{
  "best_global_step": 38800,
  "best_metric": 0.49414047598838806,
  "best_model_checkpoint": "saves/prompt-tuning/gemma-3-1b-it/train_record_1745950247/checkpoint-38800",
  "epoch": 1.2803277639075603,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016004097048844505,
      "grad_norm": 3.0578761100769043,
      "learning_rate": 0.29999999259779675,
      "loss": 12.1064,
      "num_input_tokens_seen": 7056,
      "step": 5
    },
    {
      "epoch": 0.0003200819409768901,
      "grad_norm": 2.0636138916015625,
      "learning_rate": 0.29999996252634736,
      "loss": 8.9711,
      "num_input_tokens_seen": 13952,
      "step": 10
    },
    {
      "epoch": 0.0004801229114653351,
      "grad_norm": 1.8197300434112549,
      "learning_rate": 0.2999999093230187,
      "loss": 7.1025,
      "num_input_tokens_seen": 20960,
      "step": 15
    },
    {
      "epoch": 0.0006401638819537802,
      "grad_norm": 1.62192702293396,
      "learning_rate": 0.299999832987819,
      "loss": 5.7283,
      "num_input_tokens_seen": 27968,
      "step": 20
    },
    {
      "epoch": 0.0008002048524422252,
      "grad_norm": 1.4166529178619385,
      "learning_rate": 0.29999973352076004,
      "loss": 6.0356,
      "num_input_tokens_seen": 34800,
      "step": 25
    },
    {
      "epoch": 0.0009602458229306702,
      "grad_norm": 1.4591604471206665,
      "learning_rate": 0.2999996109218572,
      "loss": 5.3845,
      "num_input_tokens_seen": 42112,
      "step": 30
    },
    {
      "epoch": 0.0011202867934191152,
      "grad_norm": 1.574499249458313,
      "learning_rate": 0.2999994651911293,
      "loss": 4.5428,
      "num_input_tokens_seen": 49136,
      "step": 35
    },
    {
      "epoch": 0.0012803277639075604,
      "grad_norm": 0.999565839767456,
      "learning_rate": 0.2999992963285989,
      "loss": 3.1188,
      "num_input_tokens_seen": 56064,
      "step": 40
    },
    {
      "epoch": 0.0014403687343960054,
      "grad_norm": 0.8974927067756653,
      "learning_rate": 0.29999910433429194,
      "loss": 3.5795,
      "num_input_tokens_seen": 63056,
      "step": 45
    },
    {
      "epoch": 0.0016004097048844504,
      "grad_norm": 1.1658265590667725,
      "learning_rate": 0.29999888920823814,
      "loss": 3.27,
      "num_input_tokens_seen": 69600,
      "step": 50
    },
    {
      "epoch": 0.0017604506753728954,
      "grad_norm": 0.9101086258888245,
      "learning_rate": 0.29999865095047057,
      "loss": 2.7005,
      "num_input_tokens_seen": 76496,
      "step": 55
    },
    {
      "epoch": 0.0019204916458613404,
      "grad_norm": 0.7330014705657959,
      "learning_rate": 0.29999838956102604,
      "loss": 3.1506,
      "num_input_tokens_seen": 83424,
      "step": 60
    },
    {
      "epoch": 0.0020805326163497854,
      "grad_norm": 0.7892524003982544,
      "learning_rate": 0.29999810503994484,
      "loss": 3.1937,
      "num_input_tokens_seen": 90352,
      "step": 65
    },
    {
      "epoch": 0.0022405735868382304,
      "grad_norm": 0.6404764652252197,
      "learning_rate": 0.29999779738727084,
      "loss": 2.186,
      "num_input_tokens_seen": 97664,
      "step": 70
    },
    {
      "epoch": 0.002400614557326676,
      "grad_norm": 0.5790537595748901,
      "learning_rate": 0.29999746660305154,
      "loss": 2.1356,
      "num_input_tokens_seen": 104352,
      "step": 75
    },
    {
      "epoch": 0.002560655527815121,
      "grad_norm": 0.771055281162262,
      "learning_rate": 0.2999971126873379,
      "loss": 2.2577,
      "num_input_tokens_seen": 111152,
      "step": 80
    },
    {
      "epoch": 0.002720696498303566,
      "grad_norm": 0.6102445721626282,
      "learning_rate": 0.2999967356401845,
      "loss": 2.7035,
      "num_input_tokens_seen": 118544,
      "step": 85
    },
    {
      "epoch": 0.002880737468792011,
      "grad_norm": 0.4069792628288269,
      "learning_rate": 0.29999633546164944,
      "loss": 1.7476,
      "num_input_tokens_seen": 125568,
      "step": 90
    },
    {
      "epoch": 0.003040778439280456,
      "grad_norm": 0.4884932339191437,
      "learning_rate": 0.29999591215179444,
      "loss": 1.9646,
      "num_input_tokens_seen": 132656,
      "step": 95
    },
    {
      "epoch": 0.003200819409768901,
      "grad_norm": 1.5889664888381958,
      "learning_rate": 0.2999954657106849,
      "loss": 2.6986,
      "num_input_tokens_seen": 139680,
      "step": 100
    },
    {
      "epoch": 0.003360860380257346,
      "grad_norm": 1.0020573139190674,
      "learning_rate": 0.2999949961383896,
      "loss": 2.219,
      "num_input_tokens_seen": 146832,
      "step": 105
    },
    {
      "epoch": 0.003520901350745791,
      "grad_norm": 0.6096405386924744,
      "learning_rate": 0.2999945034349809,
      "loss": 3.1361,
      "num_input_tokens_seen": 153760,
      "step": 110
    },
    {
      "epoch": 0.003680942321234236,
      "grad_norm": 0.536536455154419,
      "learning_rate": 0.2999939876005348,
      "loss": 3.1904,
      "num_input_tokens_seen": 160864,
      "step": 115
    },
    {
      "epoch": 0.003840983291722681,
      "grad_norm": 0.38413259387016296,
      "learning_rate": 0.29999344863513094,
      "loss": 2.6928,
      "num_input_tokens_seen": 167616,
      "step": 120
    },
    {
      "epoch": 0.004001024262211126,
      "grad_norm": 0.3247986137866974,
      "learning_rate": 0.2999928865388523,
      "loss": 2.2485,
      "num_input_tokens_seen": 174336,
      "step": 125
    },
    {
      "epoch": 0.004161065232699571,
      "grad_norm": 0.2560162842273712,
      "learning_rate": 0.29999230131178567,
      "loss": 2.1574,
      "num_input_tokens_seen": 181216,
      "step": 130
    },
    {
      "epoch": 0.004321106203188016,
      "grad_norm": 0.2367495447397232,
      "learning_rate": 0.2999916929540212,
      "loss": 2.5981,
      "num_input_tokens_seen": 188080,
      "step": 135
    },
    {
      "epoch": 0.004481147173676461,
      "grad_norm": 0.3011842966079712,
      "learning_rate": 0.29999106146565285,
      "loss": 1.5545,
      "num_input_tokens_seen": 194640,
      "step": 140
    },
    {
      "epoch": 0.004641188144164906,
      "grad_norm": 0.5006332993507385,
      "learning_rate": 0.29999040684677786,
      "loss": 2.2324,
      "num_input_tokens_seen": 201024,
      "step": 145
    },
    {
      "epoch": 0.004801229114653352,
      "grad_norm": 0.20275336503982544,
      "learning_rate": 0.2999897290974972,
      "loss": 2.0483,
      "num_input_tokens_seen": 207904,
      "step": 150
    },
    {
      "epoch": 0.004961270085141797,
      "grad_norm": 0.28025177121162415,
      "learning_rate": 0.2999890282179155,
      "loss": 2.5456,
      "num_input_tokens_seen": 214400,
      "step": 155
    },
    {
      "epoch": 0.005121311055630242,
      "grad_norm": 0.2749873399734497,
      "learning_rate": 0.29998830420814077,
      "loss": 2.6351,
      "num_input_tokens_seen": 220912,
      "step": 160
    },
    {
      "epoch": 0.005281352026118687,
      "grad_norm": 0.3283006548881531,
      "learning_rate": 0.2999875570682846,
      "loss": 1.7641,
      "num_input_tokens_seen": 227968,
      "step": 165
    },
    {
      "epoch": 0.005441392996607132,
      "grad_norm": 0.16868865489959717,
      "learning_rate": 0.2999867867984623,
      "loss": 1.4777,
      "num_input_tokens_seen": 234864,
      "step": 170
    },
    {
      "epoch": 0.005601433967095577,
      "grad_norm": 0.21636277437210083,
      "learning_rate": 0.29998599339879267,
      "loss": 1.7198,
      "num_input_tokens_seen": 242080,
      "step": 175
    },
    {
      "epoch": 0.005761474937584022,
      "grad_norm": 0.19554872810840607,
      "learning_rate": 0.29998517686939796,
      "loss": 1.544,
      "num_input_tokens_seen": 249040,
      "step": 180
    },
    {
      "epoch": 0.005921515908072467,
      "grad_norm": 0.19937533140182495,
      "learning_rate": 0.29998433721040413,
      "loss": 2.0178,
      "num_input_tokens_seen": 256080,
      "step": 185
    },
    {
      "epoch": 0.006081556878560912,
      "grad_norm": 0.18051904439926147,
      "learning_rate": 0.29998347442194073,
      "loss": 1.5629,
      "num_input_tokens_seen": 263296,
      "step": 190
    },
    {
      "epoch": 0.006241597849049357,
      "grad_norm": 0.16671404242515564,
      "learning_rate": 0.2999825885041407,
      "loss": 1.798,
      "num_input_tokens_seen": 270320,
      "step": 195
    },
    {
      "epoch": 0.006401638819537802,
      "grad_norm": 0.2789769768714905,
      "learning_rate": 0.29998167945714077,
      "loss": 1.2371,
      "num_input_tokens_seen": 277264,
      "step": 200
    },
    {
      "epoch": 0.006401638819537802,
      "eval_loss": 1.5826903581619263,
      "eval_runtime": 331.3873,
      "eval_samples_per_second": 41.903,
      "eval_steps_per_second": 20.951,
      "num_input_tokens_seen": 277264,
      "step": 200
    },
    {
      "epoch": 0.006561679790026247,
      "grad_norm": 0.21591319143772125,
      "learning_rate": 0.2999807472810811,
      "loss": 1.3141,
      "num_input_tokens_seen": 283904,
      "step": 205
    },
    {
      "epoch": 0.006721720760514692,
      "grad_norm": 0.17060643434524536,
      "learning_rate": 0.29997979197610536,
      "loss": 1.7519,
      "num_input_tokens_seen": 291024,
      "step": 210
    },
    {
      "epoch": 0.006881761731003137,
      "grad_norm": 0.10981401056051254,
      "learning_rate": 0.299978813542361,
      "loss": 1.5566,
      "num_input_tokens_seen": 297840,
      "step": 215
    },
    {
      "epoch": 0.007041802701491582,
      "grad_norm": 0.1833302527666092,
      "learning_rate": 0.2999778119799988,
      "loss": 1.7466,
      "num_input_tokens_seen": 304432,
      "step": 220
    },
    {
      "epoch": 0.007201843671980027,
      "grad_norm": 0.11496254056692123,
      "learning_rate": 0.29997678728917326,
      "loss": 1.3753,
      "num_input_tokens_seen": 311312,
      "step": 225
    },
    {
      "epoch": 0.007361884642468472,
      "grad_norm": 0.20235130190849304,
      "learning_rate": 0.2999757394700424,
      "loss": 1.6966,
      "num_input_tokens_seen": 318112,
      "step": 230
    },
    {
      "epoch": 0.007521925612956917,
      "grad_norm": 0.1977204978466034,
      "learning_rate": 0.29997466852276783,
      "loss": 1.6576,
      "num_input_tokens_seen": 324944,
      "step": 235
    },
    {
      "epoch": 0.007681966583445362,
      "grad_norm": 0.1502850502729416,
      "learning_rate": 0.29997357444751466,
      "loss": 1.4971,
      "num_input_tokens_seen": 331584,
      "step": 240
    },
    {
      "epoch": 0.007842007553933807,
      "grad_norm": 0.1365310102701187,
      "learning_rate": 0.2999724572444516,
      "loss": 1.4237,
      "num_input_tokens_seen": 338688,
      "step": 245
    },
    {
      "epoch": 0.008002048524422252,
      "grad_norm": 0.10405360907316208,
      "learning_rate": 0.29997131691375095,
      "loss": 1.3377,
      "num_input_tokens_seen": 345312,
      "step": 250
    },
    {
      "epoch": 0.008162089494910697,
      "grad_norm": 0.09966043382883072,
      "learning_rate": 0.2999701534555886,
      "loss": 1.2417,
      "num_input_tokens_seen": 352192,
      "step": 255
    },
    {
      "epoch": 0.008322130465399142,
      "grad_norm": 0.22841760516166687,
      "learning_rate": 0.2999689668701439,
      "loss": 1.3816,
      "num_input_tokens_seen": 358784,
      "step": 260
    },
    {
      "epoch": 0.008482171435887587,
      "grad_norm": 0.12493032962083817,
      "learning_rate": 0.29996775715759993,
      "loss": 1.2268,
      "num_input_tokens_seen": 365568,
      "step": 265
    },
    {
      "epoch": 0.008642212406376032,
      "grad_norm": 0.11994169652462006,
      "learning_rate": 0.2999665243181432,
      "loss": 1.202,
      "num_input_tokens_seen": 372416,
      "step": 270
    },
    {
      "epoch": 0.008802253376864477,
      "grad_norm": 0.15383580327033997,
      "learning_rate": 0.2999652683519638,
      "loss": 1.3504,
      "num_input_tokens_seen": 379456,
      "step": 275
    },
    {
      "epoch": 0.008962294347352922,
      "grad_norm": 0.0952424556016922,
      "learning_rate": 0.29996398925925544,
      "loss": 1.6937,
      "num_input_tokens_seen": 386144,
      "step": 280
    },
    {
      "epoch": 0.009122335317841367,
      "grad_norm": 0.1080094575881958,
      "learning_rate": 0.2999626870402154,
      "loss": 1.6647,
      "num_input_tokens_seen": 393008,
      "step": 285
    },
    {
      "epoch": 0.009282376288329812,
      "grad_norm": 0.1467948704957962,
      "learning_rate": 0.29996136169504445,
      "loss": 1.6002,
      "num_input_tokens_seen": 399744,
      "step": 290
    },
    {
      "epoch": 0.009442417258818258,
      "grad_norm": 0.10355555266141891,
      "learning_rate": 0.29996001322394694,
      "loss": 1.5074,
      "num_input_tokens_seen": 406576,
      "step": 295
    },
    {
      "epoch": 0.009602458229306703,
      "grad_norm": 0.10277792066335678,
      "learning_rate": 0.29995864162713093,
      "loss": 1.048,
      "num_input_tokens_seen": 413200,
      "step": 300
    },
    {
      "epoch": 0.009762499199795148,
      "grad_norm": 0.1158032938838005,
      "learning_rate": 0.2999572469048079,
      "loss": 1.4383,
      "num_input_tokens_seen": 420032,
      "step": 305
    },
    {
      "epoch": 0.009922540170283593,
      "grad_norm": 0.22755113244056702,
      "learning_rate": 0.29995582905719287,
      "loss": 1.0968,
      "num_input_tokens_seen": 426896,
      "step": 310
    },
    {
      "epoch": 0.010082581140772038,
      "grad_norm": 0.12465681880712509,
      "learning_rate": 0.2999543880845046,
      "loss": 1.2624,
      "num_input_tokens_seen": 433440,
      "step": 315
    },
    {
      "epoch": 0.010242622111260483,
      "grad_norm": 0.09946932643651962,
      "learning_rate": 0.2999529239869652,
      "loss": 1.164,
      "num_input_tokens_seen": 439936,
      "step": 320
    },
    {
      "epoch": 0.010402663081748928,
      "grad_norm": 0.09509330242872238,
      "learning_rate": 0.2999514367648005,
      "loss": 1.6065,
      "num_input_tokens_seen": 446752,
      "step": 325
    },
    {
      "epoch": 0.010562704052237373,
      "grad_norm": 0.06777254492044449,
      "learning_rate": 0.29994992641823987,
      "loss": 1.4257,
      "num_input_tokens_seen": 453216,
      "step": 330
    },
    {
      "epoch": 0.010722745022725818,
      "grad_norm": 0.051390137523412704,
      "learning_rate": 0.29994839294751613,
      "loss": 1.1843,
      "num_input_tokens_seen": 459632,
      "step": 335
    },
    {
      "epoch": 0.010882785993214263,
      "grad_norm": 0.1065596491098404,
      "learning_rate": 0.29994683635286584,
      "loss": 0.9637,
      "num_input_tokens_seen": 466448,
      "step": 340
    },
    {
      "epoch": 0.011042826963702708,
      "grad_norm": 0.08686139434576035,
      "learning_rate": 0.2999452566345291,
      "loss": 1.1693,
      "num_input_tokens_seen": 473056,
      "step": 345
    },
    {
      "epoch": 0.011202867934191153,
      "grad_norm": 0.11881335079669952,
      "learning_rate": 0.2999436537927494,
      "loss": 1.3577,
      "num_input_tokens_seen": 480080,
      "step": 350
    },
    {
      "epoch": 0.011362908904679598,
      "grad_norm": 0.06330325454473495,
      "learning_rate": 0.299942027827774,
      "loss": 1.3136,
      "num_input_tokens_seen": 487184,
      "step": 355
    },
    {
      "epoch": 0.011522949875168043,
      "grad_norm": 0.13511759042739868,
      "learning_rate": 0.29994037873985363,
      "loss": 1.4699,
      "num_input_tokens_seen": 494208,
      "step": 360
    },
    {
      "epoch": 0.011682990845656488,
      "grad_norm": 0.1339394450187683,
      "learning_rate": 0.29993870652924254,
      "loss": 1.0125,
      "num_input_tokens_seen": 501120,
      "step": 365
    },
    {
      "epoch": 0.011843031816144933,
      "grad_norm": 0.11662374436855316,
      "learning_rate": 0.29993701119619876,
      "loss": 1.2378,
      "num_input_tokens_seen": 507520,
      "step": 370
    },
    {
      "epoch": 0.012003072786633378,
      "grad_norm": 0.14825321733951569,
      "learning_rate": 0.2999352927409835,
      "loss": 1.523,
      "num_input_tokens_seen": 514480,
      "step": 375
    },
    {
      "epoch": 0.012163113757121823,
      "grad_norm": 0.06371418386697769,
      "learning_rate": 0.29993355116386194,
      "loss": 1.4331,
      "num_input_tokens_seen": 521168,
      "step": 380
    },
    {
      "epoch": 0.012323154727610268,
      "grad_norm": 0.1129341572523117,
      "learning_rate": 0.29993178646510266,
      "loss": 1.2876,
      "num_input_tokens_seen": 528064,
      "step": 385
    },
    {
      "epoch": 0.012483195698098713,
      "grad_norm": 0.2215065360069275,
      "learning_rate": 0.2999299986449777,
      "loss": 1.2215,
      "num_input_tokens_seen": 534656,
      "step": 390
    },
    {
      "epoch": 0.012643236668587158,
      "grad_norm": 0.13764750957489014,
      "learning_rate": 0.29992818770376284,
      "loss": 1.4869,
      "num_input_tokens_seen": 542080,
      "step": 395
    },
    {
      "epoch": 0.012803277639075603,
      "grad_norm": 0.0671665221452713,
      "learning_rate": 0.29992635364173725,
      "loss": 1.3444,
      "num_input_tokens_seen": 548976,
      "step": 400
    },
    {
      "epoch": 0.012803277639075603,
      "eval_loss": 1.272114872932434,
      "eval_runtime": 331.9493,
      "eval_samples_per_second": 41.832,
      "eval_steps_per_second": 20.916,
      "num_input_tokens_seen": 548976,
      "step": 400
    },
    {
      "epoch": 0.012963318609564048,
      "grad_norm": 0.0850641205906868,
      "learning_rate": 0.2999244964591839,
      "loss": 1.1269,
      "num_input_tokens_seen": 555728,
      "step": 405
    },
    {
      "epoch": 0.013123359580052493,
      "grad_norm": 0.22971424460411072,
      "learning_rate": 0.2999226161563891,
      "loss": 1.1004,
      "num_input_tokens_seen": 562704,
      "step": 410
    },
    {
      "epoch": 0.013283400550540938,
      "grad_norm": 0.08974096179008484,
      "learning_rate": 0.2999207127336429,
      "loss": 1.4887,
      "num_input_tokens_seen": 569552,
      "step": 415
    },
    {
      "epoch": 0.013443441521029383,
      "grad_norm": 0.12166355550289154,
      "learning_rate": 0.2999187861912387,
      "loss": 1.0504,
      "num_input_tokens_seen": 576336,
      "step": 420
    },
    {
      "epoch": 0.013603482491517828,
      "grad_norm": 0.09148211777210236,
      "learning_rate": 0.2999168365294737,
      "loss": 1.3945,
      "num_input_tokens_seen": 582912,
      "step": 425
    },
    {
      "epoch": 0.013763523462006273,
      "grad_norm": 0.07193322479724884,
      "learning_rate": 0.29991486374864856,
      "loss": 1.1023,
      "num_input_tokens_seen": 589856,
      "step": 430
    },
    {
      "epoch": 0.013923564432494718,
      "grad_norm": 0.06250562518835068,
      "learning_rate": 0.29991286784906745,
      "loss": 1.0762,
      "num_input_tokens_seen": 596704,
      "step": 435
    },
    {
      "epoch": 0.014083605402983163,
      "grad_norm": 0.12213505059480667,
      "learning_rate": 0.2999108488310382,
      "loss": 1.0846,
      "num_input_tokens_seen": 603520,
      "step": 440
    },
    {
      "epoch": 0.014243646373471608,
      "grad_norm": 0.1024893969297409,
      "learning_rate": 0.29990880669487213,
      "loss": 1.1661,
      "num_input_tokens_seen": 610384,
      "step": 445
    },
    {
      "epoch": 0.014403687343960053,
      "grad_norm": 0.08478960394859314,
      "learning_rate": 0.29990674144088425,
      "loss": 1.4844,
      "num_input_tokens_seen": 616912,
      "step": 450
    },
    {
      "epoch": 0.014563728314448498,
      "grad_norm": 0.10282246023416519,
      "learning_rate": 0.299904653069393,
      "loss": 1.3283,
      "num_input_tokens_seen": 623904,
      "step": 455
    },
    {
      "epoch": 0.014723769284936943,
      "grad_norm": 0.09214538335800171,
      "learning_rate": 0.29990254158072044,
      "loss": 1.2538,
      "num_input_tokens_seen": 631232,
      "step": 460
    },
    {
      "epoch": 0.014883810255425388,
      "grad_norm": 0.12646439671516418,
      "learning_rate": 0.2999004069751921,
      "loss": 1.1003,
      "num_input_tokens_seen": 638592,
      "step": 465
    },
    {
      "epoch": 0.015043851225913833,
      "grad_norm": 0.08143655210733414,
      "learning_rate": 0.2998982492531373,
      "loss": 1.378,
      "num_input_tokens_seen": 645792,
      "step": 470
    },
    {
      "epoch": 0.015203892196402278,
      "grad_norm": 0.06771702319383621,
      "learning_rate": 0.2998960684148887,
      "loss": 1.0724,
      "num_input_tokens_seen": 653344,
      "step": 475
    },
    {
      "epoch": 0.015363933166890723,
      "grad_norm": 0.09527411311864853,
      "learning_rate": 0.29989386446078264,
      "loss": 1.5387,
      "num_input_tokens_seen": 659760,
      "step": 480
    },
    {
      "epoch": 0.015523974137379168,
      "grad_norm": 0.09126506000757217,
      "learning_rate": 0.299891637391159,
      "loss": 1.0737,
      "num_input_tokens_seen": 666480,
      "step": 485
    },
    {
      "epoch": 0.015684015107867613,
      "grad_norm": 0.08925271034240723,
      "learning_rate": 0.2998893872063612,
      "loss": 1.5689,
      "num_input_tokens_seen": 673152,
      "step": 490
    },
    {
      "epoch": 0.01584405607835606,
      "grad_norm": 0.131103977560997,
      "learning_rate": 0.2998871139067363,
      "loss": 1.3667,
      "num_input_tokens_seen": 679936,
      "step": 495
    },
    {
      "epoch": 0.016004097048844503,
      "grad_norm": 0.07525502890348434,
      "learning_rate": 0.2998848174926348,
      "loss": 1.1148,
      "num_input_tokens_seen": 686576,
      "step": 500
    },
    {
      "epoch": 0.01616413801933295,
      "grad_norm": 0.0761452466249466,
      "learning_rate": 0.2998824979644109,
      "loss": 1.2237,
      "num_input_tokens_seen": 693392,
      "step": 505
    },
    {
      "epoch": 0.016324178989821393,
      "grad_norm": 0.06278499215841293,
      "learning_rate": 0.29988015532242224,
      "loss": 1.3243,
      "num_input_tokens_seen": 700720,
      "step": 510
    },
    {
      "epoch": 0.01648421996030984,
      "grad_norm": 0.07430978864431381,
      "learning_rate": 0.29987778956703015,
      "loss": 0.8821,
      "num_input_tokens_seen": 708064,
      "step": 515
    },
    {
      "epoch": 0.016644260930798283,
      "grad_norm": 0.14992351830005646,
      "learning_rate": 0.2998754006985994,
      "loss": 1.3585,
      "num_input_tokens_seen": 715616,
      "step": 520
    },
    {
      "epoch": 0.01680430190128673,
      "grad_norm": 0.10392703115940094,
      "learning_rate": 0.29987298871749846,
      "loss": 1.158,
      "num_input_tokens_seen": 722352,
      "step": 525
    },
    {
      "epoch": 0.016964342871775173,
      "grad_norm": 0.07877684384584427,
      "learning_rate": 0.2998705536240992,
      "loss": 1.254,
      "num_input_tokens_seen": 729408,
      "step": 530
    },
    {
      "epoch": 0.01712438384226362,
      "grad_norm": 0.10572336614131927,
      "learning_rate": 0.2998680954187772,
      "loss": 1.0777,
      "num_input_tokens_seen": 736480,
      "step": 535
    },
    {
      "epoch": 0.017284424812752063,
      "grad_norm": 0.08524316549301147,
      "learning_rate": 0.2998656141019115,
      "loss": 1.1476,
      "num_input_tokens_seen": 743280,
      "step": 540
    },
    {
      "epoch": 0.01744446578324051,
      "grad_norm": 0.10895151644945145,
      "learning_rate": 0.2998631096738848,
      "loss": 1.2941,
      "num_input_tokens_seen": 750128,
      "step": 545
    },
    {
      "epoch": 0.017604506753728953,
      "grad_norm": 0.07000859081745148,
      "learning_rate": 0.29986058213508326,
      "loss": 1.0749,
      "num_input_tokens_seen": 757152,
      "step": 550
    },
    {
      "epoch": 0.0177645477242174,
      "grad_norm": 0.07217040657997131,
      "learning_rate": 0.29985803148589674,
      "loss": 1.1063,
      "num_input_tokens_seen": 763648,
      "step": 555
    },
    {
      "epoch": 0.017924588694705843,
      "grad_norm": 0.1985628455877304,
      "learning_rate": 0.2998554577267185,
      "loss": 1.6178,
      "num_input_tokens_seen": 770928,
      "step": 560
    },
    {
      "epoch": 0.01808462966519429,
      "grad_norm": 0.08160335570573807,
      "learning_rate": 0.2998528608579455,
      "loss": 1.4384,
      "num_input_tokens_seen": 778128,
      "step": 565
    },
    {
      "epoch": 0.018244670635682733,
      "grad_norm": 0.06385325640439987,
      "learning_rate": 0.2998502408799781,
      "loss": 0.7822,
      "num_input_tokens_seen": 784848,
      "step": 570
    },
    {
      "epoch": 0.01840471160617118,
      "grad_norm": 0.10723140090703964,
      "learning_rate": 0.2998475977932205,
      "loss": 1.0497,
      "num_input_tokens_seen": 791744,
      "step": 575
    },
    {
      "epoch": 0.018564752576659623,
      "grad_norm": 0.081060029566288,
      "learning_rate": 0.29984493159808023,
      "loss": 1.1167,
      "num_input_tokens_seen": 798640,
      "step": 580
    },
    {
      "epoch": 0.01872479354714807,
      "grad_norm": 0.09432626515626907,
      "learning_rate": 0.29984224229496836,
      "loss": 1.1808,
      "num_input_tokens_seen": 805840,
      "step": 585
    },
    {
      "epoch": 0.018884834517636517,
      "grad_norm": 0.05939660221338272,
      "learning_rate": 0.2998395298842998,
      "loss": 1.0824,
      "num_input_tokens_seen": 812560,
      "step": 590
    },
    {
      "epoch": 0.01904487548812496,
      "grad_norm": 0.077802874147892,
      "learning_rate": 0.29983679436649263,
      "loss": 1.2309,
      "num_input_tokens_seen": 819616,
      "step": 595
    },
    {
      "epoch": 0.019204916458613407,
      "grad_norm": 0.07756383717060089,
      "learning_rate": 0.2998340357419689,
      "loss": 1.5009,
      "num_input_tokens_seen": 826016,
      "step": 600
    },
    {
      "epoch": 0.019204916458613407,
      "eval_loss": 1.2181637287139893,
      "eval_runtime": 331.6775,
      "eval_samples_per_second": 41.866,
      "eval_steps_per_second": 20.933,
      "num_input_tokens_seen": 826016,
      "step": 600
    },
    {
      "epoch": 0.01936495742910185,
      "grad_norm": 0.10048788040876389,
      "learning_rate": 0.29983125401115385,
      "loss": 0.9545,
      "num_input_tokens_seen": 832736,
      "step": 605
    },
    {
      "epoch": 0.019524998399590297,
      "grad_norm": 0.07206173986196518,
      "learning_rate": 0.29982844917447654,
      "loss": 1.2676,
      "num_input_tokens_seen": 840304,
      "step": 610
    },
    {
      "epoch": 0.01968503937007874,
      "grad_norm": 0.15476635098457336,
      "learning_rate": 0.2998256212323695,
      "loss": 0.9404,
      "num_input_tokens_seen": 846784,
      "step": 615
    },
    {
      "epoch": 0.019845080340567187,
      "grad_norm": 0.14916111528873444,
      "learning_rate": 0.29982277018526887,
      "loss": 1.3091,
      "num_input_tokens_seen": 853536,
      "step": 620
    },
    {
      "epoch": 0.02000512131105563,
      "grad_norm": 0.07826250791549683,
      "learning_rate": 0.2998198960336143,
      "loss": 1.0542,
      "num_input_tokens_seen": 860304,
      "step": 625
    },
    {
      "epoch": 0.020165162281544077,
      "grad_norm": 0.0930299311876297,
      "learning_rate": 0.299816998777849,
      "loss": 1.1148,
      "num_input_tokens_seen": 867088,
      "step": 630
    },
    {
      "epoch": 0.02032520325203252,
      "grad_norm": 0.09595726430416107,
      "learning_rate": 0.2998140784184197,
      "loss": 0.8027,
      "num_input_tokens_seen": 873600,
      "step": 635
    },
    {
      "epoch": 0.020485244222520967,
      "grad_norm": 0.14486631751060486,
      "learning_rate": 0.2998111349557769,
      "loss": 1.3052,
      "num_input_tokens_seen": 880720,
      "step": 640
    },
    {
      "epoch": 0.02064528519300941,
      "grad_norm": 0.23855800926685333,
      "learning_rate": 0.29980816839037444,
      "loss": 1.3085,
      "num_input_tokens_seen": 887760,
      "step": 645
    },
    {
      "epoch": 0.020805326163497857,
      "grad_norm": 0.06722548604011536,
      "learning_rate": 0.2998051787226698,
      "loss": 0.9656,
      "num_input_tokens_seen": 894800,
      "step": 650
    },
    {
      "epoch": 0.0209653671339863,
      "grad_norm": 0.07688195258378983,
      "learning_rate": 0.29980216595312403,
      "loss": 1.1687,
      "num_input_tokens_seen": 901856,
      "step": 655
    },
    {
      "epoch": 0.021125408104474747,
      "grad_norm": 0.06659464538097382,
      "learning_rate": 0.29979913008220177,
      "loss": 1.148,
      "num_input_tokens_seen": 908240,
      "step": 660
    },
    {
      "epoch": 0.02128544907496319,
      "grad_norm": 0.05235271528363228,
      "learning_rate": 0.2997960711103711,
      "loss": 1.1978,
      "num_input_tokens_seen": 915312,
      "step": 665
    },
    {
      "epoch": 0.021445490045451637,
      "grad_norm": 0.11381974816322327,
      "learning_rate": 0.29979298903810386,
      "loss": 1.7557,
      "num_input_tokens_seen": 922336,
      "step": 670
    },
    {
      "epoch": 0.02160553101594008,
      "grad_norm": 0.08986150473356247,
      "learning_rate": 0.29978988386587524,
      "loss": 1.2403,
      "num_input_tokens_seen": 929120,
      "step": 675
    },
    {
      "epoch": 0.021765571986428527,
      "grad_norm": 0.10677788406610489,
      "learning_rate": 0.2997867555941642,
      "loss": 1.3024,
      "num_input_tokens_seen": 935952,
      "step": 680
    },
    {
      "epoch": 0.02192561295691697,
      "grad_norm": 0.10724510252475739,
      "learning_rate": 0.299783604223453,
      "loss": 0.9819,
      "num_input_tokens_seen": 942608,
      "step": 685
    },
    {
      "epoch": 0.022085653927405417,
      "grad_norm": 0.12156035006046295,
      "learning_rate": 0.29978042975422786,
      "loss": 1.4512,
      "num_input_tokens_seen": 950080,
      "step": 690
    },
    {
      "epoch": 0.02224569489789386,
      "grad_norm": 0.24857783317565918,
      "learning_rate": 0.29977723218697816,
      "loss": 1.6885,
      "num_input_tokens_seen": 956912,
      "step": 695
    },
    {
      "epoch": 0.022405735868382307,
      "grad_norm": 0.09824664890766144,
      "learning_rate": 0.299774011522197,
      "loss": 1.3298,
      "num_input_tokens_seen": 963296,
      "step": 700
    },
    {
      "epoch": 0.02256577683887075,
      "grad_norm": 0.1406419575214386,
      "learning_rate": 0.29977076776038114,
      "loss": 1.3731,
      "num_input_tokens_seen": 969744,
      "step": 705
    },
    {
      "epoch": 0.022725817809359197,
      "grad_norm": 0.10532058030366898,
      "learning_rate": 0.2997675009020307,
      "loss": 1.2839,
      "num_input_tokens_seen": 976512,
      "step": 710
    },
    {
      "epoch": 0.02288585877984764,
      "grad_norm": 0.1382746547460556,
      "learning_rate": 0.2997642109476496,
      "loss": 1.1751,
      "num_input_tokens_seen": 983792,
      "step": 715
    },
    {
      "epoch": 0.023045899750336087,
      "grad_norm": 0.16744109988212585,
      "learning_rate": 0.299760897897745,
      "loss": 1.3156,
      "num_input_tokens_seen": 990784,
      "step": 720
    },
    {
      "epoch": 0.02320594072082453,
      "grad_norm": 0.06063464283943176,
      "learning_rate": 0.29975756175282803,
      "loss": 0.9766,
      "num_input_tokens_seen": 997936,
      "step": 725
    },
    {
      "epoch": 0.023365981691312977,
      "grad_norm": 0.1420663744211197,
      "learning_rate": 0.29975420251341306,
      "loss": 1.5646,
      "num_input_tokens_seen": 1004832,
      "step": 730
    },
    {
      "epoch": 0.02352602266180142,
      "grad_norm": 0.06001969426870346,
      "learning_rate": 0.29975082018001814,
      "loss": 1.1448,
      "num_input_tokens_seen": 1011744,
      "step": 735
    },
    {
      "epoch": 0.023686063632289867,
      "grad_norm": 0.13449528813362122,
      "learning_rate": 0.2997474147531648,
      "loss": 1.1679,
      "num_input_tokens_seen": 1018208,
      "step": 740
    },
    {
      "epoch": 0.02384610460277831,
      "grad_norm": 0.059380821883678436,
      "learning_rate": 0.29974398623337833,
      "loss": 1.3072,
      "num_input_tokens_seen": 1024960,
      "step": 745
    },
    {
      "epoch": 0.024006145573266757,
      "grad_norm": 0.07175704836845398,
      "learning_rate": 0.2997405346211873,
      "loss": 1.4348,
      "num_input_tokens_seen": 1031856,
      "step": 750
    },
    {
      "epoch": 0.0241661865437552,
      "grad_norm": 0.0536097027361393,
      "learning_rate": 0.2997370599171241,
      "loss": 0.9496,
      "num_input_tokens_seen": 1038656,
      "step": 755
    },
    {
      "epoch": 0.024326227514243647,
      "grad_norm": 0.08360739797353745,
      "learning_rate": 0.2997335621217246,
      "loss": 1.1815,
      "num_input_tokens_seen": 1045648,
      "step": 760
    },
    {
      "epoch": 0.02448626848473209,
      "grad_norm": 0.10112405568361282,
      "learning_rate": 0.29973004123552816,
      "loss": 1.3288,
      "num_input_tokens_seen": 1052720,
      "step": 765
    },
    {
      "epoch": 0.024646309455220537,
      "grad_norm": 0.05530364438891411,
      "learning_rate": 0.2997264972590777,
      "loss": 1.1739,
      "num_input_tokens_seen": 1059360,
      "step": 770
    },
    {
      "epoch": 0.02480635042570898,
      "grad_norm": 0.07933059334754944,
      "learning_rate": 0.29972293019291973,
      "loss": 1.1737,
      "num_input_tokens_seen": 1065696,
      "step": 775
    },
    {
      "epoch": 0.024966391396197427,
      "grad_norm": 0.06677951663732529,
      "learning_rate": 0.2997193400376045,
      "loss": 1.1937,
      "num_input_tokens_seen": 1072560,
      "step": 780
    },
    {
      "epoch": 0.025126432366685873,
      "grad_norm": 0.04446530342102051,
      "learning_rate": 0.2997157267936854,
      "loss": 1.1357,
      "num_input_tokens_seen": 1079232,
      "step": 785
    },
    {
      "epoch": 0.025286473337174317,
      "grad_norm": 0.11063764989376068,
      "learning_rate": 0.2997120904617199,
      "loss": 1.0302,
      "num_input_tokens_seen": 1086000,
      "step": 790
    },
    {
      "epoch": 0.025446514307662763,
      "grad_norm": 0.10790906101465225,
      "learning_rate": 0.29970843104226863,
      "loss": 1.3662,
      "num_input_tokens_seen": 1092816,
      "step": 795
    },
    {
      "epoch": 0.025606555278151207,
      "grad_norm": 0.05237681046128273,
      "learning_rate": 0.2997047485358959,
      "loss": 0.8678,
      "num_input_tokens_seen": 1099968,
      "step": 800
    },
    {
      "epoch": 0.025606555278151207,
      "eval_loss": 1.183394193649292,
      "eval_runtime": 332.2165,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 1099968,
      "step": 800
    },
    {
      "epoch": 0.025766596248639653,
      "grad_norm": 0.07519704848527908,
      "learning_rate": 0.2997010429431697,
