_tokens_seen": 21192744,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 0.24310483038425446,
      "learning_rate": 6.669031293444177e-05,
      "loss": 0.6774,
      "num_input_tokens_seen": 21195368,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.17910346388816833,
      "learning_rate": 6.65762510030436e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 21198728,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.1611778289079666,
      "learning_rate": 6.646227973894886e-05,
      "loss": 0.3275,
      "num_input_tokens_seen": 21201768,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.15818949043750763,
      "learning_rate": 6.634839916599921e-05,
      "loss": 0.3926,
      "num_input_tokens_seen": 21204520,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.1213599145412445,
      "learning_rate": 6.623460930801701e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 21207560,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.1282043755054474,
      "learning_rate": 6.61209101888059e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 21210536,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 0.1115058958530426,
      "learning_rate": 6.600730183215043e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 21214440,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 0.5200253129005432,
      "learning_rate": 6.589378426181624e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 21217128,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 0.14155489206314087,
      "learning_rate": 6.578035750154976e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 21220808,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.17278362810611725,
      "learning_rate": 6.566702157507875e-05,
      "loss": 0.2838,
      "num_input_tokens_seen": 21224456,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 0.14674106240272522,
      "learning_rate": 6.555377650611155e-05,
      "loss": 0.3263,
      "num_input_tokens_seen": 21227112,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.12997651100158691,
      "learning_rate": 6.544062231833792e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 21230152,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.45030152797698975,
      "learning_rate": 6.532755903542847e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 21232904,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.15076641738414764,
      "learning_rate": 6.521458668103458e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 21235816,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.14793530106544495,
      "learning_rate": 6.510170527878889e-05,
      "loss": 0.3335,
      "num_input_tokens_seen": 21239112,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 0.17851027846336365,
      "learning_rate": 6.498891485230491e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 21241928,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 0.25319311022758484,
      "learning_rate": 6.487621542517703e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 21248616,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 0.1862017810344696,
      "learning_rate": 6.476360702098078e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 21251464,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.1621546596288681,
      "learning_rate": 6.465108966327243e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 21254408,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.21552762389183044,
      "learning_rate": 6.453866337558939e-05,
      "loss": 0.5574,
      "num_input_tokens_seen": 21257704,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 0.15658774971961975,
      "learning_rate": 6.442632818145011e-05,
      "loss": 0.3381,
      "num_input_tokens_seen": 21260040,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.12217100709676743,
      "learning_rate": 6.431408410435352e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 21262664,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.07691247016191483,
      "learning_rate": 6.420193116778e-05,
      "loss": 0.3552,
      "num_input_tokens_seen": 21266728,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 0.16303318738937378,
      "learning_rate": 6.408986939519074e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 21269416,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.1422528773546219,
      "learning_rate": 6.397789881002752e-05,
      "loss": 0.3185,
      "num_input_tokens_seen": 21271944,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.23502717912197113,
      "learning_rate": 6.386601943571352e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 21274728,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.11451420933008194,
      "learning_rate": 6.375423129565266e-05,
      "loss": 0.369,
      "num_input_tokens_seen": 21278184,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 0.11905203759670258,
      "learning_rate": 6.36425344132296e-05,
      "loss": 0.5166,
      "num_input_tokens_seen": 21281288,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.2619273066520691,
      "learning_rate": 6.353092881181016e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 21284232,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.20513278245925903,
      "learning_rate": 6.341941451474082e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 21286472,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.26455995440483093,
      "learning_rate": 6.330799154534921e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 21289704,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.17391394078731537,
      "learning_rate": 6.319665992694368e-05,
      "loss": 0.359,
      "num_input_tokens_seen": 21292392,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.13133490085601807,
      "learning_rate": 6.308541968281373e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 21295624,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.13990579545497894,
      "learning_rate": 6.297427083622936e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 21298248,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.19056811928749084,
      "learning_rate": 6.28632134104416e-05,
      "loss": 0.5544,
      "num_input_tokens_seen": 21302120,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.23679663240909576,
      "learning_rate": 6.275224742868247e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 21304744,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.10131235420703888,
      "learning_rate": 6.264137291416477e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 21309000,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.3233333230018616,
      "learning_rate": 6.253058989008226e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 21311688,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.21275508403778076,
      "learning_rate": 6.241989837960949e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 21314568,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.2624080777168274,
      "learning_rate": 6.230929840590177e-05,
      "loss": 0.339,
      "num_input_tokens_seen": 21321512,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.13485604524612427,
      "learning_rate": 6.219878999209533e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 21324264,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.10396567732095718,
      "learning_rate": 6.208837316130733e-05,
      "loss": 0.3439,
      "num_input_tokens_seen": 21327944,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.15723049640655518,
      "learning_rate": 6.197804793663564e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 21330632,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.32037588953971863,
      "learning_rate": 6.18678143411593e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 21332712,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.2240595668554306,
      "learning_rate": 6.175767239793767e-05,
      "loss": 0.4177,
      "num_input_tokens_seen": 21336616,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.24336323142051697,
      "learning_rate": 6.164762213001112e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 21339848,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 0.1712600439786911,
      "learning_rate": 6.153766356040107e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 21342888,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.18319059908390045,
      "learning_rate": 6.142779671210951e-05,
      "loss": 0.34,
      "num_input_tokens_seen": 21346440,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.16125822067260742,
      "learning_rate": 6.131802160811956e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 21348808,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.15547814965248108,
      "learning_rate": 6.120833827139466e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 21350984,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.14280878007411957,
      "learning_rate": 6.109874672487936e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 21354888,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 0.2993508279323578,
      "learning_rate": 6.098924699149894e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 21357352,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.23472470045089722,
      "learning_rate": 6.087983909415962e-05,
      "loss": 0.6382,
      "num_input_tokens_seen": 21360104,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.165735125541687,
      "learning_rate": 6.077052305574815e-05,
      "loss": 0.3397,
      "num_input_tokens_seen": 21363144,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.1169019415974617,
      "learning_rate": 6.066129889913241e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 21367432,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.15653592348098755,
      "learning_rate": 6.0552166647160676e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 21370984,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.15371182560920715,
      "learning_rate": 6.044312632266208e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 21374024,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 0.15587425231933594,
      "learning_rate": 6.03341779484467e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 21377160,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 0.18634530901908875,
      "learning_rate": 6.022532154730537e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 21380904,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.13509753346443176,
      "learning_rate": 6.011655714200964e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 21383880,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.15272529423236847,
      "learning_rate": 6.0007884755311715e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 21387048,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 0.24760524928569794,
      "learning_rate": 5.989930440994451e-05,
      "loss": 0.2713,
      "num_input_tokens_seen": 21390568,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 0.16573162376880646,
      "learning_rate": 5.979081612862186e-05,
      "loss": 0.3346,
      "num_input_tokens_seen": 21392904,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.23356333374977112,
      "learning_rate": 5.968241993403834e-05,
      "loss": 0.3125,
      "num_input_tokens_seen": 21396072,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.15703098475933075,
      "learning_rate": 5.957411584886924e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 21399208,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 0.20517238974571228,
      "learning_rate": 5.946590389577033e-05,
      "loss": 0.3161,
      "num_input_tokens_seen": 21402248,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.20674194395542145,
      "learning_rate": 5.935778409737857e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 21405480,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.29224857687950134,
      "learning_rate": 5.92497564763112e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 21408168,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.2869126498699188,
      "learning_rate": 5.914182105516641e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 21411112,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.19021840393543243,
      "learning_rate": 5.903397785652304e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 21414856,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.22653083503246307,
      "learning_rate": 5.8926226902940804e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 21417736,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 0.1621048003435135,
      "learning_rate": 5.88185682169598e-05,
      "loss": 0.373,
      "num_input_tokens_seen": 21421352,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.09369970113039017,
      "learning_rate": 5.871100182110117e-05,
      "loss": 0.4799,
      "num_input_tokens_seen": 21425352,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.1532381922006607,
      "learning_rate": 5.8603527737866314e-05,
      "loss": 0.3505,
      "num_input_tokens_seen": 21428232,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.116681769490242,
      "learning_rate": 5.849614598973779e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 21430984,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 0.21422627568244934,
      "learning_rate": 5.8388856599178644e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 21434152,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.13492606580257416,
      "learning_rate": 5.8281659588632464e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 21438088,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.1911870539188385,
      "learning_rate": 5.817455498052382e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 21442408,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.19315452873706818,
      "learning_rate": 5.806754279725751e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 21444680,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 0.20671972632408142,
      "learning_rate": 5.796062306121947e-05,
      "loss": 0.4182,
      "num_input_tokens_seen": 21447272,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 0.1462617814540863,
      "learning_rate": 5.785379579477607e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 21450120,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.13050304353237152,
      "learning_rate": 5.774706102027427e-05,
      "loss": 0.3441,
      "num_input_tokens_seen": 21453736,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.2006949484348297,
      "learning_rate": 5.7640418760041776e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 21457512,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.27250656485557556,
      "learning_rate": 5.753386903638713e-05,
      "loss": 0.356,
      "num_input_tokens_seen": 21460104,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.09895804524421692,
      "learning_rate": 5.7427411871599e-05,
      "loss": 0.3935,
      "num_input_tokens_seen": 21463048,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.12705457210540771,
      "learning_rate": 5.7321047287947235e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 21465288,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.20880186557769775,
      "learning_rate": 5.72147753076821e-05,
      "loss": 0.6149,
      "num_input_tokens_seen": 21468552,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.720123827457428,
      "learning_rate": 5.7108595953034306e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 21471784,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.34526559710502625,
      "learning_rate": 5.7002509246215415e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 21474312,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 0.2116483747959137,
      "learning_rate": 5.68965152094178e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 21477640,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 0.13260850310325623,
      "learning_rate": 5.679061386481388e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 21480744,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.14963537454605103,
      "learning_rate": 5.6684805234557216e-05,
      "loss": 0.4379,
      "num_input_tokens_seen": 21483816,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.1444205641746521,
      "learning_rate": 5.65790893407816e-05,
      "loss": 0.3171,
      "num_input_tokens_seen": 21487080,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.4482099711894989,
      "learning_rate": 5.647346620560168e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 21490088,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.33550167083740234,
      "learning_rate": 5.6367935851112605e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 21493672,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.12845838069915771,
      "learning_rate": 5.626249829939023e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 21496776,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.1789470911026001,
      "learning_rate": 5.6157153572490795e-05,
      "loss": 0.3174,
      "num_input_tokens_seen": 21499208,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.14584848284721375,
      "learning_rate": 5.6051901692451115e-05,
      "loss": 0.3048,
      "num_input_tokens_seen": 21503368,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.17046412825584412,
      "learning_rate": 5.5946742681288744e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 21506824,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.2518687844276428,
      "learning_rate": 5.5841676561001805e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 21509384,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 0.2247796654701233,
      "learning_rate": 5.573670335356901e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 21512872,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 0.16767363250255585,
      "learning_rate": 5.563182308094933e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 21517000,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.24467255175113678,
      "learning_rate": 5.552703576508272e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 21519272,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.11859647929668427,
      "learning_rate": 5.5422341427889324e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 21522536,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 0.09523256868124008,
      "learning_rate": 5.531774009127011e-05,
      "loss": 0.5942,
      "num_input_tokens_seen": 21527240,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.1383872777223587,
      "learning_rate": 5.521323177710647e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 21530984,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 0.257706880569458,
      "learning_rate": 5.510881650726046e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 21534344,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.16152746975421906,
      "learning_rate": 5.50044943035744e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 21537576,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.14891678094863892,
      "learning_rate": 5.490026518787128e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 21540488,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 0.15794678032398224,
      "learning_rate": 5.47961291819547e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 21543592,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 0.14769822359085083,
      "learning_rate": 5.46920863076088e-05,
      "loss": 0.3816,
      "num_input_tokens_seen": 21546760,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.14900720119476318,
      "learning_rate": 5.4588136586598056e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 21549608,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 0.14366599917411804,
      "learning_rate": 5.44842800406678e-05,
      "loss": 0.3234,
      "num_input_tokens_seen": 21553736,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.1576109081506729,
      "learning_rate": 5.438051669154326e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 21556648,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.18359020352363586,
      "learning_rate": 5.427684656093074e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 21559912,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.17451313138008118,
      "learning_rate": 5.417326967051683e-05,
      "loss": 0.3621,
      "num_input_tokens_seen": 21562344,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.13069263100624084,
      "learning_rate": 5.4069786041968663e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 21565288,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.15878558158874512,
      "learning_rate": 5.39663956969339e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 21567816,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 0.10594012588262558,
      "learning_rate": 5.3863098657040534e-05,
      "loss": 0.6142,
      "num_input_tokens_seen": 21571912,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.3602842688560486,
      "learning_rate": 5.375989494389705e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 21574600,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.1502152383327484,
      "learning_rate": 5.365678457909257e-05,
      "loss": 0.3541,
      "num_input_tokens_seen": 21577864,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.1277158260345459,
      "learning_rate": 5.3553767584196555e-05,
      "loss": 0.7593,
      "num_input_tokens_seen": 21580552,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 0.21033728122711182,
      "learning_rate": 5.3450843980759166e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 21583880,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.1649927794933319,
      "learning_rate": 5.3348013790310736e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 21587432,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.3073296546936035,
      "learning_rate": 5.324527703436199e-05,
      "loss": 0.338,
      "num_input_tokens_seen": 21590376,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 0.1996297836303711,
      "learning_rate": 5.314263373440448e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 21592936,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 0.31247270107269287,
      "learning_rate": 5.304008391190995e-05,
      "loss": 0.6415,
      "num_input_tokens_seen": 21596040,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.2852339446544647,
      "learning_rate": 5.293762758833071e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 21599560,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.1218380257487297,
      "learning_rate": 5.283526478509953e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 21602344,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.11746138334274292,
      "learning_rate": 5.273299552362942e-05,
      "loss": 0.3327,
      "num_input_tokens_seen": 21607016,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.09783264994621277,
      "learning_rate": 5.263081982531381e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 21610344,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.2561156749725342,
      "learning_rate": 5.2528737711526885e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 21613288,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.17240439355373383,
      "learning_rate": 5.2426749203623005e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 21616456,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 0.215982124209404,
      "learning_rate": 5.23248543229371e-05,
      "loss": 0.6062,
      "num_input_tokens_seen": 21619272,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.14433543384075165,
      "learning_rate": 5.222305309078429e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 21622216,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.17561034858226776,
      "learning_rate": 5.2121345528460205e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 21625704,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.17785368859767914,
      "learning_rate": 5.201973165724094e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 21629288,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.17039033770561218,
      "learning_rate": 5.1918211498382996e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 21632264,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.3396422266960144,
      "learning_rate": 5.181678507312332e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 21636584,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.2275839000940323,
      "learning_rate": 5.171545240267894e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 21640168,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.43213117122650146,
      "learning_rate": 5.161421350824774e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 21643176,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.11716420203447342,
      "learning_rate": 5.151306841100756e-05,
      "loss": 0.5282,
      "num_input_tokens_seen": 21647112,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.2013532668352127,
      "learning_rate": 5.141201713211679e-05,
      "loss": 0.3469,
      "num_input_tokens_seen": 21650920,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.17933636903762817,
      "learning_rate": 5.1311059692714324e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 21653864,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 0.1833985596895218,