      "loss": 1.0188,
      "num_input_tokens_seen": 1107088,
      "step": 805
    },
    {
      "epoch": 0.025926637219128097,
      "grad_norm": 0.05799093842506409,
      "learning_rate": 0.29969731426466134,
      "loss": 1.2371,
      "num_input_tokens_seen": 1114160,
      "step": 810
    },
    {
      "epoch": 0.026086678189616543,
      "grad_norm": 0.08366210758686066,
      "learning_rate": 0.299693562500946,
      "loss": 1.1239,
      "num_input_tokens_seen": 1120784,
      "step": 815
    },
    {
      "epoch": 0.026246719160104987,
      "grad_norm": 0.08121301978826523,
      "learning_rate": 0.29968978765260207,
      "loss": 1.1718,
      "num_input_tokens_seen": 1127232,
      "step": 820
    },
    {
      "epoch": 0.026406760130593433,
      "grad_norm": 0.040331702679395676,
      "learning_rate": 0.2996859897202118,
      "loss": 1.3906,
      "num_input_tokens_seen": 1134416,
      "step": 825
    },
    {
      "epoch": 0.026566801101081877,
      "grad_norm": 0.08420103788375854,
      "learning_rate": 0.2996821687043609,
      "loss": 1.0838,
      "num_input_tokens_seen": 1141488,
      "step": 830
    },
    {
      "epoch": 0.026726842071570323,
      "grad_norm": 0.17464607954025269,
      "learning_rate": 0.2996783246056384,
      "loss": 1.1114,
      "num_input_tokens_seen": 1148608,
      "step": 835
    },
    {
      "epoch": 0.026886883042058767,
      "grad_norm": 0.04397923871874809,
      "learning_rate": 0.29967445742463744,
      "loss": 1.1308,
      "num_input_tokens_seen": 1155504,
      "step": 840
    },
    {
      "epoch": 0.027046924012547213,
      "grad_norm": 0.06807153671979904,
      "learning_rate": 0.29967056716195417,
      "loss": 0.9456,
      "num_input_tokens_seen": 1162128,
      "step": 845
    },
    {
      "epoch": 0.027206964983035656,
      "grad_norm": 0.05960332602262497,
      "learning_rate": 0.2996666538181885,
      "loss": 0.9697,
      "num_input_tokens_seen": 1169088,
      "step": 850
    },
    {
      "epoch": 0.027367005953524103,
      "grad_norm": 0.23901644349098206,
      "learning_rate": 0.29966271739394407,
      "loss": 1.3427,
      "num_input_tokens_seen": 1175840,
      "step": 855
    },
    {
      "epoch": 0.027527046924012546,
      "grad_norm": 0.03538615256547928,
      "learning_rate": 0.29965875788982776,
      "loss": 1.1363,
      "num_input_tokens_seen": 1182624,
      "step": 860
    },
    {
      "epoch": 0.027687087894500993,
      "grad_norm": 0.1038421019911766,
      "learning_rate": 0.2996547753064503,
      "loss": 1.2249,
      "num_input_tokens_seen": 1189504,
      "step": 865
    },
    {
      "epoch": 0.027847128864989436,
      "grad_norm": 0.07227155566215515,
      "learning_rate": 0.29965076964442583,
      "loss": 1.3839,
      "num_input_tokens_seen": 1196448,
      "step": 870
    },
    {
      "epoch": 0.028007169835477883,
      "grad_norm": 0.045242760330438614,
      "learning_rate": 0.299646740904372,
      "loss": 1.0784,
      "num_input_tokens_seen": 1203376,
      "step": 875
    },
    {
      "epoch": 0.028167210805966326,
      "grad_norm": 0.05637679249048233,
      "learning_rate": 0.29964268908691016,
      "loss": 1.0835,
      "num_input_tokens_seen": 1209936,
      "step": 880
    },
    {
      "epoch": 0.028327251776454773,
      "grad_norm": 0.06357412785291672,
      "learning_rate": 0.29963861419266513,
      "loss": 1.0219,
      "num_input_tokens_seen": 1216496,
      "step": 885
    },
    {
      "epoch": 0.028487292746943216,
      "grad_norm": 0.0920478105545044,
      "learning_rate": 0.29963451622226533,
      "loss": 0.8846,
      "num_input_tokens_seen": 1222928,
      "step": 890
    },
    {
      "epoch": 0.028647333717431663,
      "grad_norm": 0.06964627653360367,
      "learning_rate": 0.29963039517634277,
      "loss": 1.1858,
      "num_input_tokens_seen": 1229808,
      "step": 895
    },
    {
      "epoch": 0.028807374687920106,
      "grad_norm": 0.04664414003491402,
      "learning_rate": 0.2996262510555328,
      "loss": 0.9058,
      "num_input_tokens_seen": 1236688,
      "step": 900
    },
    {
      "epoch": 0.028967415658408553,
      "grad_norm": 0.043758492916822433,
      "learning_rate": 0.2996220838604746,
      "loss": 1.3017,
      "num_input_tokens_seen": 1243472,
      "step": 905
    },
    {
      "epoch": 0.029127456628896996,
      "grad_norm": 0.03300178050994873,
      "learning_rate": 0.29961789359181085,
      "loss": 1.0901,
      "num_input_tokens_seen": 1250592,
      "step": 910
    },
    {
      "epoch": 0.029287497599385443,
      "grad_norm": 0.0692644789814949,
      "learning_rate": 0.29961368025018764,
      "loss": 1.179,
      "num_input_tokens_seen": 1257152,
      "step": 915
    },
    {
      "epoch": 0.029447538569873886,
      "grad_norm": 0.06598510593175888,
      "learning_rate": 0.2996094438362548,
      "loss": 1.1847,
      "num_input_tokens_seen": 1263984,
      "step": 920
    },
    {
      "epoch": 0.029607579540362333,
      "grad_norm": 0.07139763981103897,
      "learning_rate": 0.2996051843506657,
      "loss": 1.1847,
      "num_input_tokens_seen": 1271216,
      "step": 925
    },
    {
      "epoch": 0.029767620510850776,
      "grad_norm": 0.0748894140124321,
      "learning_rate": 0.299600901794077,
      "loss": 1.0004,
      "num_input_tokens_seen": 1278208,
      "step": 930
    },
    {
      "epoch": 0.029927661481339223,
      "grad_norm": 0.05430731177330017,
      "learning_rate": 0.29959659616714923,
      "loss": 1.1467,
      "num_input_tokens_seen": 1284816,
      "step": 935
    },
    {
      "epoch": 0.030087702451827666,
      "grad_norm": 0.04089600220322609,
      "learning_rate": 0.2995922674705464,
      "loss": 1.2923,
      "num_input_tokens_seen": 1291296,
      "step": 940
    },
    {
      "epoch": 0.030247743422316113,
      "grad_norm": 0.05056396499276161,
      "learning_rate": 0.2995879157049361,
      "loss": 1.11,
      "num_input_tokens_seen": 1297696,
      "step": 945
    },
    {
      "epoch": 0.030407784392804556,
      "grad_norm": 0.05439203605055809,
      "learning_rate": 0.2995835408709893,
      "loss": 1.0449,
      "num_input_tokens_seen": 1304384,
      "step": 950
    },
    {
      "epoch": 0.030567825363293003,
      "grad_norm": 0.06213686987757683,
      "learning_rate": 0.29957914296938076,
      "loss": 1.3724,
      "num_input_tokens_seen": 1312176,
      "step": 955
    },
    {
      "epoch": 0.030727866333781446,
      "grad_norm": 0.04116702824831009,
      "learning_rate": 0.2995747220007886,
      "loss": 0.9826,
      "num_input_tokens_seen": 1319216,
      "step": 960
    },
    {
      "epoch": 0.030887907304269893,
      "grad_norm": 0.06491638720035553,
      "learning_rate": 0.2995702779658947,
      "loss": 1.0133,
      "num_input_tokens_seen": 1326368,
      "step": 965
    },
    {
      "epoch": 0.031047948274758336,
      "grad_norm": 0.0906352549791336,
      "learning_rate": 0.29956581086538425,
      "loss": 1.4511,
      "num_input_tokens_seen": 1333248,
      "step": 970
    },
    {
      "epoch": 0.031207989245246783,
      "grad_norm": 0.048890773206949234,
      "learning_rate": 0.2995613206999462,
      "loss": 1.0744,
      "num_input_tokens_seen": 1340032,
      "step": 975
    },
    {
      "epoch": 0.031368030215735226,
      "grad_norm": 0.04187506064772606,
      "learning_rate": 0.29955680747027297,
      "loss": 0.7962,
      "num_input_tokens_seen": 1346672,
      "step": 980
    },
    {
      "epoch": 0.03152807118622367,
      "grad_norm": 0.03373297303915024,
      "learning_rate": 0.2995522711770607,
      "loss": 0.7246,
      "num_input_tokens_seen": 1353664,
      "step": 985
    },
    {
      "epoch": 0.03168811215671212,
      "grad_norm": 0.039820268750190735,
      "learning_rate": 0.2995477118210087,
      "loss": 0.9756,
      "num_input_tokens_seen": 1360992,
      "step": 990
    },
    {
      "epoch": 0.03184815312720057,
      "grad_norm": 0.0649365484714508,
      "learning_rate": 0.29954312940282024,
      "loss": 1.3394,
      "num_input_tokens_seen": 1367648,
      "step": 995
    },
    {
      "epoch": 0.032008194097689006,
      "grad_norm": 0.09034864604473114,
      "learning_rate": 0.29953852392320196,
      "loss": 1.1076,
      "num_input_tokens_seen": 1374672,
      "step": 1000
    },
    {
      "epoch": 0.032008194097689006,
      "eval_loss": 1.060508131980896,
      "eval_runtime": 331.9135,
      "eval_samples_per_second": 41.836,
      "eval_steps_per_second": 20.918,
      "num_input_tokens_seen": 1374672,
      "step": 1000
    },
    {
      "epoch": 0.03216823506817745,
      "grad_norm": 0.06653349846601486,
      "learning_rate": 0.2995338953828641,
      "loss": 0.969,
      "num_input_tokens_seen": 1381504,
      "step": 1005
    },
    {
      "epoch": 0.0323282760386659,
      "grad_norm": 0.03621481731534004,
      "learning_rate": 0.2995292437825204,
      "loss": 1.0101,
      "num_input_tokens_seen": 1388448,
      "step": 1010
    },
    {
      "epoch": 0.03248831700915435,
      "grad_norm": 0.05216300114989281,
      "learning_rate": 0.29952456912288816,
      "loss": 0.9022,
      "num_input_tokens_seen": 1394992,
      "step": 1015
    },
    {
      "epoch": 0.032648357979642786,
      "grad_norm": 0.059891246259212494,
      "learning_rate": 0.2995198714046884,
      "loss": 1.2266,
      "num_input_tokens_seen": 1401712,
      "step": 1020
    },
    {
      "epoch": 0.03280839895013123,
      "grad_norm": 0.06922195851802826,
      "learning_rate": 0.2995151506286454,
      "loss": 1.094,
      "num_input_tokens_seen": 1408288,
      "step": 1025
    },
    {
      "epoch": 0.03296843992061968,
      "grad_norm": 0.0431087464094162,
      "learning_rate": 0.2995104067954873,
      "loss": 1.0052,
      "num_input_tokens_seen": 1415216,
      "step": 1030
    },
    {
      "epoch": 0.03312848089110813,
      "grad_norm": 0.07935462892055511,
      "learning_rate": 0.2995056399059456,
      "loss": 1.1466,
      "num_input_tokens_seen": 1422272,
      "step": 1035
    },
    {
      "epoch": 0.033288521861596566,
      "grad_norm": 0.06211337074637413,
      "learning_rate": 0.2995008499607554,
      "loss": 1.4228,
      "num_input_tokens_seen": 1429520,
      "step": 1040
    },
    {
      "epoch": 0.03344856283208501,
      "grad_norm": 0.057185523211956024,
      "learning_rate": 0.2994960369606554,
      "loss": 0.9768,
      "num_input_tokens_seen": 1436160,
      "step": 1045
    },
    {
      "epoch": 0.03360860380257346,
      "grad_norm": 0.08126810193061829,
      "learning_rate": 0.2994912009063878,
      "loss": 1.289,
      "num_input_tokens_seen": 1443120,
      "step": 1050
    },
    {
      "epoch": 0.03376864477306191,
      "grad_norm": 0.05696961656212807,
      "learning_rate": 0.29948634179869843,
      "loss": 1.184,
      "num_input_tokens_seen": 1449904,
      "step": 1055
    },
    {
      "epoch": 0.033928685743550346,
      "grad_norm": 0.058745380491018295,
      "learning_rate": 0.29948145963833656,
      "loss": 1.1047,
      "num_input_tokens_seen": 1456432,
      "step": 1060
    },
    {
      "epoch": 0.03408872671403879,
      "grad_norm": 0.050173573195934296,
      "learning_rate": 0.29947655442605514,
      "loss": 0.9561,
      "num_input_tokens_seen": 1463200,
      "step": 1065
    },
    {
      "epoch": 0.03424876768452724,
      "grad_norm": 0.058213941752910614,
      "learning_rate": 0.2994716261626106,
      "loss": 0.9739,
      "num_input_tokens_seen": 1470064,
      "step": 1070
    },
    {
      "epoch": 0.03440880865501569,
      "grad_norm": 0.04412616044282913,
      "learning_rate": 0.2994666748487629,
      "loss": 0.8766,
      "num_input_tokens_seen": 1477312,
      "step": 1075
    },
    {
      "epoch": 0.034568849625504126,
      "grad_norm": 0.1199728399515152,
      "learning_rate": 0.2994617004852756,
      "loss": 1.2738,
      "num_input_tokens_seen": 1484176,
      "step": 1080
    },
    {
      "epoch": 0.03472889059599257,
      "grad_norm": 0.04901612177491188,
      "learning_rate": 0.2994567030729159,
      "loss": 0.8992,
      "num_input_tokens_seen": 1491088,
      "step": 1085
    },
    {
      "epoch": 0.03488893156648102,
      "grad_norm": 0.04599233344197273,
      "learning_rate": 0.29945168261245436,
      "loss": 1.1529,
      "num_input_tokens_seen": 1497760,
      "step": 1090
    },
    {
      "epoch": 0.03504897253696947,
      "grad_norm": 0.06795535236597061,
      "learning_rate": 0.29944663910466524,
      "loss": 0.9293,
      "num_input_tokens_seen": 1504672,
      "step": 1095
    },
    {
      "epoch": 0.035209013507457906,
      "grad_norm": 0.05336945876479149,
      "learning_rate": 0.2994415725503263,
      "loss": 1.0234,
      "num_input_tokens_seen": 1511984,
      "step": 1100
    },
    {
      "epoch": 0.03536905447794635,
      "grad_norm": 0.11039286851882935,
      "learning_rate": 0.29943648295021885,
      "loss": 1.0535,
      "num_input_tokens_seen": 1518912,
      "step": 1105
    },
    {
      "epoch": 0.0355290954484348,
      "grad_norm": 0.04557309299707413,
      "learning_rate": 0.2994313703051278,
      "loss": 1.3539,
      "num_input_tokens_seen": 1525472,
      "step": 1110
    },
    {
      "epoch": 0.03568913641892325,
      "grad_norm": 0.03710173815488815,
      "learning_rate": 0.29942623461584156,
      "loss": 1.2005,
      "num_input_tokens_seen": 1531984,
      "step": 1115
    },
    {
      "epoch": 0.035849177389411686,
      "grad_norm": 0.07652467489242554,
      "learning_rate": 0.29942107588315214,
      "loss": 1.2288,
      "num_input_tokens_seen": 1539296,
      "step": 1120
    },
    {
      "epoch": 0.03600921835990013,
      "grad_norm": 0.04620356485247612,
      "learning_rate": 0.29941589410785513,
      "loss": 1.1613,
      "num_input_tokens_seen": 1546112,
      "step": 1125
    },
    {
      "epoch": 0.03616925933038858,
      "grad_norm": 0.08342301845550537,
      "learning_rate": 0.29941068929074954,
      "loss": 1.1751,
      "num_input_tokens_seen": 1553040,
      "step": 1130
    },
    {
      "epoch": 0.03632930030087703,
      "grad_norm": 0.059764280915260315,
      "learning_rate": 0.2994054614326381,
      "loss": 0.9019,
      "num_input_tokens_seen": 1559504,
      "step": 1135
    },
    {
      "epoch": 0.036489341271365466,
      "grad_norm": 0.0717848464846611,
      "learning_rate": 0.29940021053432686,
      "loss": 1.2005,
      "num_input_tokens_seen": 1566384,
      "step": 1140
    },
    {
      "epoch": 0.03664938224185391,
      "grad_norm": 0.07254995405673981,
      "learning_rate": 0.29939493659662575,
      "loss": 1.3211,
      "num_input_tokens_seen": 1573072,
      "step": 1145
    },
    {
      "epoch": 0.03680942321234236,
      "grad_norm": 0.053411975502967834,
      "learning_rate": 0.299389639620348,
      "loss": 1.1182,
      "num_input_tokens_seen": 1580000,
      "step": 1150
    },
    {
      "epoch": 0.03696946418283081,
      "grad_norm": 0.057727016508579254,
      "learning_rate": 0.29938431960631046,
      "loss": 1.2881,
      "num_input_tokens_seen": 1586624,
      "step": 1155
    },
    {
      "epoch": 0.037129505153319246,
      "grad_norm": 0.07785720378160477,
      "learning_rate": 0.2993789765553335,
      "loss": 1.1709,
      "num_input_tokens_seen": 1593696,
      "step": 1160
    },
    {
      "epoch": 0.03728954612380769,
      "grad_norm": 0.04368605837225914,
      "learning_rate": 0.2993736104682412,
      "loss": 1.4118,
      "num_input_tokens_seen": 1600384,
      "step": 1165
    },
    {
      "epoch": 0.03744958709429614,
      "grad_norm": 0.04937903955578804,
      "learning_rate": 0.299368221345861,
      "loss": 1.1121,
      "num_input_tokens_seen": 1606912,
      "step": 1170
    },
    {
      "epoch": 0.03760962806478459,
      "grad_norm": 0.12851306796073914,
      "learning_rate": 0.29936280918902397,
      "loss": 1.0397,
      "num_input_tokens_seen": 1613760,
      "step": 1175
    },
    {
      "epoch": 0.03776966903527303,
      "grad_norm": 0.043151240795850754,
      "learning_rate": 0.2993573739985648,
      "loss": 1.2319,
      "num_input_tokens_seen": 1620768,
      "step": 1180
    },
    {
      "epoch": 0.03792971000576147,
      "grad_norm": 0.07067602127790451,
      "learning_rate": 0.2993519157753216,
      "loss": 1.0662,
      "num_input_tokens_seen": 1627536,
      "step": 1185
    },
    {
      "epoch": 0.03808975097624992,
      "grad_norm": 0.03777698054909706,
      "learning_rate": 0.2993464345201361,
      "loss": 0.875,
      "num_input_tokens_seen": 1634432,
      "step": 1190
    },
    {
      "epoch": 0.038249791946738367,
      "grad_norm": 0.05872073024511337,
      "learning_rate": 0.2993409302338536,
      "loss": 1.0814,
      "num_input_tokens_seen": 1641296,
      "step": 1195
    },
    {
      "epoch": 0.03840983291722681,
      "grad_norm": 0.03110041841864586,
      "learning_rate": 0.2993354029173229,
      "loss": 0.9496,
      "num_input_tokens_seen": 1647936,
      "step": 1200
    },
    {
      "epoch": 0.03840983291722681,
      "eval_loss": 1.0762405395507812,
      "eval_runtime": 331.7245,
      "eval_samples_per_second": 41.86,
      "eval_steps_per_second": 20.93,
      "num_input_tokens_seen": 1647936,
      "step": 1200
    },
    {
      "epoch": 0.03856987388771525,
      "grad_norm": 0.043167658150196075,
      "learning_rate": 0.2993298525713965,
      "loss": 1.0641,
      "num_input_tokens_seen": 1654928,
      "step": 1205
    },
    {
      "epoch": 0.0387299148582037,
      "grad_norm": 0.08296706527471542,
      "learning_rate": 0.29932427919693017,
      "loss": 1.3136,
      "num_input_tokens_seen": 1661696,
      "step": 1210
    },
    {
      "epoch": 0.038889955828692147,
      "grad_norm": 0.046553414314985275,
      "learning_rate": 0.2993186827947834,
      "loss": 1.1683,
      "num_input_tokens_seen": 1668576,
      "step": 1215
    },
    {
      "epoch": 0.03904999679918059,
      "grad_norm": 0.05918470770120621,
      "learning_rate": 0.2993130633658194,
      "loss": 1.1217,
      "num_input_tokens_seen": 1675408,
      "step": 1220
    },
    {
      "epoch": 0.03921003776966903,
      "grad_norm": 0.08333911001682281,
      "learning_rate": 0.29930742091090456,
      "loss": 1.0977,
      "num_input_tokens_seen": 1682304,
      "step": 1225
    },
    {
      "epoch": 0.03937007874015748,
      "grad_norm": 0.03855828195810318,
      "learning_rate": 0.29930175543090914,
      "loss": 1.0274,
      "num_input_tokens_seen": 1689152,
      "step": 1230
    },
    {
      "epoch": 0.039530119710645927,
      "grad_norm": 0.03851839527487755,
      "learning_rate": 0.2992960669267068,
      "loss": 1.0148,
      "num_input_tokens_seen": 1696096,
      "step": 1235
    },
    {
      "epoch": 0.03969016068113437,
      "grad_norm": 0.026873165741562843,
      "learning_rate": 0.29929035539917476,
      "loss": 0.7357,
      "num_input_tokens_seen": 1702816,
      "step": 1240
    },
    {
      "epoch": 0.03985020165162281,
      "grad_norm": 0.03212342411279678,
      "learning_rate": 0.2992846208491938,
      "loss": 1.0779,
      "num_input_tokens_seen": 1709568,
      "step": 1245
    },
    {
      "epoch": 0.04001024262211126,
      "grad_norm": 0.0400514118373394,
      "learning_rate": 0.2992788632776483,
      "loss": 1.1695,
      "num_input_tokens_seen": 1716368,
      "step": 1250
    },
    {
      "epoch": 0.040170283592599706,
      "grad_norm": 0.09034435451030731,
      "learning_rate": 0.29927308268542613,
      "loss": 1.2038,
      "num_input_tokens_seen": 1723136,
      "step": 1255
    },
    {
      "epoch": 0.04033032456308815,
      "grad_norm": 0.06526792794466019,
      "learning_rate": 0.2992672790734187,
      "loss": 0.9879,
      "num_input_tokens_seen": 1730144,
      "step": 1260
    },
    {
      "epoch": 0.04049036553357659,
      "grad_norm": 0.05261319503188133,
      "learning_rate": 0.299261452442521,
      "loss": 0.9727,
      "num_input_tokens_seen": 1737072,
      "step": 1265
    },
    {
      "epoch": 0.04065040650406504,
      "grad_norm": 0.06508932262659073,
      "learning_rate": 0.29925560279363167,
      "loss": 0.9243,
      "num_input_tokens_seen": 1743728,
      "step": 1270
    },
    {
      "epoch": 0.040810447474553486,
      "grad_norm": 0.04002340883016586,
      "learning_rate": 0.29924973012765266,
      "loss": 0.9117,
      "num_input_tokens_seen": 1750560,
      "step": 1275
    },
    {
      "epoch": 0.04097048844504193,
      "grad_norm": 0.049347516149282455,
      "learning_rate": 0.29924383444548974,
      "loss": 1.0455,
      "num_input_tokens_seen": 1757248,
      "step": 1280
    },
    {
      "epoch": 0.04113052941553037,
      "grad_norm": 0.0655953586101532,
      "learning_rate": 0.299237915748052,
      "loss": 1.0634,
      "num_input_tokens_seen": 1764336,
      "step": 1285
    },
    {
      "epoch": 0.04129057038601882,
      "grad_norm": 0.029108574613928795,
      "learning_rate": 0.2992319740362522,
      "loss": 0.9713,
      "num_input_tokens_seen": 1771296,
      "step": 1290
    },
    {
      "epoch": 0.041450611356507266,
      "grad_norm": 0.04089698940515518,
      "learning_rate": 0.2992260093110066,
      "loss": 0.9085,
      "num_input_tokens_seen": 1778144,
      "step": 1295
    },
    {
      "epoch": 0.04161065232699571,
      "grad_norm": 0.04282957687973976,
      "learning_rate": 0.2992200215732352,
      "loss": 1.2368,
      "num_input_tokens_seen": 1784832,
      "step": 1300
    },
    {
      "epoch": 0.04177069329748415,
      "grad_norm": 0.0337517149746418,
      "learning_rate": 0.2992140108238611,
      "loss": 1.0009,
      "num_input_tokens_seen": 1791744,
      "step": 1305
    },
    {
      "epoch": 0.0419307342679726,
      "grad_norm": 0.04545779153704643,
      "learning_rate": 0.2992079770638115,
      "loss": 1.2712,
      "num_input_tokens_seen": 1798224,
      "step": 1310
    },
    {
      "epoch": 0.042090775238461046,
      "grad_norm": 0.042945485562086105,
      "learning_rate": 0.29920192029401677,
      "loss": 1.0574,
      "num_input_tokens_seen": 1805168,
      "step": 1315
    },
    {
      "epoch": 0.04225081620894949,
      "grad_norm": 0.067461296916008,
      "learning_rate": 0.2991958405154109,
      "loss": 0.9299,
      "num_input_tokens_seen": 1812624,
      "step": 1320
    },
    {
      "epoch": 0.04241085717943793,
      "grad_norm": 0.08728460222482681,
      "learning_rate": 0.29918973772893154,
      "loss": 1.076,
      "num_input_tokens_seen": 1819520,
      "step": 1325
    },
    {
      "epoch": 0.04257089814992638,
      "grad_norm": 0.053034331649541855,
      "learning_rate": 0.29918361193551973,
      "loss": 0.9339,
      "num_input_tokens_seen": 1826176,
      "step": 1330
    },
    {
      "epoch": 0.042730939120414826,
      "grad_norm": 0.025531893596053123,
      "learning_rate": 0.29917746313612026,
      "loss": 1.1609,
      "num_input_tokens_seen": 1833312,
      "step": 1335
    },
    {
      "epoch": 0.04289098009090327,
      "grad_norm": 0.025029171258211136,
      "learning_rate": 0.29917129133168124,
      "loss": 0.8015,
      "num_input_tokens_seen": 1839808,
      "step": 1340
    },
    {
      "epoch": 0.04305102106139171,
      "grad_norm": 0.020545821636915207,
      "learning_rate": 0.2991650965231546,
      "loss": 0.9907,
      "num_input_tokens_seen": 1846512,
      "step": 1345
    },
    {
      "epoch": 0.04321106203188016,
      "grad_norm": 0.03342806175351143,
      "learning_rate": 0.29915887871149544,
      "loss": 0.888,
      "num_input_tokens_seen": 1853696,
      "step": 1350
    },
    {
      "epoch": 0.043371103002368606,
      "grad_norm": 0.034904856234788895,
      "learning_rate": 0.2991526378976628,
      "loss": 0.9775,
      "num_input_tokens_seen": 1860352,
      "step": 1355
    },
    {
      "epoch": 0.04353114397285705,
      "grad_norm": 0.027617553249001503,
      "learning_rate": 0.29914637408261896,
      "loss": 0.6603,
      "num_input_tokens_seen": 1867008,
      "step": 1360
    },
    {
      "epoch": 0.0436911849433455,
      "grad_norm": 0.03732902184128761,
      "learning_rate": 0.29914008726733,
      "loss": 0.9573,
      "num_input_tokens_seen": 1873600,
      "step": 1365
    },
    {
      "epoch": 0.04385122591383394,
      "grad_norm": 0.04380154609680176,
      "learning_rate": 0.2991337774527653,
      "loss": 1.0548,
      "num_input_tokens_seen": 1880512,
      "step": 1370
    },
    {
      "epoch": 0.044011266884322386,
      "grad_norm": 0.030988754704594612,
      "learning_rate": 0.2991274446398981,
      "loss": 1.1723,
      "num_input_tokens_seen": 1887168,
      "step": 1375
    },
    {
      "epoch": 0.04417130785481083,
      "grad_norm": 0.042881764471530914,
      "learning_rate": 0.29912108882970484,
      "loss": 0.8024,
      "num_input_tokens_seen": 1894336,
      "step": 1380
    },
    {
      "epoch": 0.04433134882529928,
      "grad_norm": 0.027261964976787567,
      "learning_rate": 0.2991147100231657,
      "loss": 0.9786,
      "num_input_tokens_seen": 1901120,
      "step": 1385
    },
    {
      "epoch": 0.04449138979578772,
      "grad_norm": 0.04184195026755333,
      "learning_rate": 0.2991083082212644,
      "loss": 1.026,
      "num_input_tokens_seen": 1907904,
      "step": 1390
    },
    {
      "epoch": 0.044651430766276166,
      "grad_norm": 0.03985224664211273,
      "learning_rate": 0.2991018834249881,
      "loss": 1.0688,
      "num_input_tokens_seen": 1914816,
      "step": 1395
    },
    {
      "epoch": 0.04481147173676461,
      "grad_norm": 0.054830152541399,
      "learning_rate": 0.29909543563532764,
      "loss": 1.0765,
      "num_input_tokens_seen": 1921648,
      "step": 1400
    },
    {
      "epoch": 0.04481147173676461,
      "eval_loss": 0.9991708993911743,
      "eval_runtime": 332.0527,
      "eval_samples_per_second": 41.819,
      "eval_steps_per_second": 20.909,
      "num_input_tokens_seen": 1921648,
      "step": 1400
    },
    {
      "epoch": 0.04497151270725306,
      "grad_norm": 0.04970398545265198,
      "learning_rate": 0.29908896485327746,
      "loss": 0.7447,
      "num_input_tokens_seen": 1928352,
      "step": 1405
    },
    {
      "epoch": 0.0451315536777415,
      "grad_norm": 0.05061139911413193,
      "learning_rate": 0.29908247107983527,
      "loss": 0.9842,
      "num_input_tokens_seen": 1935232,
      "step": 1410
    },
    {
      "epoch": 0.045291594648229946,
      "grad_norm": 0.02771879732608795,
      "learning_rate": 0.29907595431600253,
      "loss": 0.7994,
      "num_input_tokens_seen": 1941728,
      "step": 1415
    },
    {
      "epoch": 0.04545163561871839,
      "grad_norm": 0.09495393931865692,
      "learning_rate": 0.29906941456278424,
      "loss": 1.0565,
      "num_input_tokens_seen": 1948272,
      "step": 1420
    },
    {
      "epoch": 0.04561167658920684,
      "grad_norm": 0.035785991698503494,
      "learning_rate": 0.2990628518211889,
      "loss": 0.9362,
      "num_input_tokens_seen": 1954784,
      "step": 1425
    },
    {
      "epoch": 0.04577171755969528,
      "grad_norm": 0.07509482651948929,
      "learning_rate": 0.2990562660922286,
      "loss": 0.7994,
      "num_input_tokens_seen": 1961808,
      "step": 1430
    },
    {
      "epoch": 0.045931758530183726,
      "grad_norm": 0.053934551775455475,
      "learning_rate": 0.2990496573769189,
      "loss": 1.1302,
      "num_input_tokens_seen": 1968304,
      "step": 1435
    },
    {
      "epoch": 0.04609179950067217,
      "grad_norm": 0.04055286571383476,
      "learning_rate": 0.29904302567627894,
      "loss": 1.1034,
      "num_input_tokens_seen": 1975024,
      "step": 1440
    },
    {
      "epoch": 0.04625184047116062,
      "grad_norm": 0.0563739649951458,
      "learning_rate": 0.2990363709913314,
      "loss": 1.0294,
      "num_input_tokens_seen": 1982192,
      "step": 1445
    },
    {
      "epoch": 0.04641188144164906,
      "grad_norm": 0.02800201252102852,
      "learning_rate": 0.29902969332310264,
      "loss": 1.1554,
      "num_input_tokens_seen": 1989168,
      "step": 1450
    },
    {
      "epoch": 0.046571922412137506,
      "grad_norm": 0.03550909459590912,
      "learning_rate": 0.2990229926726223,
      "loss": 0.9651,
      "num_input_tokens_seen": 1995824,
      "step": 1455
    },
    {
      "epoch": 0.04673196338262595,
      "grad_norm": 0.04200877249240875,
      "learning_rate": 0.29901626904092365,
      "loss": 0.9089,
      "num_input_tokens_seen": 2002608,
      "step": 1460
    },
    {
      "epoch": 0.0468920043531144,
      "grad_norm": 0.049336519092321396,
      "learning_rate": 0.2990095224290438,
      "loss": 0.985,
      "num_input_tokens_seen": 2009424,
      "step": 1465
    },
    {
      "epoch": 0.04705204532360284,
      "grad_norm": 0.04744068533182144,
      "learning_rate": 0.29900275283802297,
      "loss": 1.21,
      "num_input_tokens_seen": 2015968,
      "step": 1470
    },
    {
      "epoch": 0.047212086294091286,
      "grad_norm": 0.033990323543548584,
      "learning_rate": 0.2989959602689051,
      "loss": 0.9305,
      "num_input_tokens_seen": 2022896,
      "step": 1475
    },
    {
      "epoch": 0.04737212726457973,
      "grad_norm": 0.05207419395446777,
      "learning_rate": 0.2989891447227379,
      "loss": 1.0303,
      "num_input_tokens_seen": 2030080,
      "step": 1480
    },
    {
      "epoch": 0.04753216823506818,
      "grad_norm": 0.04970686882734299,
      "learning_rate": 0.29898230620057215,
      "loss": 1.1681,
      "num_input_tokens_seen": 2036608,
      "step": 1485
    },
    {
      "epoch": 0.04769220920555662,
      "grad_norm": 0.04199592396616936,
      "learning_rate": 0.2989754447034626,
      "loss": 0.7831,
      "num_input_tokens_seen": 2043264,
      "step": 1490
    },
    {
      "epoch": 0.047852250176045066,
      "grad_norm": 0.038075413554906845,
      "learning_rate": 0.2989685602324673,
      "loss": 0.8966,
      "num_input_tokens_seen": 2050000,
      "step": 1495
    },
    {
      "epoch": 0.04801229114653351,
      "grad_norm": 0.042209286242723465,
      "learning_rate": 0.298961652788648,
      "loss": 0.9866,
      "num_input_tokens_seen": 2056784,
      "step": 1500
    },
    {
      "epoch": 0.04817233211702196,
      "grad_norm": 0.039651621133089066,
      "learning_rate": 0.29895472237306986,
      "loss": 0.9251,
      "num_input_tokens_seen": 2064032,
      "step": 1505
    },
    {
      "epoch": 0.0483323730875104,
      "grad_norm": 0.14769086241722107,
      "learning_rate": 0.29894776898680164,
      "loss": 1.0203,
      "num_input_tokens_seen": 2070576,
      "step": 1510
    },
    {
      "epoch": 0.048492414057998846,
      "grad_norm": 0.058343786746263504,
      "learning_rate": 0.29894079263091566,
      "loss": 1.0991,
      "num_input_tokens_seen": 2077056,
      "step": 1515
    },
    {
      "epoch": 0.04865245502848729,
      "grad_norm": 0.02995944768190384,
      "learning_rate": 0.2989337933064877,
      "loss": 0.9292,
      "num_input_tokens_seen": 2083472,
      "step": 1520
    },
    {
      "epoch": 0.04881249599897574,
      "grad_norm": 0.030637720599770546,
      "learning_rate": 0.29892677101459725,
      "loss": 0.9518,
      "num_input_tokens_seen": 2091008,
      "step": 1525
    },
    {
      "epoch": 0.04897253696946418,
      "grad_norm": 0.027993880212306976,
      "learning_rate": 0.2989197257563272,
      "loss": 1.1096,
      "num_input_tokens_seen": 2097456,
      "step": 1530
    },
    {
      "epoch": 0.049132577939952626,
      "grad_norm": 0.03736189752817154,
      "learning_rate": 0.2989126575327639,
      "loss": 0.9493,
      "num_input_tokens_seen": 2104448,
      "step": 1535
    },
    {
      "epoch": 0.04929261891044107,
      "grad_norm": 0.047107595950365067,
      "learning_rate": 0.29890556634499754,
      "loss": 1.066,
      "num_input_tokens_seen": 2110832,
      "step": 1540
    },
    {
      "epoch": 0.04945265988092952,
      "grad_norm": 0.09245922416448593,
      "learning_rate": 0.2988984521941216,
      "loss": 1.0034,
      "num_input_tokens_seen": 2117584,
      "step": 1545
    },
    {
      "epoch": 0.04961270085141796,
      "grad_norm": 0.07592987269163132,
      "learning_rate": 0.29889131508123307,
      "loss": 1.1673,
      "num_input_tokens_seen": 2124288,
      "step": 1550
    },
    {
      "epoch": 0.049772741821906406,
      "grad_norm": 0.0364064984023571,
      "learning_rate": 0.2988841550074327,
      "loss": 0.855,
      "num_input_tokens_seen": 2131376,
      "step": 1555
    },
    {
      "epoch": 0.04993278279239485,
      "grad_norm": 0.07710907608270645,
      "learning_rate": 0.2988769719738246,
      "loss": 1.1158,
      "num_input_tokens_seen": 2138192,
      "step": 1560
    },
    {
      "epoch": 0.0500928237628833,
      "grad_norm": 0.054551512002944946,
      "learning_rate": 0.29886976598151666,
      "loss": 0.9815,
      "num_input_tokens_seen": 2144992,
      "step": 1565
    },
    {
      "epoch": 0.05025286473337175,
      "grad_norm": 0.046480029821395874,
      "learning_rate": 0.29886253703161986,
      "loss": 0.934,
      "num_input_tokens_seen": 2151792,
      "step": 1570
    },
    {
      "epoch": 0.050412905703860186,
      "grad_norm": 0.0534232035279274,
      "learning_rate": 0.29885528512524917,
      "loss": 0.8535,
      "num_input_tokens_seen": 2159040,
      "step": 1575
    },
    {
      "epoch": 0.05057294667434863,
      "grad_norm": 0.04880591854453087,
      "learning_rate": 0.29884801026352287,
      "loss": 0.8858,
      "num_input_tokens_seen": 2166032,
      "step": 1580
    },
    {
      "epoch": 0.05073298764483708,
      "grad_norm": 0.058654919266700745,
      "learning_rate": 0.2988407124475629,
      "loss": 1.1776,
      "num_input_tokens_seen": 2172944,
      "step": 1585
    },
    {
      "epoch": 0.050893028615325527,
      "grad_norm": 0.13233400881290436,
      "learning_rate": 0.2988333916784945,
      "loss": 1.0488,
      "num_input_tokens_seen": 2180560,
      "step": 1590
    },
    {
      "epoch": 0.051053069585813966,
      "grad_norm": 0.05278223380446434,
      "learning_rate": 0.2988260479574468,
      "loss": 0.8769,
      "num_input_tokens_seen": 2187616,
      "step": 1595
    },
    {
      "epoch": 0.05121311055630241,
      "grad_norm": 0.023036153987050056,
      "learning_rate": 0.2988186812855523,
      "loss": 0.8033,
      "num_input_tokens_seen": 2194448,
      "step": 1600
    },
    {
      "epoch": 0.05121311055630241,
      "eval_loss": 1.0181723833084106,
      "eval_runtime": 332.1668,
      "eval_samples_per_second": 41.804,
      "eval_steps_per_second": 20.902,
      "num_input_tokens_seen": 2194448,
      "step": 1600
    },
    {
      "epoch": 0.05137315152679086,
      "grad_norm": 0.028718333691358566,
      "learning_rate": 0.29881129166394693,