      "learning_rate": 5.1210196113919306e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 21656296,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.12762850522994995,
      "learning_rate": 5.110942641683114e-05,
      "loss": 0.3036,
      "num_input_tokens_seen": 21659240,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.18190352618694305,
      "learning_rate": 5.10087506225298e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 21661768,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.2710205614566803,
      "learning_rate": 5.090816875207543e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 21664712,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.14129574596881866,
      "learning_rate": 5.0807680826508604e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 21668008,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 0.2668319642543793,
      "learning_rate": 5.070728686685039e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 21670792,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 0.21203643083572388,
      "learning_rate": 5.0606986894101925e-05,
      "loss": 0.3324,
      "num_input_tokens_seen": 21673160,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.16453906893730164,
      "learning_rate": 5.0506780929244875e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 21677000,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 0.09915203601121902,
      "learning_rate": 5.040666899324109e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 21680424,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.1586768627166748,
      "learning_rate": 5.0306651107032916e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 21683752,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.16905954480171204,
      "learning_rate": 5.020672729154308e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 21686280,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 0.1653849184513092,
      "learning_rate": 5.010689756767428e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 21689928,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.23705323040485382,
      "learning_rate": 5.000716195630983e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 21692456,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.2562268376350403,
      "learning_rate": 4.990752047831343e-05,
      "loss": 0.3784,
      "num_input_tokens_seen": 21695112,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.14271117746829987,
      "learning_rate": 4.980797315452873e-05,
      "loss": 0.3503,
      "num_input_tokens_seen": 21698696,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.1724715679883957,
      "learning_rate": 4.9708520005779965e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 21702664,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 0.21856023371219635,
      "learning_rate": 4.960916105287172e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 21705960,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.18765057623386383,
      "learning_rate": 4.950989631658859e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 21710472,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.15538851916790009,
      "learning_rate": 4.941072581769562e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 21713288,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 0.14643383026123047,
      "learning_rate": 4.931164957693834e-05,
      "loss": 0.3085,
      "num_input_tokens_seen": 21716584,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 0.15962587296962738,
      "learning_rate": 4.921266761504217e-05,
      "loss": 0.5123,
      "num_input_tokens_seen": 21719752,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 0.18451271951198578,
      "learning_rate": 4.9113779952713155e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 21723112,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.1966363787651062,
      "learning_rate": 4.90149866106373e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 21726184,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.257609099149704,
      "learning_rate": 4.8916287609481137e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 21729416,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.13971306383609772,
      "learning_rate": 4.8817682969891366e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 21732648,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.2713840901851654,
      "learning_rate": 4.8719172712495095e-05,
      "loss": 0.31,
      "num_input_tokens_seen": 21736520,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.11089646816253662,
      "learning_rate": 4.862075685789935e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 21739464,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 0.14888165891170502,
      "learning_rate": 4.8522435426691604e-05,
      "loss": 0.2975,
      "num_input_tokens_seen": 21742312,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.22566768527030945,
      "learning_rate": 4.842420843943962e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 21744936,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.1031644344329834,
      "learning_rate": 4.8326075916691415e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 21749032,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.1029287576675415,
      "learning_rate": 4.822803787897523e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 21752744,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.23077405989170074,
      "learning_rate": 4.813009434679932e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 21756552,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.2883990705013275,
      "learning_rate": 4.803224534065259e-05,
      "loss": 0.332,
      "num_input_tokens_seen": 21758984,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.18777824938297272,
      "learning_rate": 4.793449088100371e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 21761800,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.12373790144920349,
      "learning_rate": 4.783683098830188e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 21765096,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.19725872576236725,
      "learning_rate": 4.7739265682976485e-05,
      "loss": 0.5747,
      "num_input_tokens_seen": 21768552,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.1381436586380005,
      "learning_rate": 4.764179498543714e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 21772424,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.17818881571292877,
      "learning_rate": 4.754441891607347e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 21774824,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.17276471853256226,
      "learning_rate": 4.7447137495255424e-05,
      "loss": 0.5559,
      "num_input_tokens_seen": 21778728,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 0.16241933405399323,
      "learning_rate": 4.7349950743333206e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 21782152,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.1668451726436615,
      "learning_rate": 4.725285868063722e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 21785608,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.22751487791538239,
      "learning_rate": 4.715586132747801e-05,
      "loss": 0.5471,
      "num_input_tokens_seen": 21789480,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.10244466364383698,
      "learning_rate": 4.705895870414634e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 21793192,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.18931162357330322,
      "learning_rate": 4.696215083091299e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 21795816,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.15632577240467072,
      "learning_rate": 4.686543772802915e-05,
      "loss": 0.5148,
      "num_input_tokens_seen": 21799240,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.2983201742172241,
      "learning_rate": 4.6768819415726085e-05,
      "loss": 0.3159,
      "num_input_tokens_seen": 21802184,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.2163780927658081,
      "learning_rate": 4.667229591421529e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 21805640,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.09587600082159042,
      "learning_rate": 4.657586724368845e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 21808648,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.12488976120948792,
      "learning_rate": 4.647953342431726e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 21812616,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 0.21623530983924866,
      "learning_rate": 4.63832944762535e-05,
      "loss": 0.2877,
      "num_input_tokens_seen": 21815624,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 0.1172020211815834,
      "learning_rate": 4.628715041962944e-05,
      "loss": 0.3531,
      "num_input_tokens_seen": 21818472,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 0.10845764726400375,
      "learning_rate": 4.619110127455728e-05,
      "loss": 0.2614,
      "num_input_tokens_seen": 21823208,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.1978272795677185,
      "learning_rate": 4.6095147061129494e-05,
      "loss": 0.315,
      "num_input_tokens_seen": 21826120,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.10359276831150055,
      "learning_rate": 4.599928779941853e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 21829992,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 0.381439208984375,
      "learning_rate": 4.590352350947691e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 21832968,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.13626369833946228,
      "learning_rate": 4.580785421133759e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 21836328,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.1863114982843399,
      "learning_rate": 4.571227992501348e-05,
      "loss": 0.4009,
      "num_input_tokens_seen": 21838856,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.10061265528202057,
      "learning_rate": 4.561680067049773e-05,
      "loss": 0.4984,
      "num_input_tokens_seen": 21842440,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.20729807019233704,
      "learning_rate": 4.552141646776331e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 21845128,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.291865736246109,
      "learning_rate": 4.542612733676365e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 21847528,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.10341054946184158,
      "learning_rate": 4.533093329743204e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 21851240,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 0.2902868092060089,
      "learning_rate": 4.523583436968204e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 21853672,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.16972032189369202,
      "learning_rate": 4.514083057340723e-05,
      "loss": 0.6426,
      "num_input_tokens_seen": 21857736,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.1534290611743927,
      "learning_rate": 4.504592192848151e-05,
      "loss": 0.3524,
      "num_input_tokens_seen": 21861192,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.19429710507392883,
      "learning_rate": 4.495110845475847e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 21864584,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 0.267551988363266,
      "learning_rate": 4.4856390172072045e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 21867560,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.12894201278686523,
      "learning_rate": 4.476176710023622e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 21870408,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.22681660950183868,
      "learning_rate": 4.4667239259045035e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 21873256,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.25409096479415894,
      "learning_rate": 4.457280666827285e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 21876328,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.15821272134780884,
      "learning_rate": 4.447846934767358e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 21879784,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.17614033818244934,
      "learning_rate": 4.4384227316981774e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 21882792,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 0.14444264769554138,
      "learning_rate": 4.4290080595911555e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 21886632,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 0.14242242276668549,
      "learning_rate": 4.419602920415749e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 21889544,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.19553345441818237,
      "learning_rate": 4.410207316139408e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 21893192,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.1308313012123108,
      "learning_rate": 4.400821248727577e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 21896584,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.16589295864105225,
      "learning_rate": 4.391444720143717e-05,
      "loss": 0.3781,
      "num_input_tokens_seen": 21899656,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.07993833720684052,
      "learning_rate": 4.382077732349299e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 21903688,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 0.15411600470542908,
      "learning_rate": 4.372720287303772e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 21906760,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.1752956658601761,
      "learning_rate": 4.3633723869646204e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 21909864,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 0.16251912713050842,
      "learning_rate": 4.3540340332873184e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 21912808,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.15298208594322205,
      "learning_rate": 4.3447052282253375e-05,
      "loss": 0.3074,
      "num_input_tokens_seen": 21916040,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.09755720943212509,
      "learning_rate": 4.335385973730166e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 21921064,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.17086288332939148,
      "learning_rate": 4.3260762717512726e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 21923624,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 0.10963369905948639,
      "learning_rate": 4.316776124236144e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 21927688,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 0.2971051037311554,
      "learning_rate": 4.307485533130279e-05,
      "loss": 0.5056,
      "num_input_tokens_seen": 21930248,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.15028274059295654,
      "learning_rate": 4.298204500377145e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 21933736,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.17200078070163727,
      "learning_rate": 4.288933027918235e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 21937256,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.39375028014183044,
      "learning_rate": 4.279671117693046e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 21940136,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.17711661756038666,
      "learning_rate": 4.270418771639045e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 21942888,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.17542654275894165,
      "learning_rate": 4.2611759916917345e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 21947752,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.14529961347579956,
      "learning_rate": 4.251942779784579e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 21951496,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.10709859430789948,
      "learning_rate": 4.242719137849077e-05,
      "loss": 0.3588,
      "num_input_tokens_seen": 21955048,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 0.13673005998134613,
      "learning_rate": 4.233505067814702e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 21958408,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 0.1456914097070694,
      "learning_rate": 4.224300571608947e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 21961000,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.15071216225624084,
      "learning_rate": 4.215105651157264e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 21964552,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.1548219621181488,
      "learning_rate": 4.2059203083831434e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 21967272,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.12895813584327698,
      "learning_rate": 4.196744545208042e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 21970824,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 0.335930734872818,
      "learning_rate": 4.187578363551431e-05,
      "loss": 0.4281,
      "num_input_tokens_seen": 21973768,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.1804431527853012,
      "learning_rate": 4.17842176533077e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 21976712,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.2007146030664444,
      "learning_rate": 4.1692747524615274e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 21979848,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.204485222697258,
      "learning_rate": 4.160137326857138e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 21983048,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 0.09263735264539719,
      "learning_rate": 4.151009490429042e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 21987176,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.20489653944969177,
      "learning_rate": 4.141891245086687e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 21990088,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.16715843975543976,
      "learning_rate": 4.132782592737511e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 21993928,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.16096799075603485,
      "learning_rate": 4.123683535286943e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 21998536,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.14789581298828125,
      "learning_rate": 4.114594074638395e-05,
      "loss": 0.518,
      "num_input_tokens_seen": 22002184,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 0.23054124414920807,
      "learning_rate": 4.105514212693268e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 22006504,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.17468389868736267,
      "learning_rate": 4.096443951350976e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 22009000,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.12351113557815552,
      "learning_rate": 4.0873832925089116e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 22012456,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.22480028867721558,
      "learning_rate": 4.0783322380624656e-05,
      "loss": 0.3254,
      "num_input_tokens_seen": 22015272,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.2122095376253128,
      "learning_rate": 4.069290789905023e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 22018088,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.10830177366733551,
      "learning_rate": 4.060258949927942e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 22021224,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.14300526678562164,
      "learning_rate": 4.051236720020568e-05,
      "loss": 0.335,
      "num_input_tokens_seen": 22023848,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.17077170312404633,
      "learning_rate": 4.042224102070263e-05,
      "loss": 0.6032,
      "num_input_tokens_seen": 22026568,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.15760654211044312,
      "learning_rate": 4.033221097962364e-05,
      "loss": 0.4114,
      "num_input_tokens_seen": 22029448,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.2566448152065277,
      "learning_rate": 4.0242277095802025e-05,
      "loss": 0.3916,
      "num_input_tokens_seen": 22032392,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.2796376347541809,
      "learning_rate": 4.015243938805086e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 22035208,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.2509583830833435,
      "learning_rate": 4.006269787516304e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 22038696,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.1709301471710205,
      "learning_rate": 3.9973052575911606e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 22042088,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.19269989430904388,
      "learning_rate": 3.988350350904929e-05,
      "loss": 0.3502,
      "num_input_tokens_seen": 22045064,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.14873547852039337,
      "learning_rate": 3.979405069330882e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 22049128,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 0.14840014278888702,
      "learning_rate": 3.970469414740269e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 22053128,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.15391354262828827,
      "learning_rate": 3.961543389002309e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 22056616,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.11201635003089905,
      "learning_rate": 3.952626993984243e-05,
      "loss": 0.5818,
      "num_input_tokens_seen": 22059816,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.178574800491333,
      "learning_rate": 3.943720231551268e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 22062504,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.16979075968265533,
      "learning_rate": 3.934823103566587e-05,
      "loss": 0.3241,
      "num_input_tokens_seen": 22065448,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.2056834101676941,
      "learning_rate": 3.925935611891379e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 22068840,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.13197845220565796,
      "learning_rate": 3.917057758384801e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 22071688,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.16555939614772797,
      "learning_rate": 3.9081895449039896e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 22074568,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 0.2015630453824997,
      "learning_rate": 3.899330973304083e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 22077640,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.20057733356952667,
      "learning_rate": 3.8904820454381875e-05,
      "loss": 0.566,
      "num_input_tokens_seen": 22080712,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.10929752141237259,
      "learning_rate": 3.881642763157417e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 22084360,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 0.15837925672531128,
      "learning_rate": 3.872813128310826e-05,
      "loss": 0.3075,
      "num_input_tokens_seen": 22087176,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.11516980081796646,
      "learning_rate": 3.8639931427454866e-05,
      "loss": 0.5672,
      "num_input_tokens_seen": 22090248,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.1389378160238266,
      "learning_rate": 3.855182808306429e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 22093672,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.08736956119537354,
      "learning_rate": 3.8463821268366784e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 22097128,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.12722565233707428,
      "learning_rate": 3.837591100177246e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 22099976,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.11841108649969101,
      "learning_rate": 3.828809730167099e-05,
      "loss": 0.4159,
      "num_input_tokens_seen": 22103880,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.1706533133983612,
      "learning_rate": 3.820038018643218e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 22107400,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 0.2601208984851837,
      "learning_rate": 3.8112759674405254e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 22110184,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 0.239711195230484,
      "learning_rate": 3.802523578391953e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 22112520,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 0.20228661596775055,
      "learning_rate": 3.793780853328399e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 22115752,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.13939887285232544,
      "learning_rate": 3.785047794078744e-05,
      "loss": 0.3152,
      "num_input_tokens_seen": 22120072,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 0.10268932580947876,
      "learning_rate": 3.776324402469844e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 22123400,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 0.14696282148361206,