      "loss": 0.9869,
      "num_input_tokens_seen": 2201312,
      "step": 1605
    },
    {
      "epoch": 0.051533192497279307,
      "grad_norm": 0.06482122093439102,
      "learning_rate": 0.29880387909377026,
      "loss": 1.4414,
      "num_input_tokens_seen": 2208496,
      "step": 1610
    },
    {
      "epoch": 0.051693233467767746,
      "grad_norm": 0.050805073231458664,
      "learning_rate": 0.2987964435761655,
      "loss": 1.2881,
      "num_input_tokens_seen": 2215248,
      "step": 1615
    },
    {
      "epoch": 0.05185327443825619,
      "grad_norm": 0.03578615188598633,
      "learning_rate": 0.29878898511227925,
      "loss": 1.0646,
      "num_input_tokens_seen": 2222080,
      "step": 1620
    },
    {
      "epoch": 0.05201331540874464,
      "grad_norm": 0.030217839404940605,
      "learning_rate": 0.2987815037032617,
      "loss": 1.0554,
      "num_input_tokens_seen": 2228784,
      "step": 1625
    },
    {
      "epoch": 0.052173356379233086,
      "grad_norm": 0.027946092188358307,
      "learning_rate": 0.29877399935026655,
      "loss": 0.9274,
      "num_input_tokens_seen": 2235856,
      "step": 1630
    },
    {
      "epoch": 0.052333397349721526,
      "grad_norm": 0.03949752077460289,
      "learning_rate": 0.2987664720544511,
      "loss": 0.8703,
      "num_input_tokens_seen": 2242560,
      "step": 1635
    },
    {
      "epoch": 0.05249343832020997,
      "grad_norm": 0.04093791916966438,
      "learning_rate": 0.2987589218169761,
      "loss": 0.7042,
      "num_input_tokens_seen": 2249504,
      "step": 1640
    },
    {
      "epoch": 0.05265347929069842,
      "grad_norm": 0.030638163909316063,
      "learning_rate": 0.29875134863900604,
      "loss": 0.9827,
      "num_input_tokens_seen": 2256192,
      "step": 1645
    },
    {
      "epoch": 0.052813520261186866,
      "grad_norm": 0.04729931801557541,
      "learning_rate": 0.29874375252170865,
      "loss": 0.9916,
      "num_input_tokens_seen": 2262960,
      "step": 1650
    },
    {
      "epoch": 0.052973561231675306,
      "grad_norm": 0.039900004863739014,
      "learning_rate": 0.2987361334662553,
      "loss": 1.1064,
      "num_input_tokens_seen": 2270128,
      "step": 1655
    },
    {
      "epoch": 0.05313360220216375,
      "grad_norm": 0.033136747777462006,
      "learning_rate": 0.29872849147382113,
      "loss": 0.8541,
      "num_input_tokens_seen": 2276976,
      "step": 1660
    },
    {
      "epoch": 0.0532936431726522,
      "grad_norm": 0.08802048861980438,
      "learning_rate": 0.2987208265455845,
      "loss": 0.9146,
      "num_input_tokens_seen": 2284192,
      "step": 1665
    },
    {
      "epoch": 0.053453684143140646,
      "grad_norm": 0.033150654286146164,
      "learning_rate": 0.29871313868272753,
      "loss": 0.8505,
      "num_input_tokens_seen": 2290832,
      "step": 1670
    },
    {
      "epoch": 0.053613725113629086,
      "grad_norm": 0.037632089108228683,
      "learning_rate": 0.29870542788643567,
      "loss": 1.086,
      "num_input_tokens_seen": 2297824,
      "step": 1675
    },
    {
      "epoch": 0.05377376608411753,
      "grad_norm": 0.06925234943628311,
      "learning_rate": 0.2986976941578981,
      "loss": 1.0978,
      "num_input_tokens_seen": 2304848,
      "step": 1680
    },
    {
      "epoch": 0.05393380705460598,
      "grad_norm": 0.04344506189227104,
      "learning_rate": 0.29868993749830747,
      "loss": 1.0333,
      "num_input_tokens_seen": 2312080,
      "step": 1685
    },
    {
      "epoch": 0.054093848025094426,
      "grad_norm": 0.04869559034705162,
      "learning_rate": 0.2986821579088598,
      "loss": 1.193,
      "num_input_tokens_seen": 2318848,
      "step": 1690
    },
    {
      "epoch": 0.054253888995582866,
      "grad_norm": 0.03320528194308281,
      "learning_rate": 0.29867435539075504,
      "loss": 0.9073,
      "num_input_tokens_seen": 2325760,
      "step": 1695
    },
    {
      "epoch": 0.05441392996607131,
      "grad_norm": 0.05941273644566536,
      "learning_rate": 0.2986665299451963,
      "loss": 1.2804,
      "num_input_tokens_seen": 2333408,
      "step": 1700
    },
    {
      "epoch": 0.05457397093655976,
      "grad_norm": 0.0499674454331398,
      "learning_rate": 0.29865868157339037,
      "loss": 0.8864,
      "num_input_tokens_seen": 2340448,
      "step": 1705
    },
    {
      "epoch": 0.054734011907048206,
      "grad_norm": 0.0746932253241539,
      "learning_rate": 0.2986508102765476,
      "loss": 1.2754,
      "num_input_tokens_seen": 2347696,
      "step": 1710
    },
    {
      "epoch": 0.054894052877536646,
      "grad_norm": 0.045476093888282776,
      "learning_rate": 0.2986429160558818,
      "loss": 1.1326,
      "num_input_tokens_seen": 2354656,
      "step": 1715
    },
    {
      "epoch": 0.05505409384802509,
      "grad_norm": 0.05554944649338722,
      "learning_rate": 0.2986349989126104,
      "loss": 1.2362,
      "num_input_tokens_seen": 2362224,
      "step": 1720
    },
    {
      "epoch": 0.05521413481851354,
      "grad_norm": 0.04240943863987923,
      "learning_rate": 0.29862705884795426,
      "loss": 1.0845,
      "num_input_tokens_seen": 2368896,
      "step": 1725
    },
    {
      "epoch": 0.055374175789001986,
      "grad_norm": 0.06309725344181061,
      "learning_rate": 0.2986190958631379,
      "loss": 1.1523,
      "num_input_tokens_seen": 2375872,
      "step": 1730
    },
    {
      "epoch": 0.055534216759490426,
      "grad_norm": 0.05694306269288063,
      "learning_rate": 0.29861110995938933,
      "loss": 1.1589,
      "num_input_tokens_seen": 2382832,
      "step": 1735
    },
    {
      "epoch": 0.05569425772997887,
      "grad_norm": 0.06695012748241425,
      "learning_rate": 0.29860310113794,
      "loss": 1.0639,
      "num_input_tokens_seen": 2389648,
      "step": 1740
    },
    {
      "epoch": 0.05585429870046732,
      "grad_norm": 0.07461343705654144,
      "learning_rate": 0.29859506940002506,
      "loss": 1.1733,
      "num_input_tokens_seen": 2396832,
      "step": 1745
    },
    {
      "epoch": 0.056014339670955766,
      "grad_norm": 0.04093345254659653,
      "learning_rate": 0.298587014746883,
      "loss": 0.9373,
      "num_input_tokens_seen": 2403488,
      "step": 1750
    },
    {
      "epoch": 0.05617438064144421,
      "grad_norm": 0.06338132172822952,
      "learning_rate": 0.298578937179756,
      "loss": 1.523,
      "num_input_tokens_seen": 2410384,
      "step": 1755
    },
    {
      "epoch": 0.05633442161193265,
      "grad_norm": 0.09627722203731537,
      "learning_rate": 0.29857083669988976,
      "loss": 1.2479,
      "num_input_tokens_seen": 2417168,
      "step": 1760
    },
    {
      "epoch": 0.0564944625824211,
      "grad_norm": 0.06408246606588364,
      "learning_rate": 0.29856271330853346,
      "loss": 1.0165,
      "num_input_tokens_seen": 2423664,
      "step": 1765
    },
    {
      "epoch": 0.056654503552909546,
      "grad_norm": 0.05055200681090355,
      "learning_rate": 0.2985545670069398,
      "loss": 0.9783,
      "num_input_tokens_seen": 2430640,
      "step": 1770
    },
    {
      "epoch": 0.05681454452339799,
      "grad_norm": 0.06814990192651749,
      "learning_rate": 0.29854639779636505,
      "loss": 1.2925,
      "num_input_tokens_seen": 2437504,
      "step": 1775
    },
    {
      "epoch": 0.05697458549388643,
      "grad_norm": 0.056757766753435135,
      "learning_rate": 0.298538205678069,
      "loss": 1.1426,
      "num_input_tokens_seen": 2444848,
      "step": 1780
    },
    {
      "epoch": 0.05713462646437488,
      "grad_norm": 0.13704052567481995,
      "learning_rate": 0.298529990653315,
      "loss": 1.55,
      "num_input_tokens_seen": 2451712,
      "step": 1785
    },
    {
      "epoch": 0.057294667434863326,
      "grad_norm": 0.047152671962976456,
      "learning_rate": 0.29852175272336984,
      "loss": 1.1135,
      "num_input_tokens_seen": 2458704,
      "step": 1790
    },
    {
      "epoch": 0.05745470840535177,
      "grad_norm": 0.07824893295764923,
      "learning_rate": 0.29851349188950405,
      "loss": 0.949,
      "num_input_tokens_seen": 2465088,
      "step": 1795
    },
    {
      "epoch": 0.05761474937584021,
      "grad_norm": 0.05511604622006416,
      "learning_rate": 0.2985052081529914,
      "loss": 0.9584,
      "num_input_tokens_seen": 2472048,
      "step": 1800
    },
    {
      "epoch": 0.05761474937584021,
      "eval_loss": 1.162886381149292,
      "eval_runtime": 331.5624,
      "eval_samples_per_second": 41.881,
      "eval_steps_per_second": 20.94,
      "num_input_tokens_seen": 2472048,
      "step": 1800
    },
    {
      "epoch": 0.05777479034632866,
      "grad_norm": 0.05737105384469032,
      "learning_rate": 0.29849690151510944,
      "loss": 1.094,
      "num_input_tokens_seen": 2478640,
      "step": 1805
    },
    {
      "epoch": 0.057934831316817106,
      "grad_norm": 0.03288362920284271,
      "learning_rate": 0.2984885719771392,
      "loss": 1.2216,
      "num_input_tokens_seen": 2485520,
      "step": 1810
    },
    {
      "epoch": 0.05809487228730555,
      "grad_norm": 0.04519723355770111,
      "learning_rate": 0.2984802195403651,
      "loss": 1.0431,
      "num_input_tokens_seen": 2492400,
      "step": 1815
    },
    {
      "epoch": 0.05825491325779399,
      "grad_norm": 0.06843841820955276,
      "learning_rate": 0.2984718442060752,
      "loss": 1.1691,
      "num_input_tokens_seen": 2499136,
      "step": 1820
    },
    {
      "epoch": 0.05841495422828244,
      "grad_norm": 0.03483305498957634,
      "learning_rate": 0.2984634459755611,
      "loss": 1.0184,
      "num_input_tokens_seen": 2505632,
      "step": 1825
    },
    {
      "epoch": 0.058574995198770886,
      "grad_norm": 0.03836182877421379,
      "learning_rate": 0.29845502485011793,
      "loss": 1.0497,
      "num_input_tokens_seen": 2512672,
      "step": 1830
    },
    {
      "epoch": 0.05873503616925933,
      "grad_norm": 0.04130954295396805,
      "learning_rate": 0.2984465808310444,
      "loss": 0.9259,
      "num_input_tokens_seen": 2519568,
      "step": 1835
    },
    {
      "epoch": 0.05889507713974777,
      "grad_norm": 0.03268648684024811,
      "learning_rate": 0.29843811391964253,
      "loss": 0.8441,
      "num_input_tokens_seen": 2526448,
      "step": 1840
    },
    {
      "epoch": 0.05905511811023622,
      "grad_norm": 0.04816924035549164,
      "learning_rate": 0.2984296241172182,
      "loss": 1.1265,
      "num_input_tokens_seen": 2533696,
      "step": 1845
    },
    {
      "epoch": 0.059215159080724666,
      "grad_norm": 0.03160792216658592,
      "learning_rate": 0.29842111142508043,
      "loss": 0.926,
      "num_input_tokens_seen": 2540464,
      "step": 1850
    },
    {
      "epoch": 0.05937520005121311,
      "grad_norm": 0.03407744690775871,
      "learning_rate": 0.29841257584454217,
      "loss": 1.3151,
      "num_input_tokens_seen": 2546880,
      "step": 1855
    },
    {
      "epoch": 0.05953524102170155,
      "grad_norm": 0.04182155057787895,
      "learning_rate": 0.29840401737691963,
      "loss": 0.8387,
      "num_input_tokens_seen": 2553824,
      "step": 1860
    },
    {
      "epoch": 0.05969528199219,
      "grad_norm": 0.03527449071407318,
      "learning_rate": 0.29839543602353263,
      "loss": 0.9192,
      "num_input_tokens_seen": 2560240,
      "step": 1865
    },
    {
      "epoch": 0.059855322962678446,
      "grad_norm": 0.02996794693171978,
      "learning_rate": 0.2983868317857046,
      "loss": 1.0378,
      "num_input_tokens_seen": 2567680,
      "step": 1870
    },
    {
      "epoch": 0.06001536393316689,
      "grad_norm": 0.029317706823349,
      "learning_rate": 0.2983782046647623,
      "loss": 1.2169,
      "num_input_tokens_seen": 2574352,
      "step": 1875
    },
    {
      "epoch": 0.06017540490365533,
      "grad_norm": 0.02599935606122017,
      "learning_rate": 0.2983695546620362,
      "loss": 1.0198,
      "num_input_tokens_seen": 2581648,
      "step": 1880
    },
    {
      "epoch": 0.06033544587414378,
      "grad_norm": 0.05412604659795761,
      "learning_rate": 0.2983608817788603,
      "loss": 1.0743,
      "num_input_tokens_seen": 2588384,
      "step": 1885
    },
    {
      "epoch": 0.060495486844632226,
      "grad_norm": 0.05255366489291191,
      "learning_rate": 0.29835218601657193,
      "loss": 1.2917,
      "num_input_tokens_seen": 2595024,
      "step": 1890
    },
    {
      "epoch": 0.06065552781512067,
      "grad_norm": 0.028853043913841248,
      "learning_rate": 0.29834346737651224,
      "loss": 0.8033,
      "num_input_tokens_seen": 2601360,
      "step": 1895
    },
    {
      "epoch": 0.06081556878560911,
      "grad_norm": 0.03496192768216133,
      "learning_rate": 0.29833472586002563,
      "loss": 1.0222,
      "num_input_tokens_seen": 2608544,
      "step": 1900
    },
    {
      "epoch": 0.06097560975609756,
      "grad_norm": 0.0656491294503212,
      "learning_rate": 0.29832596146846024,
      "loss": 0.8292,
      "num_input_tokens_seen": 2615216,
      "step": 1905
    },
    {
      "epoch": 0.061135650726586006,
      "grad_norm": 0.07778309285640717,
      "learning_rate": 0.2983171742031676,
      "loss": 0.8245,
      "num_input_tokens_seen": 2621904,
      "step": 1910
    },
    {
      "epoch": 0.06129569169707445,
      "grad_norm": 0.062137678265571594,
      "learning_rate": 0.2983083640655028,
      "loss": 0.7267,
      "num_input_tokens_seen": 2628896,
      "step": 1915
    },
    {
      "epoch": 0.06145573266756289,
      "grad_norm": 0.025789188221096992,
      "learning_rate": 0.29829953105682455,
      "loss": 0.9284,
      "num_input_tokens_seen": 2635568,
      "step": 1920
    },
    {
      "epoch": 0.06161577363805134,
      "grad_norm": 0.03534629940986633,
      "learning_rate": 0.29829067517849495,
      "loss": 0.9059,
      "num_input_tokens_seen": 2642016,
      "step": 1925
    },
    {
      "epoch": 0.061775814608539786,
      "grad_norm": 0.028243888169527054,
      "learning_rate": 0.2982817964318797,
      "loss": 0.9152,
      "num_input_tokens_seen": 2648944,
      "step": 1930
    },
    {
      "epoch": 0.06193585557902823,
      "grad_norm": 0.030106090009212494,
      "learning_rate": 0.298272894818348,
      "loss": 1.1941,
      "num_input_tokens_seen": 2656016,
      "step": 1935
    },
    {
      "epoch": 0.06209589654951667,
      "grad_norm": 0.0326281301677227,
      "learning_rate": 0.2982639703392726,
      "loss": 0.8372,
      "num_input_tokens_seen": 2662992,
      "step": 1940
    },
    {
      "epoch": 0.06225593752000512,
      "grad_norm": 0.025082595646381378,
      "learning_rate": 0.29825502299602974,
      "loss": 0.9782,
      "num_input_tokens_seen": 2669776,
      "step": 1945
    },
    {
      "epoch": 0.062415978490493566,
      "grad_norm": 0.022234827280044556,
      "learning_rate": 0.2982460527899993,
      "loss": 0.6815,
      "num_input_tokens_seen": 2676656,
      "step": 1950
    },
    {
      "epoch": 0.06257601946098201,
      "grad_norm": 0.028397276997566223,
      "learning_rate": 0.29823705972256453,
      "loss": 0.7929,
      "num_input_tokens_seen": 2683552,
      "step": 1955
    },
    {
      "epoch": 0.06273606043147045,
      "grad_norm": 0.03179681673645973,
      "learning_rate": 0.2982280437951123,
      "loss": 0.8025,
      "num_input_tokens_seen": 2690400,
      "step": 1960
    },
    {
      "epoch": 0.0628961014019589,
      "grad_norm": 0.03708629310131073,
      "learning_rate": 0.298219005009033,
      "loss": 0.9256,
      "num_input_tokens_seen": 2696960,
      "step": 1965
    },
    {
      "epoch": 0.06305614237244735,
      "grad_norm": 0.022542528808116913,
      "learning_rate": 0.29820994336572043,
      "loss": 0.9926,
      "num_input_tokens_seen": 2705200,
      "step": 1970
    },
    {
      "epoch": 0.06321618334293579,
      "grad_norm": 0.03364500403404236,
      "learning_rate": 0.2982008588665721,
      "loss": 0.918,
      "num_input_tokens_seen": 2712304,
      "step": 1975
    },
    {
      "epoch": 0.06337622431342424,
      "grad_norm": 0.041539739817380905,
      "learning_rate": 0.2981917515129889,
      "loss": 1.005,
      "num_input_tokens_seen": 2719136,
      "step": 1980
    },
    {
      "epoch": 0.06353626528391268,
      "grad_norm": 0.03843523934483528,
      "learning_rate": 0.2981826213063753,
      "loss": 1.1512,
      "num_input_tokens_seen": 2726208,
      "step": 1985
    },
    {
      "epoch": 0.06369630625440113,
      "grad_norm": 0.04854675382375717,
      "learning_rate": 0.2981734682481394,
      "loss": 1.0211,
      "num_input_tokens_seen": 2733312,
      "step": 1990
    },
    {
      "epoch": 0.06385634722488957,
      "grad_norm": 0.07815606147050858,
      "learning_rate": 0.29816429233969255,
      "loss": 0.76,
      "num_input_tokens_seen": 2740032,
      "step": 1995
    },
    {
      "epoch": 0.06401638819537801,
      "grad_norm": 0.05454551801085472,
      "learning_rate": 0.2981550935824499,
      "loss": 1.1006,
      "num_input_tokens_seen": 2746752,
      "step": 2000
    },
    {
      "epoch": 0.06401638819537801,
      "eval_loss": 0.9535754919052124,
      "eval_runtime": 331.6391,
      "eval_samples_per_second": 41.871,
      "eval_steps_per_second": 20.935,
      "num_input_tokens_seen": 2746752,
      "step": 2000
    },
    {
      "epoch": 0.06417642916586647,
      "grad_norm": 0.0239472184330225,
      "learning_rate": 0.29814587197783,
      "loss": 0.8408,
      "num_input_tokens_seen": 2753872,
      "step": 2005
    },
    {
      "epoch": 0.0643364701363549,
      "grad_norm": 0.05902140960097313,
      "learning_rate": 0.29813662752725495,
      "loss": 0.9517,
      "num_input_tokens_seen": 2760624,
      "step": 2010
    },
    {
      "epoch": 0.06449651110684335,
      "grad_norm": 0.02107575163245201,
      "learning_rate": 0.29812736023215025,
      "loss": 0.9404,
      "num_input_tokens_seen": 2767104,
      "step": 2015
    },
    {
      "epoch": 0.0646565520773318,
      "grad_norm": 0.028297988697886467,
      "learning_rate": 0.29811807009394514,
      "loss": 0.6535,
      "num_input_tokens_seen": 2773776,
      "step": 2020
    },
    {
      "epoch": 0.06481659304782024,
      "grad_norm": 0.06157582998275757,
      "learning_rate": 0.2981087571140723,
      "loss": 0.9417,
      "num_input_tokens_seen": 2780768,
      "step": 2025
    },
    {
      "epoch": 0.0649766340183087,
      "grad_norm": 0.031068218871951103,
      "learning_rate": 0.2980994212939678,
      "loss": 0.9961,
      "num_input_tokens_seen": 2787568,
      "step": 2030
    },
    {
      "epoch": 0.06513667498879713,
      "grad_norm": 0.03627339377999306,
      "learning_rate": 0.2980900626350715,
      "loss": 0.8241,
      "num_input_tokens_seen": 2794320,
      "step": 2035
    },
    {
      "epoch": 0.06529671595928557,
      "grad_norm": 0.03428220376372337,
      "learning_rate": 0.29808068113882646,
      "loss": 1.1318,
      "num_input_tokens_seen": 2800912,
      "step": 2040
    },
    {
      "epoch": 0.06545675692977403,
      "grad_norm": 0.05361752584576607,
      "learning_rate": 0.2980712768066795,
      "loss": 1.0286,
      "num_input_tokens_seen": 2807920,
      "step": 2045
    },
    {
      "epoch": 0.06561679790026247,
      "grad_norm": 0.02109481580555439,
      "learning_rate": 0.2980618496400809,
      "loss": 0.8364,
      "num_input_tokens_seen": 2815360,
      "step": 2050
    },
    {
      "epoch": 0.0657768388707509,
      "grad_norm": 0.0319003090262413,
      "learning_rate": 0.2980523996404844,
      "loss": 0.9583,
      "num_input_tokens_seen": 2822240,
      "step": 2055
    },
    {
      "epoch": 0.06593687984123936,
      "grad_norm": 0.027229541912674904,
      "learning_rate": 0.2980429268093473,
      "loss": 0.8465,
      "num_input_tokens_seen": 2829360,
      "step": 2060
    },
    {
      "epoch": 0.0660969208117278,
      "grad_norm": 0.039451539516448975,
      "learning_rate": 0.29803343114813047,
      "loss": 1.0162,
      "num_input_tokens_seen": 2835824,
      "step": 2065
    },
    {
      "epoch": 0.06625696178221625,
      "grad_norm": 0.03709424287080765,
      "learning_rate": 0.2980239126582983,
      "loss": 0.8117,
      "num_input_tokens_seen": 2842416,
      "step": 2070
    },
    {
      "epoch": 0.06641700275270469,
      "grad_norm": 0.046099431812763214,
      "learning_rate": 0.2980143713413186,
      "loss": 1.1047,
      "num_input_tokens_seen": 2849360,
      "step": 2075
    },
    {
      "epoch": 0.06657704372319313,
      "grad_norm": 0.028384272009134293,
      "learning_rate": 0.29800480719866274,
      "loss": 0.8644,
      "num_input_tokens_seen": 2856464,
      "step": 2080
    },
    {
      "epoch": 0.06673708469368159,
      "grad_norm": 0.05254603549838066,
      "learning_rate": 0.2979952202318057,
      "loss": 0.714,
      "num_input_tokens_seen": 2863408,
      "step": 2085
    },
    {
      "epoch": 0.06689712566417003,
      "grad_norm": 0.059367213398218155,
      "learning_rate": 0.2979856104422259,
      "loss": 1.0758,
      "num_input_tokens_seen": 2870656,
      "step": 2090
    },
    {
      "epoch": 0.06705716663465847,
      "grad_norm": 0.025883030146360397,
      "learning_rate": 0.2979759778314052,
      "loss": 1.0638,
      "num_input_tokens_seen": 2876960,
      "step": 2095
    },
    {
      "epoch": 0.06721720760514692,
      "grad_norm": 0.02196875587105751,
      "learning_rate": 0.2979663224008292,
      "loss": 0.8358,
      "num_input_tokens_seen": 2883904,
      "step": 2100
    },
    {
      "epoch": 0.06737724857563536,
      "grad_norm": 0.034532468765974045,
      "learning_rate": 0.2979566441519868,
      "loss": 1.0923,
      "num_input_tokens_seen": 2890560,
      "step": 2105
    },
    {
      "epoch": 0.06753728954612381,
      "grad_norm": 0.021536095067858696,
      "learning_rate": 0.29794694308637054,
      "loss": 0.7522,
      "num_input_tokens_seen": 2897536,
      "step": 2110
    },
    {
      "epoch": 0.06769733051661225,
      "grad_norm": 0.026271328330039978,
      "learning_rate": 0.2979372192054764,
      "loss": 1.0051,
      "num_input_tokens_seen": 2904736,
      "step": 2115
    },
    {
      "epoch": 0.06785737148710069,
      "grad_norm": 0.019160965457558632,
      "learning_rate": 0.297927472510804,
      "loss": 1.0054,
      "num_input_tokens_seen": 2911488,
      "step": 2120
    },
    {
      "epoch": 0.06801741245758915,
      "grad_norm": 0.08792518079280853,
      "learning_rate": 0.29791770300385634,
      "loss": 1.2924,
      "num_input_tokens_seen": 2918592,
      "step": 2125
    },
    {
      "epoch": 0.06817745342807759,
      "grad_norm": 0.03166602551937103,
      "learning_rate": 0.29790791068614003,
      "loss": 1.0931,
      "num_input_tokens_seen": 2925360,
      "step": 2130
    },
    {
      "epoch": 0.06833749439856603,
      "grad_norm": 0.018563084304332733,
      "learning_rate": 0.2978980955591652,
      "loss": 0.7852,
      "num_input_tokens_seen": 2932400,
      "step": 2135
    },
    {
      "epoch": 0.06849753536905448,
      "grad_norm": 0.017069144174456596,
      "learning_rate": 0.2978882576244454,
      "loss": 0.9433,
      "num_input_tokens_seen": 2939344,
      "step": 2140
    },
    {
      "epoch": 0.06865757633954292,
      "grad_norm": 0.01615871489048004,
      "learning_rate": 0.2978783968834978,
      "loss": 1.0498,
      "num_input_tokens_seen": 2945984,
      "step": 2145
    },
    {
      "epoch": 0.06881761731003137,
      "grad_norm": 0.021614765748381615,
      "learning_rate": 0.29786851333784303,
      "loss": 0.9079,
      "num_input_tokens_seen": 2952704,
      "step": 2150
    },
    {
      "epoch": 0.06897765828051981,
      "grad_norm": 0.17688217759132385,
      "learning_rate": 0.2978586069890053,
      "loss": 0.8883,
      "num_input_tokens_seen": 2959472,
      "step": 2155
    },
    {
      "epoch": 0.06913769925100825,
      "grad_norm": 0.030204644426703453,
      "learning_rate": 0.29784867783851227,
      "loss": 0.7712,
      "num_input_tokens_seen": 2966064,
      "step": 2160
    },
    {
      "epoch": 0.0692977402214967,
      "grad_norm": 0.029287612065672874,
      "learning_rate": 0.2978387258878951,
      "loss": 0.9565,
      "num_input_tokens_seen": 2973040,
      "step": 2165
    },
    {
      "epoch": 0.06945778119198515,
      "grad_norm": 0.015305745415389538,
      "learning_rate": 0.29782875113868856,
      "loss": 0.8356,
      "num_input_tokens_seen": 2979760,
      "step": 2170
    },
    {
      "epoch": 0.0696178221624736,
      "grad_norm": 0.023467538878321648,
      "learning_rate": 0.2978187535924309,
      "loss": 0.9892,
      "num_input_tokens_seen": 2986784,
      "step": 2175
    },
    {
      "epoch": 0.06977786313296204,
      "grad_norm": 0.028152013197541237,
      "learning_rate": 0.29780873325066376,
      "loss": 1.199,
      "num_input_tokens_seen": 2993392,
      "step": 2180
    },
    {
      "epoch": 0.06993790410345048,
      "grad_norm": 0.032606903463602066,
      "learning_rate": 0.2977986901149325,
      "loss": 0.8083,
      "num_input_tokens_seen": 3000112,
      "step": 2185
    },
    {
      "epoch": 0.07009794507393893,
      "grad_norm": 0.034918393939733505,
      "learning_rate": 0.29778862418678587,
      "loss": 1.0759,
      "num_input_tokens_seen": 3006544,
      "step": 2190
    },
    {
      "epoch": 0.07025798604442737,
      "grad_norm": 0.01593155600130558,
      "learning_rate": 0.29777853546777616,
      "loss": 1.0401,
      "num_input_tokens_seen": 3013072,
      "step": 2195
    },
    {
      "epoch": 0.07041802701491581,
      "grad_norm": 0.027908844873309135,
      "learning_rate": 0.2977684239594592,
      "loss": 1.0743,
      "num_input_tokens_seen": 3020144,
      "step": 2200
    },
    {
      "epoch": 0.07041802701491581,
      "eval_loss": 0.8951613903045654,
      "eval_runtime": 332.2588,
      "eval_samples_per_second": 41.793,
      "eval_steps_per_second": 20.896,
      "num_input_tokens_seen": 3020144,
      "step": 2200
    },
    {
      "epoch": 0.07057806798540427,
      "grad_norm": 0.036959387362003326,
      "learning_rate": 0.29775828966339424,
      "loss": 0.8655,
      "num_input_tokens_seen": 3027072,
      "step": 2205
    },
    {
      "epoch": 0.0707381089558927,
      "grad_norm": 0.03517621010541916,
      "learning_rate": 0.29774813258114424,
      "loss": 0.8903,
      "num_input_tokens_seen": 3033808,
      "step": 2210
    },
    {
      "epoch": 0.07089814992638116,
      "grad_norm": 0.02097235806286335,
      "learning_rate": 0.29773795271427544,
      "loss": 0.7236,
      "num_input_tokens_seen": 3040752,
      "step": 2215
    },
    {
      "epoch": 0.0710581908968696,
      "grad_norm": 0.02903757244348526,
      "learning_rate": 0.2977277500643577,
      "loss": 0.8927,
      "num_input_tokens_seen": 3047696,
      "step": 2220
    },
    {
      "epoch": 0.07121823186735804,
      "grad_norm": 0.02130020782351494,
      "learning_rate": 0.29771752463296447,
      "loss": 0.7262,
      "num_input_tokens_seen": 3054656,
      "step": 2225
    },
    {
      "epoch": 0.0713782728378465,
      "grad_norm": 0.018406519666314125,
      "learning_rate": 0.29770727642167266,
      "loss": 0.8816,
      "num_input_tokens_seen": 3061904,
      "step": 2230
    },
    {
      "epoch": 0.07153831380833493,
      "grad_norm": 0.025671251118183136,
      "learning_rate": 0.29769700543206257,
      "loss": 0.943,
      "num_input_tokens_seen": 3068704,
      "step": 2235
    },
    {
      "epoch": 0.07169835477882337,
      "grad_norm": 0.02854747325181961,
      "learning_rate": 0.2976867116657182,
      "loss": 0.8477,
      "num_input_tokens_seen": 3075280,
      "step": 2240
    },
    {
      "epoch": 0.07185839574931183,
      "grad_norm": 0.02015259675681591,
      "learning_rate": 0.2976763951242269,
      "loss": 0.7155,
      "num_input_tokens_seen": 3082384,
      "step": 2245
    },
    {
      "epoch": 0.07201843671980027,
      "grad_norm": 0.012468107044696808,
      "learning_rate": 0.29766605580917965,
      "loss": 0.9537,
      "num_input_tokens_seen": 3088848,
      "step": 2250
    },
    {
      "epoch": 0.07217847769028872,
      "grad_norm": 0.018722545355558395,
      "learning_rate": 0.29765569372217093,
      "loss": 0.8839,
      "num_input_tokens_seen": 3095648,
      "step": 2255
    },
    {
      "epoch": 0.07233851866077716,
      "grad_norm": 0.019184235483407974,
      "learning_rate": 0.2976453088647987,
      "loss": 0.831,
      "num_input_tokens_seen": 3102528,
      "step": 2260
    },
    {
      "epoch": 0.0724985596312656,
      "grad_norm": 0.017950115725398064,
      "learning_rate": 0.2976349012386644,
      "loss": 0.6468,
      "num_input_tokens_seen": 3108976,
      "step": 2265
    },
    {
      "epoch": 0.07265860060175405,
      "grad_norm": 0.022605376318097115,
      "learning_rate": 0.29762447084537297,
      "loss": 0.9304,
      "num_input_tokens_seen": 3115872,
      "step": 2270
    },
    {
      "epoch": 0.07281864157224249,
      "grad_norm": 0.02807733044028282,
      "learning_rate": 0.29761401768653306,
      "loss": 0.8816,
      "num_input_tokens_seen": 3122528,
      "step": 2275
    },
    {
      "epoch": 0.07297868254273093,
      "grad_norm": 0.025005780160427094,
      "learning_rate": 0.29760354176375653,
      "loss": 1.1423,
      "num_input_tokens_seen": 3130256,
      "step": 2280
    },
    {
      "epoch": 0.07313872351321939,
      "grad_norm": 0.025095868855714798,
      "learning_rate": 0.29759304307865897,
      "loss": 0.895,
      "num_input_tokens_seen": 3137312,
      "step": 2285
    },
    {
      "epoch": 0.07329876448370783,
      "grad_norm": 0.031532805413007736,
      "learning_rate": 0.2975825216328594,
      "loss": 0.9815,
      "num_input_tokens_seen": 3144192,
      "step": 2290
    },
    {
      "epoch": 0.07345880545419628,
      "grad_norm": 0.03918134421110153,
      "learning_rate": 0.2975719774279804,
      "loss": 1.0028,
      "num_input_tokens_seen": 3151296,
      "step": 2295
    },
    {
      "epoch": 0.07361884642468472,
      "grad_norm": 0.04288696497678757,
      "learning_rate": 0.29756141046564794,
      "loss": 0.6378,
      "num_input_tokens_seen": 3158384,
      "step": 2300
    },
    {
      "epoch": 0.07377888739517316,
      "grad_norm": 0.025210021063685417,
      "learning_rate": 0.2975508207474916,
      "loss": 0.9427,
      "num_input_tokens_seen": 3165136,
      "step": 2305
    },
    {
      "epoch": 0.07393892836566161,
      "grad_norm": 0.02535681053996086,
      "learning_rate": 0.2975402082751445,
      "loss": 0.8712,
      "num_input_tokens_seen": 3172464,
      "step": 2310
    },
    {
      "epoch": 0.07409896933615005,
      "grad_norm": 0.02362346649169922,
      "learning_rate": 0.29752957305024313,
      "loss": 0.9546,
      "num_input_tokens_seen": 3179472,
      "step": 2315
    },
    {
      "epoch": 0.07425901030663849,
      "grad_norm": 0.020337892696261406,
      "learning_rate": 0.2975189150744277,
      "loss": 0.904,
      "num_input_tokens_seen": 3186512,
      "step": 2320
    },
    {
      "epoch": 0.07441905127712695,
      "grad_norm": 0.02863241918385029,
      "learning_rate": 0.29750823434934165,
      "loss": 0.8054,
      "num_input_tokens_seen": 3193440,
      "step": 2325
    },
    {
      "epoch": 0.07457909224761539,
      "grad_norm": 0.014944222755730152,
      "learning_rate": 0.29749753087663217,
      "loss": 0.9121,
      "num_input_tokens_seen": 3200480,
      "step": 2330
    },
    {
      "epoch": 0.07473913321810384,
      "grad_norm": 0.04658471792936325,
      "learning_rate": 0.29748680465794985,
      "loss": 0.8874,
      "num_input_tokens_seen": 3207024,
      "step": 2335
    },
    {
      "epoch": 0.07489917418859228,
      "grad_norm": 0.02366526611149311,
      "learning_rate": 0.29747605569494884,
      "loss": 0.7957,
      "num_input_tokens_seen": 3213536,
      "step": 2340
    },
    {
      "epoch": 0.07505921515908072,
      "grad_norm": 0.021311555057764053,
      "learning_rate": 0.29746528398928673,
      "loss": 0.8716,
      "num_input_tokens_seen": 3220192,
      "step": 2345
    },
    {
      "epoch": 0.07521925612956917,
      "grad_norm": 0.03303728625178337,
      "learning_rate": 0.2974544895426247,
      "loss": 0.9614,
      "num_input_tokens_seen": 3227088,
      "step": 2350
    },
    {
      "epoch": 0.07537929710005761,
      "grad_norm": 0.037857428193092346,
      "learning_rate": 0.29744367235662733,
      "loss": 0.8174,
      "num_input_tokens_seen": 3233824,
      "step": 2355
    },
    {
      "epoch": 0.07553933807054607,
      "grad_norm": 0.017725912854075432,
      "learning_rate": 0.29743283243296276,
      "loss": 0.6383,
      "num_input_tokens_seen": 3240736,
      "step": 2360
    },
    {
      "epoch": 0.0756993790410345,
      "grad_norm": 0.039801765233278275,
      "learning_rate": 0.29742196977330276,
      "loss": 0.7966,
      "num_input_tokens_seen": 3247392,
      "step": 2365
    },
    {
      "epoch": 0.07585942001152295,
      "grad_norm": 0.03519587591290474,
      "learning_rate": 0.2974110843793223,
      "loss": 0.8047,
      "num_input_tokens_seen": 3254160,
      "step": 2370
    },
    {
      "epoch": 0.0760194609820114,
      "grad_norm": 0.022199038416147232,
      "learning_rate": 0.2974001762527002,
      "loss": 0.8735,
      "num_input_tokens_seen": 3260736,
      "step": 2375
    },
    {
      "epoch": 0.07617950195249984,
      "grad_norm": 0.021532388404011726,
      "learning_rate": 0.2973892453951186,
      "loss": 0.7652,
      "num_input_tokens_seen": 3268624,
      "step": 2380
    },
    {
      "epoch": 0.07633954292298828,
      "grad_norm": 0.04304944723844528,
      "learning_rate": 0.2973782918082631,
      "loss": 1.1548,
      "num_input_tokens_seen": 3276048,
      "step": 2385
    },
    {
      "epoch": 0.07649958389347673,
      "grad_norm": 0.019212951883673668,
      "learning_rate": 0.29736731549382295,
      "loss": 0.9158,
      "num_input_tokens_seen": 3282912,
      "step": 2390
    },
    {
      "epoch": 0.07665962486396517,
      "grad_norm": 0.024079786613583565,
      "learning_rate": 0.2973563164534908,
      "loss": 0.9954,
      "num_input_tokens_seen": 3290112,
      "step": 2395
    },
    {
      "epoch": 0.07681966583445363,
      "grad_norm": 0.07569558918476105,
      "learning_rate": 0.29734529468896287,
      "loss": 1.0446,
      "num_input_tokens_seen": 3296624,
      "step": 2400
    },
    {
      "epoch": 0.07681966583445363,
      "eval_loss": 0.895025908946991,
      "eval_runtime": 331.7198,
      "eval_samples_per_second": 41.861,
      "eval_steps_per_second": 20.93,
      "num_input_tokens_seen": 3296624,
      "step": 2400
    },
    {
      "epoch": 0.07697970680494207,
      "grad_norm": 0.07982786744832993,