      "learning_rate": 3.767610680326533e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 22127144,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.20866553485393524,
      "learning_rate": 3.7589066294716145e-05,
      "loss": 0.3327,
      "num_input_tokens_seen": 22129928,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.15543605387210846,
      "learning_rate": 3.750212251725882e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 22132776,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.2079925239086151,
      "learning_rate": 3.741527548908113e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 22136232,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.20202365517616272,
      "learning_rate": 3.732852522835023e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 22140168,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.17783407866954803,
      "learning_rate": 3.724187175321342e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 22143368,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.11345410346984863,
      "learning_rate": 3.71553150817977e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 22146824,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.13080620765686035,
      "learning_rate": 3.70688552322096e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 22149928,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.18028582632541656,
      "learning_rate": 3.698249222253569e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 22153768,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.20658257603645325,
      "learning_rate": 3.6896226070841863e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 22156744,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.14830563962459564,
      "learning_rate": 3.681005679517424e-05,
      "loss": 0.3622,
      "num_input_tokens_seen": 22159240,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.21252046525478363,
      "learning_rate": 3.6723984413558455e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 22162856,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 0.24783259630203247,
      "learning_rate": 3.663800894399971e-05,
      "loss": 0.3508,
      "num_input_tokens_seen": 22165736,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.13522969186306,
      "learning_rate": 3.655213040448318e-05,
      "loss": 0.3195,
      "num_input_tokens_seen": 22168488,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.14021754264831543,
      "learning_rate": 3.6466348812973723e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 22173448,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.21405918896198273,
      "learning_rate": 3.6380664187415815e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 22177096,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 0.4278430640697479,
      "learning_rate": 3.6295076545733663e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 22180168,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 0.16951018571853638,
      "learning_rate": 3.6209585905831344e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 22182920,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.2929782271385193,
      "learning_rate": 3.6124192285592385e-05,
      "loss": 0.343,
      "num_input_tokens_seen": 22185288,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 0.09972021728754044,
      "learning_rate": 3.6038895702880324e-05,
      "loss": 0.3763,
      "num_input_tokens_seen": 22190472,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 0.18252532184123993,
      "learning_rate": 3.595369617553806e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 22193960,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.14296485483646393,
      "learning_rate": 3.58685937213884e-05,
      "loss": 0.3226,
      "num_input_tokens_seen": 22196392,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.13940362632274628,
      "learning_rate": 3.5783588358233863e-05,
      "loss": 0.3541,
      "num_input_tokens_seen": 22199272,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 0.19220462441444397,
      "learning_rate": 3.569868010385663e-05,
      "loss": 0.4702,
      "num_input_tokens_seen": 22202536,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 0.21232739090919495,
      "learning_rate": 3.5613868976018535e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 22205864,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.1311037540435791,
      "learning_rate": 3.5529154992460955e-05,
      "loss": 0.3926,
      "num_input_tokens_seen": 22208520,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.10882343351840973,
      "learning_rate": 3.544453817090521e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 22211336,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.18633036315441132,
      "learning_rate": 3.536001852905207e-05,
      "loss": 0.5102,
      "num_input_tokens_seen": 22214888,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.1350817084312439,
      "learning_rate": 3.527559608458225e-05,
      "loss": 0.4068,
      "num_input_tokens_seen": 22218216,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.20310048758983612,
      "learning_rate": 3.5191270855155886e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 22221224,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.2306475043296814,
      "learning_rate": 3.510704285841293e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 22223912,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 0.21488967537879944,
      "learning_rate": 3.502291211197267e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 22227784,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.340104341506958,
      "learning_rate": 3.493887863343448e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 22230280,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.3403419256210327,
      "learning_rate": 3.485494244037718e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 22232744,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.20128527283668518,
      "learning_rate": 3.477110355035934e-05,
      "loss": 0.3208,
      "num_input_tokens_seen": 22235784,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.12981784343719482,
      "learning_rate": 3.468736198091904e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 22239912,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.14070147275924683,
      "learning_rate": 3.4603717749573935e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 22244104,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 0.20885023474693298,
      "learning_rate": 3.452017087382159e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 22247816,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.19924741983413696,
      "learning_rate": 3.443672137113901e-05,
      "loss": 0.3848,
      "num_input_tokens_seen": 22251688,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.13359296321868896,
      "learning_rate": 3.4353369258983025e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 22255144,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.2249699831008911,
      "learning_rate": 3.4270114554789735e-05,
      "loss": 0.705,
      "num_input_tokens_seen": 22258536,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.08786601573228836,
      "learning_rate": 3.41869572759751e-05,
      "loss": 0.3681,
      "num_input_tokens_seen": 22264104,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.14363159239292145,
      "learning_rate": 3.410389743993475e-05,
      "loss": 0.3085,
      "num_input_tokens_seen": 22266952,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 0.20508348941802979,
      "learning_rate": 3.40209350640438e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 22270088,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.13313525915145874,
      "learning_rate": 3.3938070165657084e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 22272680,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.3270128071308136,
      "learning_rate": 3.385530276210902e-05,
      "loss": 0.3354,
      "num_input_tokens_seen": 22276456,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 0.1340351551771164,
      "learning_rate": 3.377263287071358e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 22279784,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.16489654779434204,
      "learning_rate": 3.369006050876428e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 22282760,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 0.29359757900238037,
      "learning_rate": 3.3607585693534394e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 22286088,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.11831710487604141,
      "learning_rate": 3.352520844227663e-05,
      "loss": 0.35,
      "num_input_tokens_seen": 22290632,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.11774906516075134,
      "learning_rate": 3.344292877222355e-05,
      "loss": 0.344,
      "num_input_tokens_seen": 22293448,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 0.16985121369361877,
      "learning_rate": 3.336074670058703e-05,
      "loss": 0.5803,
      "num_input_tokens_seen": 22297704,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.24074269831180573,
      "learning_rate": 3.327866224455844e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 22300488,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 0.2094181776046753,
      "learning_rate": 3.3196675421309165e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 22303656,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.195506751537323,
      "learning_rate": 3.3114786247989744e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 22306664,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.12194539606571198,
      "learning_rate": 3.303299474173066e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 22309512,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 0.14909633994102478,
      "learning_rate": 3.2951300919641544e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 22312424,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.11365817487239838,
      "learning_rate": 3.2869704798811964e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 22316136,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.244266077876091,
      "learning_rate": 3.2788206396310746e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 22319112,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.1690843552350998,
      "learning_rate": 3.270680572918655e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 22322504,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.11245808750391006,
      "learning_rate": 3.262550281446741e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 22328744,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.1264505684375763,
      "learning_rate": 3.2544297669161115e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 22332936,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 0.18264815211296082,
      "learning_rate": 3.2463190310254633e-05,
      "loss": 0.6511,
      "num_input_tokens_seen": 22336552,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.16212956607341766,
      "learning_rate": 3.238218075471483e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 22339496,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.1580171138048172,
      "learning_rate": 3.230126901948788e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 22342696,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 0.11236253380775452,
      "learning_rate": 3.222045512149968e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 22346024,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 0.2193615436553955,
      "learning_rate": 3.213973907765566e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 22348808,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 0.1155107393860817,
      "learning_rate": 3.205912090484048e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 22352584,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.19565312564373016,
      "learning_rate": 3.1978600619918754e-05,
      "loss": 0.7423,
      "num_input_tokens_seen": 22355336,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.1844637244939804,
      "learning_rate": 3.189817823973423e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 22362664,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.18063890933990479,
      "learning_rate": 3.18178537811104e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 22365736,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.19818814098834991,
      "learning_rate": 3.1737627260850335e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 22368744,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.19376789033412933,
      "learning_rate": 3.1657498695736364e-05,
      "loss": 0.7972,
      "num_input_tokens_seen": 22371176,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.1503302901983261,
      "learning_rate": 3.157746810253054e-05,
      "loss": 0.4022,
      "num_input_tokens_seen": 22375592,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 0.1646248698234558,
      "learning_rate": 3.149753549797446e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 22379112,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.16104629635810852,
      "learning_rate": 3.141770089878887e-05,
      "loss": 0.3368,
      "num_input_tokens_seen": 22382824,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 0.15789438784122467,
      "learning_rate": 3.1337964321674354e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 22386408,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 0.21265321969985962,
      "learning_rate": 3.125832578331106e-05,
      "loss": 0.3316,
      "num_input_tokens_seen": 22389320,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.1969943344593048,
      "learning_rate": 3.117878530035823e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 22391848,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.21997900307178497,
      "learning_rate": 3.1099342889454875e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 22395304,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 0.1502523571252823,
      "learning_rate": 3.1019998567219597e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 22397960,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.19001692533493042,
      "learning_rate": 3.0940752350250125e-05,
      "loss": 0.3121,
      "num_input_tokens_seen": 22400296,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.1318349689245224,
      "learning_rate": 3.0861604255124074e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 22403432,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.1233772411942482,
      "learning_rate": 3.0782554298398045e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 22406952,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 0.11184627562761307,
      "learning_rate": 3.070360249660859e-05,
      "loss": 0.3258,
      "num_input_tokens_seen": 22410568,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.2730933725833893,
      "learning_rate": 3.0624748866271535e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 22413832,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.1334444135427475,
      "learning_rate": 3.0545993423882036e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 22417576,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 0.24707269668579102,
      "learning_rate": 3.046733618591496e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 22420456,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.20117047429084778,
      "learning_rate": 3.0388777168824434e-05,
      "loss": 0.3725,
      "num_input_tokens_seen": 22423016,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.20501424372196198,
      "learning_rate": 3.0310316389044022e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 22425544,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.16762401163578033,
      "learning_rate": 3.0231953862986995e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 22428392,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.4931029677391052,
      "eval_runtime": 13.8471,
      "eval_samples_per_second": 61.24,
      "eval_steps_per_second": 15.31,
      "num_input_tokens_seen": 22429992,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.1855165809392929,
      "learning_rate": 3.0153689607045842e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 22430632,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 0.2891322672367096,
      "learning_rate": 3.0075523637592474e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 22433864,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.15728046000003815,
      "learning_rate": 2.999745597097847e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 22437128,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.20671749114990234,
      "learning_rate": 2.9919486623534497e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 22439912,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.25747692584991455,
      "learning_rate": 2.9841615611571005e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 22442760,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.12637361884117126,
      "learning_rate": 2.9763842951377628e-05,
      "loss": 0.3037,
      "num_input_tokens_seen": 22445704,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.23384173214435577,
      "learning_rate": 2.968616865922369e-05,
      "loss": 0.5975,
      "num_input_tokens_seen": 22448808,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.18457916378974915,
      "learning_rate": 2.960859275135758e-05,
      "loss": 0.3181,
      "num_input_tokens_seen": 22451688,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 0.19517427682876587,
      "learning_rate": 2.9531115244007367e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 22454472,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 0.19037581980228424,
      "learning_rate": 2.945373615338037e-05,
      "loss": 0.5466,
      "num_input_tokens_seen": 22457224,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.27897703647613525,
      "learning_rate": 2.9376455495663534e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 22459784,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.213216170668602,
      "learning_rate": 2.9299273287023144e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 22462408,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.10806307941675186,
      "learning_rate": 2.922218954360473e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 22466152,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.15899161994457245,
      "learning_rate": 2.9145204281533278e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 22469352,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.2231459766626358,
      "learning_rate": 2.9068317516913244e-05,
      "loss": 0.3545,
      "num_input_tokens_seen": 22472488,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 0.1570112109184265,
      "learning_rate": 2.8991529265828542e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 22475272,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.16525065898895264,
      "learning_rate": 2.891483954434243e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 22478312,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.22891026735305786,
      "learning_rate": 2.8838248368497465e-05,
      "loss": 0.3584,
      "num_input_tokens_seen": 22480712,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.2149941325187683,
      "learning_rate": 2.8761755754315667e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 22485064,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 0.21092642843723297,
      "learning_rate": 2.8685361717798286e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 22488040,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.1785474717617035,
      "learning_rate": 2.8609066274926265e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 22491560,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.2080378234386444,
      "learning_rate": 2.8532869441659615e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 22494088,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.13036499917507172,
      "learning_rate": 2.8456771233937973e-05,
      "loss": 0.6348,
      "num_input_tokens_seen": 22497864,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.22414544224739075,
      "learning_rate": 2.838077166768016e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 22501320,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.1657143384218216,
      "learning_rate": 2.8304870758784296e-05,
      "loss": 0.3042,
      "num_input_tokens_seen": 22503816,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.34331414103507996,
      "learning_rate": 2.822906852312812e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 22507432,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 0.30723756551742554,
      "learning_rate": 2.8153364976568563e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 22514472,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.14607465267181396,
      "learning_rate": 2.8077760134941955e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 22517864,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.20443418622016907,
      "learning_rate": 2.800225401406392e-05,
      "loss": 0.3262,
      "num_input_tokens_seen": 22521256,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.13750848174095154,
      "learning_rate": 2.7926846629729607e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 22523624,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.14469273388385773,
      "learning_rate": 2.7851537997713174e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 22526984,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.1434602439403534,
      "learning_rate": 2.7776328133768458e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 22529960,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.2990577220916748,
      "learning_rate": 2.770121705362849e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 22532456,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.28938522934913635,
      "learning_rate": 2.7626204773005704e-05,
      "loss": 0.4136,
      "num_input_tokens_seen": 22536040,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.22389575839042664,
      "learning_rate": 2.7551291307591765e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 22539144,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.2158312350511551,
      "learning_rate": 2.7476476673057636e-05,
      "loss": 0.5041,
      "num_input_tokens_seen": 22541960,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.18798750638961792,
      "learning_rate": 2.740176088505375e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 22544968,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.14489759504795074,
      "learning_rate": 2.7327143959209765e-05,
      "loss": 0.3261,
      "num_input_tokens_seen": 22548232,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 0.42261332273483276,
      "learning_rate": 2.725262591113481e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 22550824,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.22310598194599152,
      "learning_rate": 2.7178206756417078e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 22553736,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.11772625893354416,
      "learning_rate": 2.7103886510624344e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 22556936,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 0.18590402603149414,
      "learning_rate": 2.7029665189303387e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 22560008,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 0.2231011837720871,
      "learning_rate": 2.6955542807980515e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 22562824,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.17415772378444672,
      "learning_rate": 2.688151938216138e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 22566856,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.1761215180158615,
      "learning_rate": 2.6807594927330703e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 22571528,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.19926658272743225,
      "learning_rate": 2.6733769458952727e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 22575016,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.11856850981712341,
      "learning_rate": 2.6660042992470934e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 22578920,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.2283056080341339,
      "learning_rate": 2.658641554330793e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 22585256,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.162910595536232,
      "learning_rate": 2.6512887126865782e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 22588648,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.12640711665153503,
      "learning_rate": 2.6439457758525908e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 22591176,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.10980343818664551,
      "learning_rate": 2.6366127453648748e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 22595592,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.16774222254753113,