      "learning_rate": 0.2973342502019388,
      "loss": 1.1316,
      "num_input_tokens_seen": 3303216,
      "step": 2405
    },
    {
      "epoch": 0.0771397477754305,
      "grad_norm": 0.02651214972138405,
      "learning_rate": 0.2973231829941219,
      "loss": 0.8856,
      "num_input_tokens_seen": 3309952,
      "step": 2410
    },
    {
      "epoch": 0.07729978874591896,
      "grad_norm": 0.09073583781719208,
      "learning_rate": 0.2973120930672188,
      "loss": 0.931,
      "num_input_tokens_seen": 3316688,
      "step": 2415
    },
    {
      "epoch": 0.0774598297164074,
      "grad_norm": 0.04646508768200874,
      "learning_rate": 0.2973009804229397,
      "loss": 0.8392,
      "num_input_tokens_seen": 3323840,
      "step": 2420
    },
    {
      "epoch": 0.07761987068689584,
      "grad_norm": 0.027831653133034706,
      "learning_rate": 0.29728984506299827,
      "loss": 0.9281,
      "num_input_tokens_seen": 3330864,
      "step": 2425
    },
    {
      "epoch": 0.07777991165738429,
      "grad_norm": 0.04295162484049797,
      "learning_rate": 0.2972786869891118,
      "loss": 0.9991,
      "num_input_tokens_seen": 3337520,
      "step": 2430
    },
    {
      "epoch": 0.07793995262787273,
      "grad_norm": 0.028936225920915604,
      "learning_rate": 0.29726750620300096,
      "loss": 0.7396,
      "num_input_tokens_seen": 3344896,
      "step": 2435
    },
    {
      "epoch": 0.07809999359836119,
      "grad_norm": 0.048804569989442825,
      "learning_rate": 0.29725630270639003,
      "loss": 0.7596,
      "num_input_tokens_seen": 3351584,
      "step": 2440
    },
    {
      "epoch": 0.07826003456884963,
      "grad_norm": 0.11323976516723633,
      "learning_rate": 0.2972450765010067,
      "loss": 0.9127,
      "num_input_tokens_seen": 3358528,
      "step": 2445
    },
    {
      "epoch": 0.07842007553933807,
      "grad_norm": 0.024686379358172417,
      "learning_rate": 0.29723382758858213,
      "loss": 0.9964,
      "num_input_tokens_seen": 3364960,
      "step": 2450
    },
    {
      "epoch": 0.07858011650982652,
      "grad_norm": 0.03159874677658081,
      "learning_rate": 0.29722255597085107,
      "loss": 0.8593,
      "num_input_tokens_seen": 3371904,
      "step": 2455
    },
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 0.044604748487472534,
      "learning_rate": 0.2972112616495518,
      "loss": 0.8718,
      "num_input_tokens_seen": 3379008,
      "step": 2460
    },
    {
      "epoch": 0.0789001984508034,
      "grad_norm": 0.038417693227529526,
      "learning_rate": 0.297199944626426,
      "loss": 1.0208,
      "num_input_tokens_seen": 3385952,
      "step": 2465
    },
    {
      "epoch": 0.07906023942129185,
      "grad_norm": 0.07025646418333054,
      "learning_rate": 0.2971886049032189,
      "loss": 1.1566,
      "num_input_tokens_seen": 3392960,
      "step": 2470
    },
    {
      "epoch": 0.07922028039178029,
      "grad_norm": 0.04576229304075241,
      "learning_rate": 0.29717724248167926,
      "loss": 0.8748,
      "num_input_tokens_seen": 3400048,
      "step": 2475
    },
    {
      "epoch": 0.07938032136226875,
      "grad_norm": 0.05086356773972511,
      "learning_rate": 0.29716585736355927,
      "loss": 0.8133,
      "num_input_tokens_seen": 3406576,
      "step": 2480
    },
    {
      "epoch": 0.07954036233275719,
      "grad_norm": 0.04543200135231018,
      "learning_rate": 0.2971544495506147,
      "loss": 0.8745,
      "num_input_tokens_seen": 3413232,
      "step": 2485
    },
    {
      "epoch": 0.07970040330324563,
      "grad_norm": 0.03134472668170929,
      "learning_rate": 0.2971430190446048,
      "loss": 0.9493,
      "num_input_tokens_seen": 3420400,
      "step": 2490
    },
    {
      "epoch": 0.07986044427373408,
      "grad_norm": 0.08303499221801758,
      "learning_rate": 0.2971315658472921,
      "loss": 1.1184,
      "num_input_tokens_seen": 3427680,
      "step": 2495
    },
    {
      "epoch": 0.08002048524422252,
      "grad_norm": 0.053759701550006866,
      "learning_rate": 0.2971200899604431,
      "loss": 0.9793,
      "num_input_tokens_seen": 3434400,
      "step": 2500
    },
    {
      "epoch": 0.08018052621471096,
      "grad_norm": 0.020210757851600647,
      "learning_rate": 0.29710859138582735,
      "loss": 0.746,
      "num_input_tokens_seen": 3441200,
      "step": 2505
    },
    {
      "epoch": 0.08034056718519941,
      "grad_norm": 0.02629619836807251,
      "learning_rate": 0.29709707012521813,
      "loss": 0.918,
      "num_input_tokens_seen": 3448096,
      "step": 2510
    },
    {
      "epoch": 0.08050060815568785,
      "grad_norm": 0.03507111594080925,
      "learning_rate": 0.29708552618039213,
      "loss": 0.9939,
      "num_input_tokens_seen": 3454832,
      "step": 2515
    },
    {
      "epoch": 0.0806606491261763,
      "grad_norm": 0.0183506291359663,
      "learning_rate": 0.2970739595531296,
      "loss": 0.8797,
      "num_input_tokens_seen": 3461504,
      "step": 2520
    },
    {
      "epoch": 0.08082069009666475,
      "grad_norm": 0.020519349724054337,
      "learning_rate": 0.2970623702452143,
      "loss": 0.8675,
      "num_input_tokens_seen": 3468208,
      "step": 2525
    },
    {
      "epoch": 0.08098073106715319,
      "grad_norm": 0.029565978795289993,
      "learning_rate": 0.2970507582584334,
      "loss": 0.7745,
      "num_input_tokens_seen": 3474992,
      "step": 2530
    },
    {
      "epoch": 0.08114077203764164,
      "grad_norm": 0.15644314885139465,
      "learning_rate": 0.2970391235945776,
      "loss": 1.038,
      "num_input_tokens_seen": 3482528,
      "step": 2535
    },
    {
      "epoch": 0.08130081300813008,
      "grad_norm": 0.0274698156863451,
      "learning_rate": 0.2970274662554412,
      "loss": 0.9003,
      "num_input_tokens_seen": 3489168,
      "step": 2540
    },
    {
      "epoch": 0.08146085397861853,
      "grad_norm": 0.022678952664136887,
      "learning_rate": 0.2970157862428218,
      "loss": 0.9722,
      "num_input_tokens_seen": 3495824,
      "step": 2545
    },
    {
      "epoch": 0.08162089494910697,
      "grad_norm": 0.024003585800528526,
      "learning_rate": 0.2970040835585206,
      "loss": 0.846,
      "num_input_tokens_seen": 3502448,
      "step": 2550
    },
    {
      "epoch": 0.08178093591959541,
      "grad_norm": 0.0988493338227272,
      "learning_rate": 0.2969923582043424,
      "loss": 0.9112,
      "num_input_tokens_seen": 3509072,
      "step": 2555
    },
    {
      "epoch": 0.08194097689008387,
      "grad_norm": 0.01857171393930912,
      "learning_rate": 0.2969806101820953,
      "loss": 1.0251,
      "num_input_tokens_seen": 3516160,
      "step": 2560
    },
    {
      "epoch": 0.0821010178605723,
      "grad_norm": 0.024435564875602722,
      "learning_rate": 0.2969688394935911,
      "loss": 0.6729,
      "num_input_tokens_seen": 3522704,
      "step": 2565
    },
    {
      "epoch": 0.08226105883106075,
      "grad_norm": 0.027200905606150627,
      "learning_rate": 0.2969570461406449,
      "loss": 0.7667,
      "num_input_tokens_seen": 3529856,
      "step": 2570
    },
    {
      "epoch": 0.0824210998015492,
      "grad_norm": 0.037611573934555054,
      "learning_rate": 0.29694523012507534,
      "loss": 0.9623,
      "num_input_tokens_seen": 3536912,
      "step": 2575
    },
    {
      "epoch": 0.08258114077203764,
      "grad_norm": 0.01947665959596634,
      "learning_rate": 0.2969333914487048,
      "loss": 0.8103,
      "num_input_tokens_seen": 3544480,
      "step": 2580
    },
    {
      "epoch": 0.0827411817425261,
      "grad_norm": 0.03202647343277931,
      "learning_rate": 0.2969215301133587,
      "loss": 0.7606,
      "num_input_tokens_seen": 3551280,
      "step": 2585
    },
    {
      "epoch": 0.08290122271301453,
      "grad_norm": 0.0348396971821785,
      "learning_rate": 0.29690964612086634,
      "loss": 0.7794,
      "num_input_tokens_seen": 3558288,
      "step": 2590
    },
    {
      "epoch": 0.08306126368350297,
      "grad_norm": 0.04154239594936371,
      "learning_rate": 0.2968977394730604,
      "loss": 1.1266,
      "num_input_tokens_seen": 3564992,
      "step": 2595
    },
    {
      "epoch": 0.08322130465399143,
      "grad_norm": 0.04206286370754242,
      "learning_rate": 0.296885810171777,
      "loss": 0.9012,
      "num_input_tokens_seen": 3571808,
      "step": 2600
    },
    {
      "epoch": 0.08322130465399143,
      "eval_loss": 0.9014667272567749,
      "eval_runtime": 331.5337,
      "eval_samples_per_second": 41.884,
      "eval_steps_per_second": 20.942,
      "num_input_tokens_seen": 3571808,
      "step": 2600
    },
    {
      "epoch": 0.08338134562447987,
      "grad_norm": 0.02868243306875229,
      "learning_rate": 0.2968738582188558,
      "loss": 0.8794,
      "num_input_tokens_seen": 3578656,
      "step": 2605
    },
    {
      "epoch": 0.0835413865949683,
      "grad_norm": 0.01994793303310871,
      "learning_rate": 0.2968618836161399,
      "loss": 0.7331,
      "num_input_tokens_seen": 3585776,
      "step": 2610
    },
    {
      "epoch": 0.08370142756545676,
      "grad_norm": 0.02382545918226242,
      "learning_rate": 0.296849886365476,
      "loss": 0.781,
      "num_input_tokens_seen": 3592992,
      "step": 2615
    },
    {
      "epoch": 0.0838614685359452,
      "grad_norm": 0.04853321984410286,
      "learning_rate": 0.2968378664687142,
      "loss": 0.8499,
      "num_input_tokens_seen": 3599904,
      "step": 2620
    },
    {
      "epoch": 0.08402150950643365,
      "grad_norm": 0.02400670200586319,
      "learning_rate": 0.296825823927708,
      "loss": 0.9969,
      "num_input_tokens_seen": 3607088,
      "step": 2625
    },
    {
      "epoch": 0.08418155047692209,
      "grad_norm": 0.03273465111851692,
      "learning_rate": 0.29681375874431476,
      "loss": 0.9157,
      "num_input_tokens_seen": 3614016,
      "step": 2630
    },
    {
      "epoch": 0.08434159144741053,
      "grad_norm": 0.03211615979671478,
      "learning_rate": 0.29680167092039483,
      "loss": 0.6798,
      "num_input_tokens_seen": 3620544,
      "step": 2635
    },
    {
      "epoch": 0.08450163241789899,
      "grad_norm": 0.05813964828848839,
      "learning_rate": 0.2967895604578125,
      "loss": 0.9075,
      "num_input_tokens_seen": 3628080,
      "step": 2640
    },
    {
      "epoch": 0.08466167338838743,
      "grad_norm": 0.1306462436914444,
      "learning_rate": 0.2967774273584352,
      "loss": 1.026,
      "num_input_tokens_seen": 3635264,
      "step": 2645
    },
    {
      "epoch": 0.08482171435887587,
      "grad_norm": 0.08853847533464432,
      "learning_rate": 0.2967652716241342,
      "loss": 0.9208,
      "num_input_tokens_seen": 3642304,
      "step": 2650
    },
    {
      "epoch": 0.08498175532936432,
      "grad_norm": 0.027957109734416008,
      "learning_rate": 0.29675309325678384,
      "loss": 0.8072,
      "num_input_tokens_seen": 3648832,
      "step": 2655
    },
    {
      "epoch": 0.08514179629985276,
      "grad_norm": 0.04962674155831337,
      "learning_rate": 0.29674089225826233,
      "loss": 0.9691,
      "num_input_tokens_seen": 3655568,
      "step": 2660
    },
    {
      "epoch": 0.08530183727034121,
      "grad_norm": 0.0281141996383667,
      "learning_rate": 0.29672866863045116,
      "loss": 0.7233,
      "num_input_tokens_seen": 3662032,
      "step": 2665
    },
    {
      "epoch": 0.08546187824082965,
      "grad_norm": 0.03938557952642441,
      "learning_rate": 0.2967164223752354,
      "loss": 0.9759,
      "num_input_tokens_seen": 3669088,
      "step": 2670
    },
    {
      "epoch": 0.08562191921131809,
      "grad_norm": 0.03567596524953842,
      "learning_rate": 0.2967041534945035,
      "loss": 0.8097,
      "num_input_tokens_seen": 3676272,
      "step": 2675
    },
    {
      "epoch": 0.08578196018180655,
      "grad_norm": 0.0685010701417923,
      "learning_rate": 0.2966918619901476,
      "loss": 1.2875,
      "num_input_tokens_seen": 3683184,
      "step": 2680
    },
    {
      "epoch": 0.08594200115229499,
      "grad_norm": 0.037775617092847824,
      "learning_rate": 0.2966795478640631,
      "loss": 0.9176,
      "num_input_tokens_seen": 3689872,
      "step": 2685
    },
    {
      "epoch": 0.08610204212278343,
      "grad_norm": 0.025037609040737152,
      "learning_rate": 0.29666721111814903,
      "loss": 1.0326,
      "num_input_tokens_seen": 3696432,
      "step": 2690
    },
    {
      "epoch": 0.08626208309327188,
      "grad_norm": 0.021003074944019318,
      "learning_rate": 0.2966548517543079,
      "loss": 0.9613,
      "num_input_tokens_seen": 3703648,
      "step": 2695
    },
    {
      "epoch": 0.08642212406376032,
      "grad_norm": 0.03006921522319317,
      "learning_rate": 0.29664246977444564,
      "loss": 0.954,
      "num_input_tokens_seen": 3710256,
      "step": 2700
    },
    {
      "epoch": 0.08658216503424877,
      "grad_norm": 0.023586994037032127,
      "learning_rate": 0.2966300651804717,
      "loss": 0.918,
      "num_input_tokens_seen": 3716992,
      "step": 2705
    },
    {
      "epoch": 0.08674220600473721,
      "grad_norm": 0.04569338634610176,
      "learning_rate": 0.296617637974299,
      "loss": 0.7417,
      "num_input_tokens_seen": 3723648,
      "step": 2710
    },
    {
      "epoch": 0.08690224697522565,
      "grad_norm": 0.036224089562892914,
      "learning_rate": 0.2966051881578441,
      "loss": 0.8354,
      "num_input_tokens_seen": 3730240,
      "step": 2715
    },
    {
      "epoch": 0.0870622879457141,
      "grad_norm": 0.03659806773066521,
      "learning_rate": 0.29659271573302676,
      "loss": 0.9505,
      "num_input_tokens_seen": 3737456,
      "step": 2720
    },
    {
      "epoch": 0.08722232891620255,
      "grad_norm": 0.022218452766537666,
      "learning_rate": 0.2965802207017705,
      "loss": 0.7381,
      "num_input_tokens_seen": 3744528,
      "step": 2725
    },
    {
      "epoch": 0.087382369886691,
      "grad_norm": 0.034058406949043274,
      "learning_rate": 0.2965677030660021,
      "loss": 1.055,
      "num_input_tokens_seen": 3751696,
      "step": 2730
    },
    {
      "epoch": 0.08754241085717944,
      "grad_norm": 0.03282810375094414,
      "learning_rate": 0.2965551628276521,
      "loss": 0.7754,
      "num_input_tokens_seen": 3758176,
      "step": 2735
    },
    {
      "epoch": 0.08770245182766788,
      "grad_norm": 0.04164854809641838,
      "learning_rate": 0.29654259998865423,
      "loss": 0.9314,
      "num_input_tokens_seen": 3765568,
      "step": 2740
    },
    {
      "epoch": 0.08786249279815633,
      "grad_norm": 0.03136226162314415,
      "learning_rate": 0.2965300145509458,
      "loss": 0.8162,
      "num_input_tokens_seen": 3772064,
      "step": 2745
    },
    {
      "epoch": 0.08802253376864477,
      "grad_norm": 0.01932949759066105,
      "learning_rate": 0.2965174065164678,
      "loss": 0.7528,
      "num_input_tokens_seen": 3779232,
      "step": 2750
    },
    {
      "epoch": 0.08818257473913321,
      "grad_norm": 0.020621245726943016,
      "learning_rate": 0.2965047758871644,
      "loss": 0.9765,
      "num_input_tokens_seen": 3785728,
      "step": 2755
    },
    {
      "epoch": 0.08834261570962167,
      "grad_norm": 0.025927433744072914,
      "learning_rate": 0.2964921226649835,
      "loss": 0.8323,
      "num_input_tokens_seen": 3792528,
      "step": 2760
    },
    {
      "epoch": 0.0885026566801101,
      "grad_norm": 0.032827094197273254,
      "learning_rate": 0.2964794468518763,
      "loss": 0.8333,
      "num_input_tokens_seen": 3799376,
      "step": 2765
    },
    {
      "epoch": 0.08866269765059856,
      "grad_norm": 0.0167847853153944,
      "learning_rate": 0.2964667484497977,
      "loss": 0.7174,
      "num_input_tokens_seen": 3806128,
      "step": 2770
    },
    {
      "epoch": 0.088822738621087,
      "grad_norm": 0.0198505911976099,
      "learning_rate": 0.29645402746070587,
      "loss": 0.9282,
      "num_input_tokens_seen": 3812896,
      "step": 2775
    },
    {
      "epoch": 0.08898277959157544,
      "grad_norm": 0.02394021488726139,
      "learning_rate": 0.2964412838865625,
      "loss": 0.8232,
      "num_input_tokens_seen": 3819648,
      "step": 2780
    },
    {
      "epoch": 0.08914282056206389,
      "grad_norm": 0.04552878066897392,
      "learning_rate": 0.29642851772933293,
      "loss": 1.2415,
      "num_input_tokens_seen": 3826352,
      "step": 2785
    },
    {
      "epoch": 0.08930286153255233,
      "grad_norm": 0.0195161160081625,
      "learning_rate": 0.29641572899098567,
      "loss": 0.8286,
      "num_input_tokens_seen": 3833312,
      "step": 2790
    },
    {
      "epoch": 0.08946290250304077,
      "grad_norm": 0.039463143795728683,
      "learning_rate": 0.29640291767349314,
      "loss": 1.1187,
      "num_input_tokens_seen": 3840384,
      "step": 2795
    },
    {
      "epoch": 0.08962294347352923,
      "grad_norm": 0.03657103329896927,
      "learning_rate": 0.2963900837788308,
      "loss": 1.1622,
      "num_input_tokens_seen": 3847184,
      "step": 2800
    },
    {
      "epoch": 0.08962294347352923,
      "eval_loss": 0.8871753215789795,
      "eval_runtime": 331.659,
      "eval_samples_per_second": 41.868,
      "eval_steps_per_second": 20.934,
      "num_input_tokens_seen": 3847184,
      "step": 2800
    },
    {
      "epoch": 0.08978298444401767,
      "grad_norm": 0.031838662922382355,
      "learning_rate": 0.2963772273089779,
      "loss": 0.7837,
      "num_input_tokens_seen": 3853904,
      "step": 2805
    },
    {
      "epoch": 0.08994302541450612,
      "grad_norm": 0.02805480733513832,
      "learning_rate": 0.2963643482659171,
      "loss": 0.8607,
      "num_input_tokens_seen": 3860832,
      "step": 2810
    },
    {
      "epoch": 0.09010306638499456,
      "grad_norm": 0.01604062132537365,
      "learning_rate": 0.2963514466516345,
      "loss": 0.8037,
      "num_input_tokens_seen": 3867488,
      "step": 2815
    },
    {
      "epoch": 0.090263107355483,
      "grad_norm": 0.03517954424023628,
      "learning_rate": 0.2963385224681196,
      "loss": 0.8483,
      "num_input_tokens_seen": 3874480,
      "step": 2820
    },
    {
      "epoch": 0.09042314832597145,
      "grad_norm": 0.02101394534111023,
      "learning_rate": 0.29632557571736556,
      "loss": 0.9105,
      "num_input_tokens_seen": 3881248,
      "step": 2825
    },
    {
      "epoch": 0.09058318929645989,
      "grad_norm": 0.02499464713037014,
      "learning_rate": 0.2963126064013689,
      "loss": 0.9939,
      "num_input_tokens_seen": 3888464,
      "step": 2830
    },
    {
      "epoch": 0.09074323026694833,
      "grad_norm": 0.01678423210978508,
      "learning_rate": 0.29629961452212966,
      "loss": 0.793,
      "num_input_tokens_seen": 3895296,
      "step": 2835
    },
    {
      "epoch": 0.09090327123743679,
      "grad_norm": 0.023224307224154472,
      "learning_rate": 0.2962866000816513,
      "loss": 0.9593,
      "num_input_tokens_seen": 3901968,
      "step": 2840
    },
    {
      "epoch": 0.09106331220792523,
      "grad_norm": 0.036121875047683716,
      "learning_rate": 0.2962735630819409,
      "loss": 0.8046,
      "num_input_tokens_seen": 3908864,
      "step": 2845
    },
    {
      "epoch": 0.09122335317841368,
      "grad_norm": 0.10176915675401688,
      "learning_rate": 0.2962605035250089,
      "loss": 0.9504,
      "num_input_tokens_seen": 3915536,
      "step": 2850
    },
    {
      "epoch": 0.09138339414890212,
      "grad_norm": 0.027835194021463394,
      "learning_rate": 0.29624742141286914,
      "loss": 0.8273,
      "num_input_tokens_seen": 3922464,
      "step": 2855
    },
    {
      "epoch": 0.09154343511939056,
      "grad_norm": 0.0211714468896389,
      "learning_rate": 0.29623431674753925,
      "loss": 0.7247,
      "num_input_tokens_seen": 3928992,
      "step": 2860
    },
    {
      "epoch": 0.09170347608987901,
      "grad_norm": 0.011989237740635872,
      "learning_rate": 0.29622118953103993,
      "loss": 0.8122,
      "num_input_tokens_seen": 3936112,
      "step": 2865
    },
    {
      "epoch": 0.09186351706036745,
      "grad_norm": 0.019058970734477043,
      "learning_rate": 0.2962080397653957,
      "loss": 0.9033,
      "num_input_tokens_seen": 3942848,
      "step": 2870
    },
    {
      "epoch": 0.09202355803085589,
      "grad_norm": 0.0197038184851408,
      "learning_rate": 0.29619486745263435,
      "loss": 0.9826,
      "num_input_tokens_seen": 3949856,
      "step": 2875
    },
    {
      "epoch": 0.09218359900134435,
      "grad_norm": 0.019994515925645828,
      "learning_rate": 0.2961816725947873,
      "loss": 1.0092,
      "num_input_tokens_seen": 3956240,
      "step": 2880
    },
    {
      "epoch": 0.09234363997183279,
      "grad_norm": 0.01108331885188818,
      "learning_rate": 0.29616845519388924,
      "loss": 0.8724,
      "num_input_tokens_seen": 3963440,
      "step": 2885
    },
    {
      "epoch": 0.09250368094232124,
      "grad_norm": 0.02833056077361107,
      "learning_rate": 0.2961552152519785,
      "loss": 0.6931,
      "num_input_tokens_seen": 3970000,
      "step": 2890
    },
    {
      "epoch": 0.09266372191280968,
      "grad_norm": 0.025135723873972893,
      "learning_rate": 0.29614195277109695,
      "loss": 0.7218,
      "num_input_tokens_seen": 3976720,
      "step": 2895
    },
    {
      "epoch": 0.09282376288329812,
      "grad_norm": 0.028774382546544075,
      "learning_rate": 0.2961286677532897,
      "loss": 0.8355,
      "num_input_tokens_seen": 3983280,
      "step": 2900
    },
    {
      "epoch": 0.09298380385378657,
      "grad_norm": 0.025496425107121468,
      "learning_rate": 0.2961153602006055,
      "loss": 0.9355,
      "num_input_tokens_seen": 3990096,
      "step": 2905
    },
    {
      "epoch": 0.09314384482427501,
      "grad_norm": 0.04050411283969879,
      "learning_rate": 0.29610203011509656,
      "loss": 0.8148,
      "num_input_tokens_seen": 3997152,
      "step": 2910
    },
    {
      "epoch": 0.09330388579476345,
      "grad_norm": 0.024634752422571182,
      "learning_rate": 0.29608867749881856,
      "loss": 0.6357,
      "num_input_tokens_seen": 4003792,
      "step": 2915
    },
    {
      "epoch": 0.0934639267652519,
      "grad_norm": 0.028475254774093628,
      "learning_rate": 0.29607530235383067,
      "loss": 0.7625,
      "num_input_tokens_seen": 4010704,
      "step": 2920
    },
    {
      "epoch": 0.09362396773574035,
      "grad_norm": 0.020152267068624496,
      "learning_rate": 0.2960619046821954,
      "loss": 1.0075,
      "num_input_tokens_seen": 4017264,
      "step": 2925
    },
    {
      "epoch": 0.0937840087062288,
      "grad_norm": 0.0196086373180151,
      "learning_rate": 0.2960484844859789,
      "loss": 0.8422,
      "num_input_tokens_seen": 4023920,
      "step": 2930
    },
    {
      "epoch": 0.09394404967671724,
      "grad_norm": 0.016282029449939728,
      "learning_rate": 0.29603504176725076,
      "loss": 0.8319,
      "num_input_tokens_seen": 4030720,
      "step": 2935
    },
    {
      "epoch": 0.09410409064720568,
      "grad_norm": 0.01432620920240879,
      "learning_rate": 0.296021576528084,
      "loss": 0.7597,
      "num_input_tokens_seen": 4037376,
      "step": 2940
    },
    {
      "epoch": 0.09426413161769413,
      "grad_norm": 0.025762319564819336,
      "learning_rate": 0.29600808877055507,
      "loss": 0.8781,
      "num_input_tokens_seen": 4044352,
      "step": 2945
    },
    {
      "epoch": 0.09442417258818257,
      "grad_norm": 0.018573876470327377,
      "learning_rate": 0.29599457849674404,
      "loss": 0.6757,
      "num_input_tokens_seen": 4051264,
      "step": 2950
    },
    {
      "epoch": 0.09458421355867103,
      "grad_norm": 0.024991434067487717,
      "learning_rate": 0.2959810457087343,
      "loss": 0.9095,
      "num_input_tokens_seen": 4058032,
      "step": 2955
    },
    {
      "epoch": 0.09474425452915947,
      "grad_norm": 0.01955263875424862,
      "learning_rate": 0.2959674904086128,
      "loss": 0.7309,
      "num_input_tokens_seen": 4064768,
      "step": 2960
    },
    {
      "epoch": 0.0949042954996479,
      "grad_norm": 0.03437419608235359,
      "learning_rate": 0.2959539125984699,
      "loss": 0.8615,
      "num_input_tokens_seen": 4071728,
      "step": 2965
    },
    {
      "epoch": 0.09506433647013636,
      "grad_norm": 0.032900370657444,
      "learning_rate": 0.2959403122803996,
      "loss": 0.6606,
      "num_input_tokens_seen": 4078464,
      "step": 2970
    },
    {
      "epoch": 0.0952243774406248,
      "grad_norm": 0.09383966028690338,
      "learning_rate": 0.2959266894564991,
      "loss": 0.8925,
      "num_input_tokens_seen": 4085632,
      "step": 2975
    },
    {
      "epoch": 0.09538441841111324,
      "grad_norm": 0.0730961412191391,
      "learning_rate": 0.2959130441288692,
      "loss": 0.7327,
      "num_input_tokens_seen": 4092912,
      "step": 2980
    },
    {
      "epoch": 0.09554445938160169,
      "grad_norm": 0.015902100130915642,
      "learning_rate": 0.2958993762996143,
      "loss": 0.8534,
      "num_input_tokens_seen": 4100384,
      "step": 2985
    },
    {
      "epoch": 0.09570450035209013,
      "grad_norm": 0.016379261389374733,
      "learning_rate": 0.2958856859708421,
      "loss": 1.0322,
      "num_input_tokens_seen": 4107376,
      "step": 2990
    },
    {
      "epoch": 0.09586454132257859,
      "grad_norm": 0.01341941673308611,
      "learning_rate": 0.2958719731446638,
      "loss": 0.8292,
      "num_input_tokens_seen": 4113936,
      "step": 2995
    },
    {
      "epoch": 0.09602458229306703,
      "grad_norm": 0.017522580921649933,
      "learning_rate": 0.29585823782319404,
      "loss": 0.9041,
      "num_input_tokens_seen": 4121024,
      "step": 3000
    },
    {
      "epoch": 0.09602458229306703,
      "eval_loss": 0.84299635887146,
      "eval_runtime": 331.8504,
      "eval_samples_per_second": 41.844,
      "eval_steps_per_second": 20.922,
      "num_input_tokens_seen": 4121024,
      "step": 3000
    },
    {
      "epoch": 0.09618462326355547,
      "grad_norm": 0.021503068506717682,
      "learning_rate": 0.2958444800085511,
      "loss": 0.8588,
      "num_input_tokens_seen": 4127872,
      "step": 3005
    },
    {
      "epoch": 0.09634466423404392,
      "grad_norm": 0.009541040286421776,
      "learning_rate": 0.2958306997028565,
      "loss": 0.6378,
      "num_input_tokens_seen": 4134560,
      "step": 3010
    },
    {
      "epoch": 0.09650470520453236,
      "grad_norm": 0.024257687851786613,
      "learning_rate": 0.2958168969082354,
      "loss": 0.882,
      "num_input_tokens_seen": 4141904,
      "step": 3015
    },
    {
      "epoch": 0.0966647461750208,
      "grad_norm": 0.018324250355362892,
      "learning_rate": 0.2958030716268164,
      "loss": 0.7304,
      "num_input_tokens_seen": 4148640,
      "step": 3020
    },
    {
      "epoch": 0.09682478714550925,
      "grad_norm": 0.02150866575539112,
      "learning_rate": 0.2957892238607314,
      "loss": 0.8907,
      "num_input_tokens_seen": 4155392,
      "step": 3025
    },
    {
      "epoch": 0.09698482811599769,
      "grad_norm": 0.01507558859884739,
      "learning_rate": 0.2957753536121161,
      "loss": 0.8506,
      "num_input_tokens_seen": 4162288,
      "step": 3030
    },
    {
      "epoch": 0.09714486908648615,
      "grad_norm": 0.022267062216997147,
      "learning_rate": 0.29576146088310923,
      "loss": 0.7469,
      "num_input_tokens_seen": 4169408,
      "step": 3035
    },
    {
      "epoch": 0.09730491005697459,
      "grad_norm": 0.014250526204705238,
      "learning_rate": 0.2957475456758533,
      "loss": 0.985,
      "num_input_tokens_seen": 4176128,
      "step": 3040
    },
    {
      "epoch": 0.09746495102746303,
      "grad_norm": 0.015031605027616024,
      "learning_rate": 0.2957336079924944,
      "loss": 0.8296,
      "num_input_tokens_seen": 4182960,
      "step": 3045
    },
    {
      "epoch": 0.09762499199795148,
      "grad_norm": 0.01880693808197975,
      "learning_rate": 0.2957196478351816,
      "loss": 0.8519,
      "num_input_tokens_seen": 4189584,
      "step": 3050
    },
    {
      "epoch": 0.09778503296843992,
      "grad_norm": 0.050490766763687134,
      "learning_rate": 0.295705665206068,
      "loss": 0.8593,
      "num_input_tokens_seen": 4196144,
      "step": 3055
    },
    {
      "epoch": 0.09794507393892836,
      "grad_norm": 0.0556550994515419,
      "learning_rate": 0.2956916601073097,
      "loss": 0.9532,
      "num_input_tokens_seen": 4203264,
      "step": 3060
    },
    {
      "epoch": 0.09810511490941681,
      "grad_norm": 0.02442692220211029,
      "learning_rate": 0.29567763254106655,
      "loss": 0.8183,
      "num_input_tokens_seen": 4210048,
      "step": 3065
    },
    {
      "epoch": 0.09826515587990525,
      "grad_norm": 0.01975967176258564,
      "learning_rate": 0.29566358250950175,
      "loss": 1.0519,
      "num_input_tokens_seen": 4216448,
      "step": 3070
    },
    {
      "epoch": 0.0984251968503937,
      "grad_norm": 0.06205729395151138,
      "learning_rate": 0.295649510014782,
      "loss": 0.7959,
      "num_input_tokens_seen": 4223296,
      "step": 3075
    },
    {
      "epoch": 0.09858523782088215,
      "grad_norm": 0.03214986249804497,
      "learning_rate": 0.2956354150590775,
      "loss": 1.1572,
      "num_input_tokens_seen": 4230256,
      "step": 3080
    },
    {
      "epoch": 0.09874527879137059,
      "grad_norm": 0.04632348567247391,
      "learning_rate": 0.2956212976445618,
      "loss": 0.9187,
      "num_input_tokens_seen": 4236976,
      "step": 3085
    },
    {
      "epoch": 0.09890531976185904,
      "grad_norm": 0.026367774233222008,
      "learning_rate": 0.295607157773412,
      "loss": 0.7598,
      "num_input_tokens_seen": 4243584,
      "step": 3090
    },
    {
      "epoch": 0.09906536073234748,
      "grad_norm": 0.024985596537590027,
      "learning_rate": 0.2955929954478087,
      "loss": 0.9047,
      "num_input_tokens_seen": 4250352,
      "step": 3095
    },
    {
      "epoch": 0.09922540170283592,
      "grad_norm": 0.07687194645404816,
      "learning_rate": 0.29557881066993585,
      "loss": 0.9337,
      "num_input_tokens_seen": 4257072,
      "step": 3100
    },
    {
      "epoch": 0.09938544267332437,
      "grad_norm": 0.02786894515156746,
      "learning_rate": 0.29556460344198093,
      "loss": 0.7735,
      "num_input_tokens_seen": 4264416,
      "step": 3105
    },
    {
      "epoch": 0.09954548364381281,
      "grad_norm": 0.026888268068432808,
      "learning_rate": 0.29555037376613486,
      "loss": 1.1086,
      "num_input_tokens_seen": 4271104,
      "step": 3110
    },
    {
      "epoch": 0.09970552461430127,
      "grad_norm": 0.020158063620328903,
      "learning_rate": 0.29553612164459203,
      "loss": 0.9575,
      "num_input_tokens_seen": 4278352,
      "step": 3115
    },
    {
      "epoch": 0.0998655655847897,
      "grad_norm": 0.023630354553461075,
      "learning_rate": 0.29552184707955037,
      "loss": 0.9047,
      "num_input_tokens_seen": 4285056,
      "step": 3120
    },
    {
      "epoch": 0.10002560655527815,
      "grad_norm": 0.03509252518415451,
      "learning_rate": 0.29550755007321117,
      "loss": 1.0027,
      "num_input_tokens_seen": 4292096,
      "step": 3125
    },
    {
      "epoch": 0.1001856475257666,
      "grad_norm": 0.014360545203089714,
      "learning_rate": 0.29549323062777916,
      "loss": 0.8683,
      "num_input_tokens_seen": 4298960,
      "step": 3130
    },
    {
      "epoch": 0.10034568849625504,
      "grad_norm": 0.013630741275846958,
      "learning_rate": 0.29547888874546263,
      "loss": 0.94,
      "num_input_tokens_seen": 4305600,
      "step": 3135
    },
    {
      "epoch": 0.1005057294667435,
      "grad_norm": 0.020953712984919548,
      "learning_rate": 0.2954645244284732,
      "loss": 0.7957,
      "num_input_tokens_seen": 4312832,
      "step": 3140
    },
    {
      "epoch": 0.10066577043723193,
      "grad_norm": 0.011418617330491543,
      "learning_rate": 0.2954501376790261,
      "loss": 0.8224,
      "num_input_tokens_seen": 4319776,
      "step": 3145
    },
    {
      "epoch": 0.10082581140772037,
      "grad_norm": 0.01779334805905819,
      "learning_rate": 0.29543572849933997,
      "loss": 0.7782,
      "num_input_tokens_seen": 4326992,
      "step": 3150
    },
    {
      "epoch": 0.10098585237820883,
      "grad_norm": 0.02925274148583412,
      "learning_rate": 0.2954212968916368,
      "loss": 0.752,
      "num_input_tokens_seen": 4334160,
      "step": 3155
    },
    {
      "epoch": 0.10114589334869727,
      "grad_norm": 0.02301093004643917,
      "learning_rate": 0.29540684285814217,
      "loss": 0.8967,
      "num_input_tokens_seen": 4341104,
      "step": 3160
    },
    {
      "epoch": 0.1013059343191857,
      "grad_norm": 0.02037692815065384,
      "learning_rate": 0.2953923664010851,
      "loss": 0.8745,
      "num_input_tokens_seen": 4348192,
      "step": 3165
    },
    {
      "epoch": 0.10146597528967416,
      "grad_norm": 0.02021806873381138,
      "learning_rate": 0.295377867522698,
      "loss": 0.7265,
      "num_input_tokens_seen": 4355024,
      "step": 3170
    },
    {
      "epoch": 0.1016260162601626,
      "grad_norm": 0.013971948064863682,
      "learning_rate": 0.2953633462252168,
      "loss": 0.8071,
      "num_input_tokens_seen": 4361744,
      "step": 3175
    },
    {
      "epoch": 0.10178605723065105,
      "grad_norm": 0.030126402154564857,
      "learning_rate": 0.2953488025108809,
      "loss": 0.704,
      "num_input_tokens_seen": 4369024,
      "step": 3180
    },
    {
      "epoch": 0.10194609820113949,
      "grad_norm": 0.020049095153808594,
      "learning_rate": 0.295334236381933,
      "loss": 1.0849,
      "num_input_tokens_seen": 4376224,
      "step": 3185
    },
    {
      "epoch": 0.10210613917162793,
      "grad_norm": 0.021965760737657547,
      "learning_rate": 0.29531964784061954,
      "loss": 0.9825,
      "num_input_tokens_seen": 4383248,
      "step": 3190
    },
    {
      "epoch": 0.10226618014211639,
      "grad_norm": 0.01918606460094452,
      "learning_rate": 0.2953050368891902,
      "loss": 0.8587,
      "num_input_tokens_seen": 4390032,
      "step": 3195
    },
    {
      "epoch": 0.10242622111260483,
      "grad_norm": 0.011563633568584919,
      "learning_rate": 0.29529040352989805,
      "loss": 0.8787,
      "num_input_tokens_seen": 4396880,
      "step": 3200
    },
    {
      "epoch": 0.10242622111260483,
      "eval_loss": 0.8599197268486023,
      "eval_runtime": 331.5871,
      "eval_samples_per_second": 41.877,
      "eval_steps_per_second": 20.939,
      "num_input_tokens_seen": 4396880,
      "step": 3200
    },
    {
      "epoch": 0.10258626208309327,
      "grad_norm": 0.028652679175138474,
      "learning_rate": 0.29527574776499993,
      "loss": 0.7455,
      "num_input_tokens_seen": 4403616,
      "step": 3205
    },
    {
      "epoch": 0.10274630305358172,
      "grad_norm": 0.025051867589354515,
      "learning_rate": 0.2952610695967558,
      "loss": 0.8843,
      "num_input_tokens_seen": 4410528,
      "step": 3210
    },
    {
      "epoch": 0.10290634402407016,
      "grad_norm": 0.020329970866441727,
      "learning_rate": 0.29524636902742935,
      "loss": 1.0053,
      "num_input_tokens_seen": 4417792,
      "step": 3215
    },
    {
      "epoch": 0.10306638499455861,
      "grad_norm": 0.021442372351884842,
      "learning_rate": 0.2952316460592875,
      "loss": 0.8837,
      "num_input_tokens_seen": 4424752,
      "step": 3220
    },
    {
      "epoch": 0.10322642596504705,
      "grad_norm": 0.0299655981361866,
      "learning_rate": 0.29521690069460066,
      "loss": 1.3002,
      "num_input_tokens_seen": 4431408,
      "step": 3225
    },
    {
      "epoch": 0.10338646693553549,
      "grad_norm": 0.021624721586704254,
      "learning_rate": 0.29520213293564285,
      "loss": 1.1113,
      "num_input_tokens_seen": 4438240,
      "step": 3230
    },
    {
      "epoch": 0.10354650790602395,
      "grad_norm": 0.02684534713625908,
      "learning_rate": 0.29518734278469144,
      "loss": 0.9633,
      "num_input_tokens_seen": 4445056,
      "step": 3235
    },
    {
      "epoch": 0.10370654887651239,
      "grad_norm": 0.016264580190181732,
      "learning_rate": 0.29517253024402723,
      "loss": 0.8046,
      "num_input_tokens_seen": 4451808,
      "step": 3240
    },
    {
      "epoch": 0.10386658984700083,
      "grad_norm": 0.017269233241677284,
      "learning_rate": 0.2951576953159345,
      "loss": 0.9532,
      "num_input_tokens_seen": 4458784,
      "step": 3245
    },
    {
      "epoch": 0.10402663081748928,
      "grad_norm": 0.02150384895503521,
      "learning_rate": 0.29514283800270097,
      "loss": 1.2124,
      "num_input_tokens_seen": 4465648,
      "step": 3250
    },
    {
      "epoch": 0.10418667178797772,
      "grad_norm": 0.02519349940121174,
      "learning_rate": 0.2951279583066179,
      "loss": 0.8925,
      "num_input_tokens_seen": 4472352,
      "step": 3255
    },
    {
      "epoch": 0.10434671275846617,
      "grad_norm": 0.024772528558969498,
      "learning_rate": 0.2951130562299798,
      "loss": 1.026,
      "num_input_tokens_seen": 4479200,
      "step": 3260
    },
    {
      "epoch": 0.10450675372895461,
      "grad_norm": 0.05027385056018829,
      "learning_rate": 0.29509813177508487,
      "loss": 0.968,
      "num_input_tokens_seen": 4485904,
      "step": 3265
    },
    {
      "epoch": 0.10466679469944305,
      "grad_norm": 0.026262173429131508,
      "learning_rate": 0.2950831849442346,
      "loss": 0.836,
      "num_input_tokens_seen": 4492608,
      "step": 3270
    },
    {
      "epoch": 0.1048268356699315,
      "grad_norm": 0.02261030301451683,
      "learning_rate": 0.2950682157397339,
      "loss": 0.8749,
      "num_input_tokens_seen": 4499424,
      "step": 3275
    },
    {
      "epoch": 0.10498687664041995,
      "grad_norm": 0.025274818763136864,
      "learning_rate": 0.2950532241638914,
      "loss": 0.9462,
      "num_input_tokens_seen": 4506288,
      "step": 3280
    },
    {
      "epoch": 0.10514691761090839,
      "grad_norm": 0.030555788427591324,
      "learning_rate": 0.2950382102190188,
      "loss": 0.9158,
      "num_input_tokens_seen": 4512672,
      "step": 3285
    },
    {
      "epoch": 0.10530695858139684,
      "grad_norm": 0.01953103020787239,
      "learning_rate": 0.2950231739074316,
      "loss": 0.7343,
      "num_input_tokens_seen": 4519376,
      "step": 3290
    },
    {
      "epoch": 0.10546699955188528,
      "grad_norm": 0.024239560589194298,
      "learning_rate": 0.29500811523144843,
      "loss": 0.7253,
      "num_input_tokens_seen": 4526352,
      "step": 3295
    },
    {
      "epoch": 0.10562704052237373,
      "grad_norm": 0.02974797412753105,
      "learning_rate": 0.2949930341933917,
      "loss": 1.0255,
      "num_input_tokens_seen": 4533264,
      "step": 3300
    },
    {
      "epoch": 0.10578708149286217,
      "grad_norm": 0.44687438011169434,
      "learning_rate": 0.29497793079558693,
      "loss": 0.8985,
      "num_input_tokens_seen": 4540016,
      "step": 3305
    },
    {
      "epoch": 0.10594712246335061,
      "grad_norm": 0.02079041302204132,
      "learning_rate": 0.2949628050403633,
      "loss": 0.9074,
      "num_input_tokens_seen": 4546544,
      "step": 3310
    },
    {
      "epoch": 0.10610716343383907,
      "grad_norm": 0.015733834356069565,
      "learning_rate": 0.2949476569300535,
      "loss": 0.6591,
      "num_input_tokens_seen": 4553520,
      "step": 3315
    },
    {
      "epoch": 0.1062672044043275,
      "grad_norm": 0.06179334223270416,
      "learning_rate": 0.29493248646699344,
      "loss": 0.8144,
      "num_input_tokens_seen": 4560832,
      "step": 3320
    },
    {
      "epoch": 0.10642724537481596,
      "grad_norm": 0.015170645900070667,
      "learning_rate": 0.29491729365352265,
      "loss": 0.8025,
      "num_input_tokens_seen": 4567824,
      "step": 3325
    },
    {
      "epoch": 0.1065872863453044,
      "grad_norm": 0.03188195079565048,
      "learning_rate": 0.29490207849198397,
      "loss": 0.7625,
      "num_input_tokens_seen": 4574608,
      "step": 3330
    },
    {
      "epoch": 0.10674732731579284,
      "grad_norm": 0.0673358142375946,
      "learning_rate": 0.29488684098472384,
      "loss": 0.8825,
      "num_input_tokens_seen": 4581296,
      "step": 3335
    },
    {
      "epoch": 0.10690736828628129,
      "grad_norm": 0.09976831078529358,
      "learning_rate": 0.2948715811340921,
      "loss": 1.2925,
      "num_input_tokens_seen": 4588320,
      "step": 3340
    },
    {
      "epoch": 0.10706740925676973,
      "grad_norm": 0.021256951615214348,
      "learning_rate": 0.294856298942442,
      "loss": 0.8324,
      "num_input_tokens_seen": 4595408,
      "step": 3345
    },
    {
      "epoch": 0.10722745022725817,
      "grad_norm": 0.015081590041518211,
      "learning_rate": 0.2948409944121302,
      "loss": 0.7859,
      "num_input_tokens_seen": 4602992,
      "step": 3350
    },
    {
      "epoch": 0.10738749119774663,
      "grad_norm": 0.0473429411649704,
      "learning_rate": 0.29482566754551687,
      "loss": 0.7306,
      "num_input_tokens_seen": 4609648,
      "step": 3355
    },
    {
      "epoch": 0.10754753216823507,
      "grad_norm": 0.031014898791909218,
      "learning_rate": 0.2948103183449656,
      "loss": 0.9314,
      "num_input_tokens_seen": 4616432,
      "step": 3360
    },
    {
      "epoch": 0.10770757313872352,
      "grad_norm": 0.019322555512189865,
      "learning_rate": 0.2947949468128435,
      "loss": 0.7397,
      "num_input_tokens_seen": 4623168,
      "step": 3365
    },
    {
      "epoch": 0.10786761410921196,
      "grad_norm": 0.022394292056560516,
      "learning_rate": 0.2947795529515209,
      "loss": 1.1533,
      "num_input_tokens_seen": 4630064,
      "step": 3370
    },
    {
      "epoch": 0.1080276550797004,
      "grad_norm": 0.0173397995531559,
      "learning_rate": 0.29476413676337193,
      "loss": 1.0416,
      "num_input_tokens_seen": 4637008,
      "step": 3375
    },
    {
      "epoch": 0.10818769605018885,
      "grad_norm": 0.016513651236891747,
      "learning_rate": 0.2947486982507738,
      "loss": 0.7118,
      "num_input_tokens_seen": 4643792,
      "step": 3380
    },
    {
      "epoch": 0.10834773702067729,
      "grad_norm": 0.015620708465576172,
      "learning_rate": 0.29473323741610735,
      "loss": 0.7791,
      "num_input_tokens_seen": 4650464,
      "step": 3385
    },
    {
      "epoch": 0.10850777799116573,
      "grad_norm": 0.030179603025317192,
      "learning_rate": 0.2947177542617569,
      "loss": 0.7584,
      "num_input_tokens_seen": 4657248,
      "step": 3390
    },
    {
      "epoch": 0.10866781896165419,
      "grad_norm": 0.013304522261023521,
      "learning_rate": 0.2947022487901101,
      "loss": 0.9252,
      "num_input_tokens_seen": 4663696,
      "step": 3395
    },
    {
      "epoch": 0.10882785993214263,
      "grad_norm": 0.018777184188365936,
      "learning_rate": 0.2946867210035581,
      "loss": 0.8314,
      "num_input_tokens_seen": 4671152,
      "step": 3400
    },
    {
      "epoch": 0.10882785993214263,
      "eval_loss": 0.8540629148483276,
      "eval_runtime": 331.5506,
      "eval_samples_per_second": 41.882,
      "eval_steps_per_second": 20.941,
      "num_input_tokens_seen": 4671152,
      "step": 3400
    },
    {
      "epoch": 0.10898790090263108,
      "grad_norm": 0.023990405723452568,
      "learning_rate": 0.2946711709044954,
      "loss": 0.9067,
      "num_input_tokens_seen": 4677840,
      "step": 3405
    },
    {
      "epoch": 0.10914794187311952,
      "grad_norm": 0.02437015436589718,
      "learning_rate": 0.2946555984953202,
      "loss": 0.7994,
      "num_input_tokens_seen": 4684432,
      "step": 3410
    },
    {
      "epoch": 0.10930798284360796,
      "grad_norm": 0.02323911525309086,
      "learning_rate": 0.2946400037784338,
      "loss": 0.7543,
      "num_input_tokens_seen": 4691552,
      "step": 3415
    },
    {
      "epoch": 0.10946802381409641,
      "grad_norm": 0.025276200845837593,
      "learning_rate": 0.29462438675624114,
      "loss": 0.8009,
      "num_input_tokens_seen": 4698528,
      "step": 3420
    },
    {
      "epoch": 0.10962806478458485,
      "grad_norm": 0.029562344774603844,
      "learning_rate": 0.2946087474311506,
      "loss": 0.7907,
      "num_input_tokens_seen": 4705728,
      "step": 3425
    },
    {
      "epoch": 0.10978810575507329,
      "grad_norm": 0.030119413509964943,
      "learning_rate": 0.294593085805574,
      "loss": 0.7954,
      "num_input_tokens_seen": 4712240,
      "step": 3430
    },
    {
      "epoch": 0.10994814672556175,
      "grad_norm": 0.02752627059817314,
      "learning_rate": 0.2945774018819264,
      "loss": 0.9172,
      "num_input_tokens_seen": 4720000,
      "step": 3435
    },
    {
      "epoch": 0.11010818769605019,
      "grad_norm": 0.013486248441040516,
      "learning_rate": 0.2945616956626266,
      "loss": 1.1039,
      "num_input_tokens_seen": 4727648,
      "step": 3440
    },
    {
      "epoch": 0.11026822866653864,
      "grad_norm": 0.05222737789154053,
      "learning_rate": 0.2945459671500966,
      "loss": 1.0049,
      "num_input_tokens_seen": 4734128,
      "step": 3445
    },
    {
      "epoch": 0.11042826963702708,
      "grad_norm": 0.0212120171636343,
      "learning_rate": 0.2945302163467621,
      "loss": 0.8558,
      "num_input_tokens_seen": 4741088,
      "step": 3450
    },
    {
      "epoch": 0.11058831060751552,
      "grad_norm": 0.017955109477043152,
      "learning_rate": 0.2945144432550519,
      "loss": 0.9521,
      "num_input_tokens_seen": 4748064,
      "step": 3455
    },
    {
      "epoch": 0.11074835157800397,
      "grad_norm": 0.026706678792834282,
      "learning_rate": 0.29449864787739843,
      "loss": 0.8405,
      "num_input_tokens_seen": 4755104,
      "step": 3460
    },
    {
      "epoch": 0.11090839254849241,
      "grad_norm": 0.0167748611420393,
      "learning_rate": 0.2944828302162376,
      "loss": 0.8068,
      "num_input_tokens_seen": 4762240,
      "step": 3465
    },
    {
      "epoch": 0.11106843351898085,
      "grad_norm": 0.02341376058757305,
      "learning_rate": 0.2944669902740087,
      "loss": 0.8256,
      "num_input_tokens_seen": 4768672,
      "step": 3470
    },
    {
      "epoch": 0.1112284744894693,
      "grad_norm": 0.021351661533117294,
      "learning_rate": 0.2944511280531544,
      "loss": 0.7523,
      "num_input_tokens_seen": 4775456,
      "step": 3475
    },
    {
      "epoch": 0.11138851545995775,
      "grad_norm": 0.011645836755633354,
      "learning_rate": 0.29443524355612083,
      "loss": 0.725,
      "num_input_tokens_seen": 4782512,
      "step": 3480
    },
    {
      "epoch": 0.1115485564304462,
      "grad_norm": 0.02421445958316326,
      "learning_rate": 0.29441933678535764,
      "loss": 0.8363,
      "num_input_tokens_seen": 4789376,
      "step": 3485
    },
    {
      "epoch": 0.11170859740093464,
      "grad_norm": 0.06420710682868958,
      "learning_rate": 0.29440340774331786,
      "loss": 0.8837,
      "num_input_tokens_seen": 4796336,
      "step": 3490
    },
    {
      "epoch": 0.11186863837142308,
      "grad_norm": 0.030908100306987762,
      "learning_rate": 0.2943874564324579,
      "loss": 0.7262,
      "num_input_tokens_seen": 4803168,
      "step": 3495
    },
    {
      "epoch": 0.11202867934191153,
      "grad_norm": 0.03238767012953758,
      "learning_rate": 0.2943714828552376,
      "loss": 0.9058,
      "num_input_tokens_seen": 4810368,
      "step": 3500
    },
    {
      "epoch": 0.11218872031239997,
      "grad_norm": 0.01430445071309805,
      "learning_rate": 0.29435548701412045,
      "loss": 0.7746,
      "num_input_tokens_seen": 4817712,
      "step": 3505
    },
    {
      "epoch": 0.11234876128288843,
      "grad_norm": 0.010222128592431545,
      "learning_rate": 0.2943394689115731,
      "loss": 0.8239,
      "num_input_tokens_seen": 4824640,
      "step": 3510
    },
    {
      "epoch": 0.11250880225337687,
      "grad_norm": 0.018444610759615898,
      "learning_rate": 0.29432342855006577,
      "loss": 0.7657,
      "num_input_tokens_seen": 4831520,
      "step": 3515
    },
    {
      "epoch": 0.1126688432238653,
      "grad_norm": 0.02543519251048565,
      "learning_rate": 0.294307365932072,
      "loss": 0.8009,
      "num_input_tokens_seen": 4838064,
      "step": 3520
    },
    {
      "epoch": 0.11282888419435376,
      "grad_norm": 0.026901086792349815,
      "learning_rate": 0.294291281060069,
      "loss": 0.9748,
      "num_input_tokens_seen": 4845824,
      "step": 3525
    },
    {
      "epoch": 0.1129889251648422,
      "grad_norm": 0.046859197318553925,
      "learning_rate": 0.29427517393653724,
      "loss": 1.1513,
      "num_input_tokens_seen": 4852768,
      "step": 3530
    },
    {
      "epoch": 0.11314896613533064,
      "grad_norm": 0.018387749791145325,
      "learning_rate": 0.29425904456396046,
      "loss": 0.684,
      "num_input_tokens_seen": 4859728,
      "step": 3535
    },
    {
      "epoch": 0.11330900710581909,
      "grad_norm": 0.01531580276787281,
      "learning_rate": 0.2942428929448262,
      "loss": 0.727,
      "num_input_tokens_seen": 4866256,
      "step": 3540
    },
    {
      "epoch": 0.11346904807630753,
      "grad_norm": 0.06913020461797714,
      "learning_rate": 0.2942267190816252,
      "loss": 1.7166,
      "num_input_tokens_seen": 4876560,
      "step": 3545
    },
    {
      "epoch": 0.11362908904679599,
      "grad_norm": 0.027299843728542328,
      "learning_rate": 0.2942105229768516,
      "loss": 0.8683,
      "num_input_tokens_seen": 4883216,
      "step": 3550
    },
    {
      "epoch": 0.11378913001728443,
      "grad_norm": 0.02372070960700512,
      "learning_rate": 0.29419430463300306,
      "loss": 0.7582,
      "num_input_tokens_seen": 4890224,
      "step": 3555
    },
    {
      "epoch": 0.11394917098777287,
      "grad_norm": 0.02146444469690323,
      "learning_rate": 0.2941780640525808,
      "loss": 0.7593,
      "num_input_tokens_seen": 4896944,
      "step": 3560
    },
    {
      "epoch": 0.11410921195826132,
      "grad_norm": 0.028577176854014397,
      "learning_rate": 0.2941618012380891,
      "loss": 0.9695,
      "num_input_tokens_seen": 4903920,
      "step": 3565
    },
    {
      "epoch": 0.11426925292874976,
      "grad_norm": 0.05316779762506485,
      "learning_rate": 0.29414551619203605,
      "loss": 0.908,
      "num_input_tokens_seen": 4910560,
      "step": 3570
    },
    {
      "epoch": 0.1144292938992382,
      "grad_norm": 0.01737932488322258,
      "learning_rate": 0.29412920891693295,
      "loss": 0.8701,
      "num_input_tokens_seen": 4917296,
      "step": 3575
    },
    {
      "epoch": 0.11458933486972665,
      "grad_norm": 0.01699652336537838,
      "learning_rate": 0.2941128794152946,
      "loss": 0.771,
      "num_input_tokens_seen": 4923920,
      "step": 3580
    },
    {
      "epoch": 0.11474937584021509,
      "grad_norm": 0.02327611856162548,
      "learning_rate": 0.2940965276896392,
      "loss": 0.9542,
      "num_input_tokens_seen": 4930672,
      "step": 3585
    },
    {
      "epoch": 0.11490941681070355,
      "grad_norm": 0.015447588637471199,
      "learning_rate": 0.2940801537424884,
      "loss": 0.8875,
      "num_input_tokens_seen": 4937536,
      "step": 3590
    },
    {
      "epoch": 0.11506945778119199,
      "grad_norm": 0.01985803060233593,
      "learning_rate": 0.2940637575763673,
      "loss": 0.8832,
      "num_input_tokens_seen": 4944144,
      "step": 3595
    },
    {
      "epoch": 0.11522949875168043,
      "grad_norm": 0.027470547705888748,
      "learning_rate": 0.2940473391938043,
      "loss": 0.7955,
      "num_input_tokens_seen": 4950800,
      "step": 3600
    },
    {
      "epoch": 0.11522949875168043,
      "eval_loss": 0.85062575340271,
      "eval_runtime": 331.5899,
      "eval_samples_per_second": 41.877,
      "eval_steps_per_second": 20.939,
      "num_input_tokens_seen": 4950800,
      "step": 3600
    },
    {
      "epoch": 0.11538953972216888,
      "grad_norm": 0.01958993822336197,
      "learning_rate": 0.29403089859733145,
      "loss": 0.9329,
      "num_input_tokens_seen": 4957856,
      "step": 3605
    },
    {
      "epoch": 0.11554958069265732,
      "grad_norm": 0.010106900706887245,
      "learning_rate": 0.294014435789484,
      "loss": 0.8032,
      "num_input_tokens_seen": 4965168,
      "step": 3610
    },
    {
      "epoch": 0.11570962166314576,
      "grad_norm": 0.02829821966588497,
      "learning_rate": 0.2939979507728007,
      "loss": 0.6,
      "num_input_tokens_seen": 4971968,
      "step": 3615
    },
    {
      "epoch": 0.11586966263363421,
      "grad_norm": 0.02448991872370243,
      "learning_rate": 0.2939814435498239,
      "loss": 1.0541,
      "num_input_tokens_seen": 4979104,
      "step": 3620
    },
    {
      "epoch": 0.11602970360412265,
      "grad_norm": 0.021043168380856514,
      "learning_rate": 0.29396491412309905,
      "loss": 0.6968,
      "num_input_tokens_seen": 4986144,
      "step": 3625
    },
    {
      "epoch": 0.1161897445746111,
      "grad_norm": 0.015849949792027473,
      "learning_rate": 0.2939483624951753,
      "loss": 0.7519,
      "num_input_tokens_seen": 4992752,
      "step": 3630
    },
    {
      "epoch": 0.11634978554509955,
      "grad_norm": 0.020781276747584343,
      "learning_rate": 0.2939317886686051,
      "loss": 0.6866,
      "num_input_tokens_seen": 5000144,
      "step": 3635
    },
    {
      "epoch": 0.11650982651558799,
      "grad_norm": 0.02884560078382492,
      "learning_rate": 0.2939151926459443,
      "loss": 0.8832,
      "num_input_tokens_seen": 5006992,
      "step": 3640
    },
    {
      "epoch": 0.11666986748607644,
      "grad_norm": 0.026714343577623367,
      "learning_rate": 0.2938985744297522,
      "loss": 0.8463,
      "num_input_tokens_seen": 5013696,
      "step": 3645
    },
    {
      "epoch": 0.11682990845656488,
      "grad_norm": 0.019191361963748932,
      "learning_rate": 0.29388193402259166,
      "loss": 0.8192,
      "num_input_tokens_seen": 5020912,
      "step": 3650
    },
    {
      "epoch": 0.11698994942705332,
      "grad_norm": 0.06706949323415756,
      "learning_rate": 0.29386527142702873,
      "loss": 1.0456,
      "num_input_tokens_seen": 5027488,
      "step": 3655
    },
    {
      "epoch": 0.11714999039754177,
      "grad_norm": 0.025216246023774147,
      "learning_rate": 0.293848586645633,
      "loss": 0.8798,
      "num_input_tokens_seen": 5034160,
      "step": 3660
    },
    {
      "epoch": 0.11731003136803021,
      "grad_norm": 0.01919614151120186,
      "learning_rate": 0.2938318796809775,
      "loss": 0.8803,
      "num_input_tokens_seen": 5041168,
      "step": 3665
    },
    {
      "epoch": 0.11747007233851867,
      "grad_norm": 0.07523654401302338,
      "learning_rate": 0.29381515053563867,
      "loss": 0.7953,
      "num_input_tokens_seen": 5048432,
      "step": 3670
    },
    {
      "epoch": 0.1176301133090071,
      "grad_norm": 0.02204674482345581,
      "learning_rate": 0.29379839921219636,
      "loss": 0.8772,
      "num_input_tokens_seen": 5055552,
      "step": 3675
    },
    {
      "epoch": 0.11779015427949555,
      "grad_norm": 0.023394528776407242,
      "learning_rate": 0.2937816257132338,
      "loss": 0.9127,
      "num_input_tokens_seen": 5062704,
      "step": 3680
    },
    {
      "epoch": 0.117950195249984,
      "grad_norm": 0.01688675954937935,
      "learning_rate": 0.2937648300413376,
      "loss": 0.8217,
      "num_input_tokens_seen": 5070528,
      "step": 3685
    },
    {
      "epoch": 0.11811023622047244,
      "grad_norm": 0.011811976321041584,
      "learning_rate": 0.293748012199098,
      "loss": 0.8057,
      "num_input_tokens_seen": 5077552,
      "step": 3690
    },
    {
      "epoch": 0.11827027719096088,
      "grad_norm": 0.018397657200694084,
      "learning_rate": 0.29373117218910844,
      "loss": 0.8777,
      "num_input_tokens_seen": 5084768,
      "step": 3695
    },
    {
      "epoch": 0.11843031816144933,
      "grad_norm": 0.012376165948808193,
      "learning_rate": 0.2937143100139659,
      "loss": 0.8981,
      "num_input_tokens_seen": 5091536,
      "step": 3700
    },
    {
      "epoch": 0.11859035913193777,
      "grad_norm": 0.012822093442082405,
      "learning_rate": 0.29369742567627083,
      "loss": 0.7215,
      "num_input_tokens_seen": 5098112,
      "step": 3705
    },
    {
      "epoch": 0.11875040010242623,
      "grad_norm": 0.01409594714641571,
      "learning_rate": 0.29368051917862675,
      "loss": 0.8043,
      "num_input_tokens_seen": 5104528,
      "step": 3710
    },
    {
      "epoch": 0.11891044107291467,
      "grad_norm": 0.011159508489072323,
      "learning_rate": 0.2936635905236411,
      "loss": 0.8296,
      "num_input_tokens_seen": 5111568,
      "step": 3715
    },
    {
      "epoch": 0.1190704820434031,
      "grad_norm": 0.011981361545622349,
      "learning_rate": 0.2936466397139244,
      "loss": 0.8767,
      "num_input_tokens_seen": 5118336,
      "step": 3720
    },
    {
      "epoch": 0.11923052301389156,
      "grad_norm": 0.02121446095407009,
      "learning_rate": 0.2936296667520907,
      "loss": 0.8741,
      "num_input_tokens_seen": 5125392,
      "step": 3725
    },
    {
      "epoch": 0.11939056398438,
      "grad_norm": 0.013244383037090302,
      "learning_rate": 0.2936126716407574,
      "loss": 0.6399,
      "num_input_tokens_seen": 5131920,
      "step": 3730
    },
    {
      "epoch": 0.11955060495486845,
      "grad_norm": 0.021896518766880035,
      "learning_rate": 0.29359565438254537,
      "loss": 0.8432,
      "num_input_tokens_seen": 5138880,
      "step": 3735
    },
    {
      "epoch": 0.11971064592535689,
      "grad_norm": 0.023376574739813805,
      "learning_rate": 0.29357861498007887,
      "loss": 0.7772,
      "num_input_tokens_seen": 5145808,
      "step": 3740
    },
    {
      "epoch": 0.11987068689584533,
      "grad_norm": 0.013369983062148094,
      "learning_rate": 0.29356155343598567,
      "loss": 0.8647,
      "num_input_tokens_seen": 5152672,
      "step": 3745
    },
    {
      "epoch": 0.12003072786633379,
      "grad_norm": 0.02116667665541172,
      "learning_rate": 0.2935444697528968,
      "loss": 0.9498,
      "num_input_tokens_seen": 5159824,
      "step": 3750
    },
    {
      "epoch": 0.12019076883682223,
      "grad_norm": 0.017712300643324852,
      "learning_rate": 0.2935273639334468,
      "loss": 0.7835,
      "num_input_tokens_seen": 5166752,
      "step": 3755
    },
    {
      "epoch": 0.12035080980731067,
      "grad_norm": 0.017742367461323738,
      "learning_rate": 0.29351023598027365,
      "loss": 0.8158,
      "num_input_tokens_seen": 5173744,
      "step": 3760
    },
    {
      "epoch": 0.12051085077779912,
      "grad_norm": 0.016643978655338287,
      "learning_rate": 0.2934930858960186,
      "loss": 0.8112,
      "num_input_tokens_seen": 5181040,
      "step": 3765
    },
    {
      "epoch": 0.12067089174828756,
      "grad_norm": 0.025628434494137764,
      "learning_rate": 0.29347591368332643,
      "loss": 0.8769,
      "num_input_tokens_seen": 5187584,
      "step": 3770
    },
    {
      "epoch": 0.12083093271877601,
      "grad_norm": 0.0438547357916832,
      "learning_rate": 0.2934587193448454,
      "loss": 0.7903,
      "num_input_tokens_seen": 5194544,
      "step": 3775
    },
    {
      "epoch": 0.12099097368926445,
      "grad_norm": 0.010861962102353573,
      "learning_rate": 0.29344150288322696,
      "loss": 0.8931,
      "num_input_tokens_seen": 5201408,
      "step": 3780
    },
    {
      "epoch": 0.12115101465975289,
      "grad_norm": 0.014039445668458939,
      "learning_rate": 0.2934242643011263,
      "loss": 0.8257,
      "num_input_tokens_seen": 5208064,
      "step": 3785
    },
    {
      "epoch": 0.12131105563024135,
      "grad_norm": 0.011654184199869633,
      "learning_rate": 0.2934070036012016,
      "loss": 0.6162,
      "num_input_tokens_seen": 5214752,
      "step": 3790
    },
    {
      "epoch": 0.12147109660072979,
      "grad_norm": 0.02554907090961933,
      "learning_rate": 0.29338972078611475,
      "loss": 0.7524,
      "num_input_tokens_seen": 5221536,
      "step": 3795
    },
    {
      "epoch": 0.12163113757121823,
      "grad_norm": 0.01490397285670042,
      "learning_rate": 0.2933724158585311,
      "loss": 0.7693,
      "num_input_tokens_seen": 5228512,
      "step": 3800
    },
    {
      "epoch": 0.12163113757121823,
      "eval_loss": 0.8172886967658997,
      "eval_runtime": 331.8456,
      "eval_samples_per_second": 41.845,
      "eval_steps_per_second": 20.922,
      "num_input_tokens_seen": 5228512,
      "step": 3800
    },
    {
      "epoch": 0.12179117854170668,
      "grad_norm": 0.018447382375597954,
      "learning_rate": 0.29335508882111916,
      "loss": 0.9661,
      "num_input_tokens_seen": 5235344,
      "step": 3805
    },
    {
      "epoch": 0.12195121951219512,
      "grad_norm": 0.0125975850969553,
      "learning_rate": 0.29333773967655097,
      "loss": 0.7487,
      "num_input_tokens_seen": 5242176,
      "step": 3810
    },
    {
      "epoch": 0.12211126048268357,
      "grad_norm": 0.016813907772302628,
      "learning_rate": 0.2933203684275021,
      "loss": 0.925,
      "num_input_tokens_seen": 5248992,
      "step": 3815
    },
    {
      "epoch": 0.12227130145317201,
      "grad_norm": 0.017779188230633736,
      "learning_rate": 0.2933029750766513,
      "loss": 1.0048,
      "num_input_tokens_seen": 5256080,
      "step": 3820
    },
    {
      "epoch": 0.12243134242366045,
      "grad_norm": 0.0102293835952878,
      "learning_rate": 0.2932855596266809,
      "loss": 0.6936,
      "num_input_tokens_seen": 5262688,
      "step": 3825
    },
    {
      "epoch": 0.1225913833941489,
      "grad_norm": 0.018354211002588272,
      "learning_rate": 0.2932681220802765,
      "loss": 0.7498,
      "num_input_tokens_seen": 5269616,
      "step": 3830
    },
    {
      "epoch": 0.12275142436463735,
      "grad_norm": 0.012362288311123848,
      "learning_rate": 0.2932506624401274,
      "loss": 0.8527,
      "num_input_tokens_seen": 5277264,
      "step": 3835
    },
    {
      "epoch": 0.12291146533512579,
      "grad_norm": 0.01163034699857235,
      "learning_rate": 0.29323318070892584,
      "loss": 0.9322,
      "num_input_tokens_seen": 5283952,
      "step": 3840
    },
    {
      "epoch": 0.12307150630561424,
      "grad_norm": 0.019077887758612633,
      "learning_rate": 0.29321567688936784,
      "loss": 0.827,
      "num_input_tokens_seen": 5290928,
      "step": 3845
    },
    {
      "epoch": 0.12323154727610268,
      "grad_norm": 0.02152593433856964,
      "learning_rate": 0.29319815098415275,
      "loss": 0.8392,
      "num_input_tokens_seen": 5297584,
      "step": 3850
    },
    {
      "epoch": 0.12339158824659113,
      "grad_norm": 0.02257768251001835,
      "learning_rate": 0.2931806029959832,
      "loss": 0.7697,
      "num_input_tokens_seen": 5304208,
      "step": 3855
    },
    {
      "epoch": 0.12355162921707957,
      "grad_norm": 0.013133706524968147,
      "learning_rate": 0.29316303292756535,
      "loss": 0.9261,
      "num_input_tokens_seen": 5310880,
      "step": 3860
    },
    {
      "epoch": 0.12371167018756801,
      "grad_norm": 0.021953968331217766,
      "learning_rate": 0.29314544078160876,
      "loss": 0.6039,
      "num_input_tokens_seen": 5317600,
      "step": 3865
    },
    {
      "epoch": 0.12387171115805647,
      "grad_norm": 0.01908229850232601,
      "learning_rate": 0.2931278265608263,
      "loss": 0.7737,
      "num_input_tokens_seen": 5324800,
      "step": 3870
    },
    {
      "epoch": 0.1240317521285449,
      "grad_norm": 0.01715894229710102,
      "learning_rate": 0.29311019026793433,
      "loss": 0.8791,
      "num_input_tokens_seen": 5331472,
      "step": 3875
    },
    {
      "epoch": 0.12419179309903335,
      "grad_norm": 0.011560235172510147,
      "learning_rate": 0.29309253190565254,
      "loss": 0.6207,
      "num_input_tokens_seen": 5337744,
      "step": 3880
    },
    {
      "epoch": 0.1243518340695218,
      "grad_norm": 0.01966211386024952,
      "learning_rate": 0.2930748514767042,
      "loss": 0.7404,
      "num_input_tokens_seen": 5344624,
      "step": 3885
    },
    {
      "epoch": 0.12451187504001024,
      "grad_norm": 0.0157721359282732,
      "learning_rate": 0.29305714898381574,
      "loss": 0.7653,
      "num_input_tokens_seen": 5351472,
      "step": 3890
    },
    {
      "epoch": 0.12467191601049869,
      "grad_norm": 0.012889767996966839,
      "learning_rate": 0.29303942442971714,
      "loss": 0.9049,
      "num_input_tokens_seen": 5358400,
      "step": 3895
    },
    {
      "epoch": 0.12483195698098713,
      "grad_norm": 0.027433138340711594,
      "learning_rate": 0.2930216778171417,
      "loss": 0.9847,
      "num_input_tokens_seen": 5365472,
      "step": 3900
    },
    {
      "epoch": 0.12499199795147557,
      "grad_norm": 0.01750747486948967,
      "learning_rate": 0.2930039091488263,
      "loss": 0.7575,
      "num_input_tokens_seen": 5372464,
      "step": 3905
    },
    {
      "epoch": 0.12515203892196403,
      "grad_norm": 0.035390522330999374,
      "learning_rate": 0.29298611842751093,
      "loss": 0.9368,
      "num_input_tokens_seen": 5379840,
      "step": 3910
    },
    {
      "epoch": 0.12531207989245247,
      "grad_norm": 0.12549471855163574,
      "learning_rate": 0.29296830565593923,
      "loss": 0.9518,
      "num_input_tokens_seen": 5386976,
      "step": 3915
    },
    {
      "epoch": 0.1254721208629409,
      "grad_norm": 0.06145868077874184,
      "learning_rate": 0.2929504708368582,
      "loss": 1.154,
      "num_input_tokens_seen": 5393712,
      "step": 3920
    },
    {
      "epoch": 0.12563216183342935,
      "grad_norm": 0.014726854860782623,
      "learning_rate": 0.29293261397301806,
      "loss": 0.9891,
      "num_input_tokens_seen": 5400528,
      "step": 3925
    },
    {
      "epoch": 0.1257922028039178,
      "grad_norm": 0.02214822731912136,
      "learning_rate": 0.29291473506717275,
      "loss": 1.0261,
      "num_input_tokens_seen": 5407216,
      "step": 3930
    },
    {
      "epoch": 0.12595224377440625,
      "grad_norm": 0.04739654064178467,
      "learning_rate": 0.29289683412207923,
      "loss": 1.2516,
      "num_input_tokens_seen": 5414368,
      "step": 3935
    },
    {
      "epoch": 0.1261122847448947,
      "grad_norm": 0.03204645216464996,
      "learning_rate": 0.29287891114049813,
      "loss": 1.095,
      "num_input_tokens_seen": 5421504,
      "step": 3940
    },
    {
      "epoch": 0.12627232571538313,
      "grad_norm": 0.028620868921279907,
      "learning_rate": 0.29286096612519347,
      "loss": 0.9362,
      "num_input_tokens_seen": 5428016,
      "step": 3945
    },
    {
      "epoch": 0.12643236668587157,
      "grad_norm": 0.07020670175552368,
      "learning_rate": 0.2928429990789325,
      "loss": 1.2039,
      "num_input_tokens_seen": 5435440,
      "step": 3950
    },
    {
      "epoch": 0.12659240765636004,
      "grad_norm": 0.030824657529592514,
      "learning_rate": 0.29282501000448596,
      "loss": 0.9708,
      "num_input_tokens_seen": 5442384,
      "step": 3955
    },
    {
      "epoch": 0.12675244862684848,
      "grad_norm": 0.01846209540963173,
      "learning_rate": 0.2928069989046281,
      "loss": 1.1273,
      "num_input_tokens_seen": 5449200,
      "step": 3960
    },
    {
      "epoch": 0.12691248959733692,
      "grad_norm": 0.014661205001175404,
      "learning_rate": 0.2927889657821363,
      "loss": 0.7096,
      "num_input_tokens_seen": 5456064,
      "step": 3965
    },
    {
      "epoch": 0.12707253056782536,
      "grad_norm": 0.019266992807388306,
      "learning_rate": 0.2927709106397916,
      "loss": 0.9408,
      "num_input_tokens_seen": 5463200,
      "step": 3970
    },
    {
      "epoch": 0.1272325715383138,
      "grad_norm": 0.032656073570251465,
      "learning_rate": 0.29275283348037834,
      "loss": 0.7648,
      "num_input_tokens_seen": 5470304,
      "step": 3975
    },
    {
      "epoch": 0.12739261250880227,
      "grad_norm": 0.030997617170214653,
      "learning_rate": 0.29273473430668423,
      "loss": 0.7623,
      "num_input_tokens_seen": 5477152,
      "step": 3980
    },
    {
      "epoch": 0.1275526534792907,
      "grad_norm": 0.016278227791190147,
      "learning_rate": 0.2927166131215003,
      "loss": 0.7458,
      "num_input_tokens_seen": 5483984,
      "step": 3985
    },
    {
      "epoch": 0.12771269444977915,
      "grad_norm": 0.018898416310548782,
      "learning_rate": 0.2926984699276212,
      "loss": 0.9006,
      "num_input_tokens_seen": 5491024,
      "step": 3990
    },
    {
      "epoch": 0.12787273542026759,
      "grad_norm": 0.013471771031618118,
      "learning_rate": 0.29268030472784473,
      "loss": 0.9445,
      "num_input_tokens_seen": 5497968,
      "step": 3995
    },
    {
      "epoch": 0.12803277639075603,
      "grad_norm": 0.02179572731256485,
      "learning_rate": 0.2926621175249723,
      "loss": 0.8709,
      "num_input_tokens_seen": 5504608,
      "step": 4000
    },
    {
      "epoch": 0.12803277639075603,
      "eval_loss": 0.8573163151741028,
      "eval_runtime": 332.2487,
      "eval_samples_per_second": 41.794,
      "eval_steps_per_second": 20.897,
      "num_input_tokens_seen": 5504608,
      "step": 4000
    },
    {
      "epoch": 0.12819281736124447,
      "grad_norm": 0.01561743300408125,
      "learning_rate": 0.29264390832180853,