      "learning_rate": 2.6292896227574303e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 22597928,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.1818038523197174,
      "learning_rate": 2.6219764095621547e-05,
      "loss": 0.639,
      "num_input_tokens_seen": 22600840,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.1031910702586174,
      "learning_rate": 2.6146731073088958e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 22603624,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.19689813256263733,
      "learning_rate": 2.607379717525432e-05,
      "loss": 0.5868,
      "num_input_tokens_seen": 22605928,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 0.15679362416267395,
      "learning_rate": 2.6000962417374373e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 22609960,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 0.1256701499223709,
      "learning_rate": 2.5928226814685485e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 22613672,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.21008779108524323,
      "learning_rate": 2.585559038240304e-05,
      "loss": 0.5347,
      "num_input_tokens_seen": 22616552,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 0.24788309633731842,
      "learning_rate": 2.5783053135721714e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 22620392,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.2063319832086563,
      "learning_rate": 2.571061508981565e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 22623208,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.2061099410057068,
      "learning_rate": 2.5638276259837778e-05,
      "loss": 0.537,
      "num_input_tokens_seen": 22626248,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 0.20470239222049713,
      "learning_rate": 2.5566036660920778e-05,
      "loss": 0.4926,
      "num_input_tokens_seen": 22629352,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 0.1812644600868225,
      "learning_rate": 2.5493896308176223e-05,
      "loss": 0.3422,
      "num_input_tokens_seen": 22632712,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.2186390906572342,
      "learning_rate": 2.542185521669521e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 22636648,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.1043945923447609,
      "learning_rate": 2.534991340154774e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 22643752,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.1743871122598648,
      "learning_rate": 2.5278070877783332e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 22647336,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 0.15672555565834045,
      "learning_rate": 2.520632766043052e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 22650024,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.17338702082633972,
      "learning_rate": 2.513468376449729e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 22652840,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 0.17940428853034973,
      "learning_rate": 2.506313920497061e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 22655560,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.13570700585842133,
      "learning_rate": 2.4991693996816888e-05,
      "loss": 0.3528,
      "num_input_tokens_seen": 22658920,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.1666480004787445,
      "learning_rate": 2.4920348154981677e-05,
      "loss": 0.3299,
      "num_input_tokens_seen": 22661960,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.18759405612945557,
      "learning_rate": 2.4849101694389477e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 22664840,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.14697666466236115,
      "learning_rate": 2.4777954629944478e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 22667464,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.16858980059623718,
      "learning_rate": 2.4706906976529718e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 22671176,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.19016508758068085,
      "learning_rate": 2.4635958749007648e-05,
      "loss": 0.3581,
      "num_input_tokens_seen": 22674888,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.21270734071731567,
      "learning_rate": 2.456510996221978e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 22678408,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.126927450299263,
      "learning_rate": 2.4494360630986756e-05,
      "loss": 0.6222,
      "num_input_tokens_seen": 22682056,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.1510840654373169,
      "learning_rate": 2.4423710770108687e-05,
      "loss": 0.517,
      "num_input_tokens_seen": 22685320,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.17239296436309814,
      "learning_rate": 2.435316039436464e-05,
      "loss": 0.3386,
      "num_input_tokens_seen": 22688616,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.17283965647220612,
      "learning_rate": 2.428270951851297e-05,
      "loss": 0.3327,
      "num_input_tokens_seen": 22690952,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 0.11105705052614212,
      "learning_rate": 2.421235815729128e-05,
      "loss": 0.349,
      "num_input_tokens_seen": 22694472,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.1042785793542862,
      "learning_rate": 2.414210632541619e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 22697832,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.11891894787549973,
      "learning_rate": 2.40719540375835e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 22701992,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 0.2592790126800537,
      "learning_rate": 2.4001901308468353e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 22704808,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.13293564319610596,
      "learning_rate": 2.3931948152724982e-05,
      "loss": 0.3735,
      "num_input_tokens_seen": 22708008,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.18382956087589264,
      "learning_rate": 2.386209458498684e-05,
      "loss": 0.3788,
      "num_input_tokens_seen": 22711496,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.3066089451313019,
      "learning_rate": 2.3792340619866458e-05,
      "loss": 0.3022,
      "num_input_tokens_seen": 22715400,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 0.13416482508182526,
      "learning_rate": 2.3722686271955507e-05,
      "loss": 0.5922,
      "num_input_tokens_seen": 22718728,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.16009555757045746,
      "learning_rate": 2.365313155582488e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 22721928,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.14111478626728058,
      "learning_rate": 2.358367648602472e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 22725128,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.2568966746330261,
      "learning_rate": 2.3514321077084234e-05,
      "loss": 0.2987,
      "num_input_tokens_seen": 22729224,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.14950641989707947,
      "learning_rate": 2.3445065343511763e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 22731880,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.13384760916233063,
      "learning_rate": 2.3375909299794717e-05,
      "loss": 0.3214,
      "num_input_tokens_seen": 22735272,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.12437179684638977,
      "learning_rate": 2.330685296039986e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 22738824,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.18276569247245789,
      "learning_rate": 2.3237896339772914e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 22741384,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.15694132447242737,
      "learning_rate": 2.3169039452338892e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 22744840,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.14275380969047546,
      "learning_rate": 2.3100282312501886e-05,
      "loss": 0.5875,
      "num_input_tokens_seen": 22750024,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 0.17618587613105774,
      "learning_rate": 2.3031624934645113e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 22752904,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.17407751083374023,
      "learning_rate": 2.296306733313075e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 22755624,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 0.16904081404209137,
      "learning_rate": 2.289460952230038e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 22758664,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.1888609379529953,
      "learning_rate": 2.2826251516474604e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 22761288,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 0.13119062781333923,
      "learning_rate": 2.2757993329953152e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 22763688,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.15113471448421478,
      "learning_rate": 2.2689834977014822e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 22766280,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.11286121606826782,
      "learning_rate": 2.2621776471917598e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 22769800,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.12476212531328201,
      "learning_rate": 2.255381782889848e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 22773128,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.13399286568164825,
      "learning_rate": 2.2485959062173654e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 22776072,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 0.1540578007698059,
      "learning_rate": 2.2418200185938485e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 22779976,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 0.22782227396965027,
      "learning_rate": 2.235054121436725e-05,
      "loss": 0.3495,
      "num_input_tokens_seen": 22782600,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.22518831491470337,
      "learning_rate": 2.2282982161613562e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 22785224,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.3446803689002991,
      "learning_rate": 2.221552304180985e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 22788168,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.17994728684425354,
      "learning_rate": 2.2148163869067816e-05,
      "loss": 0.591,
      "num_input_tokens_seen": 22791464,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.13964030146598816,
      "learning_rate": 2.2080904657478305e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 22794952,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.3216710686683655,
      "learning_rate": 2.201374542111123e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 22797896,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.13924312591552734,
      "learning_rate": 2.1946686174015407e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 22801000,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 0.15653444826602936,
      "learning_rate": 2.1879726930218946e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 22804712,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.09790589660406113,
      "learning_rate": 2.1812867703728866e-05,
      "loss": 0.4102,
      "num_input_tokens_seen": 22808936,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 0.35635581612586975,
      "learning_rate": 2.1746108508531426e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 22811272,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 0.17799846827983856,
      "learning_rate": 2.1679449358591952e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 22814280,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.11551640182733536,
      "learning_rate": 2.1612890267854624e-05,
      "loss": 0.3861,
      "num_input_tokens_seen": 22818888,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.2342590093612671,
      "learning_rate": 2.1546431250242914e-05,
      "loss": 0.4978,
      "num_input_tokens_seen": 22821352,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.17530664801597595,
      "learning_rate": 2.1480072319659427e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 22825448,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.2290903627872467,
      "learning_rate": 2.1413813489985447e-05,
      "loss": 0.3248,
      "num_input_tokens_seen": 22828936,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.17072775959968567,
      "learning_rate": 2.1347654775081716e-05,
      "loss": 0.2779,
      "num_input_tokens_seen": 22831752,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 0.23330576717853546,
      "learning_rate": 2.1281596188787834e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 22834760,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.4654647409915924,
      "learning_rate": 2.121563774492252e-05,
      "loss": 0.5098,
      "num_input_tokens_seen": 22837064,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.1489732414484024,
      "learning_rate": 2.114977945728358e-05,
      "loss": 0.4986,
      "num_input_tokens_seen": 22840232,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.15034353733062744,
      "learning_rate": 2.1084021339647707e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 22843752,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.11379547417163849,
      "learning_rate": 2.1018363405770792e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 22847336,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.1343124508857727,
      "learning_rate": 2.095280566938784e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 22850792,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.13867850601673126,
      "learning_rate": 2.0887348144212615e-05,
      "loss": 0.3535,
      "num_input_tokens_seen": 22853768,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.09067701548337936,
      "learning_rate": 2.08219908439381e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 22858024,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.1456032693386078,
      "learning_rate": 2.075673378223647e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 22861416,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.17319491505622864,
      "learning_rate": 2.069157697275853e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 22864712,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 0.13432489335536957,
      "learning_rate": 2.0626520429134543e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 22867272,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.21626149117946625,
      "learning_rate": 2.0561564164973458e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 22869640,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.10203959792852402,
      "learning_rate": 2.04967081938634e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 22874152,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.10709547996520996,
      "learning_rate": 2.043195252937152e-05,
      "loss": 0.3661,
      "num_input_tokens_seen": 22877736,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.14798897504806519,
      "learning_rate": 2.0367297185044043e-05,
      "loss": 0.5159,
      "num_input_tokens_seen": 22880264,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.22824281454086304,
      "learning_rate": 2.030274217440603e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 22884520,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 0.4498204290866852,
      "learning_rate": 2.0238287510961628e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 22887624,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.1632859855890274,
      "learning_rate": 2.017393320819405e-05,
      "loss": 0.3581,
      "num_input_tokens_seen": 22891752,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.37549805641174316,
      "learning_rate": 2.010967927956553e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 22894184,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.2044895887374878,
      "learning_rate": 2.00455257385172e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 22896968,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.14861270785331726,
      "learning_rate": 1.9981472598469386e-05,
      "loss": 0.3566,
      "num_input_tokens_seen": 22899848,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.17612238228321075,
      "learning_rate": 1.9917519872821142e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 22902312,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.15583381056785583,
      "learning_rate": 1.9853667574950605e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 22905128,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.10962918400764465,
      "learning_rate": 1.978991571821498e-05,
      "loss": 0.3504,
      "num_input_tokens_seen": 22908488,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.1879458874464035,
      "learning_rate": 1.972626431595048e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 22911592,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 0.11442931741476059,
      "learning_rate": 1.9662713381472295e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 22914376,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.13337668776512146,
      "learning_rate": 1.959926292807451e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 22918056,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.11966341733932495,
      "learning_rate": 1.9535912969030178e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 22921992,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.07565189898014069,
      "learning_rate": 1.947266351759136e-05,
      "loss": 0.3444,
      "num_input_tokens_seen": 22926024,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 0.18069538474082947,
      "learning_rate": 1.940951458698925e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 22928936,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.17994935810565948,
      "learning_rate": 1.9346466190433842e-05,
      "loss": 0.3597,
      "num_input_tokens_seen": 22932168,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 0.2113357037305832,
      "learning_rate": 1.9283518341114136e-05,
      "loss": 0.4468,
      "num_input_tokens_seen": 22934728,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 0.1890542358160019,
      "learning_rate": 1.9220671052198047e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 22937736,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.17475919425487518,
      "learning_rate": 1.9157924336832556e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 22941256,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.16302737593650818,
      "learning_rate": 1.909527820814355e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 22943784,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.18868179619312286,
      "learning_rate": 1.9032732679235886e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 22946312,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 0.1166311576962471,
      "learning_rate": 1.8970287763193428e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 22949576,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.11901483684778214,
      "learning_rate": 1.8907943473078892e-05,
      "loss": 0.3545,
      "num_input_tokens_seen": 22952712,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.16784577071666718,
      "learning_rate": 1.884569982193396e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 22955400,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 0.2677464187145233,
      "learning_rate": 1.8783556822779267e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 22958312,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.13567668199539185,
      "learning_rate": 1.8721514488614532e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 22961192,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 0.1582242250442505,
      "learning_rate": 1.8659572832418315e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 22964040,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 0.17530059814453125,
      "learning_rate": 1.8597731867148026e-05,
      "loss": 0.4448,
      "num_input_tokens_seen": 22966440,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.2132526934146881,
      "learning_rate": 1.8535991605740043e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 22970344,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.1906210333108902,
      "learning_rate": 1.8474352061109757e-05,
      "loss": 0.3346,
      "num_input_tokens_seen": 22973896,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 0.19536015391349792,
      "learning_rate": 1.8412813246151515e-05,
      "loss": 0.3076,
      "num_input_tokens_seen": 22977352,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.07855857163667679,
      "learning_rate": 1.8351375173738584e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 22981064,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.16457638144493103,
      "learning_rate": 1.829003785672295e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 22984040,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 0.13017655909061432,
      "learning_rate": 1.8228801307935806e-05,
      "loss": 0.3396,
      "num_input_tokens_seen": 22987080,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.1721695214509964,
      "learning_rate": 1.8167665540187063e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 22989992,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.1450960636138916,
      "learning_rate": 1.8106630566265604e-05,
      "loss": 0.304,
      "num_input_tokens_seen": 22993512,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.27007439732551575,
      "learning_rate": 1.8045696398939326e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 22996232,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.2421991229057312,
      "learning_rate": 1.7984863050955036e-05,
      "loss": 0.5528,
      "num_input_tokens_seen": 22998856,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.28060972690582275,
      "learning_rate": 1.7924130535038162e-05,
      "loss": 0.5102,
      "num_input_tokens_seen": 23002312,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.262511283159256,
      "learning_rate": 1.7863498863893433e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 23004808,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.2012605220079422,
      "learning_rate": 1.7802968050204203e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 23008328,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.28892120718955994,
      "learning_rate": 1.7742538106632844e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 23011176,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.14702463150024414,
      "learning_rate": 1.7682209045820684e-05,
      "loss": 0.459,
      "num_input_tokens_seen": 23014792,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.15153133869171143,
      "learning_rate": 1.76219808803878e-05,
      "loss": 0.299,
      "num_input_tokens_seen": 23017640,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 0.17070819437503815,
      "learning_rate": 1.7561853622933278e-05,
      "loss": 0.3247,
      "num_input_tokens_seen": 23021128,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.20084337890148163,
      "learning_rate": 1.7501827286035e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 23026184,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.11814521998167038,
      "learning_rate": 1.7441901882249754e-05,
      "loss": 0.4513,
      "num_input_tokens_seen": 23029768,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.14401401579380035,
      "learning_rate": 1.7382077424113464e-05,
      "loss": 0.3845,
      "num_input_tokens_seen": 23033096,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.3694028854370117,
      "learning_rate": 1.7322353924140498e-05,
      "loss": 0.2512,
      "num_input_tokens_seen": 23035528,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.19472914934158325,
      "learning_rate": 1.7262731394824372e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 23038536,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.1365070790052414,
      "learning_rate": 1.7203209848637603e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 23041672,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.11912477761507034,
      "learning_rate": 1.7143789298031175e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 23045928,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 0.10479530692100525,
      "learning_rate": 1.708446975543537e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 23049448,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 0.19859342277050018,
      "learning_rate": 1.7025251233259098e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 23052200,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.10911912471055984,
      "learning_rate": 1.6966133743890166e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 23056072,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.18123920261859894,
      "learning_rate": 1.690711729969535e-05,
      "loss": 0.6121,
      "num_input_tokens_seen": 23059208,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.1594647616147995,
      "learning_rate": 1.684820191302022e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 23061928,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.21269646286964417,
      "learning_rate": 1.6789387596189087e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 23064904,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 0.22354836761951447,
      "learning_rate": 1.6730674361505382e-05,
      "loss": 0.3247,
      "num_input_tokens_seen": 23068552,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 0.13825413584709167,
      "learning_rate": 1.6672062221251117e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 23071240,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.4899066090583801,
      "learning_rate": 1.6613551187687314e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 23075016,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.14319130778312683,
      "learning_rate": 1.6555141273053907e-05,
      "loss": 0.3201,
      "num_input_tokens_seen": 23078152,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.12622922658920288,
      "learning_rate": 1.6496832489569457e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 23081128,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.13551068305969238,
      "learning_rate": 1.643862484943165e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 23084488,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.22954273223876953,
      "learning_rate": 1.6380518364816687e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 23087144,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.27263832092285156,
      "learning_rate": 1.632251304787985e-05,
      "loss": 0.3585,
      "num_input_tokens_seen": 23090152,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.33294227719306946,
      "learning_rate": 1.6264608910755153e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 23093992,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.2277754545211792,
      "learning_rate": 1.6206805965555627e-05,
      "loss": 0.3441,
      "num_input_tokens_seen": 23096808,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.20901769399642944,
      "learning_rate": 1.614910422437288e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 23099144,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.17872971296310425,
      "learning_rate": 1.6091503699277477e-05,
      "loss": 0.2904,
      "num_input_tokens_seen": 23102440,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 0.13922424614429474,
      "learning_rate": 1.6034004402318726e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 23106376,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.15561123192310333,
      "learning_rate": 1.5976606345524836e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 23109448,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 0.23890332877635956,
      "learning_rate": 1.5919309540902927e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 23113160,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.2456153780221939,
      "learning_rate": 1.5862114000438797e-05,
      "loss": 0.4059,
      "num_input_tokens_seen": 23116136,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.14507479965686798,
      "learning_rate": 1.5805019736097104e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 23119112,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.14177098870277405,
      "learning_rate": 1.5748026759821232e-05,
      "loss": 0.3732,
      "num_input_tokens_seen": 23121704,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 0.15818758308887482,
      "learning_rate": 1.5691135083533537e-05,
      "loss": 0.3488,
      "num_input_tokens_seen": 23125256,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.25421932339668274,
      "learning_rate": 1.5634344719135052e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 23128360,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.1103968471288681,
      "learning_rate": 1.5577655678505776e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 23132648,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.2012850046157837,
      "learning_rate": 1.5521067973504442e-05,
      "loss": 0.3089,
      "num_input_tokens_seen": 23135112,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.2320774495601654,
      "learning_rate": 1.546458161596831e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 23138088,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 0.2187204658985138,
      "learning_rate": 1.5408196617713866e-05,
      "loss": 0.592,
      "num_input_tokens_seen": 23140616,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.2020856887102127,
      "learning_rate": 1.5351912990536175e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 23143624,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 0.37205663323402405,
      "learning_rate": 1.5295730746209103e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 23146568,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.2545333504676819,
      "learning_rate": 1.5239649896485463e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 23149352,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.270510196685791,
      "learning_rate": 1.5183670453096598e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 23151944,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.18802864849567413,
      "learning_rate": 1.5127792427752696e-05,
      "loss": 0.4136,
      "num_input_tokens_seen": 23155496,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.17963309586048126,
      "learning_rate": 1.507201583214296e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 23158440,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.14823903143405914,
      "learning_rate": 1.5016340677935169e-05,
      "loss": 0.3198,
      "num_input_tokens_seen": 23162120,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.3238062858581543,
      "learning_rate": 1.4960766976775953e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 23165128,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.136344775557518,
      "learning_rate": 1.4905294740290677e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 23168456,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.14153073728084564,
      "learning_rate": 1.4849923980083391e-05,
      "loss": 0.3768,
      "num_input_tokens_seen": 23172520,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.14452682435512543,
      "learning_rate": 1.4794654707737164e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 23175816,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.24011506140232086,
      "learning_rate": 1.4739486934813696e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 23179048,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.12574176490306854,
      "learning_rate": 1.468442067285336e-05,
      "loss": 0.3457,
      "num_input_tokens_seen": 23182248,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.14936180412769318,
      "learning_rate": 1.4629455933375501e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 23185224,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.11796744167804718,
      "learning_rate": 1.4574592727878089e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 23188552,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.162204310297966,
      "learning_rate": 1.4519831067837774e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 23192392,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.1905370056629181,
      "learning_rate": 1.4465170964710172e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 23195656,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 0.1542830467224121,
      "learning_rate": 1.441061242992947e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 23198568,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.18546095490455627,
      "learning_rate": 1.4356155474908871e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 23201768,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.16190478205680847,
      "learning_rate": 1.4301800111039986e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 23205064,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.22300559282302856,
      "learning_rate": 1.4247546349693386e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 23209512,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.36011070013046265,
      "learning_rate": 1.4193394202218268e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 23213544,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.10266387462615967,
      "learning_rate": 1.413934367994274e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 23216904,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 0.2175936996936798,
      "learning_rate": 1.408539479417359e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 23219624,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.21052224934101105,
      "learning_rate": 1.4031547556196178e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 23222568,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.20155668258666992,
      "learning_rate": 1.3977801977274828e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 23226600,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.25484898686408997,
      "learning_rate": 1.3924158068652437e-05,
      "loss": 0.5309,
      "num_input_tokens_seen": 23230216,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.15615616738796234,
      "learning_rate": 1.3870615841550693e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 23232968,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 0.19416052103042603,
      "learning_rate": 1.3817175307170138e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 23236264,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 0.20424307882785797,
      "learning_rate": 1.3763836476689828e-05,
      "loss": 0.3239,
      "num_input_tokens_seen": 23241896,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.14073695242404938,
      "learning_rate": 1.3710599361267617e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 23244616,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.18624205887317657,
      "learning_rate": 1.3657463972040207e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 23248520,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.2586056888103485,
      "learning_rate": 1.3604430320122762e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 23251720,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.2274068146944046,
      "learning_rate": 1.355149841660941e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 23254760,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.23699498176574707,
      "learning_rate": 1.3498668272572955e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 23258056,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.36338916420936584,
      "learning_rate": 1.3445939899064729e-05,
      "loss": 0.3914,
      "num_input_tokens_seen": 23260776,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.24685275554656982,
      "learning_rate": 1.3393313307115019e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 23263976,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.2236342579126358,
      "learning_rate": 1.3340788507732626e-05,
      "loss": 0.4227,
      "num_input_tokens_seen": 23267400,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.23364095389842987,
      "learning_rate": 1.328836551190521e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 23271240,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.13508664071559906,
      "learning_rate": 1.323604433059905e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 23274248,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.3205825686454773,
      "learning_rate": 1.3183824974759063e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 23278152,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.18478980660438538,
      "learning_rate": 1.3131707455309006e-05,
      "loss": 0.5904,
      "num_input_tokens_seen": 23281640,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.2324880063533783,
      "learning_rate": 1.307969178315127e-05,
      "loss": 0.492,
      "num_input_tokens_seen": 23284680,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 0.16563667356967926,
      "learning_rate": 1.3027777969166932e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 23287496,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.18921568989753723,
      "learning_rate": 1.2975966024215746e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 23290248,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.15262743830680847,
      "learning_rate": 1.2924255959136267e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 23294440,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.13748043775558472,
      "learning_rate": 1.2872647784745561e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 23298536,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.13979105651378632,
      "learning_rate": 1.282114151183944e-05,
      "loss": 0.5258,
      "num_input_tokens_seen": 23301224,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 0.11631079763174057,
      "learning_rate": 1.2769737151192562e-05,
      "loss": 0.5909,
      "num_input_tokens_seen": 23305384,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.22379247844219208,
      "learning_rate": 1.2718434713558047e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 23308488,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.21059027314186096,
      "learning_rate": 1.2667234209667755e-05,
      "loss": 0.6777,
      "num_input_tokens_seen": 23311528,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.22651849687099457,
      "learning_rate": 1.2616135650232286e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 23314568,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.23007513582706451,
      "learning_rate": 1.2565139045940866e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 23317288,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.13363929092884064,
      "learning_rate": 1.2514244407461462e-05,
      "loss": 0.2819,
      "num_input_tokens_seen": 23320200,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.15146790444850922,
      "learning_rate": 1.2463451745440501e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 23323816,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.12569354474544525,
      "learning_rate": 1.241276107050343e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 23327432,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.18611302971839905,
      "learning_rate": 1.236217239325399e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 23330408,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.17175284028053284,
      "learning_rate": 1.2311685724274768e-05,
      "loss": 0.3104,
      "num_input_tokens_seen": 23333224,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 0.13615643978118896,
      "learning_rate": 1.2261301074127096e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 23338248,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.1599036306142807,
      "learning_rate": 1.2211018453350874e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 23340648,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.18163365125656128,
      "learning_rate": 1.2160837872464581e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 23344072,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 0.17827452719211578,
      "learning_rate": 1.2110759341965428e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 23346760,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.1359497606754303,
      "learning_rate": 1.2060782872329256e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 23349768,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 0.1380023956298828,
      "learning_rate": 1.2010908474010595e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 23353608,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.14433059096336365,
      "learning_rate": 1.1961136157442654e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 23356776,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.15529952943325043,
      "learning_rate": 1.1911465933037214e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 23360552,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.14879584312438965,
      "learning_rate": 1.1861897811184686e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 23364264,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.14993064105510712,
      "learning_rate": 1.1812431802254109e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 23367176,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 0.19048336148262024,
      "learning_rate": 1.1763067916593262e-05,
      "loss": 0.362,
      "num_input_tokens_seen": 23371464,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 0.1053682342171669,
      "learning_rate": 1.1713806164528496e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 23374504,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.32986849546432495,
      "learning_rate": 1.1664646556364844e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 23377352,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.1509605497121811,
      "learning_rate": 1.161558910238597e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 23381288,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.1568446010351181,
      "learning_rate": 1.1566633812854e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 23384456,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.18277594447135925,
      "learning_rate": 1.151778069800985e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 23388232,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.14220401644706726,
      "learning_rate": 1.1469029768073125e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 23391240,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.10299301147460938,
      "learning_rate": 1.1420381033241889e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 23394856,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 0.2537440061569214,
      "learning_rate": 1.1371834503693002e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 23398472,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.21726979315280914,
      "learning_rate": 1.1323390189581784e-05,
      "loss": 0.3616,
      "num_input_tokens_seen": 23401640,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.15720190107822418,
      "learning_rate": 1.127504810104213e-05,
      "loss": 0.3258,
      "num_input_tokens_seen": 23404776,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.1390436589717865,
      "learning_rate": 1.1226808248186782e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 23407944,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.23306827247142792,
      "learning_rate": 1.1178670641106891e-05,
      "loss": 0.4702,
      "num_input_tokens_seen": 23411496,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.193431094288826,
      "learning_rate": 1.1130635289872403e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 23414440,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.20571961998939514,
      "learning_rate": 1.1082702204531725e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 23418152,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.20399607717990875,
      "learning_rate": 1.1034871395111778e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 23421320,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.13380391895771027,
      "learning_rate": 1.0987142871618394e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 23424968,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 0.25615450739860535,
      "learning_rate": 1.0939516644035696e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 23427592,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 0.18310369551181793,
      "learning_rate": 1.0891992722326716e-05,
      "loss": 0.5272,
      "num_input_tokens_seen": 23431080,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.13264212012290955,
      "learning_rate": 1.0844571116432778e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 23434728,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.1965257227420807,
      "learning_rate": 1.0797251836274003e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 23439112,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.0955861434340477,
      "learning_rate": 1.0750034891748972e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 23442344,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.13321605324745178,
      "learning_rate": 1.0702920292735009e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 23445640,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.14671945571899414,
      "learning_rate": 1.0655908049087893e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 23448264,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.2448928952217102,
      "learning_rate": 1.0608998170642149e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 23451272,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.1331969052553177,
      "learning_rate": 1.0562190667210703e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 23453832,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.2905953526496887,
      "learning_rate": 1.0515485548585113e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 23456520,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 0.28623029589653015,
      "learning_rate": 1.0468882824535676e-05,
      "loss": 0.3608,
      "num_input_tokens_seen": 23459624,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 0.17610014975070953,
      "learning_rate": 1.0422382504811034e-05,
      "loss": 0.3168,
      "num_input_tokens_seen": 23463720,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.10059921443462372,
      "learning_rate": 1.0375984599138633e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 23467592,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.2197076827287674,
      "learning_rate": 1.0329689117224261e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 23470024,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.31795018911361694,
      "learning_rate": 1.0283496068752507e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 23472264,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.18017710745334625,
      "learning_rate": 1.0237405463386418e-05,
      "loss": 0.4993,
      "num_input_tokens_seen": 23475208,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.13812997937202454,
      "learning_rate": 1.0191417310767503e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 23478472,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.19653576612472534,
      "learning_rate": 1.0145531620516179e-05,
      "loss": 0.2936,
      "num_input_tokens_seen": 23482056,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.4552341103553772,
      "learning_rate": 1.009974840223099e-05,
      "loss": 0.3089,
      "num_input_tokens_seen": 23484200,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.0975058376789093,
      "learning_rate": 1.0054067665489386e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 23488616,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.1607268750667572,
      "learning_rate": 1.0008489419847278e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 23491944,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 0.12499398738145828,
      "learning_rate": 9.963013674839038e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 23496008,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.14335346221923828,
      "learning_rate": 9.91764043997767e-06,
      "loss": 0.4354,
      "num_input_tokens_seen": 23498888,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.15893903374671936,
      "learning_rate": 9.872369724754804e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 23501896,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.18764974176883698,
      "learning_rate": 9.827201538640473e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 23504840,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.17190995812416077,
      "learning_rate": 9.782135891083455e-06,
      "loss": 0.3839,
      "num_input_tokens_seen": 23507336,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.11147291213274002,
      "learning_rate": 9.737172791510873e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 23510536,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.21879243850708008,
      "learning_rate": 9.692312249328483e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 23513576,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 0.11018979549407959,
      "learning_rate": 9.647554273920722e-06,
      "loss": 0.6944,
      "num_input_tokens_seen": 23516744,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.17133565247058868,
      "learning_rate": 9.602898874650323e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 23519720,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 0.1869552731513977,
      "learning_rate": 9.558346060858759e-06,
      "loss": 0.6306,
      "num_input_tokens_seen": 23522696,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.2892177104949951,
      "learning_rate": 9.51389584186596e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 23525160,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.1605701744556427,
      "learning_rate": 9.469548226970326e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 23527848,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.17305505275726318,
      "learning_rate": 9.425303225448989e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 23530824,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.3159346878528595,