      "loss": 0.7983,
      "num_input_tokens_seen": 5511520,
      "step": 4005
    },
    {
      "epoch": 0.12835285833173293,
      "grad_norm": 0.018042705953121185,
      "learning_rate": 0.29262567712116144,
      "loss": 0.9869,
      "num_input_tokens_seen": 5518112,
      "step": 4010
    },
    {
      "epoch": 0.12851289930222137,
      "grad_norm": 0.011006519198417664,
      "learning_rate": 0.29260742392584266,
      "loss": 0.7036,
      "num_input_tokens_seen": 5524720,
      "step": 4015
    },
    {
      "epoch": 0.1286729402727098,
      "grad_norm": 0.019443951547145844,
      "learning_rate": 0.292589148738667,
      "loss": 0.8101,
      "num_input_tokens_seen": 5531968,
      "step": 4020
    },
    {
      "epoch": 0.12883298124319825,
      "grad_norm": 0.016778424382209778,
      "learning_rate": 0.2925708515624527,
      "loss": 0.7953,
      "num_input_tokens_seen": 5538544,
      "step": 4025
    },
    {
      "epoch": 0.1289930222136867,
      "grad_norm": 0.019470561295747757,
      "learning_rate": 0.29255253240002144,
      "loss": 0.8433,
      "num_input_tokens_seen": 5545328,
      "step": 4030
    },
    {
      "epoch": 0.12915306318417516,
      "grad_norm": 0.01075194962322712,
      "learning_rate": 0.2925341912541983,
      "loss": 1.0235,
      "num_input_tokens_seen": 5551936,
      "step": 4035
    },
    {
      "epoch": 0.1293131041546636,
      "grad_norm": 0.026279840618371964,
      "learning_rate": 0.2925158281278116,
      "loss": 0.9483,
      "num_input_tokens_seen": 5558224,
      "step": 4040
    },
    {
      "epoch": 0.12947314512515204,
      "grad_norm": 0.016100440174341202,
      "learning_rate": 0.29249744302369324,
      "loss": 0.7253,
      "num_input_tokens_seen": 5564832,
      "step": 4045
    },
    {
      "epoch": 0.12963318609564048,
      "grad_norm": 0.013285326771438122,
      "learning_rate": 0.29247903594467844,
      "loss": 0.9195,
      "num_input_tokens_seen": 5571120,
      "step": 4050
    },
    {
      "epoch": 0.12979322706612892,
      "grad_norm": 0.012391169555485249,
      "learning_rate": 0.2924606068936058,
      "loss": 0.7575,
      "num_input_tokens_seen": 5578032,
      "step": 4055
    },
    {
      "epoch": 0.1299532680366174,
      "grad_norm": 0.0105628352612257,
      "learning_rate": 0.2924421558733173,
      "loss": 0.5437,
      "num_input_tokens_seen": 5584576,
      "step": 4060
    },
    {
      "epoch": 0.13011330900710583,
      "grad_norm": 0.012182592414319515,
      "learning_rate": 0.2924236828866583,
      "loss": 0.7297,
      "num_input_tokens_seen": 5591872,
      "step": 4065
    },
    {
      "epoch": 0.13027334997759427,
      "grad_norm": 0.015593967400491238,
      "learning_rate": 0.29240518793647763,
      "loss": 0.9342,
      "num_input_tokens_seen": 5598416,
      "step": 4070
    },
    {
      "epoch": 0.1304333909480827,
      "grad_norm": 0.012053494341671467,
      "learning_rate": 0.29238667102562743,
      "loss": 0.8121,
      "num_input_tokens_seen": 5605600,
      "step": 4075
    },
    {
      "epoch": 0.13059343191857115,
      "grad_norm": 0.020935354754328728,
      "learning_rate": 0.29236813215696317,
      "loss": 0.931,
      "num_input_tokens_seen": 5612272,
      "step": 4080
    },
    {
      "epoch": 0.13075347288905959,
      "grad_norm": 0.012881615199148655,
      "learning_rate": 0.2923495713333439,
      "loss": 0.7691,
      "num_input_tokens_seen": 5618800,
      "step": 4085
    },
    {
      "epoch": 0.13091351385954805,
      "grad_norm": 0.01491338387131691,
      "learning_rate": 0.29233098855763173,
      "loss": 0.6842,
      "num_input_tokens_seen": 5625664,
      "step": 4090
    },
    {
      "epoch": 0.1310735548300365,
      "grad_norm": 0.01067566592246294,
      "learning_rate": 0.29231238383269254,
      "loss": 0.7552,
      "num_input_tokens_seen": 5632560,
      "step": 4095
    },
    {
      "epoch": 0.13123359580052493,
      "grad_norm": 0.026599271222949028,
      "learning_rate": 0.2922937571613954,
      "loss": 0.8265,
      "num_input_tokens_seen": 5639696,
      "step": 4100
    },
    {
      "epoch": 0.13139363677101337,
      "grad_norm": 0.018459642305970192,
      "learning_rate": 0.29227510854661265,
      "loss": 0.8431,
      "num_input_tokens_seen": 5646208,
      "step": 4105
    },
    {
      "epoch": 0.1315536777415018,
      "grad_norm": 0.010795938782393932,
      "learning_rate": 0.29225643799122025,
      "loss": 0.7685,
      "num_input_tokens_seen": 5653312,
      "step": 4110
    },
    {
      "epoch": 0.13171371871199028,
      "grad_norm": 0.05027012899518013,
      "learning_rate": 0.2922377454980974,
      "loss": 0.7663,
      "num_input_tokens_seen": 5659888,
      "step": 4115
    },
    {
      "epoch": 0.13187375968247872,
      "grad_norm": 0.013603785075247288,
      "learning_rate": 0.29221903107012676,
      "loss": 0.7955,
      "num_input_tokens_seen": 5666592,
      "step": 4120
    },
    {
      "epoch": 0.13203380065296716,
      "grad_norm": 0.10084132105112076,
      "learning_rate": 0.29220029471019426,
      "loss": 1.0335,
      "num_input_tokens_seen": 5673728,
      "step": 4125
    },
    {
      "epoch": 0.1321938416234556,
      "grad_norm": 0.013515281490981579,
      "learning_rate": 0.2921815364211893,
      "loss": 0.7823,
      "num_input_tokens_seen": 5680656,
      "step": 4130
    },
    {
      "epoch": 0.13235388259394404,
      "grad_norm": 0.014379197731614113,
      "learning_rate": 0.29216275620600474,
      "loss": 0.8557,
      "num_input_tokens_seen": 5687920,
      "step": 4135
    },
    {
      "epoch": 0.1325139235644325,
      "grad_norm": 0.019892407581210136,
      "learning_rate": 0.29214395406753657,
      "loss": 0.8814,
      "num_input_tokens_seen": 5694912,
      "step": 4140
    },
    {
      "epoch": 0.13267396453492095,
      "grad_norm": 0.02304846979677677,
      "learning_rate": 0.2921251300086844,
      "loss": 0.8071,
      "num_input_tokens_seen": 5701840,
      "step": 4145
    },
    {
      "epoch": 0.13283400550540939,
      "grad_norm": 0.012230420485138893,
      "learning_rate": 0.2921062840323511,
      "loss": 0.9713,
      "num_input_tokens_seen": 5708640,
      "step": 4150
    },
    {
      "epoch": 0.13299404647589783,
      "grad_norm": 0.020873988047242165,
      "learning_rate": 0.29208741614144307,
      "loss": 0.7195,
      "num_input_tokens_seen": 5715632,
      "step": 4155
    },
    {
      "epoch": 0.13315408744638627,
      "grad_norm": 0.13864214718341827,
      "learning_rate": 0.2920685263388698,
      "loss": 0.9059,
      "num_input_tokens_seen": 5722176,
      "step": 4160
    },
    {
      "epoch": 0.13331412841687473,
      "grad_norm": 0.012311822734773159,
      "learning_rate": 0.2920496146275445,
      "loss": 0.7265,
      "num_input_tokens_seen": 5728672,
      "step": 4165
    },
    {
      "epoch": 0.13347416938736317,
      "grad_norm": 0.010608458891510963,
      "learning_rate": 0.29203068101038343,
      "loss": 0.8507,
      "num_input_tokens_seen": 5735840,
      "step": 4170
    },
    {
      "epoch": 0.1336342103578516,
      "grad_norm": 0.01331733912229538,
      "learning_rate": 0.2920117254903065,
      "loss": 0.6744,
      "num_input_tokens_seen": 5742400,
      "step": 4175
    },
    {
      "epoch": 0.13379425132834005,
      "grad_norm": 0.008516818284988403,
      "learning_rate": 0.29199274807023695,
      "loss": 0.7587,
      "num_input_tokens_seen": 5749264,
      "step": 4180
    },
    {
      "epoch": 0.1339542922988285,
      "grad_norm": 0.01280166581273079,
      "learning_rate": 0.29197374875310117,
      "loss": 0.724,
      "num_input_tokens_seen": 5756752,
      "step": 4185
    },
    {
      "epoch": 0.13411433326931693,
      "grad_norm": 0.017884088680148125,
      "learning_rate": 0.2919547275418292,
      "loss": 0.7297,
      "num_input_tokens_seen": 5763760,
      "step": 4190
    },
    {
      "epoch": 0.1342743742398054,
      "grad_norm": 0.015991367399692535,
      "learning_rate": 0.29193568443935436,
      "loss": 1.0254,
      "num_input_tokens_seen": 5770880,
      "step": 4195
    },
    {
      "epoch": 0.13443441521029384,
      "grad_norm": 0.011029266752302647,
      "learning_rate": 0.2919166194486133,
      "loss": 0.9853,
      "num_input_tokens_seen": 5778176,
      "step": 4200
    },
    {
      "epoch": 0.13443441521029384,
      "eval_loss": 0.8288453817367554,
      "eval_runtime": 331.6799,
      "eval_samples_per_second": 41.866,
      "eval_steps_per_second": 20.933,
      "num_input_tokens_seen": 5778176,
      "step": 4200
    },
    {
      "epoch": 0.13459445618078228,
      "grad_norm": 0.012600620277225971,
      "learning_rate": 0.2918975325725461,
      "loss": 0.6922,
      "num_input_tokens_seen": 5784832,
      "step": 4205
    },
    {
      "epoch": 0.13475449715127072,
      "grad_norm": 0.011402672156691551,
      "learning_rate": 0.29187842381409607,
      "loss": 0.7796,
      "num_input_tokens_seen": 5791904,
      "step": 4210
    },
    {
      "epoch": 0.13491453812175916,
      "grad_norm": 0.02454453334212303,
      "learning_rate": 0.29185929317621023,
      "loss": 0.967,
      "num_input_tokens_seen": 5798512,
      "step": 4215
    },
    {
      "epoch": 0.13507457909224763,
      "grad_norm": 0.011667240411043167,
      "learning_rate": 0.29184014066183867,
      "loss": 0.6293,
      "num_input_tokens_seen": 5805616,
      "step": 4220
    },
    {
      "epoch": 0.13523462006273607,
      "grad_norm": 0.008083803579211235,
      "learning_rate": 0.2918209662739349,
      "loss": 0.8193,
      "num_input_tokens_seen": 5812544,
      "step": 4225
    },
    {
      "epoch": 0.1353946610332245,
      "grad_norm": 0.01618681661784649,
      "learning_rate": 0.29180177001545593,
      "loss": 0.7002,
      "num_input_tokens_seen": 5819744,
      "step": 4230
    },
    {
      "epoch": 0.13555470200371295,
      "grad_norm": 0.016135545447468758,
      "learning_rate": 0.29178255188936203,
      "loss": 0.768,
      "num_input_tokens_seen": 5826960,
      "step": 4235
    },
    {
      "epoch": 0.13571474297420139,
      "grad_norm": 0.013479883782565594,
      "learning_rate": 0.2917633118986169,
      "loss": 0.8748,
      "num_input_tokens_seen": 5833872,
      "step": 4240
    },
    {
      "epoch": 0.13587478394468985,
      "grad_norm": 0.0193397868424654,
      "learning_rate": 0.2917440500461875,
      "loss": 0.7458,
      "num_input_tokens_seen": 5840944,
      "step": 4245
    },
    {
      "epoch": 0.1360348249151783,
      "grad_norm": 0.017314299941062927,
      "learning_rate": 0.29172476633504435,
      "loss": 0.8038,
      "num_input_tokens_seen": 5847952,
      "step": 4250
    },
    {
      "epoch": 0.13619486588566673,
      "grad_norm": 0.012270876206457615,
      "learning_rate": 0.2917054607681612,
      "loss": 0.8733,
      "num_input_tokens_seen": 5854704,
      "step": 4255
    },
    {
      "epoch": 0.13635490685615517,
      "grad_norm": 0.01894206553697586,
      "learning_rate": 0.29168613334851523,
      "loss": 0.9074,
      "num_input_tokens_seen": 5862240,
      "step": 4260
    },
    {
      "epoch": 0.1365149478266436,
      "grad_norm": 0.011730012483894825,
      "learning_rate": 0.2916667840790869,
      "loss": 0.7397,
      "num_input_tokens_seen": 5868960,
      "step": 4265
    },
    {
      "epoch": 0.13667498879713205,
      "grad_norm": 0.018833430483937263,
      "learning_rate": 0.2916474129628603,
      "loss": 0.7751,
      "num_input_tokens_seen": 5875472,
      "step": 4270
    },
    {
      "epoch": 0.13683502976762052,
      "grad_norm": 0.016992000862956047,
      "learning_rate": 0.29162802000282245,
      "loss": 0.7013,
      "num_input_tokens_seen": 5881888,
      "step": 4275
    },
    {
      "epoch": 0.13699507073810896,
      "grad_norm": 0.005631069652736187,
      "learning_rate": 0.2916086052019642,
      "loss": 0.654,
      "num_input_tokens_seen": 5888960,
      "step": 4280
    },
    {
      "epoch": 0.1371551117085974,
      "grad_norm": 0.012740897946059704,
      "learning_rate": 0.2915891685632794,
      "loss": 0.6325,
      "num_input_tokens_seen": 5895696,
      "step": 4285
    },
    {
      "epoch": 0.13731515267908584,
      "grad_norm": 0.012844252400100231,
      "learning_rate": 0.29156971008976545,
      "loss": 0.7998,
      "num_input_tokens_seen": 5902832,
      "step": 4290
    },
    {
      "epoch": 0.13747519364957428,
      "grad_norm": 0.015334940515458584,
      "learning_rate": 0.2915502297844232,
      "loss": 1.0256,
      "num_input_tokens_seen": 5909984,
      "step": 4295
    },
    {
      "epoch": 0.13763523462006275,
      "grad_norm": 0.01812255196273327,
      "learning_rate": 0.2915307276502566,
      "loss": 0.6865,
      "num_input_tokens_seen": 5916624,
      "step": 4300
    },
    {
      "epoch": 0.1377952755905512,
      "grad_norm": 0.053552959114313126,
      "learning_rate": 0.29151120369027334,
      "loss": 1.1652,
      "num_input_tokens_seen": 5923760,
      "step": 4305
    },
    {
      "epoch": 0.13795531656103963,
      "grad_norm": 0.014496814459562302,
      "learning_rate": 0.29149165790748405,
      "loss": 0.8588,
      "num_input_tokens_seen": 5930736,
      "step": 4310
    },
    {
      "epoch": 0.13811535753152807,
      "grad_norm": 0.018492834642529488,
      "learning_rate": 0.291472090304903,
      "loss": 0.8324,
      "num_input_tokens_seen": 5937696,
      "step": 4315
    },
    {
      "epoch": 0.1382753985020165,
      "grad_norm": 0.018123546615242958,
      "learning_rate": 0.2914525008855478,
      "loss": 0.8091,
      "num_input_tokens_seen": 5944768,
      "step": 4320
    },
    {
      "epoch": 0.13843543947250497,
      "grad_norm": 0.10142970085144043,
      "learning_rate": 0.2914328896524394,
      "loss": 0.7821,
      "num_input_tokens_seen": 5952208,
      "step": 4325
    },
    {
      "epoch": 0.1385954804429934,
      "grad_norm": 0.025013364851474762,
      "learning_rate": 0.291413256608602,
      "loss": 0.716,
      "num_input_tokens_seen": 5959488,
      "step": 4330
    },
    {
      "epoch": 0.13875552141348185,
      "grad_norm": 0.01654975116252899,
      "learning_rate": 0.29139360175706336,
      "loss": 0.7018,
      "num_input_tokens_seen": 5966080,
      "step": 4335
    },
    {
      "epoch": 0.1389155623839703,
      "grad_norm": 0.018071841448545456,
      "learning_rate": 0.2913739251008544,
      "loss": 0.7564,
      "num_input_tokens_seen": 5973184,
      "step": 4340
    },
    {
      "epoch": 0.13907560335445873,
      "grad_norm": 0.017684560269117355,
      "learning_rate": 0.29135422664300964,
      "loss": 0.9258,
      "num_input_tokens_seen": 5979872,
      "step": 4345
    },
    {
      "epoch": 0.1392356443249472,
      "grad_norm": 0.020020833238959312,
      "learning_rate": 0.29133450638656677,
      "loss": 0.6741,
      "num_input_tokens_seen": 5986704,
      "step": 4350
    },
    {
      "epoch": 0.13939568529543564,
      "grad_norm": 0.01477797981351614,
      "learning_rate": 0.2913147643345669,
      "loss": 0.9302,
      "num_input_tokens_seen": 5993728,
      "step": 4355
    },
    {
      "epoch": 0.13955572626592408,
      "grad_norm": 0.013320963829755783,
      "learning_rate": 0.29129500049005447,
      "loss": 0.6499,
      "num_input_tokens_seen": 6000544,
      "step": 4360
    },
    {
      "epoch": 0.13971576723641252,
      "grad_norm": 0.025219688192009926,
      "learning_rate": 0.2912752148560773,
      "loss": 0.8157,
      "num_input_tokens_seen": 6007552,
      "step": 4365
    },
    {
      "epoch": 0.13987580820690096,
      "grad_norm": 0.011474275961518288,
      "learning_rate": 0.2912554074356866,
      "loss": 0.6219,
      "num_input_tokens_seen": 6015424,
      "step": 4370
    },
    {
      "epoch": 0.1400358491773894,
      "grad_norm": 0.010486027225852013,
      "learning_rate": 0.2912355782319371,
      "loss": 0.7586,
      "num_input_tokens_seen": 6022224,
      "step": 4375
    },
    {
      "epoch": 0.14019589014787787,
      "grad_norm": 0.009824017062783241,
      "learning_rate": 0.2912157272478864,
      "loss": 0.7323,
      "num_input_tokens_seen": 6028880,
      "step": 4380
    },
    {
      "epoch": 0.1403559311183663,
      "grad_norm": 0.01726258173584938,
      "learning_rate": 0.291195854486596,
      "loss": 0.8881,
      "num_input_tokens_seen": 6035872,
      "step": 4385
    },
    {
      "epoch": 0.14051597208885475,
      "grad_norm": 0.012011564336717129,
      "learning_rate": 0.2911759599511305,
      "loss": 0.7144,
      "num_input_tokens_seen": 6042480,
      "step": 4390
    },
    {
      "epoch": 0.14067601305934319,
      "grad_norm": 0.03181210160255432,
      "learning_rate": 0.29115604364455777,
      "loss": 0.7822,
      "num_input_tokens_seen": 6049104,
      "step": 4395
    },
    {
      "epoch": 0.14083605402983163,
      "grad_norm": 0.05834176018834114,
      "learning_rate": 0.2911361055699493,
      "loss": 1.0905,
      "num_input_tokens_seen": 6055712,
      "step": 4400
    },
    {
      "epoch": 0.14083605402983163,
      "eval_loss": 0.8487065434455872,
      "eval_runtime": 331.994,
      "eval_samples_per_second": 41.826,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 6055712,
      "step": 4400
    },
    {
      "epoch": 0.1409960950003201,
      "grad_norm": 0.011283257976174355,
      "learning_rate": 0.2911161457303797,
      "loss": 0.8051,
      "num_input_tokens_seen": 6062800,
      "step": 4405
    },
    {
      "epoch": 0.14115613597080853,
      "grad_norm": 0.011924389749765396,
      "learning_rate": 0.291096164128927,
      "loss": 0.7574,
      "num_input_tokens_seen": 6069744,
      "step": 4410
    },
    {
      "epoch": 0.14131617694129697,
      "grad_norm": 0.01807926408946514,
      "learning_rate": 0.2910761607686727,
      "loss": 0.9341,
      "num_input_tokens_seen": 6076928,
      "step": 4415
    },
    {
      "epoch": 0.1414762179117854,
      "grad_norm": 0.014111528173089027,
      "learning_rate": 0.2910561356527016,
      "loss": 0.9625,
      "num_input_tokens_seen": 6084400,
      "step": 4420
    },
    {
      "epoch": 0.14163625888227385,
      "grad_norm": 0.007742114365100861,
      "learning_rate": 0.2910360887841017,
      "loss": 1.124,
      "num_input_tokens_seen": 6091328,
      "step": 4425
    },
    {
      "epoch": 0.14179629985276232,
      "grad_norm": 0.008953777141869068,
      "learning_rate": 0.2910160201659645,
      "loss": 0.9131,
      "num_input_tokens_seen": 6098032,
      "step": 4430
    },
    {
      "epoch": 0.14195634082325076,
      "grad_norm": 0.007890722714364529,
      "learning_rate": 0.29099592980138494,
      "loss": 0.6938,
      "num_input_tokens_seen": 6104832,
      "step": 4435
    },
    {
      "epoch": 0.1421163817937392,
      "grad_norm": 0.012505297549068928,
      "learning_rate": 0.29097581769346115,
      "loss": 0.7637,
      "num_input_tokens_seen": 6111392,
      "step": 4440
    },
    {
      "epoch": 0.14227642276422764,
      "grad_norm": 0.017935264855623245,
      "learning_rate": 0.29095568384529463,
      "loss": 0.8296,
      "num_input_tokens_seen": 6117872,
      "step": 4445
    },
    {
      "epoch": 0.14243646373471608,
      "grad_norm": 0.009602420963346958,
      "learning_rate": 0.2909355282599903,
      "loss": 0.779,
      "num_input_tokens_seen": 6124624,
      "step": 4450
    },
    {
      "epoch": 0.14259650470520452,
      "grad_norm": 0.011581191793084145,
      "learning_rate": 0.29091535094065635,
      "loss": 0.8388,
      "num_input_tokens_seen": 6131568,
      "step": 4455
    },
    {
      "epoch": 0.142756545675693,
      "grad_norm": 0.012436475604772568,
      "learning_rate": 0.2908951518904045,
      "loss": 0.8126,
      "num_input_tokens_seen": 6138784,
      "step": 4460
    },
    {
      "epoch": 0.14291658664618143,
      "grad_norm": 0.01009361632168293,
      "learning_rate": 0.29087493111234963,
      "loss": 0.9244,
      "num_input_tokens_seen": 6145904,
      "step": 4465
    },
    {
      "epoch": 0.14307662761666987,
      "grad_norm": 0.011279134079813957,
      "learning_rate": 0.29085468860961,
      "loss": 0.7118,
      "num_input_tokens_seen": 6152848,
      "step": 4470
    },
    {
      "epoch": 0.1432366685871583,
      "grad_norm": 0.01147132646292448,
      "learning_rate": 0.2908344243853073,
      "loss": 0.6177,
      "num_input_tokens_seen": 6159568,
      "step": 4475
    },
    {
      "epoch": 0.14339670955764675,
      "grad_norm": 0.05423408001661301,
      "learning_rate": 0.2908141384425666,
      "loss": 0.7654,
      "num_input_tokens_seen": 6166448,
      "step": 4480
    },
    {
      "epoch": 0.1435567505281352,
      "grad_norm": 0.008657590486109257,
      "learning_rate": 0.2907938307845161,
      "loss": 0.6749,
      "num_input_tokens_seen": 6172832,
      "step": 4485
    },
    {
      "epoch": 0.14371679149862365,
      "grad_norm": 0.008594881743192673,
      "learning_rate": 0.2907735014142876,
      "loss": 0.7225,
      "num_input_tokens_seen": 6180176,
      "step": 4490
    },
    {
      "epoch": 0.1438768324691121,
      "grad_norm": 0.024600133299827576,
      "learning_rate": 0.2907531503350161,
      "loss": 0.9895,
      "num_input_tokens_seen": 6187168,
      "step": 4495
    },
    {
      "epoch": 0.14403687343960053,
      "grad_norm": 0.009852620773017406,
      "learning_rate": 0.29073277754983995,
      "loss": 0.8866,
      "num_input_tokens_seen": 6194928,
      "step": 4500
    },
    {
      "epoch": 0.14419691441008897,
      "grad_norm": 0.00954913068562746,
      "learning_rate": 0.290712383061901,
      "loss": 1.0168,
      "num_input_tokens_seen": 6201648,
      "step": 4505
    },
    {
      "epoch": 0.14435695538057744,
      "grad_norm": 0.010376465506851673,
      "learning_rate": 0.2906919668743443,
      "loss": 0.7407,
      "num_input_tokens_seen": 6208224,
      "step": 4510
    },
    {
      "epoch": 0.14451699635106588,
      "grad_norm": 0.013114885427057743,
      "learning_rate": 0.29067152899031823,
      "loss": 0.885,
      "num_input_tokens_seen": 6214928,
      "step": 4515
    },
    {
      "epoch": 0.14467703732155432,
      "grad_norm": 0.009842830710113049,
      "learning_rate": 0.2906510694129746,
      "loss": 0.7869,
      "num_input_tokens_seen": 6221744,
      "step": 4520
    },
    {
      "epoch": 0.14483707829204276,
      "grad_norm": 0.006115804426372051,
      "learning_rate": 0.2906305881454685,
      "loss": 0.7465,
      "num_input_tokens_seen": 6228880,
      "step": 4525
    },
    {
      "epoch": 0.1449971192625312,
      "grad_norm": 0.010713424533605576,
      "learning_rate": 0.2906100851909585,
      "loss": 0.8935,
      "num_input_tokens_seen": 6235616,
      "step": 4530
    },
    {
      "epoch": 0.14515716023301967,
      "grad_norm": 0.008180897682905197,
      "learning_rate": 0.29058956055260626,
      "loss": 0.7337,
      "num_input_tokens_seen": 6242848,
      "step": 4535
    },
    {
      "epoch": 0.1453172012035081,
      "grad_norm": 0.008185225538909435,
      "learning_rate": 0.2905690142335771,
      "loss": 0.8499,
      "num_input_tokens_seen": 6249616,
      "step": 4540
    },
    {
      "epoch": 0.14547724217399655,
      "grad_norm": 0.018000105395913124,
      "learning_rate": 0.29054844623703946,
      "loss": 0.9619,
      "num_input_tokens_seen": 6256352,
      "step": 4545
    },
    {
      "epoch": 0.14563728314448499,
      "grad_norm": 0.008042537607252598,
      "learning_rate": 0.2905278565661651,
      "loss": 0.6768,
      "num_input_tokens_seen": 6263120,
      "step": 4550
    },
    {
      "epoch": 0.14579732411497343,
      "grad_norm": 0.007113131228834391,
      "learning_rate": 0.2905072452241293,
      "loss": 0.6581,
      "num_input_tokens_seen": 6269840,
      "step": 4555
    },
    {
      "epoch": 0.14595736508546187,
      "grad_norm": 0.04037633165717125,
      "learning_rate": 0.2904866122141106,
      "loss": 0.944,
      "num_input_tokens_seen": 6276720,
      "step": 4560
    },
    {
      "epoch": 0.14611740605595033,
      "grad_norm": 0.012480218894779682,
      "learning_rate": 0.2904659575392908,
      "loss": 0.7462,
      "num_input_tokens_seen": 6283408,
      "step": 4565
    },
    {
      "epoch": 0.14627744702643877,
      "grad_norm": 0.0114600183442235,
      "learning_rate": 0.2904452812028551,
      "loss": 0.8574,
      "num_input_tokens_seen": 6290112,
      "step": 4570
    },
    {
      "epoch": 0.1464374879969272,
      "grad_norm": 0.014767049811780453,
      "learning_rate": 0.2904245832079922,
      "loss": 0.8341,
      "num_input_tokens_seen": 6297088,
      "step": 4575
    },
    {
      "epoch": 0.14659752896741565,
      "grad_norm": 0.016923412680625916,
      "learning_rate": 0.29040386355789377,
      "loss": 0.8368,
      "num_input_tokens_seen": 6304144,
      "step": 4580
    },
    {
      "epoch": 0.1467575699379041,
      "grad_norm": 0.009206613525748253,
      "learning_rate": 0.29038312225575524,
      "loss": 0.8817,
      "num_input_tokens_seen": 6311040,
      "step": 4585
    },
    {
      "epoch": 0.14691761090839256,
      "grad_norm": 0.012517068535089493,
      "learning_rate": 0.29036235930477505,
      "loss": 0.8249,
      "num_input_tokens_seen": 6317920,
      "step": 4590
    },
    {
      "epoch": 0.147077651878881,
      "grad_norm": 0.00710597587749362,
      "learning_rate": 0.29034157470815514,
      "loss": 0.6661,
      "num_input_tokens_seen": 6324544,
      "step": 4595
    },
    {
      "epoch": 0.14723769284936944,
      "grad_norm": 0.008931620977818966,
      "learning_rate": 0.2903207684691008,
      "loss": 0.846,
      "num_input_tokens_seen": 6331680,
      "step": 4600
    },
    {
      "epoch": 0.14723769284936944,
      "eval_loss": 0.8267067074775696,
      "eval_runtime": 331.5649,
      "eval_samples_per_second": 41.88,
      "eval_steps_per_second": 20.94,
      "num_input_tokens_seen": 6331680,
      "step": 4600
    },
    {
      "epoch": 0.14739773381985788,
      "grad_norm": 0.008797709830105305,
      "learning_rate": 0.29029994059082054,
      "loss": 0.794,
      "num_input_tokens_seen": 6338304,
      "step": 4605
    },
    {
      "epoch": 0.14755777479034632,
      "grad_norm": 0.007075788453221321,
      "learning_rate": 0.2902790910765264,
      "loss": 0.7686,
      "num_input_tokens_seen": 6345456,
      "step": 4610
    },
    {
      "epoch": 0.1477178157608348,
      "grad_norm": 0.009171213954687119,
      "learning_rate": 0.29025821992943346,
      "loss": 0.8787,
      "num_input_tokens_seen": 6351984,
      "step": 4615
    },
    {
      "epoch": 0.14787785673132323,
      "grad_norm": 0.010541302151978016,
      "learning_rate": 0.29023732715276046,
      "loss": 0.9245,
      "num_input_tokens_seen": 6358944,
      "step": 4620
    },
    {
      "epoch": 0.14803789770181167,
      "grad_norm": 0.009613308124244213,
      "learning_rate": 0.2902164127497293,
      "loss": 0.9031,
      "num_input_tokens_seen": 6365840,
      "step": 4625
    },
    {
      "epoch": 0.1481979386723001,
      "grad_norm": 0.008268777281045914,
      "learning_rate": 0.2901954767235652,
      "loss": 0.7588,
      "num_input_tokens_seen": 6373168,
      "step": 4630
    },
    {
      "epoch": 0.14835797964278855,
      "grad_norm": 0.008914311416447163,
      "learning_rate": 0.2901745190774968,
      "loss": 0.8275,
      "num_input_tokens_seen": 6379632,
      "step": 4635
    },
    {
      "epoch": 0.14851802061327699,
      "grad_norm": 0.008368615992367268,
      "learning_rate": 0.290153539814756,
      "loss": 0.6825,
      "num_input_tokens_seen": 6386272,
      "step": 4640
    },
    {
      "epoch": 0.14867806158376545,
      "grad_norm": 0.021809889003634453,
      "learning_rate": 0.2901325389385781,
      "loss": 0.8365,
      "num_input_tokens_seen": 6393280,
      "step": 4645
    },
    {
      "epoch": 0.1488381025542539,
      "grad_norm": 0.029776757583022118,
      "learning_rate": 0.2901115164522016,
      "loss": 0.8691,
      "num_input_tokens_seen": 6400032,
      "step": 4650
    },
    {
      "epoch": 0.14899814352474233,
      "grad_norm": 0.01189467404037714,
      "learning_rate": 0.29009047235886865,
      "loss": 0.6326,
      "num_input_tokens_seen": 6406496,
      "step": 4655
    },
    {
      "epoch": 0.14915818449523077,
      "grad_norm": 0.008430328220129013,
      "learning_rate": 0.2900694066618243,
      "loss": 0.9843,
      "num_input_tokens_seen": 6413856,
      "step": 4660
    },
    {
      "epoch": 0.1493182254657192,
      "grad_norm": 0.008506018668413162,
      "learning_rate": 0.2900483193643172,
      "loss": 0.7129,
      "num_input_tokens_seen": 6420256,
      "step": 4665
    },
    {
      "epoch": 0.14947826643620768,
      "grad_norm": 0.017168166115880013,
      "learning_rate": 0.29002721046959934,
      "loss": 0.8408,
      "num_input_tokens_seen": 6426960,
      "step": 4670
    },
    {
      "epoch": 0.14963830740669612,
      "grad_norm": 0.026883108541369438,
      "learning_rate": 0.29000607998092587,
      "loss": 0.8086,
      "num_input_tokens_seen": 6434048,
      "step": 4675
    },
    {
      "epoch": 0.14979834837718456,
      "grad_norm": 0.021148962900042534,
      "learning_rate": 0.2899849279015555,
      "loss": 0.9572,
      "num_input_tokens_seen": 6440736,
      "step": 4680
    },
    {
      "epoch": 0.149958389347673,
      "grad_norm": 0.017451144754886627,
      "learning_rate": 0.28996375423475007,
      "loss": 0.7554,
      "num_input_tokens_seen": 6447344,
      "step": 4685
    },
    {
      "epoch": 0.15011843031816144,
      "grad_norm": 0.020323246717453003,
      "learning_rate": 0.28994255898377486,
      "loss": 0.8484,
      "num_input_tokens_seen": 6454112,
      "step": 4690
    },
    {
      "epoch": 0.1502784712886499,
      "grad_norm": 0.011208818294107914,
      "learning_rate": 0.2899213421518984,
      "loss": 0.7319,
      "num_input_tokens_seen": 6461424,
      "step": 4695
    },
    {
      "epoch": 0.15043851225913835,
      "grad_norm": 0.014643831178545952,
      "learning_rate": 0.2899001037423926,
      "loss": 0.6548,
      "num_input_tokens_seen": 6467872,
      "step": 4700
    },
    {
      "epoch": 0.15059855322962679,
      "grad_norm": 0.013093414716422558,
      "learning_rate": 0.28987884375853273,
      "loss": 0.7262,
      "num_input_tokens_seen": 6474512,
      "step": 4705
    },
    {
      "epoch": 0.15075859420011523,
      "grad_norm": 0.01956348866224289,
      "learning_rate": 0.2898575622035974,
      "loss": 0.6472,
      "num_input_tokens_seen": 6481248,
      "step": 4710
    },
    {
      "epoch": 0.15091863517060367,
      "grad_norm": 0.03502310812473297,
      "learning_rate": 0.2898362590808683,
      "loss": 1.1539,
      "num_input_tokens_seen": 6487920,
      "step": 4715
    },
    {
      "epoch": 0.15107867614109213,
      "grad_norm": 0.025840604677796364,
      "learning_rate": 0.2898149343936308,
      "loss": 1.0588,
      "num_input_tokens_seen": 6495056,
      "step": 4720
    },
    {
      "epoch": 0.15123871711158057,
      "grad_norm": 0.009915349073708057,
      "learning_rate": 0.2897935881451734,
      "loss": 0.881,
      "num_input_tokens_seen": 6502448,
      "step": 4725
    },
    {
      "epoch": 0.151398758082069,
      "grad_norm": 0.0214952751994133,
      "learning_rate": 0.28977222033878797,
      "loss": 0.666,
      "num_input_tokens_seen": 6509280,
      "step": 4730
    },
    {
      "epoch": 0.15155879905255745,
      "grad_norm": 0.009903205558657646,
      "learning_rate": 0.28975083097776966,
      "loss": 0.9564,
      "num_input_tokens_seen": 6516224,
      "step": 4735
    },
    {
      "epoch": 0.1517188400230459,
      "grad_norm": 0.01715020090341568,
      "learning_rate": 0.28972942006541696,
      "loss": 0.8414,
      "num_input_tokens_seen": 6522912,
      "step": 4740
    },
    {
      "epoch": 0.15187888099353433,
      "grad_norm": 0.009732005186378956,
      "learning_rate": 0.2897079876050318,
      "loss": 0.8277,
      "num_input_tokens_seen": 6529680,
      "step": 4745
    },
    {
      "epoch": 0.1520389219640228,
      "grad_norm": 0.01412002183496952,
      "learning_rate": 0.2896865335999192,
      "loss": 0.7047,
      "num_input_tokens_seen": 6536352,
      "step": 4750
    },
    {
      "epoch": 0.15219896293451124,
      "grad_norm": 0.01316248532384634,
      "learning_rate": 0.28966505805338777,
      "loss": 0.6823,
      "num_input_tokens_seen": 6543056,
      "step": 4755
    },
    {
      "epoch": 0.15235900390499968,
      "grad_norm": 0.00977543368935585,
      "learning_rate": 0.2896435609687492,
      "loss": 0.88,
      "num_input_tokens_seen": 6549504,
      "step": 4760
    },
    {
      "epoch": 0.15251904487548812,
      "grad_norm": 0.01832258515059948,
      "learning_rate": 0.2896220423493187,
      "loss": 0.8647,
      "num_input_tokens_seen": 6556384,
      "step": 4765
    },
    {
      "epoch": 0.15267908584597656,
      "grad_norm": 0.014380532316863537,
      "learning_rate": 0.28960050219841466,
      "loss": 0.8197,
      "num_input_tokens_seen": 6563024,
      "step": 4770
    },
    {
      "epoch": 0.15283912681646503,
      "grad_norm": 0.008163737133145332,
      "learning_rate": 0.28957894051935884,
      "loss": 0.8141,
      "num_input_tokens_seen": 6569712,
      "step": 4775
    },
    {
      "epoch": 0.15299916778695347,
      "grad_norm": 0.017179204151034355,
      "learning_rate": 0.2895573573154764,
      "loss": 0.9,
      "num_input_tokens_seen": 6576576,
      "step": 4780
    },
    {
      "epoch": 0.1531592087574419,
      "grad_norm": 0.0116822998970747,
      "learning_rate": 0.28953575259009556,
      "loss": 0.7184,
      "num_input_tokens_seen": 6583696,
      "step": 4785
    },
    {
      "epoch": 0.15331924972793035,
      "grad_norm": 0.015331893227994442,
      "learning_rate": 0.2895141263465482,
      "loss": 0.8442,
      "num_input_tokens_seen": 6590608,
      "step": 4790
    },
    {
      "epoch": 0.15347929069841879,
      "grad_norm": 0.012621820904314518,
      "learning_rate": 0.28949247858816934,
      "loss": 0.8584,
      "num_input_tokens_seen": 6597776,
      "step": 4795
    },
    {
      "epoch": 0.15363933166890725,
      "grad_norm": 0.007776450365781784,
      "learning_rate": 0.2894708093182973,
      "loss": 0.7685,
      "num_input_tokens_seen": 6604544,
      "step": 4800
    },
    {
      "epoch": 0.15363933166890725,
      "eval_loss": 0.8237047791481018,
      "eval_runtime": 331.4774,
      "eval_samples_per_second": 41.891,
      "eval_steps_per_second": 20.946,
      "num_input_tokens_seen": 6604544,
      "step": 4800
    },
    {
      "epoch": 0.1537993726393957,
      "grad_norm": 0.010900052264332771,
      "learning_rate": 0.2894491185402737,
      "loss": 0.8364,
      "num_input_tokens_seen": 6611472,
      "step": 4805
    },
    {
      "epoch": 0.15395941360988413,
      "grad_norm": 0.031344443559646606,
      "learning_rate": 0.2894274062574437,
      "loss": 0.965,
      "num_input_tokens_seen": 6618912,
      "step": 4810
    },
    {
      "epoch": 0.15411945458037257,
      "grad_norm": 0.014543842524290085,
      "learning_rate": 0.2894056724731554,
      "loss": 0.8366,
      "num_input_tokens_seen": 6625712,
      "step": 4815
    },
    {
      "epoch": 0.154279495550861,
      "grad_norm": 0.04437090829014778,
      "learning_rate": 0.28938391719076056,
      "loss": 0.8798,
      "num_input_tokens_seen": 6632432,
      "step": 4820
    },
    {
      "epoch": 0.15443953652134945,
      "grad_norm": 0.013954692520201206,
      "learning_rate": 0.28936214041361413,
      "loss": 0.6264,
      "num_input_tokens_seen": 6639392,
      "step": 4825
    },
    {
      "epoch": 0.15459957749183792,
      "grad_norm": 0.021928558126091957,
      "learning_rate": 0.2893403421450743,
      "loss": 0.6813,
      "num_input_tokens_seen": 6646224,
      "step": 4830
    },
    {
      "epoch": 0.15475961846232636,
      "grad_norm": 0.027742795646190643,
      "learning_rate": 0.2893185223885026,
      "loss": 0.994,
      "num_input_tokens_seen": 6653040,
      "step": 4835
    },
    {
      "epoch": 0.1549196594328148,
      "grad_norm": 0.01090640015900135,
      "learning_rate": 0.289296681147264,
      "loss": 0.9579,
      "num_input_tokens_seen": 6659968,
      "step": 4840
    },
    {
      "epoch": 0.15507970040330324,
      "grad_norm": 0.01365073211491108,
      "learning_rate": 0.28927481842472663,
      "loss": 0.7154,
      "num_input_tokens_seen": 6666912,
      "step": 4845
    },
    {
      "epoch": 0.15523974137379168,
      "grad_norm": 0.008858074434101582,
      "learning_rate": 0.28925293422426207,
      "loss": 0.611,
      "num_input_tokens_seen": 6673632,
      "step": 4850
    },
    {
      "epoch": 0.15539978234428015,
      "grad_norm": 0.03988828510046005,
      "learning_rate": 0.28923102854924504,
      "loss": 0.8318,
      "num_input_tokens_seen": 6680832,
      "step": 4855
    },
    {
      "epoch": 0.15555982331476859,
      "grad_norm": 0.008512957021594048,
      "learning_rate": 0.2892091014030537,
      "loss": 0.7525,
      "num_input_tokens_seen": 6687440,
      "step": 4860
    },
    {
      "epoch": 0.15571986428525703,
      "grad_norm": 0.01773846708238125,
      "learning_rate": 0.2891871527890696,
      "loss": 0.7346,
      "num_input_tokens_seen": 6694208,
      "step": 4865
    },
    {
      "epoch": 0.15587990525574547,
      "grad_norm": 0.008881255984306335,
      "learning_rate": 0.2891651827106773,
      "loss": 0.649,
      "num_input_tokens_seen": 6701168,
      "step": 4870
    },
    {
      "epoch": 0.1560399462262339,
      "grad_norm": 0.010856064967811108,
      "learning_rate": 0.2891431911712651,
      "loss": 0.823,
      "num_input_tokens_seen": 6707808,
      "step": 4875
    },
    {
      "epoch": 0.15619998719672237,
      "grad_norm": 0.009464248083531857,
      "learning_rate": 0.2891211781742241,
      "loss": 0.9859,
      "num_input_tokens_seen": 6714400,
      "step": 4880
    },
    {
      "epoch": 0.1563600281672108,
      "grad_norm": 0.011720051057636738,
      "learning_rate": 0.2890991437229492,
      "loss": 0.8416,
      "num_input_tokens_seen": 6721632,
      "step": 4885
    },
    {
      "epoch": 0.15652006913769925,
      "grad_norm": 0.07432214170694351,
      "learning_rate": 0.2890770878208383,
      "loss": 0.9962,
      "num_input_tokens_seen": 6729392,
      "step": 4890
    },
    {
      "epoch": 0.1566801101081877,
      "grad_norm": 0.12682794034481049,
      "learning_rate": 0.28905501047129273,
      "loss": 0.8453,
      "num_input_tokens_seen": 6736320,
      "step": 4895
    },
    {
      "epoch": 0.15684015107867613,
      "grad_norm": 0.03122575953602791,
      "learning_rate": 0.289032911677717,
      "loss": 0.9871,
      "num_input_tokens_seen": 6743664,
      "step": 4900
    },
    {
      "epoch": 0.1570001920491646,
      "grad_norm": 0.015755267813801765,
      "learning_rate": 0.28901079144351915,
      "loss": 0.9584,
      "num_input_tokens_seen": 6750672,
      "step": 4905
    },
    {
      "epoch": 0.15716023301965304,
      "grad_norm": 0.02658943273127079,
      "learning_rate": 0.2889886497721103,
      "loss": 0.9589,
      "num_input_tokens_seen": 6757488,
      "step": 4910
    },
    {
      "epoch": 0.15732027399014148,
      "grad_norm": 0.016869015991687775,
      "learning_rate": 0.28896648666690505,
      "loss": 0.847,
      "num_input_tokens_seen": 6764224,
      "step": 4915
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 0.032933540642261505,
      "learning_rate": 0.2889443021313212,
      "loss": 0.9978,
      "num_input_tokens_seen": 6770976,
      "step": 4920
    },
    {
      "epoch": 0.15764035593111836,
      "grad_norm": 0.013297748751938343,
      "learning_rate": 0.28892209616877984,
      "loss": 0.7923,
      "num_input_tokens_seen": 6777776,
      "step": 4925
    },
    {
      "epoch": 0.1578003969016068,
      "grad_norm": 0.015221870504319668,
      "learning_rate": 0.28889986878270546,
      "loss": 0.6964,
      "num_input_tokens_seen": 6784944,
      "step": 4930
    },
    {
      "epoch": 0.15796043787209527,
      "grad_norm": 0.048350848257541656,
      "learning_rate": 0.28887761997652583,
      "loss": 0.8735,
      "num_input_tokens_seen": 6791856,
      "step": 4935
    },
    {
      "epoch": 0.1581204788425837,
      "grad_norm": 0.014669148251414299,
      "learning_rate": 0.2888553497536719,
      "loss": 0.8889,
      "num_input_tokens_seen": 6798400,
      "step": 4940
    },
    {
      "epoch": 0.15828051981307215,
      "grad_norm": 0.008156539872288704,
      "learning_rate": 0.2888330581175781,
      "loss": 0.7466,
      "num_input_tokens_seen": 6805552,
      "step": 4945
    },
    {
      "epoch": 0.15844056078356059,
      "grad_norm": 0.015058329328894615,
      "learning_rate": 0.28881074507168203,
      "loss": 0.8238,
      "num_input_tokens_seen": 6812720,
      "step": 4950
    },
    {
      "epoch": 0.15860060175404903,
      "grad_norm": 0.01166201289743185,
      "learning_rate": 0.2887884106194247,
      "loss": 0.8245,
      "num_input_tokens_seen": 6819568,
      "step": 4955
    },
    {
      "epoch": 0.1587606427245375,
      "grad_norm": 0.014331943355500698,
      "learning_rate": 0.28876605476425027,
      "loss": 0.8846,
      "num_input_tokens_seen": 6826256,
      "step": 4960
    },
    {
      "epoch": 0.15892068369502593,
      "grad_norm": 0.059522729367017746,
      "learning_rate": 0.2887436775096064,
      "loss": 0.8323,
      "num_input_tokens_seen": 6833568,
      "step": 4965
    },
    {
      "epoch": 0.15908072466551437,
      "grad_norm": 0.015250430442392826,
      "learning_rate": 0.2887212788589439,
      "loss": 1.004,
      "num_input_tokens_seen": 6840208,
      "step": 4970
    },
    {
      "epoch": 0.1592407656360028,
      "grad_norm": 0.012666290625929832,
      "learning_rate": 0.2886988588157169,
      "loss": 0.9049,
      "num_input_tokens_seen": 6847408,
      "step": 4975
    },
    {
      "epoch": 0.15940080660649125,
      "grad_norm": 0.012473070062696934,
      "learning_rate": 0.28867641738338284,
      "loss": 0.9348,
      "num_input_tokens_seen": 6854496,
      "step": 4980
    },
    {
      "epoch": 0.15956084757697972,
      "grad_norm": 0.03070804290473461,
      "learning_rate": 0.2886539545654026,
      "loss": 0.8848,
      "num_input_tokens_seen": 6861392,
      "step": 4985
    },
    {
      "epoch": 0.15972088854746816,
      "grad_norm": 0.008796554058790207,
      "learning_rate": 0.28863147036524006,
      "loss": 0.7215,
      "num_input_tokens_seen": 6868368,
      "step": 4990
    },
    {
      "epoch": 0.1598809295179566,
      "grad_norm": 0.0094990199431777,
      "learning_rate": 0.2886089647863626,
      "loss": 0.8134,
      "num_input_tokens_seen": 6875584,
      "step": 4995
    },
    {
      "epoch": 0.16004097048844504,
      "grad_norm": 0.02198227308690548,
      "learning_rate": 0.288586437832241,
      "loss": 0.8457,
      "num_input_tokens_seen": 6882256,
      "step": 5000
    },
    {
      "epoch": 0.16004097048844504,
      "eval_loss": 0.8813772797584534,
      "eval_runtime": 332.1618,
      "eval_samples_per_second": 41.805,
      "eval_steps_per_second": 20.902,
      "num_input_tokens_seen": 6882256,
      "step": 5000
    },
    {
      "epoch": 0.16020101145893348,
      "grad_norm": 0.05035294592380524,
      "learning_rate": 0.28856388950634904,
      "loss": 0.8566,
      "num_input_tokens_seen": 6889504,
      "step": 5005
    },
    {
      "epoch": 0.16036105242942192,
      "grad_norm": 0.010688668116927147,
      "learning_rate": 0.288541319812164,
      "loss": 0.8136,
      "num_input_tokens_seen": 6896608,
      "step": 5010
    },
    {
      "epoch": 0.1605210933999104,
      "grad_norm": 0.010319394059479237,
      "learning_rate": 0.2885187287531665,
      "loss": 0.8221,
      "num_input_tokens_seen": 6903216,
      "step": 5015
    },
    {
      "epoch": 0.16068113437039883,
      "grad_norm": 0.011671096086502075,
      "learning_rate": 0.2884961163328402,
      "loss": 0.8593,
      "num_input_tokens_seen": 6910912,
      "step": 5020
    },
    {
      "epoch": 0.16084117534088727,
      "grad_norm": 0.010354382917284966,
      "learning_rate": 0.28847348255467237,
      "loss": 0.8014,
      "num_input_tokens_seen": 6917616,
      "step": 5025
    },
    {
      "epoch": 0.1610012163113757,
      "grad_norm": 0.011349499225616455,
      "learning_rate": 0.28845082742215333,
      "loss": 0.7714,
      "num_input_tokens_seen": 6924624,
      "step": 5030
    },
    {
      "epoch": 0.16116125728186415,
      "grad_norm": 0.008388606831431389,
      "learning_rate": 0.2884281509387769,
      "loss": 0.6245,
      "num_input_tokens_seen": 6931520,
      "step": 5035
    },
    {
      "epoch": 0.1613212982523526,
      "grad_norm": 0.012093384750187397,
      "learning_rate": 0.2884054531080399,
      "loss": 0.741,
      "num_input_tokens_seen": 6938480,
      "step": 5040
    },
    {
      "epoch": 0.16148133922284105,
      "grad_norm": 0.011984056793153286,
      "learning_rate": 0.28838273393344277,
      "loss": 0.7265,
      "num_input_tokens_seen": 6944880,
      "step": 5045
    },
    {
      "epoch": 0.1616413801933295,
      "grad_norm": 0.018424617126584053,
      "learning_rate": 0.288359993418489,
      "loss": 0.9842,
      "num_input_tokens_seen": 6951840,
      "step": 5050
    },
    {
      "epoch": 0.16180142116381793,
      "grad_norm": 0.008880564011633396,
      "learning_rate": 0.28833723156668556,
      "loss": 0.8492,
      "num_input_tokens_seen": 6958832,
      "step": 5055
    },
    {
      "epoch": 0.16196146213430637,
      "grad_norm": 0.00882524810731411,
      "learning_rate": 0.2883144483815425,
      "loss": 0.8867,
      "num_input_tokens_seen": 6965744,
      "step": 5060
    },
    {
      "epoch": 0.16212150310479484,
      "grad_norm": 0.010971170850098133,
      "learning_rate": 0.28829164386657335,
      "loss": 0.7779,
      "num_input_tokens_seen": 6972528,
      "step": 5065
    },
    {
      "epoch": 0.16228154407528328,
      "grad_norm": 0.012244435958564281,
      "learning_rate": 0.28826881802529486,
      "loss": 0.7931,
      "num_input_tokens_seen": 6979520,
      "step": 5070
    },
    {
      "epoch": 0.16244158504577172,
      "grad_norm": 0.010120471008121967,
      "learning_rate": 0.28824597086122705,
      "loss": 0.7401,
      "num_input_tokens_seen": 6986624,
      "step": 5075
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 0.010900245979428291,
      "learning_rate": 0.28822310237789317,
      "loss": 0.811,
      "num_input_tokens_seen": 6993584,
      "step": 5080
    },
    {
      "epoch": 0.1627616669867486,
      "grad_norm": 0.010918364860117435,
      "learning_rate": 0.2882002125788199,
      "loss": 0.7463,
      "num_input_tokens_seen": 7000544,
      "step": 5085
    },
    {
      "epoch": 0.16292170795723707,
      "grad_norm": 0.006258894223719835,
      "learning_rate": 0.2881773014675371,
      "loss": 0.7308,
      "num_input_tokens_seen": 7007536,
      "step": 5090
    },
    {
      "epoch": 0.1630817489277255,
      "grad_norm": 0.009459131397306919,
      "learning_rate": 0.288154369047578,
      "loss": 0.5567,
      "num_input_tokens_seen": 7013952,
      "step": 5095
    },
    {
      "epoch": 0.16324178989821395,
      "grad_norm": 0.015001785941421986,
      "learning_rate": 0.28813141532247905,
      "loss": 0.7898,
      "num_input_tokens_seen": 7021152,
      "step": 5100
    },
    {
      "epoch": 0.16340183086870239,
      "grad_norm": 0.012704001739621162,
      "learning_rate": 0.28810844029578,
      "loss": 0.7916,
      "num_input_tokens_seen": 7027760,
      "step": 5105
    },
    {
      "epoch": 0.16356187183919083,
      "grad_norm": 0.010093996301293373,
      "learning_rate": 0.2880854439710238,
      "loss": 0.8427,
      "num_input_tokens_seen": 7034976,
      "step": 5110
    },
    {
      "epoch": 0.16372191280967927,
      "grad_norm": 0.010959685780107975,
      "learning_rate": 0.28806242635175694,
      "loss": 0.7639,
      "num_input_tokens_seen": 7041728,
      "step": 5115
    },
    {
      "epoch": 0.16388195378016773,
      "grad_norm": 0.014068378135561943,
      "learning_rate": 0.2880393874415289,
      "loss": 0.7735,
      "num_input_tokens_seen": 7048560,
      "step": 5120
    },
    {
      "epoch": 0.16404199475065617,
      "grad_norm": 0.013648380525410175,
      "learning_rate": 0.2880163272438926,
      "loss": 0.6753,
      "num_input_tokens_seen": 7055040,
      "step": 5125
    },
    {
      "epoch": 0.1642020357211446,
      "grad_norm": 0.04287423938512802,
      "learning_rate": 0.2879932457624042,
      "loss": 1.015,
      "num_input_tokens_seen": 7062272,
      "step": 5130
    },
    {
      "epoch": 0.16436207669163305,
      "grad_norm": 0.0089618731290102,
      "learning_rate": 0.2879701430006232,
      "loss": 0.8025,
      "num_input_tokens_seen": 7069584,
      "step": 5135
    },
    {
      "epoch": 0.1645221176621215,
      "grad_norm": 0.014820919372141361,
      "learning_rate": 0.28794701896211233,
      "loss": 0.8088,
      "num_input_tokens_seen": 7076352,
      "step": 5140
    },
    {
      "epoch": 0.16468215863260996,
      "grad_norm": 0.013601605780422688,
      "learning_rate": 0.28792387365043753,
      "loss": 0.9999,
      "num_input_tokens_seen": 7083168,
      "step": 5145
    },
    {
      "epoch": 0.1648421996030984,
      "grad_norm": 0.006415276322513819,
      "learning_rate": 0.28790070706916815,
      "loss": 0.7378,
      "num_input_tokens_seen": 7090320,
      "step": 5150
    },
    {
      "epoch": 0.16500224057358684,
      "grad_norm": 0.008592803962528706,
      "learning_rate": 0.2878775192218768,
      "loss": 0.7596,
      "num_input_tokens_seen": 7097408,
      "step": 5155
    },
    {
      "epoch": 0.16516228154407528,
      "grad_norm": 0.013355889357626438,
      "learning_rate": 0.2878543101121393,
      "loss": 0.8063,
      "num_input_tokens_seen": 7104272,
      "step": 5160
    },
    {
      "epoch": 0.16532232251456372,
      "grad_norm": 0.011040344834327698,
      "learning_rate": 0.28783107974353483,
      "loss": 0.8223,
      "num_input_tokens_seen": 7111552,
      "step": 5165
    },
    {
      "epoch": 0.1654823634850522,
      "grad_norm": 0.01381966844201088,
      "learning_rate": 0.2878078281196457,
      "loss": 0.7132,
      "num_input_tokens_seen": 7118464,
      "step": 5170
    },
    {
      "epoch": 0.16564240445554063,
      "grad_norm": 0.00914669968187809,
      "learning_rate": 0.28778455524405777,
      "loss": 0.8069,
      "num_input_tokens_seen": 7125072,
      "step": 5175
    },
    {
      "epoch": 0.16580244542602907,
      "grad_norm": 0.03353132680058479,
      "learning_rate": 0.2877612611203598,
      "loss": 0.8955,
      "num_input_tokens_seen": 7131584,
      "step": 5180
    },
    {
      "epoch": 0.1659624863965175,
      "grad_norm": 0.006941665895283222,
      "learning_rate": 0.28773794575214423,
      "loss": 0.8117,
      "num_input_tokens_seen": 7138448,
      "step": 5185
    },
    {
      "epoch": 0.16612252736700595,
      "grad_norm": 0.017157338559627533,
      "learning_rate": 0.28771460914300645,
      "loss": 0.7347,
      "num_input_tokens_seen": 7145024,
      "step": 5190
    },
    {
      "epoch": 0.16628256833749439,
      "grad_norm": 0.016397545114159584,
      "learning_rate": 0.2876912512965454,
      "loss": 0.7867,
      "num_input_tokens_seen": 7152064,
      "step": 5195
    },
    {
      "epoch": 0.16644260930798285,
      "grad_norm": 0.021546920761466026,
      "learning_rate": 0.287667872216363,
      "loss": 0.9768,
      "num_input_tokens_seen": 7159072,
      "step": 5200
    },
    {
      "epoch": 0.16644260930798285,
      "eval_loss": 0.7992191314697266,
      "eval_runtime": 331.9249,
      "eval_samples_per_second": 41.835,
      "eval_steps_per_second": 20.917,
      "num_input_tokens_seen": 7159072,
      "step": 5200
    },
    {
      "epoch": 0.1666026502784713,
      "grad_norm": 0.010833574458956718,
      "learning_rate": 0.2876444719060647,
      "loss": 0.7096,
      "num_input_tokens_seen": 7165856,
      "step": 5205
    },
    {
      "epoch": 0.16676269124895973,
      "grad_norm": 0.011087434366345406,
      "learning_rate": 0.287621050369259,
      "loss": 0.9237,
      "num_input_tokens_seen": 7172848,
      "step": 5210
    },
    {
      "epoch": 0.16692273221944817,
      "grad_norm": 0.019307490438222885,
      "learning_rate": 0.28759760760955794,
      "loss": 0.8968,
      "num_input_tokens_seen": 7179600,
      "step": 5215
    },
    {
      "epoch": 0.1670827731899366,
      "grad_norm": 0.011238383129239082,
      "learning_rate": 0.2875741436305766,
      "loss": 0.6919,
      "num_input_tokens_seen": 7186480,
      "step": 5220
    },
    {
      "epoch": 0.16724281416042508,
      "grad_norm": 0.010440182872116566,
      "learning_rate": 0.28755065843593347,
      "loss": 0.5764,
      "num_input_tokens_seen": 7193456,
      "step": 5225
    },
    {
      "epoch": 0.16740285513091352,
      "grad_norm": 0.011546267196536064,
      "learning_rate": 0.2875271520292502,
      "loss": 0.9466,
      "num_input_tokens_seen": 7200704,
      "step": 5230
    },
    {
      "epoch": 0.16756289610140196,
      "grad_norm": 0.013081572018563747,
      "learning_rate": 0.28750362441415184,
      "loss": 0.5937,
      "num_input_tokens_seen": 7207152,
      "step": 5235
    },
    {
      "epoch": 0.1677229370718904,
      "grad_norm": 0.014713803306221962,
      "learning_rate": 0.28748007559426664,
      "loss": 0.9746,
      "num_input_tokens_seen": 7214304,
      "step": 5240
    },
    {
      "epoch": 0.16788297804237884,
      "grad_norm": 0.009363919496536255,
      "learning_rate": 0.2874565055732261,
      "loss": 0.8533,
      "num_input_tokens_seen": 7221008,
      "step": 5245
    },
    {
      "epoch": 0.1680430190128673,
      "grad_norm": 0.011203356087207794,
      "learning_rate": 0.28743291435466495,
      "loss": 0.8335,
      "num_input_tokens_seen": 7227504,
      "step": 5250
    },
    {
      "epoch": 0.16820305998335575,
      "grad_norm": 0.020099224522709846,
      "learning_rate": 0.2874093019422214,
      "loss": 0.7332,
      "num_input_tokens_seen": 7234064,
      "step": 5255
    },
    {
      "epoch": 0.16836310095384419,
      "grad_norm": 0.008676381781697273,
      "learning_rate": 0.28738566833953666,
      "loss": 1.0214,
      "num_input_tokens_seen": 7240848,
      "step": 5260
    },
    {
      "epoch": 0.16852314192433263,
      "grad_norm": 0.015950772911310196,
      "learning_rate": 0.28736201355025537,
      "loss": 1.0045,
      "num_input_tokens_seen": 7247744,
      "step": 5265
    },
    {
      "epoch": 0.16868318289482107,
      "grad_norm": 0.007409088313579559,
      "learning_rate": 0.28733833757802535,
      "loss": 0.9564,
      "num_input_tokens_seen": 7254544,
      "step": 5270
    },
    {
      "epoch": 0.16884322386530953,
      "grad_norm": 0.013851269148290157,
      "learning_rate": 0.28731464042649785,
      "loss": 0.8014,
      "num_input_tokens_seen": 7261904,
      "step": 5275
    },
    {
      "epoch": 0.16900326483579797,
      "grad_norm": 0.013606789521872997,
      "learning_rate": 0.2872909220993271,
      "loss": 0.6754,
      "num_input_tokens_seen": 7268896,
      "step": 5280
    },
    {
      "epoch": 0.1691633058062864,
      "grad_norm": 0.00848496425896883,
      "learning_rate": 0.287267182600171,
      "loss": 0.6831,
      "num_input_tokens_seen": 7275648,
      "step": 5285
    },
    {
      "epoch": 0.16932334677677485,
      "grad_norm": 0.00897847767919302,
      "learning_rate": 0.2872434219326902,
      "loss": 0.8122,
      "num_input_tokens_seen": 7282512,
      "step": 5290
    },
    {
      "epoch": 0.1694833877472633,
      "grad_norm": 0.010995623655617237,
      "learning_rate": 0.28721964010054907,
      "loss": 0.7832,
      "num_input_tokens_seen": 7289120,
      "step": 5295
    },
    {
      "epoch": 0.16964342871775173,
      "grad_norm": 0.011939791962504387,
      "learning_rate": 0.28719583710741503,
      "loss": 0.9049,
      "num_input_tokens_seen": 7296112,
      "step": 5300
    },
    {
      "epoch": 0.1698034696882402,
      "grad_norm": 0.008808690123260021,
      "learning_rate": 0.28717201295695877,
      "loss": 0.9114,
      "num_input_tokens_seen": 7302720,
      "step": 5305
    },
    {
      "epoch": 0.16996351065872864,
      "grad_norm": 0.007598648779094219,
      "learning_rate": 0.28714816765285434,
      "loss": 0.8101,
      "num_input_tokens_seen": 7309920,
      "step": 5310
    },
    {
      "epoch": 0.17012355162921708,
      "grad_norm": 0.019338253885507584,
      "learning_rate": 0.28712430119877896,
      "loss": 0.7775,
      "num_input_tokens_seen": 7317008,
      "step": 5315
    },
    {
      "epoch": 0.17028359259970552,
      "grad_norm": 0.011641616933047771,
      "learning_rate": 0.28710041359841304,
      "loss": 0.8802,
      "num_input_tokens_seen": 7323824,
      "step": 5320
    },
    {
      "epoch": 0.17044363357019396,
      "grad_norm": 0.013782523572444916,
      "learning_rate": 0.28707650485544056,
      "loss": 0.7825,
      "num_input_tokens_seen": 7330592,
      "step": 5325
    },
    {
      "epoch": 0.17060367454068243,
      "grad_norm": 0.009150613099336624,
      "learning_rate": 0.28705257497354836,
      "loss": 0.6298,
      "num_input_tokens_seen": 7337488,
      "step": 5330
    },
    {
      "epoch": 0.17076371551117087,
      "grad_norm": 0.008995797485113144,
      "learning_rate": 0.28702862395642675,
      "loss": 1.0123,
      "num_input_tokens_seen": 7344592,
      "step": 5335
    },
    {
      "epoch": 0.1709237564816593,
      "grad_norm": 0.009745310060679913,
      "learning_rate": 0.28700465180776935,
      "loss": 0.7806,
      "num_input_tokens_seen": 7351184,
      "step": 5340
    },
    {
      "epoch": 0.17108379745214775,
      "grad_norm": 0.008739744313061237,
      "learning_rate": 0.2869806585312729,
      "loss": 0.7867,
      "num_input_tokens_seen": 7357984,
      "step": 5345
    },
    {
      "epoch": 0.17124383842263619,
      "grad_norm": 0.01056212279945612,
      "learning_rate": 0.28695664413063754,
      "loss": 0.9625,
      "num_input_tokens_seen": 7364816,
      "step": 5350
    },
    {
      "epoch": 0.17140387939312465,
      "grad_norm": 0.010947045870125294,
      "learning_rate": 0.28693260860956654,
      "loss": 0.7483,
      "num_input_tokens_seen": 7371904,
      "step": 5355
    },
    {
      "epoch": 0.1715639203636131,
      "grad_norm": 0.008496479131281376,
      "learning_rate": 0.2869085519717665,
      "loss": 0.7799,
      "num_input_tokens_seen": 7378368,
      "step": 5360
    },
    {
      "epoch": 0.17172396133410153,
      "grad_norm": 0.00893363542854786,
      "learning_rate": 0.28688447422094726,
      "loss": 0.8969,
      "num_input_tokens_seen": 7385296,
      "step": 5365
    },
    {
      "epoch": 0.17188400230458997,
      "grad_norm": 0.00760342413559556,
      "learning_rate": 0.2868603753608219,
      "loss": 0.7805,
      "num_input_tokens_seen": 7392160,
      "step": 5370
    },
    {
      "epoch": 0.1720440432750784,
      "grad_norm": 0.01226919423788786,
      "learning_rate": 0.28683625539510665,
      "loss": 0.6209,
      "num_input_tokens_seen": 7398864,
      "step": 5375
    },
    {
      "epoch": 0.17220408424556685,
      "grad_norm": 0.01045202650129795,
      "learning_rate": 0.28681211432752135,
      "loss": 0.6911,
      "num_input_tokens_seen": 7405536,
      "step": 5380
    },
    {
      "epoch": 0.17236412521605532,
      "grad_norm": 0.023395206779241562,
      "learning_rate": 0.2867879521617887,
      "loss": 0.7538,
      "num_input_tokens_seen": 7412288,
      "step": 5385
    },
    {
      "epoch": 0.17252416618654376,
      "grad_norm": 0.014803316444158554,
      "learning_rate": 0.28676376890163485,
      "loss": 0.7876,
      "num_input_tokens_seen": 7419536,
      "step": 5390
    },
    {
      "epoch": 0.1726842071570322,
      "grad_norm": 0.010549250990152359,
      "learning_rate": 0.2867395645507891,
      "loss": 0.7116,
      "num_input_tokens_seen": 7426672,
      "step": 5395
    },
    {
      "epoch": 0.17284424812752064,
      "grad_norm": 0.01450091041624546,
      "learning_rate": 0.2867153391129842,
      "loss": 0.8985,
      "num_input_tokens_seen": 7433136,
      "step": 5400
    },
    {
      "epoch": 0.17284424812752064,
      "eval_loss": 0.7898479104042053,
      "eval_runtime": 331.8277,
      "eval_samples_per_second": 41.847,
      "eval_steps_per_second": 20.924,
      "num_input_tokens_seen": 7433136,
      "step": 5400
    },
    {
      "epoch": 0.17300428909800908,
      "grad_norm": 0.01259658858180046,
      "learning_rate": 0.28669109259195585,
      "loss": 0.7864,
      "num_input_tokens_seen": 7439920,
      "step": 5405
    },
    {
      "epoch": 0.17316433006849755,
      "grad_norm": 0.012842411175370216,
      "learning_rate": 0.2866668249914433,
      "loss": 0.7486,
      "num_input_tokens_seen": 7446432,
      "step": 5410
    },
    {
      "epoch": 0.17332437103898599,
      "grad_norm": 0.008732498623430729,
      "learning_rate": 0.2866425363151889,
      "loss": 0.7054,
      "num_input_tokens_seen": 7453632,
      "step": 5415
    },
    {
      "epoch": 0.17348441200947443,
      "grad_norm": 0.006760074757039547,
      "learning_rate": 0.2866182265669382,
      "loss": 0.6597,
      "num_input_tokens_seen": 7460160,
      "step": 5420
    },
    {
      "epoch": 0.17364445297996287,
      "grad_norm": 0.014200127683579922,
      "learning_rate": 0.28659389575044014,
      "loss": 0.5342,
      "num_input_tokens_seen": 7466928,
      "step": 5425
    },
    {
      "epoch": 0.1738044939504513,
      "grad_norm": 0.04077059403061867,
      "learning_rate": 0.28656954386944683,
      "loss": 0.8325,
      "num_input_tokens_seen": 7473680,
      "step": 5430
    },
    {
      "epoch": 0.17396453492093977,
      "grad_norm": 0.009243271313607693,
      "learning_rate": 0.28654517092771353,
      "loss": 0.7361,
      "num_input_tokens_seen": 7480784,
      "step": 5435
    },
    {
      "epoch": 0.1741245758914282,
      "grad_norm": 0.010319502092897892,
      "learning_rate": 0.286520776928999,
      "loss": 0.8245,
      "num_input_tokens_seen": 7487552,
      "step": 5440
    },
    {
      "epoch": 0.17428461686191665,
      "grad_norm": 0.02237994410097599,
      "learning_rate": 0.286496361877065,
      "loss": 0.6912,
      "num_input_tokens_seen": 7494240,
      "step": 5445
    },
    {
      "epoch": 0.1744446578324051,
      "grad_norm": 0.012394995428621769,
      "learning_rate": 0.28647192577567676,
      "loss": 0.9609,
      "num_input_tokens_seen": 7501216,
      "step": 5450
    },
    {
      "epoch": 0.17460469880289353,
      "grad_norm": 0.00992842111736536,
      "learning_rate": 0.28644746862860254,
      "loss": 0.8105,
      "num_input_tokens_seen": 7508064,
      "step": 5455
    },
    {
      "epoch": 0.174764739773382,
      "grad_norm": 0.008821669034659863,
      "learning_rate": 0.2864229904396139,
      "loss": 0.7469,
      "num_input_tokens_seen": 7515104,
      "step": 5460
    },
    {
      "epoch": 0.17492478074387044,
      "grad_norm": 0.008749863132834435,
      "learning_rate": 0.28639849121248573,
      "loss": 0.6603,
      "num_input_tokens_seen": 7522016,
      "step": 5465
    },
    {
      "epoch": 0.17508482171435888,
      "grad_norm": 0.015245339833199978,
      "learning_rate": 0.28637397095099615,
      "loss": 0.8663,
      "num_input_tokens_seen": 7528928,
      "step": 5470
    },
    {
      "epoch": 0.17524486268484732,
      "grad_norm": 0.01587899588048458,
      "learning_rate": 0.28634942965892646,
      "loss": 0.89,
      "num_input_tokens_seen": 7535888,
      "step": 5475
    },
    {
      "epoch": 0.17540490365533576,
      "grad_norm": 0.01176057755947113,
      "learning_rate": 0.28632486734006124,
      "loss": 0.9064,
      "num_input_tokens_seen": 7542960,
      "step": 5480
    },
    {
      "epoch": 0.1755649446258242,
      "grad_norm": 0.03438275679945946,
      "learning_rate": 0.28630028399818835,
      "loss": 0.8077,
      "num_input_tokens_seen": 7549888,
      "step": 5485
    },
    {
      "epoch": 0.17572498559631267,
      "grad_norm": 0.009365938603878021,
      "learning_rate": 0.2862756796370987,
      "loss": 0.6869,
      "num_input_tokens_seen": 7557200,
      "step": 5490
    },
    {
      "epoch": 0.1758850265668011,
      "grad_norm": 0.012230523861944675,
      "learning_rate": 0.2862510542605868,
      "loss": 0.713,
      "num_input_tokens_seen": 7564336,
      "step": 5495
    },
    {
      "epoch": 0.17604506753728955,
      "grad_norm": 0.014344689436256886,
      "learning_rate": 0.2862264078724501,
      "loss": 0.7746,
      "num_input_tokens_seen": 7571184,
      "step": 5500
    },
    {
      "epoch": 0.17620510850777799,
      "grad_norm": 0.009814325720071793,
      "learning_rate": 0.28620174047648933,
      "loss": 0.7657,
      "num_input_tokens_seen": 7578048,
      "step": 5505
    },
    {
      "epoch": 0.17636514947826643,
      "grad_norm": 0.013154823333024979,
      "learning_rate": 0.2861770520765086,
      "loss": 0.7164,
      "num_input_tokens_seen": 7584864,
      "step": 5510
    },
    {
      "epoch": 0.1765251904487549,
      "grad_norm": 0.012708524242043495,
      "learning_rate": 0.2861523426763151,
      "loss": 0.8285,
      "num_input_tokens_seen": 7592016,
      "step": 5515
    },
    {
      "epoch": 0.17668523141924333,
      "grad_norm": 0.008033432997763157,
      "learning_rate": 0.2861276122797194,
      "loss": 0.6629,
      "num_input_tokens_seen": 7598624,
      "step": 5520
    },
    {
      "epoch": 0.17684527238973177,
      "grad_norm": 0.008680678904056549,
      "learning_rate": 0.28610286089053516,
      "loss": 0.8846,
      "num_input_tokens_seen": 7605648,
      "step": 5525
    },
    {
      "epoch": 0.1770053133602202,
      "grad_norm": 0.027956534177064896,
      "learning_rate": 0.28607808851257943,
      "loss": 1.1125,
      "num_input_tokens_seen": 7612608,
      "step": 5530
    },
    {
      "epoch": 0.17716535433070865,
      "grad_norm": 0.03366744518280029,
      "learning_rate": 0.28605329514967237,
      "loss": 0.8853,
      "num_input_tokens_seen": 7619264,
      "step": 5535
    },
    {
      "epoch": 0.17732539530119712,
      "grad_norm": 0.010157633572816849,
      "learning_rate": 0.2860284808056374,
      "loss": 0.7991,
      "num_input_tokens_seen": 7625936,
      "step": 5540
    },
    {
      "epoch": 0.17748543627168556,
      "grad_norm": 0.019094770774245262,
      "learning_rate": 0.28600364548430135,
      "loss": 0.8817,
      "num_input_tokens_seen": 7633008,
      "step": 5545
    },
    {
      "epoch": 0.177645477242174,
      "grad_norm": 0.011116916313767433,
      "learning_rate": 0.28597878918949393,
      "loss": 0.6735,
      "num_input_tokens_seen": 7639568,
      "step": 5550
    },
    {
      "epoch": 0.17780551821266244,
      "grad_norm": 0.013075451366603374,
      "learning_rate": 0.2859539119250485,
      "loss": 0.7667,
      "num_input_tokens_seen": 7646304,
      "step": 5555
    },
    {
      "epoch": 0.17796555918315088,
      "grad_norm": 0.008818255737423897,
      "learning_rate": 0.2859290136948013,
      "loss": 0.7456,
      "num_input_tokens_seen": 7653152,
      "step": 5560
    },
    {
      "epoch": 0.17812560015363932,
      "grad_norm": 0.009566454216837883,
      "learning_rate": 0.28590409450259197,
      "loss": 0.6353,
      "num_input_tokens_seen": 7660320,
      "step": 5565
    },
    {
      "epoch": 0.17828564112412779,
      "grad_norm": 0.01785111427307129,
      "learning_rate": 0.28587915435226346,
      "loss": 0.7583,
      "num_input_tokens_seen": 7666672,
      "step": 5570
    },
    {
      "epoch": 0.17844568209461623,
      "grad_norm": 0.013420413248240948,
      "learning_rate": 0.2858541932476617,
      "loss": 0.8782,
      "num_input_tokens_seen": 7673840,
      "step": 5575
    },
    {
      "epoch": 0.17860572306510467,
      "grad_norm": 0.007515890523791313,
      "learning_rate": 0.2858292111926361,
      "loss": 0.7117,
      "num_input_tokens_seen": 7680544,
      "step": 5580
    },
    {
      "epoch": 0.1787657640355931,
      "grad_norm": 0.008866403251886368,
      "learning_rate": 0.28580420819103924,
      "loss": 0.7731,
      "num_input_tokens_seen": 7687488,
      "step": 5585
    },
    {
      "epoch": 0.17892580500608155,
      "grad_norm": 0.012017983011901379,
      "learning_rate": 0.2857791842467269,
      "loss": 0.9638,
      "num_input_tokens_seen": 7694160,
      "step": 5590
    },
    {
      "epoch": 0.17908584597657,
      "grad_norm": 0.012977841310203075,
      "learning_rate": 0.2857541393635579,
      "loss": 0.8017,
      "num_input_tokens_seen": 7701120,
      "step": 5595
    },
    {
      "epoch": 0.17924588694705845,
      "grad_norm": 0.01576343923807144,
      "learning_rate": 0.2857290735453948,
      "loss": 0.9597,
      "num_input_tokens_seen": 7707776,
      "step": 5600
    },
    {
      "epoch": 0.17924588694705845,
      "eval_loss": 0.792561411857605,
      "eval_runtime": 332.0843,
      "eval_samples_per_second": 41.815,
      "eval_steps_per_second": 20.907,
      "num_input_tokens_seen": 7707776,
      "step": 5600
    },
    {
      "epoch": 0.1794059279175469,
      "grad_norm": 0.014379989355802536,
      "learning_rate": 0.28570398679610276,