      "learning_rate": 9.381160846557435e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 23533480,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 0.23282261192798615,
      "learning_rate": 9.337121099529722e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 23535688,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.11582294851541519,
      "learning_rate": 9.293183993578535e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 23539016,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.15800157189369202,
      "learning_rate": 9.249349537894968e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 23542152,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.15629443526268005,
      "learning_rate": 9.205617741648686e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 23545096,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.16521908342838287,
      "learning_rate": 9.161988613987982e-06,
      "loss": 0.259,
      "num_input_tokens_seen": 23548392,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 0.2421003133058548,
      "learning_rate": 9.118462164039387e-06,
      "loss": 0.421,
      "num_input_tokens_seen": 23550984,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.25621670484542847,
      "learning_rate": 9.07503840090823e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 23553608,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.3971928060054779,
      "learning_rate": 9.031717333678303e-06,
      "loss": 0.4779,
      "num_input_tokens_seen": 23556360,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 0.301770955324173,
      "learning_rate": 8.988498971411851e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 23559656,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.15425007045269012,
      "learning_rate": 8.945383323149647e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 23563080,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.14660774171352386,
      "learning_rate": 8.902370397911031e-06,
      "loss": 0.557,
      "num_input_tokens_seen": 23566056,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.168472021818161,
      "learning_rate": 8.859460204693748e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 23569480,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.277244508266449,
      "learning_rate": 8.816652752474175e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 23573704,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.15963692963123322,
      "learning_rate": 8.773948050207148e-06,
      "loss": 0.4292,
      "num_input_tokens_seen": 23577032,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.3006892502307892,
      "learning_rate": 8.731346106826021e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 23580232,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.130867138504982,
      "learning_rate": 8.6888469312425e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 23583816,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 0.29145655035972595,
      "learning_rate": 8.646450532347083e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 23587208,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.11941487342119217,
      "learning_rate": 8.604156919008565e-06,
      "loss": 0.5228,
      "num_input_tokens_seen": 23590920,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.29303058981895447,
      "learning_rate": 8.561966100074258e-06,
      "loss": 0.6584,
      "num_input_tokens_seen": 23593992,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.14334027469158173,
      "learning_rate": 8.519878084370048e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 23597512,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 0.18805712461471558,
      "learning_rate": 8.477892880700222e-06,
      "loss": 0.4015,
      "num_input_tokens_seen": 23600232,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.23226068913936615,
      "learning_rate": 8.436010497847646e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 23602920,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.20713433623313904,
      "learning_rate": 8.39423094457359e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 23605960,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 0.11275664716959,
      "learning_rate": 8.352554229617892e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 23609384,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.303833931684494,
      "learning_rate": 8.310980361698861e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 23612520,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.22582051157951355,
      "learning_rate": 8.26950934951326e-06,
      "loss": 0.4879,
      "num_input_tokens_seen": 23615784,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.09566082060337067,
      "learning_rate": 8.22814120173626e-06,
      "loss": 0.4344,
      "num_input_tokens_seen": 23619144,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.17963749170303345,
      "learning_rate": 8.186875927021775e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 23621992,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.12958139181137085,
      "learning_rate": 8.145713534001897e-06,
      "loss": 0.5596,
      "num_input_tokens_seen": 23625416,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.12808336317539215,
      "learning_rate": 8.104654031287406e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 23629448,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.21344414353370667,
      "learning_rate": 8.063697427467486e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 23632232,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 0.0863402932882309,
      "learning_rate": 8.022843731109675e-06,
      "loss": 0.36,
      "num_input_tokens_seen": 23635880,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.15456055104732513,
      "learning_rate": 7.982092950760245e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 23639016,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 0.1967790424823761,
      "learning_rate": 7.941445094943711e-06,
      "loss": 0.3947,
      "num_input_tokens_seen": 23642376,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.15060468018054962,
      "learning_rate": 7.900900172163107e-06,
      "loss": 0.3904,
      "num_input_tokens_seen": 23645896,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 0.128324493765831,
      "learning_rate": 7.860458190900144e-06,
      "loss": 0.505,
      "num_input_tokens_seen": 23649608,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.21097926795482635,
      "learning_rate": 7.820119159614669e-06,
      "loss": 0.4933,
      "num_input_tokens_seen": 23652808,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.08396576344966888,
      "learning_rate": 7.779883086745098e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 23656776,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.1949942409992218,
      "learning_rate": 7.739749980708533e-06,
      "loss": 0.4872,
      "num_input_tokens_seen": 23660584,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.12622672319412231,
      "learning_rate": 7.6997198499002e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 23664296,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.26880761981010437,
      "learning_rate": 7.659792702694068e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 23667048,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.1494639813899994,
      "learning_rate": 7.619968547442346e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 23670536,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.49396204948425293,
      "eval_runtime": 13.7006,
      "eval_samples_per_second": 61.895,
      "eval_steps_per_second": 15.474,
      "num_input_tokens_seen": 23671264,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.22105488181114197,
      "learning_rate": 7.580247392475926e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 23673120,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.14056064188480377,
      "learning_rate": 7.540629246103825e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 23675872,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 0.1833309531211853,
      "learning_rate": 7.501114116613861e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 23678560,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 0.157954603433609,
      "learning_rate": 7.461702012272087e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 23682208,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.21690939366817474,
      "learning_rate": 7.422392941323075e-06,
      "loss": 0.5508,
      "num_input_tokens_seen": 23685472,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 0.17387805879116058,
      "learning_rate": 7.383186911989858e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 23688064,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.18940132856369019,
      "learning_rate": 7.344083932473822e-06,
      "loss": 0.3515,
      "num_input_tokens_seen": 23690848,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.19645382463932037,
      "learning_rate": 7.305084010954976e-06,
      "loss": 0.4146,
      "num_input_tokens_seen": 23693952,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.20537424087524414,
      "learning_rate": 7.26618715559152e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 23698016,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.31263047456741333,
      "learning_rate": 7.227393374520386e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 23702976,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.13161338865756989,
      "learning_rate": 7.18870267585664e-06,
      "loss": 0.3987,
      "num_input_tokens_seen": 23706464,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.08579953014850616,
      "learning_rate": 7.150115067694085e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 23710464,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.1578620970249176,
      "learning_rate": 7.111630558104653e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 23713632,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.22468307614326477,
      "learning_rate": 7.073249155138961e-06,
      "loss": 0.6009,
      "num_input_tokens_seen": 23716192,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.14432649314403534,
      "learning_rate": 7.0349708668259736e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 23719424,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 0.22165478765964508,
      "learning_rate": 6.996795701173009e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 23722400,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.27770188450813293,
      "learning_rate": 6.958723666165901e-06,
      "loss": 0.5132,
      "num_input_tokens_seen": 23725792,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.19882969558238983,
      "learning_rate": 6.9207547697689446e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 23728800,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.19363047182559967,
      "learning_rate": 6.882889019924676e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 23732192,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.09683024883270264,
      "learning_rate": 6.84512642455426e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 23735072,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.15153451263904572,
      "learning_rate": 6.8074669915572095e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 23738144,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.2504647672176361,
      "learning_rate": 6.769910728811391e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 23740768,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.20423240959644318,
      "learning_rate": 6.73245764417324e-06,
      "loss": 0.4834,
      "num_input_tokens_seen": 23743456,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.23698431253433228,
      "learning_rate": 6.695107745477435e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 23746528,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.32792896032333374,
      "learning_rate": 6.657861040537117e-06,
      "loss": 0.5694,
      "num_input_tokens_seen": 23749024,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.30880284309387207,
      "learning_rate": 6.620717537143994e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 23751552,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.1250816136598587,
      "learning_rate": 6.583677243067965e-06,
      "loss": 0.4066,
      "num_input_tokens_seen": 23754784,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.12027736008167267,
      "learning_rate": 6.546740166057441e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 23761888,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.15024591982364655,
      "learning_rate": 6.5099063138392975e-06,
      "loss": 0.5157,
      "num_input_tokens_seen": 23765728,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.16601765155792236,
      "learning_rate": 6.473175694118705e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 23769536,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.15866181254386902,
      "learning_rate": 6.436548314579349e-06,
      "loss": 0.453,
      "num_input_tokens_seen": 23772512,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.12611669301986694,
      "learning_rate": 6.400024182883158e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 23776128,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 0.10850350558757782,
      "learning_rate": 6.363603306670629e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 23779936,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 0.10774760693311691,
      "learning_rate": 6.327285693560614e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 23783808,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.19312427937984467,
      "learning_rate": 6.2910713511503125e-06,
      "loss": 0.3823,
      "num_input_tokens_seen": 23787872,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.13499245047569275,
      "learning_rate": 6.254960287015332e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 23791296,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.2005978673696518,
      "learning_rate": 6.218952508709741e-06,
      "loss": 0.5059,
      "num_input_tokens_seen": 23795456,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.13748490810394287,
      "learning_rate": 6.183048023765903e-06,
      "loss": 0.9545,
      "num_input_tokens_seen": 23798080,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.2047869861125946,
      "learning_rate": 6.147246839694698e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 23801216,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 0.18929341435432434,
      "learning_rate": 6.111548963985247e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 23804320,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.2951078414916992,
      "learning_rate": 6.075954404105188e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 23807584,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.09407731890678406,
      "learning_rate": 6.040463167500509e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 23812320,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.15398062765598297,
      "learning_rate": 6.005075261595494e-06,
      "loss": 0.4501,
      "num_input_tokens_seen": 23815296,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.12519794702529907,
      "learning_rate": 5.969790693792998e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 23818880,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.21900847554206848,
      "learning_rate": 5.9346094714740615e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 23822112,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 0.1863218992948532,
      "learning_rate": 5.8995316019982425e-06,
      "loss": 0.3513,
      "num_input_tokens_seen": 23824928,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.19592559337615967,
      "learning_rate": 5.8645570927034485e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 23828512,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 0.22650186717510223,
      "learning_rate": 5.8296859509058275e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 23831616,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.12459415197372437,
      "learning_rate": 5.794918183900155e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 23835648,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 0.08872047066688538,
      "learning_rate": 5.760253798959447e-06,
      "loss": 0.3552,
      "num_input_tokens_seen": 23839424,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.2698991596698761,
      "learning_rate": 5.725692803335015e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 23842432,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.1994176059961319,
      "learning_rate": 5.691235204256739e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 23846016,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.17810937762260437,
      "learning_rate": 5.65688100893258e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 23850144,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.1494041383266449,
      "learning_rate": 5.622630224549174e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 23853216,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.3847842216491699,
      "learning_rate": 5.588482858271404e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 23856864,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.12668104469776154,
      "learning_rate": 5.554438917242444e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 23860192,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 0.15756602585315704,
      "learning_rate": 5.520498408583985e-06,
      "loss": 0.4384,
      "num_input_tokens_seen": 23863456,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 0.2185998409986496,
      "learning_rate": 5.486661339395904e-06,
      "loss": 0.3376,
      "num_input_tokens_seen": 23866432,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.2309594601392746,
      "learning_rate": 5.452927716756595e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 23869632,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.1298162192106247,
      "learning_rate": 5.419297547722635e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 23872864,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.22570419311523438,
      "learning_rate": 5.385770839329229e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 23876000,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.12770922482013702,
      "learning_rate": 5.352347598589713e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 23879616,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.1590583324432373,
      "learning_rate": 5.319027832495826e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 23882624,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.19354885816574097,
      "learning_rate": 5.285811548017661e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 23886304,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.1263309121131897,
      "learning_rate": 5.252698752103713e-06,
      "loss": 0.397,
      "num_input_tokens_seen": 23890624,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.12593795359134674,
      "learning_rate": 5.219689451680832e-06,
      "loss": 0.4618,
      "num_input_tokens_seen": 23893376,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.11057619750499725,
      "learning_rate": 5.186783653654214e-06,
      "loss": 0.4258,
      "num_input_tokens_seen": 23896608,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.15896178781986237,
      "learning_rate": 5.153981364907245e-06,
      "loss": 0.3835,
      "num_input_tokens_seen": 23903872,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.09045697748661041,
      "learning_rate": 5.121282592301935e-06,
      "loss": 0.3508,
      "num_input_tokens_seen": 23906912,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.20032541453838348,
      "learning_rate": 5.088687342678422e-06,
      "loss": 0.4343,
      "num_input_tokens_seen": 23909888,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.16046489775180817,
      "learning_rate": 5.056195622855253e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 23913568,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.1800174117088318,
      "learning_rate": 5.023807439629324e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 23916928,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.1280038058757782,
      "learning_rate": 4.991522799775938e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 23919840,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 0.09710245579481125,
      "learning_rate": 4.9593417100485816e-06,
      "loss": 0.4377,
      "num_input_tokens_seen": 23922944,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.19808687269687653,
      "learning_rate": 4.927264177179258e-06,
      "loss": 0.4592,
      "num_input_tokens_seen": 23926592,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 0.20424969494342804,
      "learning_rate": 4.895290207878156e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 23929632,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.17838475108146667,
      "learning_rate": 4.863419808833924e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 23933216,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.19473998248577118,
      "learning_rate": 4.831652986713453e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 23936416,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 0.13191746175289154,
      "learning_rate": 4.799989748161926e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 23939264,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.19761838018894196,
      "learning_rate": 4.768430099803101e-06,
      "loss": 0.3326,
      "num_input_tokens_seen": 23944192,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.20709697902202606,
      "learning_rate": 4.736974048238696e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 23946880,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.2615889608860016,
      "learning_rate": 4.705621600049115e-06,
      "loss": 0.5137,
      "num_input_tokens_seen": 23949184,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.24738973379135132,
      "learning_rate": 4.674372761792889e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 23952256,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 0.2048969566822052,
      "learning_rate": 4.6432275400069e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 23954912,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.1466105431318283,
      "learning_rate": 4.6121859412063264e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 23958496,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.238631010055542,
      "learning_rate": 4.581247971884861e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 23961408,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 0.17875885963439941,
      "learning_rate": 4.550413638514217e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 23964160,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.18050068616867065,
      "learning_rate": 4.519682947544679e-06,
      "loss": 0.5501,
      "num_input_tokens_seen": 23966784,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.3514525592327118,
      "learning_rate": 4.489055905404715e-06,
      "loss": 0.5166,
      "num_input_tokens_seen": 23969664,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.3662398159503937,
      "learning_rate": 4.4585325185012014e-06,
      "loss": 0.3685,
      "num_input_tokens_seen": 23972736,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 0.14580301940441132,
      "learning_rate": 4.428112793219197e-06,
      "loss": 0.3192,
      "num_input_tokens_seen": 23976064,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.22935588657855988,
      "learning_rate": 4.397796735922277e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 23978976,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.24870896339416504,
      "learning_rate": 4.367584352952092e-06,
      "loss": 0.4896,
      "num_input_tokens_seen": 23981984,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.1804845631122589,
      "learning_rate": 4.337475650628808e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 23984864,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.17422884702682495,
      "learning_rate": 4.307470635250832e-06,
      "loss": 0.5824,
      "num_input_tokens_seen": 23988288,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.1720520704984665,
      "learning_rate": 4.277569313094809e-06,
      "loss": 0.3417,
      "num_input_tokens_seen": 23991808,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.20809929072856903,
      "learning_rate": 4.247771690415791e-06,
      "loss": 0.4659,
      "num_input_tokens_seen": 23995840,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.1782568246126175,
      "learning_rate": 4.218077773447071e-06,
      "loss": 0.4566,
      "num_input_tokens_seen": 23998304,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.27651333808898926,
      "learning_rate": 4.1884875684003455e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 24000608,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.1585761457681656,
      "learning_rate": 4.159001081465497e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 24003808,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.255752295255661,
      "learning_rate": 4.129618318810702e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 24006784,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.18534301221370697,
      "learning_rate": 4.100339286582655e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 24009376,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.1810142993927002,
      "learning_rate": 4.071163990906068e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 24013120,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.21697211265563965,
      "learning_rate": 4.042092437884115e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 24016384,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 0.16040997207164764,
      "learning_rate": 4.013124633598264e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 24020416,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.15836523473262787,
      "learning_rate": 3.984260584108168e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 24024096,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.2414993941783905,
      "learning_rate": 3.955500295451942e-06,
      "loss": 0.3077,
      "num_input_tokens_seen": 24026336,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.19799691438674927,
      "learning_rate": 3.926843773645883e-06,
      "loss": 0.2307,
      "num_input_tokens_seen": 24028736,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.13238023221492767,
      "learning_rate": 3.8982910246846415e-06,
      "loss": 0.4735,
      "num_input_tokens_seen": 24031936,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.2556564211845398,
      "learning_rate": 3.869842054541051e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 24034752,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.13602302968502045,
      "learning_rate": 3.84149686916635e-06,
      "loss": 0.4016,
      "num_input_tokens_seen": 24038208,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.18317270278930664,
      "learning_rate": 3.8132554744900183e-06,
      "loss": 0.4909,
      "num_input_tokens_seen": 24041312,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.17022660374641418,
      "learning_rate": 3.7851178764198302e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 24044160,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.12827907502651215,
      "learning_rate": 3.7570840808419104e-06,
      "loss": 0.3412,
      "num_input_tokens_seen": 24047040,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.17092514038085938,
      "learning_rate": 3.729154093620568e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 24050720,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.1900731474161148,
      "learning_rate": 3.7013279205984073e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 24054048,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 0.1340666562318802,
      "learning_rate": 3.6736055675963275e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 24057152,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.13006903231143951,
      "learning_rate": 3.645987040413634e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 24060800,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.09185352921485901,
      "learning_rate": 3.6184723448277056e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 24064032,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.15630127489566803,
      "learning_rate": 3.5910614865943826e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 24067808,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.10104400664567947,
      "learning_rate": 3.563754471447689e-06,
      "loss": 0.4656,
      "num_input_tokens_seen": 24071680,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.23546016216278076,
      "learning_rate": 3.5365513050998334e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 24074336,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 0.12856914103031158,
      "learning_rate": 3.5094519932415414e-06,
      "loss": 0.6446,
      "num_input_tokens_seen": 24077152,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.22319717705249786,
      "learning_rate": 3.4824565415416123e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 24079744,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 0.23675483465194702,
      "learning_rate": 3.4555649556471946e-06,
      "loss": 0.5341,
      "num_input_tokens_seen": 24083008,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 0.21992507576942444,
      "learning_rate": 3.4287772411837338e-06,
      "loss": 0.3709,
      "num_input_tokens_seen": 24085984,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 0.11667782068252563,
      "learning_rate": 3.402093403754858e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 24089024,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.20676949620246887,
      "learning_rate": 3.375513448942602e-06,
      "loss": 0.3902,
      "num_input_tokens_seen": 24093056,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.23271280527114868,
      "learning_rate": 3.349037382307074e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 24096448,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.28450679779052734,
      "learning_rate": 3.322665209386899e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 24098592,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.15830527245998383,
      "learning_rate": 3.29639693569872e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 24100864,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 0.24224984645843506,
      "learning_rate": 3.270232566737641e-06,
      "loss": 0.428,
      "num_input_tokens_seen": 24103680,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.24536632001399994,
      "learning_rate": 3.244172107976895e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 24106784,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.13300561904907227,
      "learning_rate": 3.2182155648680657e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 24110208,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.14044110476970673,
      "learning_rate": 3.1923629428409205e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 24113216,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.1355981081724167,
      "learning_rate": 3.166614247303634e-06,
      "loss": 0.4586,
      "num_input_tokens_seen": 24116192,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.24901317059993744,
      "learning_rate": 3.140969483642453e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 24119200,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.20377673208713531,
      "learning_rate": 3.1154286572219747e-06,
      "loss": 0.5045,
      "num_input_tokens_seen": 24122912,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.15094886720180511,
      "learning_rate": 3.089991773385037e-06,
      "loss": 0.5067,
      "num_input_tokens_seen": 24125760,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.16629861295223236,
      "learning_rate": 3.064658837452772e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 24129280,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.18647156655788422,
      "learning_rate": 3.0394298547246068e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 24132000,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.08430737257003784,
      "learning_rate": 3.0143048304779875e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 24135808,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.1551714390516281,
      "learning_rate": 2.989283769968987e-06,
      "loss": 0.5266,
      "num_input_tokens_seen": 24139040,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.24949508905410767,
      "learning_rate": 2.964366678431585e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 24141760,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 0.1824297457933426,
      "learning_rate": 2.9395535610781678e-06,
      "loss": 0.4753,
      "num_input_tokens_seen": 24145216,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.1498030722141266,
      "learning_rate": 2.9148444230994166e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 24148320,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 0.21627452969551086,
      "learning_rate": 2.890239269664141e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 24150816,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.3864562213420868,
      "learning_rate": 2.8657381059194466e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 24153760,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 0.25205475091934204,
      "learning_rate": 2.8413409369907885e-06,
      "loss": 0.5218,
      "num_input_tokens_seen": 24156800,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 0.16661427915096283,
      "learning_rate": 2.817047767981695e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 24160736,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.23069363832473755,
      "learning_rate": 2.7928586039740466e-06,
      "loss": 0.3649,
      "num_input_tokens_seen": 24163424,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.14070814847946167,
      "learning_rate": 2.7687734500279615e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 24166880,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.26423338055610657,
      "learning_rate": 2.744792311181743e-06,
      "loss": 0.4761,
      "num_input_tokens_seen": 24169984,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 0.1545620709657669,
      "learning_rate": 2.720915192451989e-06,
      "loss": 0.5184,
      "num_input_tokens_seen": 24173248,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.1652277410030365,
      "learning_rate": 2.697142098833538e-06,
      "loss": 0.3636,
      "num_input_tokens_seen": 24176512,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.14182408154010773,
      "learning_rate": 2.6734730352993563e-06,
      "loss": 0.4628,
      "num_input_tokens_seen": 24179296,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.11164254695177078,
      "learning_rate": 2.649908006800872e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 24183008,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.1284722089767456,
      "learning_rate": 2.626447018267586e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 24186048,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.23248930275440216,
      "learning_rate": 2.603090074607184e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 24188640,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 0.18661947548389435,
      "learning_rate": 2.579837180705813e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 24191040,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.20522159337997437,
      "learning_rate": 2.556688341427582e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 24194016,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.14947739243507385,
      "learning_rate": 2.5336435616150066e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 24196320,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.17660829424858093,
      "learning_rate": 2.510702846088786e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 24199456,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.15956218540668488,
      "learning_rate": 2.487866199647915e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 24202848,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.16804397106170654,
      "learning_rate": 2.4651336270695156e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 24206016,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.3661581873893738,
      "learning_rate": 2.442505133108952e-06,
      "loss": 0.5147,
      "num_input_tokens_seen": 24208768,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.16480129957199097,
      "learning_rate": 2.419980722499937e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 24212320,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.18704599142074585,
      "learning_rate": 2.397560399954202e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 24215232,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.10539701581001282,
      "learning_rate": 2.3752441701618833e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 24218336,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.12373995035886765,
      "learning_rate": 2.3530320377913027e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 24222176,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.17896835505962372,
      "learning_rate": 2.3309240074890213e-06,
      "loss": 0.5679,
      "num_input_tokens_seen": 24225536,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.12759369611740112,
      "learning_rate": 2.3089200838796176e-06,
      "loss": 0.5769,
      "num_input_tokens_seen": 24228352,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.2135486751794815,
      "learning_rate": 2.2870202715662426e-06,
      "loss": 0.5192,
      "num_input_tokens_seen": 24231168,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.11573293805122375,
      "learning_rate": 2.265224575130009e-06,
      "loss": 0.3147,
      "num_input_tokens_seen": 24237184,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.3304905891418457,
      "learning_rate": 2.2435329991303268e-06,
      "loss": 0.4688,
      "num_input_tokens_seen": 24240672,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.1317165344953537,
      "learning_rate": 2.2219455481047868e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 24244032,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.2885829210281372,
      "learning_rate": 2.2004622265693886e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 24246496,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.21875135600566864,
      "learning_rate": 2.179083039018037e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 24249376,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.1117856502532959,
      "learning_rate": 2.157807989923044e-06,
      "loss": 0.3681,
      "num_input_tokens_seen": 24252288,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.28152090311050415,
      "learning_rate": 2.1366370837349603e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 24254432,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.33978062868118286,
      "learning_rate": 2.1155703248825207e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 24257408,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.31521913409233093,
      "learning_rate": 2.094607717772534e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 24260224,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 0.10185521095991135,
      "learning_rate": 2.0737492667902702e-06,
      "loss": 0.4835,
      "num_input_tokens_seen": 24263200,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.15079210698604584,
      "learning_rate": 2.0529949762989608e-06,
      "loss": 0.4686,
      "num_input_tokens_seen": 24266240,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.09136363863945007,
      "learning_rate": 2.032344850640244e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 24270656,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.17015603184700012,
      "learning_rate": 2.011798894133887e-06,
      "loss": 0.4146,
      "num_input_tokens_seen": 24274016,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 0.13885654509067535,
      "learning_rate": 1.9913571110777852e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 24277536,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.16909579932689667,
      "learning_rate": 1.971019505748295e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 24280192,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 0.15813784301280975,
      "learning_rate": 1.9507860823996803e-06,
      "loss": 0.4423,
      "num_input_tokens_seen": 24283232,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.17102308571338654,
      "learning_rate": 1.9306568452645e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 24286368,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.19333699345588684,
      "learning_rate": 1.910631798553664e-06,
      "loss": 0.404,
      "num_input_tokens_seen": 24289952,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.13937750458717346,
      "learning_rate": 1.8907109464562088e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 24293952,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.16601718962192535,
      "learning_rate": 1.870894293139247e-06,
      "loss": 0.397,
      "num_input_tokens_seen": 24297216,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.19538354873657227,
      "learning_rate": 1.8511818427482396e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 24300288,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.1674666553735733,
      "learning_rate": 1.8315735994068327e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 24303328,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.19254465401172638,
      "learning_rate": 1.8120695672168009e-06,
      "loss": 0.6623,
      "num_input_tokens_seen": 24306560,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.1609247475862503,
      "learning_rate": 1.792669750258158e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 24309760,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.1669510304927826,
      "learning_rate": 1.7733741525892134e-06,
      "loss": 0.5406,
      "num_input_tokens_seen": 24313504,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.18035106360912323,
      "learning_rate": 1.7541827782462937e-06,
      "loss": 0.4763,
      "num_input_tokens_seen": 24316800,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 0.1270327866077423,
      "learning_rate": 1.7350956312440768e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 24319616,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.14794157445430756,
      "learning_rate": 1.716112715575313e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 24323744,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.14432576298713684,
      "learning_rate": 1.6972340352110481e-06,
      "loss": 0.375,
      "num_input_tokens_seen": 24327456,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.09550423920154572,
      "learning_rate": 1.6784595941004565e-06,
      "loss": 0.4771,
      "num_input_tokens_seen": 24330400,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.12297312170267105,
      "learning_rate": 1.659789396171063e-06,
      "loss": 0.4735,
      "num_input_tokens_seen": 24333696,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.17826983332633972,
      "learning_rate": 1.6412234453282993e-06,
      "loss": 0.587,
      "num_input_tokens_seen": 24337216,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.12537159025669098,
      "learning_rate": 1.622761745456003e-06,
      "loss": 0.3825,
      "num_input_tokens_seen": 24340384,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 0.19271674752235413,
      "learning_rate": 1.6044043004161958e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 24343008,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.1334904134273529,
      "learning_rate": 1.5861511140489725e-06,
      "loss": 0.2716,
      "num_input_tokens_seen": 24345088,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 0.19589683413505554,
      "learning_rate": 1.5680021901727237e-06,
      "loss": 0.6099,
      "num_input_tokens_seen": 24348384,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 0.14634065330028534,
      "learning_rate": 1.5499575325840232e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 24351840,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 0.173519104719162,
      "learning_rate": 1.5320171450576293e-06,
      "loss": 0.4665,
      "num_input_tokens_seen": 24355232,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.16396592557430267,
      "learning_rate": 1.5141810313463733e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 24359200,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.17990873754024506,
      "learning_rate": 1.4964491951814374e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 24362272,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.1959076076745987,
      "learning_rate": 1.4788216402720766e-06,
      "loss": 0.5016,
      "num_input_tokens_seen": 24365632,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.19788800179958344,
      "learning_rate": 1.4612983703058413e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 24368256,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.18767480552196503,
      "learning_rate": 1.4438793889483549e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 24371904,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.10475440323352814,
      "learning_rate": 1.4265646998434246e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 24376640,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.2161829024553299,
      "learning_rate": 1.409354306613153e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 24380288,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.17482958734035492,
      "learning_rate": 1.3922482128577718e-06,
      "loss": 0.4975,
      "num_input_tokens_seen": 24383328,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.1636354774236679,
      "learning_rate": 1.3752464221556404e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 24386464,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 0.17654536664485931,
      "learning_rate": 1.358348938063303e-06,
      "loss": 0.3508,
      "num_input_tokens_seen": 24389440,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 0.13563226163387299,
      "learning_rate": 1.341555764115543e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 24393088,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.2539350688457489,
      "learning_rate": 1.3248669038253835e-06,
      "loss": 0.4432,
      "num_input_tokens_seen": 24396224,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.3448690176010132,
      "learning_rate": 1.3082823606838656e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 24398592,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.1848663091659546,
      "learning_rate": 1.2918021381603251e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 24403136,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 0.11027737706899643,
      "learning_rate": 1.275426239702171e-06,
      "loss": 0.3107,
      "num_input_tokens_seen": 24406048,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.1514858454465866,
      "learning_rate": 1.2591546687351073e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 24408864,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.15174615383148193,
      "learning_rate": 1.2429874286629673e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 24412128,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.21766778826713562,
      "learning_rate": 1.2269245228677116e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 24415488,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.13625144958496094,
      "learning_rate": 1.210965954709542e-06,
      "loss": 0.4132,
      "num_input_tokens_seen": 24418080,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 0.1599377542734146,
      "learning_rate": 1.1951117275268431e-06,
      "loss": 0.5161,
      "num_input_tokens_seen": 24421280,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.1856415569782257,
      "learning_rate": 1.1793618446360732e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 24424032,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.30534887313842773,
      "learning_rate": 1.1637163093319303e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 24426528,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.3310985863208771,
      "learning_rate": 1.1481751248874072e-06,
      "loss": 0.234,
      "num_input_tokens_seen": 24429632,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.13999901711940765,
      "learning_rate": 1.1327382945533482e-06,
      "loss": 0.4158,
      "num_input_tokens_seen": 24432832,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.29402437806129456,
      "learning_rate": 1.1174058215591143e-06,
      "loss": 0.5055,
      "num_input_tokens_seen": 24435712,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.3035901188850403,
      "learning_rate": 1.1021777091119732e-06,
      "loss": 0.652,
      "num_input_tokens_seen": 24438496,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.17551009356975555,
      "learning_rate": 1.0870539603975994e-06,
      "loss": 0.5132,
      "num_input_tokens_seen": 24441312,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.202798530459404,
      "learning_rate": 1.0720345785795727e-06,
      "loss": 0.4341,
      "num_input_tokens_seen": 24443840,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.2309851348400116,
      "learning_rate": 1.0571195667998802e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 24447520,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.16568708419799805,
      "learning_rate": 1.042308928178526e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 24450720,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.11912406235933304,
      "learning_rate": 1.0276026658137538e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 24455328,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.38300687074661255,
      "learning_rate": 1.013000782781881e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 24458496,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.11906341463327408,
      "learning_rate": 9.985032821375195e-07,
      "loss": 0.5322,
      "num_input_tokens_seen": 24462432,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.15839840471744537,
      "learning_rate": 9.841101669134101e-07,
      "loss": 0.4428,
      "num_input_tokens_seen": 24466240,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.042330946773290634,
      "learning_rate": 9.69821440120311e-07,
      "loss": 0.2781,
      "num_input_tokens_seen": 24472064,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 0.15654954314231873,
      "learning_rate": 9.556371047473866e-07,
      "loss": 0.3946,
      "num_input_tokens_seen": 24474496,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.21145907044410706,
      "learning_rate": 9.415571637617082e-07,
      "loss": 0.4054,
      "num_input_tokens_seen": 24478016,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.15233923494815826,
      "learning_rate": 9.275816201087528e-07,
      "loss": 0.442,
      "num_input_tokens_seen": 24481024,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.1608203500509262,
      "learning_rate": 9.137104767120153e-07,
      "loss": 0.407,
      "num_input_tokens_seen": 24484864,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.1649097502231598,
      "learning_rate": 8.999437364731189e-07,
      "loss": 0.4181,
      "num_input_tokens_seen": 24488480,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.1850004494190216,
      "learning_rate": 8.862814022720378e-07,
      "loss": 0.5786,
      "num_input_tokens_seen": 24491136,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 0.10563254356384277,
      "learning_rate": 8.727234769666526e-07,
      "loss": 0.4672,
      "num_input_tokens_seen": 24494944,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.13176818192005157,
      "learning_rate": 8.592699633931389e-07,
      "loss": 0.3444,
      "num_input_tokens_seen": 24497920,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.14913611114025116,
      "learning_rate": 8.459208643659122e-07,
      "loss": 0.4458,
      "num_input_tokens_seen": 24500800,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 0.291004478931427,
      "learning_rate": 8.326761826773499e-07,
      "loss": 0.5145,
      "num_input_tokens_seen": 24504416,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.14589574933052063,
      "learning_rate": 8.195359210981246e-07,
      "loss": 0.4289,
      "num_input_tokens_seen": 24508000,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.2228497713804245,
      "learning_rate": 8.065000823770929e-07,
      "loss": 0.4948,
      "num_input_tokens_seen": 24513280,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.23056542873382568,
      "learning_rate": 7.935686692410737e-07,
      "loss": 0.4286,
      "num_input_tokens_seen": 24516000,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.17513436079025269,
      "learning_rate": 7.807416843952364e-07,
      "loss": 0.3642,
      "num_input_tokens_seen": 24518624,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.11355561763048172,
      "learning_rate": 7.68019130522879e-07,
      "loss": 0.4408,
      "num_input_tokens_seen": 24522144,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.13470211625099182,
      "learning_rate": 7.554010102853726e-07,
      "loss": 0.3726,
      "num_input_tokens_seen": 24525312,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.15231144428253174,
      "learning_rate": 7.428873263223279e-07,
      "loss": 0.5036,
      "num_input_tokens_seen": 24528288,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.1583266258239746,
      "learning_rate": 7.304780812513734e-07,
      "loss": 0.4381,
      "num_input_tokens_seen": 24531040,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.2189723700284958,
      "learning_rate": 7.181732776684325e-07,
      "loss": 0.4578,
      "num_input_tokens_seen": 24534240,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.20927555859088898,
      "learning_rate": 7.059729181475572e-07,
      "loss": 0.4476,
      "num_input_tokens_seen": 24536864,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.11965298652648926,
      "learning_rate": 6.938770052409282e-07,
      "loss": 0.4015,
      "num_input_tokens_seen": 24540160,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.20257636904716492,
      "learning_rate": 6.81885541478855e-07,
      "loss": 0.4856,
      "num_input_tokens_seen": 24543040,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.16759350895881653,
      "learning_rate": 6.699985293697197e-07,
      "loss": 0.379,
      "num_input_tokens_seen": 24545792,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 0.14201411604881287,
      "learning_rate": 6.582159714003111e-07,
      "loss": 0.4576,
      "num_input_tokens_seen": 24548512,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 0.15954437851905823,
      "learning_rate": 6.465378700352687e-07,
      "loss": 0.3742,
      "num_input_tokens_seen": 24551200,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.24207822978496552,
      "learning_rate": 6.349642277176382e-07,
      "loss": 0.4091,
      "num_input_tokens_seen": 24554016,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.2457657754421234,
      "learning_rate": 6.23495046868372e-07,
      "loss": 0.4998,
      "num_input_tokens_seen": 24557952,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.12471891939640045,
      "learning_rate": 6.121303298868286e-07,
      "loss": 0.44,
      "num_input_tokens_seen": 24561408,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.11656949669122696,
      "learning_rate": 6.008700791502175e-07,
      "loss": 0.4321,
      "num_input_tokens_seen": 24564352,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.28370389342308044,
      "learning_rate": 5.8971429701421e-07,
      "loss": 0.4027,
      "num_input_tokens_seen": 24567008,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.17205551266670227,
      "learning_rate": 5.786629858123283e-07,
      "loss": 0.3835,
      "num_input_tokens_seen": 24569824,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 0.15395085513591766,
      "learning_rate": 5.677161478565008e-07,
      "loss": 0.4211,
      "num_input_tokens_seen": 24572960,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.11937592178583145,
      "learning_rate": 5.56873785436618e-07,
      "loss": 0.473,
      "num_input_tokens_seen": 24576864,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.2179117053747177,
      "learning_rate": 5.4613590082081e-07,
      "loss": 0.4471,
      "num_input_tokens_seen": 24579520,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.356344610452652,
      "learning_rate": 5.355024962552801e-07,
      "loss": 0.3851,
      "num_input_tokens_seen": 24583136,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.13342596590518951,
      "learning_rate": 5.249735739644157e-07,
      "loss": 0.2345,
      "num_input_tokens_seen": 24586528,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.17094610631465912,
      "learning_rate": 5.145491361508436e-07,
      "loss": 0.4754,
      "num_input_tokens_seen": 24589696,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 0.27702540159225464,
      "learning_rate": 5.042291849950975e-07,
      "loss": 0.5156,
      "num_input_tokens_seen": 24594208,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.14024178683757782,
      "learning_rate": 4.940137226560615e-07,
      "loss": 0.4156,
      "num_input_tokens_seen": 24597312,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.2813471555709839,
      "learning_rate": 4.839027512706928e-07,
      "loss": 0.3407,
      "num_input_tokens_seen": 24600096,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.24946722388267517,
      "learning_rate": 4.7389627295407743e-07,
      "loss": 0.5401,
      "num_input_tokens_seen": 24603328,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.25111570954322815,
      "learning_rate": 4.6399428979948534e-07,
      "loss": 0.3149,
      "num_input_tokens_seen": 24606944,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.3386770784854889,
      "learning_rate": 4.541968038782596e-07,
      "loss": 0.4805,
      "num_input_tokens_seen": 24610304,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.13860714435577393,
      "learning_rate": 4.445038172399829e-07,
      "loss": 0.3042,
      "num_input_tokens_seen": 24612992,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.24716387689113617,
      "learning_rate": 4.3491533191225563e-07,
      "loss": 0.543,
      "num_input_tokens_seen": 24616448,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.25174522399902344,
      "learning_rate": 4.254313499009177e-07,
      "loss": 0.4542,
      "num_input_tokens_seen": 24618816,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 0.16383862495422363,
      "learning_rate": 4.1605187318982664e-07,
      "loss": 0.4084,
      "num_input_tokens_seen": 24621632,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.17475636303424835,
      "learning_rate": 4.067769037411906e-07,
      "loss": 0.4402,
      "num_input_tokens_seen": 24625280,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.2595260739326477,
      "learning_rate": 3.9760644349517984e-07,
      "loss": 0.3894,
      "num_input_tokens_seen": 24627936,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.3650907278060913,
      "learning_rate": 3.885404943700932e-07,
      "loss": 0.4336,
      "num_input_tokens_seen": 24631776,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.2505047619342804,
      "learning_rate": 3.795790582624692e-07,
      "loss": 0.4126,
      "num_input_tokens_seen": 24634272,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.2891180217266083,
      "learning_rate": 3.707221370469749e-07,
      "loss": 0.4443,
      "num_input_tokens_seen": 24637056,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.11598588526248932,
      "learning_rate": 3.6196973257629494e-07,
      "loss": 0.371,
      "num_input_tokens_seen": 24640320,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.21159310638904572,
      "learning_rate": 3.533218466813537e-07,
      "loss": 0.3698,
      "num_input_tokens_seen": 24643456,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.24044454097747803,
      "learning_rate": 3.447784811712595e-07,
      "loss": 0.5827,
      "num_input_tokens_seen": 24646144,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.2558457851409912,
      "learning_rate": 3.363396378331385e-07,
      "loss": 0.4773,
      "num_input_tokens_seen": 24648736,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 0.14332616329193115,
      "learning_rate": 3.280053184323006e-07,
      "loss": 0.3514,
      "num_input_tokens_seen": 24651456,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.16168320178985596,
      "learning_rate": 3.1977552471218476e-07,
      "loss": 0.4711,
      "num_input_tokens_seen": 24654560,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.3368915319442749,
      "learning_rate": 3.116502583943581e-07,
      "loss": 0.4066,
      "num_input_tokens_seen": 24658144,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.2197243869304657,
      "learning_rate": 3.036295211785722e-07,
      "loss": 0.4813,
      "num_input_tokens_seen": 24664576,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.17096352577209473,
      "learning_rate": 2.957133147425961e-07,
      "loss": 0.4699,
      "num_input_tokens_seen": 24668096,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.25579649209976196,
      "learning_rate": 2.879016407425494e-07,
      "loss": 0.2858,
      "num_input_tokens_seen": 24670592,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.16511240601539612,
      "learning_rate": 2.8019450081240295e-07,
      "loss": 0.501,
      "num_input_tokens_seen": 24674208,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.13115844130516052,
      "learning_rate": 2.7259189656447803e-07,
      "loss": 0.389,
      "num_input_tokens_seen": 24676672,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.16442370414733887,
      "learning_rate": 2.650938295891692e-07,
      "loss": 0.4555,
      "num_input_tokens_seen": 24679264,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.24077151715755463,
      "learning_rate": 2.5770030145494395e-07,
      "loss": 0.3648,
      "num_input_tokens_seen": 24682080,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.1172136440873146,
      "learning_rate": 2.504113137083985e-07,
      "loss": 0.4169,
      "num_input_tokens_seen": 24685888,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.1866813749074936,
      "learning_rate": 2.4322686787442425e-07,
      "loss": 0.3683,
      "num_input_tokens_seen": 24688800,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.20567236840724945,
      "learning_rate": 2.3614696545581904e-07,
      "loss": 0.4222,
      "num_input_tokens_seen": 24692416,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.17565147578716278,
      "learning_rate": 2.2917160793367585e-07,
      "loss": 0.4183,
      "num_input_tokens_seen": 24695296,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.1452907919883728,
      "learning_rate": 2.2230079676716086e-07,
      "loss": 0.3854,
      "num_input_tokens_seen": 24698016,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.11418203264474869,
      "learning_rate": 2.1553453339356875e-07,
      "loss": 0.4183,
      "num_input_tokens_seen": 24701696,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.1847374141216278,
      "learning_rate": 2.0887281922826738e-07,
      "loss": 0.425,
      "num_input_tokens_seen": 24705760,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.15746036171913147,
      "learning_rate": 2.023156556648642e-07,
      "loss": 0.4983,
      "num_input_tokens_seen": 24709504,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.13505154848098755,
      "learning_rate": 1.9586304407503975e-07,
      "loss": 0.5899,
      "num_input_tokens_seen": 24712640,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.2584210932254791,
      "learning_rate": 1.8951498580860315e-07,
      "loss": 0.6207,
      "num_input_tokens_seen": 24715808,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.3641274869441986,
      "learning_rate": 1.832714821934922e-07,
      "loss": 0.3819,
      "num_input_tokens_seen": 24718848,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.22611215710639954,
      "learning_rate": 1.7713253453577328e-07,
      "loss": 0.3563,
      "num_input_tokens_seen": 24722016,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.1432391107082367,
      "learning_rate": 1.7109814411964132e-07,
      "loss": 0.4327,
      "num_input_tokens_seen": 24725920,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 0.13512781262397766,
      "learning_rate": 1.651683122074754e-07,
      "loss": 0.4288,
      "num_input_tokens_seen": 24730112,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 0.31681668758392334,
      "learning_rate": 1.5934304003961675e-07,
      "loss": 0.4968,
      "num_input_tokens_seen": 24733248,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.13433755934238434,
      "learning_rate": 1.5362232883475713e-07,
      "loss": 0.4249,
      "num_input_tokens_seen": 24736576,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.16536754369735718,
      "learning_rate": 1.4800617978949492e-07,
      "loss": 0.404,
      "num_input_tokens_seen": 24740544,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.1717861294746399,
      "learning_rate": 1.424945940787792e-07,
      "loss": 0.382,
      "num_input_tokens_seen": 24744128,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.09889072179794312,
      "learning_rate": 1.3708757285552098e-07,
      "loss": 0.4799,
      "num_input_tokens_seen": 24747104,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.2700487971305847,
      "learning_rate": 1.3178511725076004e-07,
      "loss": 0.4759,
      "num_input_tokens_seen": 24750432,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.18476885557174683,
      "learning_rate": 1.265872283738312e-07,
      "loss": 0.309,
      "num_input_tokens_seen": 24752896,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 0.229990616440773,
      "learning_rate": 1.2149390731192033e-07,
      "loss": 0.5311,
      "num_input_tokens_seen": 24756096,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.1555647850036621,
      "learning_rate": 1.1650515513061955e-07,
      "loss": 0.4147,
      "num_input_tokens_seen": 24759776,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.15977948904037476,
      "learning_rate": 1.1162097287342743e-07,
      "loss": 0.379,
      "num_input_tokens_seen": 24762720,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 0.1055770143866539,
      "learning_rate": 1.0684136156213775e-07,
      "loss": 0.3369,
      "num_input_tokens_seen": 24766528,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.12459831684827805,
      "learning_rate": 1.0216632219650634e-07,
      "loss": 0.4381,
      "num_input_tokens_seen": 24770496,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.178396075963974,
      "learning_rate": 9.759585575458418e-08,
      "loss": 0.5184,
      "num_input_tokens_seen": 24773024,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.16018778085708618,
      "learning_rate": 9.312996319238432e-08,
      "loss": 0.3776,
      "num_input_tokens_seen": 24776064,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.16925157606601715,
      "learning_rate": 8.876864544421493e-08,
      "loss": 0.4356,
      "num_input_tokens_seen": 24779072,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.14546671509742737,
      "learning_rate": 8.451190342229077e-08,
      "loss": 0.4974,
      "num_input_tokens_seen": 24782560,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.13588188588619232,
      "learning_rate": 8.035973801717722e-08,
      "loss": 0.4209,
      "num_input_tokens_seen": 24785312,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.2963770627975464,
      "learning_rate": 7.631215009740178e-08,
      "loss": 0.4835,
      "num_input_tokens_seen": 24788128,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.1795395314693451,
      "learning_rate": 7.236914050973153e-08,
      "loss": 0.5781,
      "num_input_tokens_seen": 24790496,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 0.15691818296909332,
      "learning_rate": 6.853071007895117e-08,
      "loss": 0.3987,
      "num_input_tokens_seen": 24793792,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 0.19001618027687073,
      "learning_rate": 6.479685960797399e-08,
      "loss": 0.3117,
      "num_input_tokens_seen": 24796992,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.1911657601594925,
      "learning_rate": 6.116758987800841e-08,
      "loss": 0.4335,
      "num_input_tokens_seen": 24800864,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.14153854548931122,
      "learning_rate": 5.7642901648113924e-08,
      "loss": 0.3512,
      "num_input_tokens_seen": 24804160,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.13877765834331512,
      "learning_rate": 5.422279565570065e-08,
      "loss": 0.4339,
      "num_input_tokens_seen": 24807072,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.09994948655366898,
      "learning_rate": 5.090727261619632e-08,
      "loss": 0.3335,
      "num_input_tokens_seen": 24811584,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.2761865556240082,
      "learning_rate": 4.7696333223212765e-08,
      "loss": 0.2925,
      "num_input_tokens_seen": 24814464,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.1316370666027069,
      "learning_rate": 4.4589978148323884e-08,
      "loss": 0.3734,
      "num_input_tokens_seen": 24817920,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.13063929975032806,
      "learning_rate": 4.1588208041454244e-08,
      "loss": 0.3705,
      "num_input_tokens_seen": 24821312,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 0.19802699983119965,
      "learning_rate": 3.8691023530545985e-08,
      "loss": 0.4274,
      "num_input_tokens_seen": 24824672,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.16922105848789215,
      "learning_rate": 3.589842522155884e-08,
      "loss": 0.478,
      "num_input_tokens_seen": 24827584,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.13217662274837494,
      "learning_rate": 3.321041369874767e-08,
      "loss": 0.5188,
      "num_input_tokens_seen": 24831680,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.7297108173370361,
      "learning_rate": 3.0626989524384916e-08,
      "loss": 0.3802,
      "num_input_tokens_seen": 24835168,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.1826811134815216,
      "learning_rate": 2.8148153238927167e-08,
      "loss": 0.4624,
      "num_input_tokens_seen": 24838432,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.1092541292309761,
      "learning_rate": 2.5773905360904072e-08,
      "loss": 0.3854,
      "num_input_tokens_seen": 24844096,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.10257282108068466,
      "learning_rate": 2.3504246386918392e-08,
      "loss": 0.4422,
      "num_input_tokens_seen": 24847872,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.3357459008693695,
      "learning_rate": 2.133917679186803e-08,
      "loss": 0.423,
      "num_input_tokens_seen": 24851520,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.1614718735218048,
      "learning_rate": 1.9278697028557447e-08,
      "loss": 0.4226,
      "num_input_tokens_seen": 24854656,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.16263256967067719,
      "learning_rate": 1.7322807528086238e-08,
      "loss": 0.4599,
      "num_input_tokens_seen": 24857952,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.24631844460964203,
      "learning_rate": 1.547150869957159e-08,
      "loss": 0.5442,
      "num_input_tokens_seen": 24860704,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.14438633620738983,
      "learning_rate": 1.3724800930314806e-08,
      "loss": 0.4286,
      "num_input_tokens_seen": 24863712,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.2679991126060486,
      "learning_rate": 1.2082684585634773e-08,
      "loss": 0.3009,
      "num_input_tokens_seen": 24866752,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.17259760200977325,
      "learning_rate": 1.0545160009145516e-08,
      "loss": 0.3789,
      "num_input_tokens_seen": 24871360,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.1875828057527542,
      "learning_rate": 9.112227522423133e-09,
      "loss": 0.5999,
      "num_input_tokens_seen": 24875488,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.18374858796596527,
      "learning_rate": 7.783887425172331e-09,
      "loss": 0.4515,
      "num_input_tokens_seen": 24878528,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.2457098364830017,
      "learning_rate": 6.560139995392955e-09,
      "loss": 0.365,
      "num_input_tokens_seen": 24881792,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.22209003567695618,
      "learning_rate": 5.440985488935901e-09,
      "loss": 0.3327,
      "num_input_tokens_seen": 24884928,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.21122117340564728,
      "learning_rate": 4.426424140058227e-09,
      "loss": 0.33,
      "num_input_tokens_seen": 24887488,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 0.2121138721704483,
      "learning_rate": 3.5164561608680424e-09,
      "loss": 0.4976,
      "num_input_tokens_seen": 24890368,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 0.2112082839012146,
      "learning_rate": 2.7110817417685953e-09,
      "loss": 0.4701,
      "num_input_tokens_seen": 24892416,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.1217755377292633,
      "learning_rate": 2.010301051291741e-09,
      "loss": 0.3155,
      "num_input_tokens_seen": 24895552,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.3486655652523041,
      "learning_rate": 1.4141142359314074e-09,
      "loss": 0.4155,
      "num_input_tokens_seen": 24899296,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.11582214385271072,
      "learning_rate": 9.22521420476663e-10,
      "loss": 0.386,
      "num_input_tokens_seen": 24902432,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 0.16183632612228394,
      "learning_rate": 5.355227077341596e-10,
      "loss": 0.4007,
      "num_input_tokens_seen": 24905696,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.10266292095184326,
      "learning_rate": 2.5311817863915566e-10,
      "loss": 0.3734,
      "num_input_tokens_seen": 24908672,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.23758825659751892,
      "learning_rate": 7.53078923110273e-11,
      "loss": 0.496,
      "num_input_tokens_seen": 24911552,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.3404746353626251,
      "learning_rate": 2.0918859422458526e-12,
      "loss": 0.3567,
      "num_input_tokens_seen": 24913312,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.49418532848358154,
      "eval_runtime": 13.7364,
      "eval_samples_per_second": 61.734,
      "eval_steps_per_second": 15.433,
      "num_input_tokens_seen": 24913312,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24913312,
      "step": 38160,
      "total_flos": 1.121835179340202e+18,
      "train_loss": 0.459870044740241,
      "train_runtime": 5880.3455,
      "train_samples_per_second": 25.947,
      "train_steps_per_second": 6.489
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24913312,
  "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": 1.121835179340202e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}