      "loss": 0.8207,
      "num_input_tokens_seen": 7714304,
      "step": 5605
    },
    {
      "epoch": 0.17956596888803533,
      "grad_norm": 0.00976470299065113,
      "learning_rate": 0.2856788791195506,
      "loss": 0.8834,
      "num_input_tokens_seen": 7721120,
      "step": 5610
    },
    {
      "epoch": 0.17972600985852377,
      "grad_norm": 0.011679790914058685,
      "learning_rate": 0.28565375051961023,
      "loss": 0.7171,
      "num_input_tokens_seen": 7728240,
      "step": 5615
    },
    {
      "epoch": 0.17988605082901224,
      "grad_norm": 0.013179700821638107,
      "learning_rate": 0.28562860100015686,
      "loss": 0.8032,
      "num_input_tokens_seen": 7734864,
      "step": 5620
    },
    {
      "epoch": 0.18004609179950068,
      "grad_norm": 0.009916502982378006,
      "learning_rate": 0.2856034305650687,
      "loss": 0.7377,
      "num_input_tokens_seen": 7741472,
      "step": 5625
    },
    {
      "epoch": 0.18020613276998912,
      "grad_norm": 0.011823792941868305,
      "learning_rate": 0.28557823921822756,
      "loss": 0.9799,
      "num_input_tokens_seen": 7748064,
      "step": 5630
    },
    {
      "epoch": 0.18036617374047756,
      "grad_norm": 0.02206702157855034,
      "learning_rate": 0.2855530269635181,
      "loss": 0.7239,
      "num_input_tokens_seen": 7754784,
      "step": 5635
    },
    {
      "epoch": 0.180526214710966,
      "grad_norm": 0.014659448526799679,
      "learning_rate": 0.2855277938048284,
      "loss": 0.8203,
      "num_input_tokens_seen": 7762080,
      "step": 5640
    },
    {
      "epoch": 0.18068625568145444,
      "grad_norm": 0.012173742055892944,
      "learning_rate": 0.2855025397460498,
      "loss": 0.727,
      "num_input_tokens_seen": 7768944,
      "step": 5645
    },
    {
      "epoch": 0.1808462966519429,
      "grad_norm": 0.01069759763777256,
      "learning_rate": 0.28547726479107666,
      "loss": 0.9863,
      "num_input_tokens_seen": 7776080,
      "step": 5650
    },
    {
      "epoch": 0.18100633762243135,
      "grad_norm": 0.009413531981408596,
      "learning_rate": 0.2854519689438068,
      "loss": 0.7612,
      "num_input_tokens_seen": 7783024,
      "step": 5655
    },
    {
      "epoch": 0.18116637859291979,
      "grad_norm": 0.007362894248217344,
      "learning_rate": 0.2854266522081412,
      "loss": 0.7019,
      "num_input_tokens_seen": 7789680,
      "step": 5660
    },
    {
      "epoch": 0.18132641956340823,
      "grad_norm": 0.00517791323363781,
      "learning_rate": 0.28540131458798385,
      "loss": 0.823,
      "num_input_tokens_seen": 7796656,
      "step": 5665
    },
    {
      "epoch": 0.18148646053389667,
      "grad_norm": 0.012055383995175362,
      "learning_rate": 0.28537595608724226,
      "loss": 0.6707,
      "num_input_tokens_seen": 7803968,
      "step": 5670
    },
    {
      "epoch": 0.18164650150438513,
      "grad_norm": 0.009848789311945438,
      "learning_rate": 0.28535057670982705,
      "loss": 0.8876,
      "num_input_tokens_seen": 7811536,
      "step": 5675
    },
    {
      "epoch": 0.18180654247487357,
      "grad_norm": 0.010252759791910648,
      "learning_rate": 0.285325176459652,
      "loss": 0.8527,
      "num_input_tokens_seen": 7818656,
      "step": 5680
    },
    {
      "epoch": 0.181966583445362,
      "grad_norm": 0.012056167237460613,
      "learning_rate": 0.28529975534063406,
      "loss": 0.8137,
      "num_input_tokens_seen": 7825328,
      "step": 5685
    },
    {
      "epoch": 0.18212662441585045,
      "grad_norm": 0.010974193923175335,
      "learning_rate": 0.2852743133566936,
      "loss": 1.0412,
      "num_input_tokens_seen": 7832288,
      "step": 5690
    },
    {
      "epoch": 0.1822866653863389,
      "grad_norm": 0.006355539429932833,
      "learning_rate": 0.2852488505117541,
      "loss": 0.7443,
      "num_input_tokens_seen": 7838832,
      "step": 5695
    },
    {
      "epoch": 0.18244670635682736,
      "grad_norm": 0.005794823169708252,
      "learning_rate": 0.28522336680974214,
      "loss": 0.7212,
      "num_input_tokens_seen": 7845840,
      "step": 5700
    },
    {
      "epoch": 0.1826067473273158,
      "grad_norm": 0.008441930636763573,
      "learning_rate": 0.2851978622545877,
      "loss": 0.671,
      "num_input_tokens_seen": 7852592,
      "step": 5705
    },
    {
      "epoch": 0.18276678829780424,
      "grad_norm": 0.007322201505303383,
      "learning_rate": 0.285172336850224,
      "loss": 0.6924,
      "num_input_tokens_seen": 7859856,
      "step": 5710
    },
    {
      "epoch": 0.18292682926829268,
      "grad_norm": 0.008273741230368614,
      "learning_rate": 0.2851467906005871,
      "loss": 0.8861,
      "num_input_tokens_seen": 7867136,
      "step": 5715
    },
    {
      "epoch": 0.18308687023878112,
      "grad_norm": 0.008188673295080662,
      "learning_rate": 0.28512122350961683,
      "loss": 0.7872,
      "num_input_tokens_seen": 7874240,
      "step": 5720
    },
    {
      "epoch": 0.1832469112092696,
      "grad_norm": 0.020428519695997238,
      "learning_rate": 0.2850956355812559,
      "loss": 0.951,
      "num_input_tokens_seen": 7881200,
      "step": 5725
    },
    {
      "epoch": 0.18340695217975803,
      "grad_norm": 0.013689405284821987,
      "learning_rate": 0.28507002681945015,
      "loss": 0.841,
      "num_input_tokens_seen": 7888192,
      "step": 5730
    },
    {
      "epoch": 0.18356699315024647,
      "grad_norm": 0.01801970787346363,
      "learning_rate": 0.28504439722814895,
      "loss": 0.7024,
      "num_input_tokens_seen": 7894912,
      "step": 5735
    },
    {
      "epoch": 0.1837270341207349,
      "grad_norm": 0.008634842000901699,
      "learning_rate": 0.28501874681130457,
      "loss": 0.6951,
      "num_input_tokens_seen": 7902048,
      "step": 5740
    },
    {
      "epoch": 0.18388707509122335,
      "grad_norm": 0.013264563865959644,
      "learning_rate": 0.2849930755728727,
      "loss": 0.7747,
      "num_input_tokens_seen": 7908720,
      "step": 5745
    },
    {
      "epoch": 0.18404711606171179,
      "grad_norm": 0.008827313780784607,
      "learning_rate": 0.28496738351681217,
      "loss": 0.8395,
      "num_input_tokens_seen": 7915488,
      "step": 5750
    },
    {
      "epoch": 0.18420715703220025,
      "grad_norm": 0.012480462901294231,
      "learning_rate": 0.284941670647085,
      "loss": 0.7105,
      "num_input_tokens_seen": 7922384,
      "step": 5755
    },
    {
      "epoch": 0.1843671980026887,
      "grad_norm": 0.012444916181266308,
      "learning_rate": 0.2849159369676563,
      "loss": 0.9001,
      "num_input_tokens_seen": 7929424,
      "step": 5760
    },
    {
      "epoch": 0.18452723897317713,
      "grad_norm": 0.014350485056638718,
      "learning_rate": 0.2848901824824948,
      "loss": 0.7389,
      "num_input_tokens_seen": 7936352,
      "step": 5765
    },
    {
      "epoch": 0.18468727994366557,
      "grad_norm": 0.012010304257273674,
      "learning_rate": 0.284864407195572,
      "loss": 0.8934,
      "num_input_tokens_seen": 7943264,
      "step": 5770
    },
    {
      "epoch": 0.184847320914154,
      "grad_norm": 0.01509664487093687,
      "learning_rate": 0.28483861111086284,
      "loss": 0.7668,
      "num_input_tokens_seen": 7950480,
      "step": 5775
    },
    {
      "epoch": 0.18500736188464248,
      "grad_norm": 0.008751357905566692,
      "learning_rate": 0.2848127942323453,
      "loss": 0.7324,
      "num_input_tokens_seen": 7957488,
      "step": 5780
    },
    {
      "epoch": 0.18516740285513092,
      "grad_norm": 0.008762643672525883,
      "learning_rate": 0.2847869565640007,
      "loss": 0.6989,
      "num_input_tokens_seen": 7964352,
      "step": 5785
    },
    {
      "epoch": 0.18532744382561936,
      "grad_norm": 0.009045968763530254,
      "learning_rate": 0.2847610981098136,
      "loss": 0.6123,
      "num_input_tokens_seen": 7971136,
      "step": 5790
    },
    {
      "epoch": 0.1854874847961078,
      "grad_norm": 0.00798843614757061,
      "learning_rate": 0.2847352188737716,
      "loss": 0.6785,
      "num_input_tokens_seen": 7978432,
      "step": 5795
    },
    {
      "epoch": 0.18564752576659624,
      "grad_norm": 0.01043709833174944,
      "learning_rate": 0.2847093188598658,
      "loss": 0.7812,
      "num_input_tokens_seen": 7985472,
      "step": 5800
    },
    {
      "epoch": 0.18564752576659624,
      "eval_loss": 0.7971620559692383,
      "eval_runtime": 331.6761,
      "eval_samples_per_second": 41.866,
      "eval_steps_per_second": 20.933,
      "num_input_tokens_seen": 7985472,
      "step": 5800
    },
    {
      "epoch": 0.1858075667370847,
      "grad_norm": 0.013805422931909561,
      "learning_rate": 0.28468339807209003,
      "loss": 0.8916,
      "num_input_tokens_seen": 7992160,
      "step": 5805
    },
    {
      "epoch": 0.18596760770757315,
      "grad_norm": 0.0070000034756958485,
      "learning_rate": 0.2846574565144418,
      "loss": 0.8286,
      "num_input_tokens_seen": 7999776,
      "step": 5810
    },
    {
      "epoch": 0.18612764867806159,
      "grad_norm": 0.014910203404724598,
      "learning_rate": 0.28463149419092154,
      "loss": 0.7885,
      "num_input_tokens_seen": 8006816,
      "step": 5815
    },
    {
      "epoch": 0.18628768964855003,
      "grad_norm": 0.015081504359841347,
      "learning_rate": 0.284605511105533,
      "loss": 0.8396,
      "num_input_tokens_seen": 8013632,
      "step": 5820
    },
    {
      "epoch": 0.18644773061903847,
      "grad_norm": 0.00915027130395174,
      "learning_rate": 0.28457950726228315,
      "loss": 0.8729,
      "num_input_tokens_seen": 8020672,
      "step": 5825
    },
    {
      "epoch": 0.1866077715895269,
      "grad_norm": 0.01736372895538807,
      "learning_rate": 0.28455348266518193,
      "loss": 0.7714,
      "num_input_tokens_seen": 8027376,
      "step": 5830
    },
    {
      "epoch": 0.18676781256001537,
      "grad_norm": 0.0084817660972476,
      "learning_rate": 0.28452743731824287,
      "loss": 0.7137,
      "num_input_tokens_seen": 8034704,
      "step": 5835
    },
    {
      "epoch": 0.1869278535305038,
      "grad_norm": 0.009476709179580212,
      "learning_rate": 0.28450137122548236,
      "loss": 0.6248,
      "num_input_tokens_seen": 8041392,
      "step": 5840
    },
    {
      "epoch": 0.18708789450099225,
      "grad_norm": 0.009985042735934258,
      "learning_rate": 0.2844752843909201,
      "loss": 0.9443,
      "num_input_tokens_seen": 8048400,
      "step": 5845
    },
    {
      "epoch": 0.1872479354714807,
      "grad_norm": 0.01168100256472826,
      "learning_rate": 0.28444917681857923,
      "loss": 0.7024,
      "num_input_tokens_seen": 8055808,
      "step": 5850
    },
    {
      "epoch": 0.18740797644196913,
      "grad_norm": 0.010368086397647858,
      "learning_rate": 0.28442304851248557,
      "loss": 0.8829,
      "num_input_tokens_seen": 8062480,
      "step": 5855
    },
    {
      "epoch": 0.1875680174124576,
      "grad_norm": 0.009838102385401726,
      "learning_rate": 0.2843968994766686,
      "loss": 0.6965,
      "num_input_tokens_seen": 8069376,
      "step": 5860
    },
    {
      "epoch": 0.18772805838294604,
      "grad_norm": 0.006517419591546059,
      "learning_rate": 0.28437072971516075,
      "loss": 0.7493,
      "num_input_tokens_seen": 8076096,
      "step": 5865
    },
    {
      "epoch": 0.18788809935343448,
      "grad_norm": 0.009209977462887764,
      "learning_rate": 0.2843445392319979,
      "loss": 0.9289,
      "num_input_tokens_seen": 8082912,
      "step": 5870
    },
    {
      "epoch": 0.18804814032392292,
      "grad_norm": 0.011852574534714222,
      "learning_rate": 0.28431832803121865,
      "loss": 0.7921,
      "num_input_tokens_seen": 8089568,
      "step": 5875
    },
    {
      "epoch": 0.18820818129441136,
      "grad_norm": 0.006033288314938545,
      "learning_rate": 0.28429209611686534,
      "loss": 0.5362,
      "num_input_tokens_seen": 8096416,
      "step": 5880
    },
    {
      "epoch": 0.18836822226489983,
      "grad_norm": 0.015508797951042652,
      "learning_rate": 0.28426584349298323,
      "loss": 0.9015,
      "num_input_tokens_seen": 8103088,
      "step": 5885
    },
    {
      "epoch": 0.18852826323538827,
      "grad_norm": 0.007764117326587439,
      "learning_rate": 0.2842395701636207,
      "loss": 0.6714,
      "num_input_tokens_seen": 8109520,
      "step": 5890
    },
    {
      "epoch": 0.1886883042058767,
      "grad_norm": 0.010040780529379845,
      "learning_rate": 0.28421327613282954,
      "loss": 0.7668,
      "num_input_tokens_seen": 8116176,
      "step": 5895
    },
    {
      "epoch": 0.18884834517636515,
      "grad_norm": 0.011722122319042683,
      "learning_rate": 0.28418696140466454,
      "loss": 0.738,
      "num_input_tokens_seen": 8123088,
      "step": 5900
    },
    {
      "epoch": 0.18900838614685359,
      "grad_norm": 0.02336014434695244,
      "learning_rate": 0.2841606259831838,
      "loss": 0.7866,
      "num_input_tokens_seen": 8129984,
      "step": 5905
    },
    {
      "epoch": 0.18916842711734205,
      "grad_norm": 0.010388502851128578,
      "learning_rate": 0.2841342698724486,
      "loss": 0.8262,
      "num_input_tokens_seen": 8137136,
      "step": 5910
    },
    {
      "epoch": 0.1893284680878305,
      "grad_norm": 0.01115521602332592,
      "learning_rate": 0.28410789307652334,
      "loss": 0.8952,
      "num_input_tokens_seen": 8143760,
      "step": 5915
    },
    {
      "epoch": 0.18948850905831893,
      "grad_norm": 0.013136297464370728,
      "learning_rate": 0.2840814955994756,
      "loss": 0.7407,
      "num_input_tokens_seen": 8150048,
      "step": 5920
    },
    {
      "epoch": 0.18964855002880737,
      "grad_norm": 0.009208026342093945,
      "learning_rate": 0.2840550774453763,
      "loss": 1.0135,
      "num_input_tokens_seen": 8156896,
      "step": 5925
    },
    {
      "epoch": 0.1898085909992958,
      "grad_norm": 0.01711205393075943,
      "learning_rate": 0.28402863861829947,
      "loss": 0.8952,
      "num_input_tokens_seen": 8163584,
      "step": 5930
    },
    {
      "epoch": 0.18996863196978425,
      "grad_norm": 0.008764347061514854,
      "learning_rate": 0.2840021791223222,
      "loss": 1.0945,
      "num_input_tokens_seen": 8170352,
      "step": 5935
    },
    {
      "epoch": 0.19012867294027272,
      "grad_norm": 0.02911754511296749,
      "learning_rate": 0.2839756989615249,
      "loss": 0.851,
      "num_input_tokens_seen": 8177392,
      "step": 5940
    },
    {
      "epoch": 0.19028871391076116,
      "grad_norm": 0.015279463492333889,
      "learning_rate": 0.28394919813999125,
      "loss": 0.7663,
      "num_input_tokens_seen": 8184576,
      "step": 5945
    },
    {
      "epoch": 0.1904487548812496,
      "grad_norm": 0.008098023943603039,
      "learning_rate": 0.28392267666180787,
      "loss": 0.7229,
      "num_input_tokens_seen": 8191200,
      "step": 5950
    },
    {
      "epoch": 0.19060879585173804,
      "grad_norm": 0.009516614489257336,
      "learning_rate": 0.2838961345310648,
      "loss": 0.7355,
      "num_input_tokens_seen": 8198144,
      "step": 5955
    },
    {
      "epoch": 0.19076883682222648,
      "grad_norm": 0.006894606631249189,
      "learning_rate": 0.2838695717518552,
      "loss": 0.557,
      "num_input_tokens_seen": 8204704,
      "step": 5960
    },
    {
      "epoch": 0.19092887779271495,
      "grad_norm": 0.009172776713967323,
      "learning_rate": 0.28384298832827526,
      "loss": 0.9178,
      "num_input_tokens_seen": 8211584,
      "step": 5965
    },
    {
      "epoch": 0.19108891876320339,
      "grad_norm": 0.00960515160113573,
      "learning_rate": 0.28381638426442457,
      "loss": 0.6366,
      "num_input_tokens_seen": 8218336,
      "step": 5970
    },
    {
      "epoch": 0.19124895973369183,
      "grad_norm": 0.014276502653956413,
      "learning_rate": 0.2837897595644057,
      "loss": 0.6316,
      "num_input_tokens_seen": 8225392,
      "step": 5975
    },
    {
      "epoch": 0.19140900070418027,
      "grad_norm": 0.007636105641722679,
      "learning_rate": 0.28376311423232475,
      "loss": 0.906,
      "num_input_tokens_seen": 8232464,
      "step": 5980
    },
    {
      "epoch": 0.1915690416746687,
      "grad_norm": 0.008522714488208294,
      "learning_rate": 0.2837364482722905,
      "loss": 0.8267,
      "num_input_tokens_seen": 8239216,
      "step": 5985
    },
    {
      "epoch": 0.19172908264515717,
      "grad_norm": 0.007837316021323204,
      "learning_rate": 0.28370976168841533,
      "loss": 0.7075,
      "num_input_tokens_seen": 8245696,
      "step": 5990
    },
    {
      "epoch": 0.1918891236156456,
      "grad_norm": 0.008158092387020588,
      "learning_rate": 0.2836830544848146,
      "loss": 0.6882,
      "num_input_tokens_seen": 8252432,
      "step": 5995
    },
    {
      "epoch": 0.19204916458613405,
      "grad_norm": 0.007684811018407345,
      "learning_rate": 0.2836563266656069,
      "loss": 0.8175,
      "num_input_tokens_seen": 8259552,
      "step": 6000
    },
    {
      "epoch": 0.19204916458613405,
      "eval_loss": 0.806366503238678,
      "eval_runtime": 332.2181,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 8259552,
      "step": 6000
    },
    {
      "epoch": 0.1922092055566225,
      "grad_norm": 0.009947127662599087,
      "learning_rate": 0.283629578234914,
      "loss": 0.5732,
      "num_input_tokens_seen": 8266208,
      "step": 6005
    },
    {
      "epoch": 0.19236924652711093,
      "grad_norm": 0.006386372726410627,
      "learning_rate": 0.2836028091968608,
      "loss": 0.7884,
      "num_input_tokens_seen": 8273216,
      "step": 6010
    },
    {
      "epoch": 0.19252928749759937,
      "grad_norm": 0.007369407918304205,
      "learning_rate": 0.28357601955557554,
      "loss": 0.7339,
      "num_input_tokens_seen": 8280032,
      "step": 6015
    },
    {
      "epoch": 0.19268932846808784,
      "grad_norm": 0.005831725895404816,
      "learning_rate": 0.2835492093151894,
      "loss": 0.7559,
      "num_input_tokens_seen": 8286928,
      "step": 6020
    },
    {
      "epoch": 0.19284936943857628,
      "grad_norm": 0.017431432381272316,
      "learning_rate": 0.2835223784798369,
      "loss": 0.7718,
      "num_input_tokens_seen": 8293968,
      "step": 6025
    },
    {
      "epoch": 0.19300941040906472,
      "grad_norm": 0.006531710736453533,
      "learning_rate": 0.2834955270536557,
      "loss": 0.5225,
      "num_input_tokens_seen": 8301024,
      "step": 6030
    },
    {
      "epoch": 0.19316945137955316,
      "grad_norm": 0.007463822141289711,
      "learning_rate": 0.2834686550407866,
      "loss": 0.9827,
      "num_input_tokens_seen": 8307504,
      "step": 6035
    },
    {
      "epoch": 0.1933294923500416,
      "grad_norm": 0.007513865362852812,
      "learning_rate": 0.28344176244537367,
      "loss": 0.8413,
      "num_input_tokens_seen": 8314560,
      "step": 6040
    },
    {
      "epoch": 0.19348953332053007,
      "grad_norm": 0.004949146881699562,
      "learning_rate": 0.28341484927156396,
      "loss": 0.7892,
      "num_input_tokens_seen": 8321392,
      "step": 6045
    },
    {
      "epoch": 0.1936495742910185,
      "grad_norm": 0.006282504182308912,
      "learning_rate": 0.28338791552350795,
      "loss": 0.6863,
      "num_input_tokens_seen": 8328256,
      "step": 6050
    },
    {
      "epoch": 0.19380961526150695,
      "grad_norm": 0.0069512613117694855,
      "learning_rate": 0.28336096120535914,
      "loss": 0.728,
      "num_input_tokens_seen": 8335072,
      "step": 6055
    },
    {
      "epoch": 0.19396965623199539,
      "grad_norm": 0.005990547128021717,
      "learning_rate": 0.2833339863212741,
      "loss": 0.6259,
      "num_input_tokens_seen": 8342096,
      "step": 6060
    },
    {
      "epoch": 0.19412969720248383,
      "grad_norm": 0.007733451668173075,
      "learning_rate": 0.28330699087541283,
      "loss": 0.7994,
      "num_input_tokens_seen": 8349232,
      "step": 6065
    },
    {
      "epoch": 0.1942897381729723,
      "grad_norm": 0.005815812386572361,
      "learning_rate": 0.2832799748719384,
      "loss": 0.8963,
      "num_input_tokens_seen": 8356256,
      "step": 6070
    },
    {
      "epoch": 0.19444977914346073,
      "grad_norm": 0.019544191658496857,
      "learning_rate": 0.28325293831501686,
      "loss": 0.9036,
      "num_input_tokens_seen": 8362752,
      "step": 6075
    },
    {
      "epoch": 0.19460982011394917,
      "grad_norm": 0.01030771154910326,
      "learning_rate": 0.2832258812088177,
      "loss": 0.6945,
      "num_input_tokens_seen": 8369696,
      "step": 6080
    },
    {
      "epoch": 0.1947698610844376,
      "grad_norm": 0.023467596620321274,
      "learning_rate": 0.2831988035575134,
      "loss": 0.7441,
      "num_input_tokens_seen": 8376352,
      "step": 6085
    },
    {
      "epoch": 0.19492990205492605,
      "grad_norm": 0.01523843314498663,
      "learning_rate": 0.28317170536527975,
      "loss": 0.8041,
      "num_input_tokens_seen": 8383120,
      "step": 6090
    },
    {
      "epoch": 0.19508994302541452,
      "grad_norm": 0.017630906775593758,
      "learning_rate": 0.2831445866362956,
      "loss": 0.6496,
      "num_input_tokens_seen": 8389920,
      "step": 6095
    },
    {
      "epoch": 0.19524998399590296,
      "grad_norm": 0.019563674926757812,
      "learning_rate": 0.2831174473747429,
      "loss": 0.9057,
      "num_input_tokens_seen": 8396928,
      "step": 6100
    },
    {
      "epoch": 0.1954100249663914,
      "grad_norm": 0.008782369084656239,
      "learning_rate": 0.2830902875848071,
      "loss": 0.6663,
      "num_input_tokens_seen": 8404288,
      "step": 6105
    },
    {
      "epoch": 0.19557006593687984,
      "grad_norm": 0.011541181243956089,
      "learning_rate": 0.28306310727067635,
      "loss": 0.7259,
      "num_input_tokens_seen": 8411104,
      "step": 6110
    },
    {
      "epoch": 0.19573010690736828,
      "grad_norm": 0.013921769335865974,
      "learning_rate": 0.2830359064365423,
      "loss": 0.8496,
      "num_input_tokens_seen": 8418368,
      "step": 6115
    },
    {
      "epoch": 0.19589014787785672,
      "grad_norm": 0.009961257688701153,
      "learning_rate": 0.28300868508659965,
      "loss": 0.9464,
      "num_input_tokens_seen": 8425248,
      "step": 6120
    },
    {
      "epoch": 0.19605018884834519,
      "grad_norm": 0.014244655147194862,
      "learning_rate": 0.28298144322504626,
      "loss": 0.7298,
      "num_input_tokens_seen": 8432144,
      "step": 6125
    },
    {
      "epoch": 0.19621022981883363,
      "grad_norm": 0.00933198444545269,
      "learning_rate": 0.2829541808560832,
      "loss": 0.7472,
      "num_input_tokens_seen": 8439328,
      "step": 6130
    },
    {
      "epoch": 0.19637027078932207,
      "grad_norm": 0.008798523806035519,
      "learning_rate": 0.2829268979839146,
      "loss": 0.9636,
      "num_input_tokens_seen": 8445856,
      "step": 6135
    },
    {
      "epoch": 0.1965303117598105,
      "grad_norm": 0.010534323751926422,
      "learning_rate": 0.2828995946127479,
      "loss": 0.6896,
      "num_input_tokens_seen": 8452880,
      "step": 6140
    },
    {
      "epoch": 0.19669035273029895,
      "grad_norm": 0.009840717539191246,
      "learning_rate": 0.2828722707467936,
      "loss": 0.7137,
      "num_input_tokens_seen": 8460320,
      "step": 6145
    },
    {
      "epoch": 0.1968503937007874,
      "grad_norm": 0.006330485921353102,
      "learning_rate": 0.2828449263902653,
      "loss": 0.6806,
      "num_input_tokens_seen": 8467456,
      "step": 6150
    },
    {
      "epoch": 0.19701043467127585,
      "grad_norm": 0.008410368114709854,
      "learning_rate": 0.28281756154738,
      "loss": 0.7667,
      "num_input_tokens_seen": 8474080,
      "step": 6155
    },
    {
      "epoch": 0.1971704756417643,
      "grad_norm": 0.010771447792649269,
      "learning_rate": 0.28279017622235764,
      "loss": 0.6686,
      "num_input_tokens_seen": 8481168,
      "step": 6160
    },
    {
      "epoch": 0.19733051661225273,
      "grad_norm": 0.008318839594721794,
      "learning_rate": 0.28276277041942127,
      "loss": 0.8071,
      "num_input_tokens_seen": 8488080,
      "step": 6165
    },
    {
      "epoch": 0.19749055758274117,
      "grad_norm": 0.0066185747273266315,
      "learning_rate": 0.2827353441427974,
      "loss": 0.7167,
      "num_input_tokens_seen": 8494880,
      "step": 6170
    },
    {
      "epoch": 0.19765059855322964,
      "grad_norm": 0.008826283738017082,
      "learning_rate": 0.2827078973967153,
      "loss": 0.7378,
      "num_input_tokens_seen": 8501824,
      "step": 6175
    },
    {
      "epoch": 0.19781063952371808,
      "grad_norm": 0.010029410012066364,
      "learning_rate": 0.2826804301854078,
      "loss": 0.7635,
      "num_input_tokens_seen": 8508448,
      "step": 6180
    },
    {
      "epoch": 0.19797068049420652,
      "grad_norm": 0.009373391978442669,
      "learning_rate": 0.2826529425131105,
      "loss": 0.8131,
      "num_input_tokens_seen": 8515200,
      "step": 6185
    },
    {
      "epoch": 0.19813072146469496,
      "grad_norm": 0.011569296009838581,
      "learning_rate": 0.2826254343840625,
      "loss": 0.8744,
      "num_input_tokens_seen": 8522016,
      "step": 6190
    },
    {
      "epoch": 0.1982907624351834,
      "grad_norm": 0.007960840128362179,
      "learning_rate": 0.2825979058025059,
      "loss": 0.8172,
      "num_input_tokens_seen": 8529344,
      "step": 6195
    },
    {
      "epoch": 0.19845080340567184,
      "grad_norm": 0.005088028497993946,
      "learning_rate": 0.2825703567726858,
      "loss": 0.8294,
      "num_input_tokens_seen": 8535952,
      "step": 6200
    },
    {
      "epoch": 0.19845080340567184,
      "eval_loss": 0.7867940068244934,
      "eval_runtime": 332.2562,
      "eval_samples_per_second": 41.793,
      "eval_steps_per_second": 20.897,
      "num_input_tokens_seen": 8535952,
      "step": 6200
    },
    {
      "epoch": 0.1986108443761603,
      "grad_norm": 0.014200182631611824,
      "learning_rate": 0.2825427872988508,
      "loss": 0.6257,
      "num_input_tokens_seen": 8542912,
      "step": 6205
    },
    {
      "epoch": 0.19877088534664875,
      "grad_norm": 0.009906553663313389,
      "learning_rate": 0.28251519738525227,
      "loss": 0.8163,
      "num_input_tokens_seen": 8549504,
      "step": 6210
    },
    {
      "epoch": 0.19893092631713719,
      "grad_norm": 0.008634925819933414,
      "learning_rate": 0.28248758703614507,
      "loss": 0.7778,
      "num_input_tokens_seen": 8556352,
      "step": 6215
    },
    {
      "epoch": 0.19909096728762563,
      "grad_norm": 0.008659744635224342,
      "learning_rate": 0.28245995625578696,
      "loss": 0.7621,
      "num_input_tokens_seen": 8563472,
      "step": 6220
    },
    {
      "epoch": 0.19925100825811407,
      "grad_norm": 0.008418464101850986,
      "learning_rate": 0.282432305048439,
      "loss": 0.6444,
      "num_input_tokens_seen": 8570368,
      "step": 6225
    },
    {
      "epoch": 0.19941104922860253,
      "grad_norm": 0.0062132831662893295,
      "learning_rate": 0.28240463341836536,
      "loss": 0.7287,
      "num_input_tokens_seen": 8577168,
      "step": 6230
    },
    {
      "epoch": 0.19957109019909097,
      "grad_norm": 0.005742646753787994,
      "learning_rate": 0.2823769413698334,
      "loss": 0.6806,
      "num_input_tokens_seen": 8583872,
      "step": 6235
    },
    {
      "epoch": 0.1997311311695794,
      "grad_norm": 0.012476149946451187,
      "learning_rate": 0.2823492289071135,
      "loss": 0.8304,
      "num_input_tokens_seen": 8591264,
      "step": 6240
    },
    {
      "epoch": 0.19989117214006785,
      "grad_norm": 0.010852538980543613,
      "learning_rate": 0.2823214960344793,
      "loss": 0.7905,
      "num_input_tokens_seen": 8598016,
      "step": 6245
    },
    {
      "epoch": 0.2000512131105563,
      "grad_norm": 0.006673283409327269,
      "learning_rate": 0.28229374275620756,
      "loss": 0.6916,
      "num_input_tokens_seen": 8604544,
      "step": 6250
    },
    {
      "epoch": 0.20021125408104476,
      "grad_norm": 0.007689275313168764,
      "learning_rate": 0.28226596907657814,
      "loss": 0.7755,
      "num_input_tokens_seen": 8611696,
      "step": 6255
    },
    {
      "epoch": 0.2003712950515332,
      "grad_norm": 0.00955126341432333,
      "learning_rate": 0.28223817499987414,
      "loss": 0.7509,
      "num_input_tokens_seen": 8618416,
      "step": 6260
    },
    {
      "epoch": 0.20053133602202164,
      "grad_norm": 0.08759411424398422,
      "learning_rate": 0.2822103605303818,
      "loss": 0.86,
      "num_input_tokens_seen": 8625504,
      "step": 6265
    },
    {
      "epoch": 0.20069137699251008,
      "grad_norm": 0.0056748720817267895,
      "learning_rate": 0.2821825256723903,
      "loss": 0.7353,
      "num_input_tokens_seen": 8632096,
      "step": 6270
    },
    {
      "epoch": 0.20085141796299852,
      "grad_norm": 0.007184438873082399,
      "learning_rate": 0.2821546704301923,
      "loss": 0.7548,
      "num_input_tokens_seen": 8638736,
      "step": 6275
    },
    {
      "epoch": 0.201011458933487,
      "grad_norm": 0.0062657808884978294,
      "learning_rate": 0.2821267948080834,
      "loss": 0.6908,
      "num_input_tokens_seen": 8645536,
      "step": 6280
    },
    {
      "epoch": 0.20117149990397543,
      "grad_norm": 0.012838541530072689,
      "learning_rate": 0.28209889881036226,
      "loss": 0.948,
      "num_input_tokens_seen": 8652448,
      "step": 6285
    },
    {
      "epoch": 0.20133154087446387,
      "grad_norm": 0.004281749948859215,
      "learning_rate": 0.28207098244133094,
      "loss": 0.6897,
      "num_input_tokens_seen": 8659024,
      "step": 6290
    },
    {
      "epoch": 0.2014915818449523,
      "grad_norm": 0.004750012885779142,
      "learning_rate": 0.2820430457052943,
      "loss": 0.6646,
      "num_input_tokens_seen": 8665872,
      "step": 6295
    },
    {
      "epoch": 0.20165162281544075,
      "grad_norm": 0.01493923831731081,
      "learning_rate": 0.28201508860656077,
      "loss": 0.7743,
      "num_input_tokens_seen": 8672336,
      "step": 6300
    },
    {
      "epoch": 0.20181166378592919,
      "grad_norm": 0.0072026425041258335,
      "learning_rate": 0.2819871111494415,
      "loss": 0.6446,
      "num_input_tokens_seen": 8679392,
      "step": 6305
    },
    {
      "epoch": 0.20197170475641765,
      "grad_norm": 0.00747298588976264,
      "learning_rate": 0.28195911333825113,
      "loss": 0.7045,
      "num_input_tokens_seen": 8686384,
      "step": 6310
    },
    {
      "epoch": 0.2021317457269061,
      "grad_norm": 0.007071210071444511,
      "learning_rate": 0.28193109517730713,
      "loss": 0.8799,
      "num_input_tokens_seen": 8693328,
      "step": 6315
    },
    {
      "epoch": 0.20229178669739453,
      "grad_norm": 0.013279836624860764,
      "learning_rate": 0.2819030566709303,
      "loss": 0.9632,
      "num_input_tokens_seen": 8700208,
      "step": 6320
    },
    {
      "epoch": 0.20245182766788297,
      "grad_norm": 0.010061192326247692,
      "learning_rate": 0.2818749978234445,
      "loss": 0.7925,
      "num_input_tokens_seen": 8707344,
      "step": 6325
    },
    {
      "epoch": 0.2026118686383714,
      "grad_norm": 0.006635295692831278,
      "learning_rate": 0.2818469186391768,
      "loss": 0.6489,
      "num_input_tokens_seen": 8714272,
      "step": 6330
    },
    {
      "epoch": 0.20277190960885988,
      "grad_norm": 0.007113542873412371,
      "learning_rate": 0.28181881912245743,
      "loss": 0.8632,
      "num_input_tokens_seen": 8720976,
      "step": 6335
    },
    {
      "epoch": 0.20293195057934832,
      "grad_norm": 0.008498596958816051,
      "learning_rate": 0.2817906992776195,
      "loss": 0.8926,
      "num_input_tokens_seen": 8727744,
      "step": 6340
    },
    {
      "epoch": 0.20309199154983676,
      "grad_norm": 0.005939749535173178,
      "learning_rate": 0.28176255910899967,
      "loss": 0.6871,
      "num_input_tokens_seen": 8734608,
      "step": 6345
    },
    {
      "epoch": 0.2032520325203252,
      "grad_norm": 0.005801820196211338,
      "learning_rate": 0.2817343986209373,
      "loss": 0.9131,
      "num_input_tokens_seen": 8741408,
      "step": 6350
    },
    {
      "epoch": 0.20341207349081364,
      "grad_norm": 0.00677253445610404,
      "learning_rate": 0.2817062178177753,
      "loss": 0.7354,
      "num_input_tokens_seen": 8748544,
      "step": 6355
    },
    {
      "epoch": 0.2035721144613021,
      "grad_norm": 0.008374691009521484,
      "learning_rate": 0.2816780167038593,
      "loss": 0.756,
      "num_input_tokens_seen": 8755216,
      "step": 6360
    },
    {
      "epoch": 0.20373215543179055,
      "grad_norm": 0.008978335186839104,
      "learning_rate": 0.28164979528353834,
      "loss": 0.8076,
      "num_input_tokens_seen": 8761952,
      "step": 6365
    },
    {
      "epoch": 0.20389219640227899,
      "grad_norm": 0.008097509853541851,
      "learning_rate": 0.28162155356116453,
      "loss": 0.8045,
      "num_input_tokens_seen": 8768608,
      "step": 6370
    },
    {
      "epoch": 0.20405223737276743,
      "grad_norm": 0.005467125680297613,
      "learning_rate": 0.28159329154109314,
      "loss": 0.7168,
      "num_input_tokens_seen": 8775840,
      "step": 6375
    },
    {
      "epoch": 0.20421227834325587,
      "grad_norm": 0.005683799274265766,
      "learning_rate": 0.28156500922768246,
      "loss": 0.6958,
      "num_input_tokens_seen": 8782736,
      "step": 6380
    },
    {
      "epoch": 0.2043723193137443,
      "grad_norm": 0.010263610631227493,
      "learning_rate": 0.28153670662529406,
      "loss": 0.8302,
      "num_input_tokens_seen": 8789392,
      "step": 6385
    },
    {
      "epoch": 0.20453236028423277,
      "grad_norm": 0.010885009542107582,
      "learning_rate": 0.28150838373829246,
      "loss": 0.7098,
      "num_input_tokens_seen": 8796368,
      "step": 6390
    },
    {
      "epoch": 0.2046924012547212,
      "grad_norm": 0.008399107493460178,
      "learning_rate": 0.2814800405710455,
      "loss": 0.738,
      "num_input_tokens_seen": 8803088,
      "step": 6395
    },
    {
      "epoch": 0.20485244222520965,
      "grad_norm": 0.009569660760462284,
      "learning_rate": 0.2814516771279239,
      "loss": 0.7955,
      "num_input_tokens_seen": 8809968,
      "step": 6400
    },
    {
      "epoch": 0.20485244222520965,
      "eval_loss": 0.7670645713806152,
      "eval_runtime": 332.0325,
      "eval_samples_per_second": 41.821,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 8809968,
      "step": 6400
    },
    {
      "epoch": 0.2050124831956981,
      "grad_norm": 0.011988811194896698,
      "learning_rate": 0.28142329341330186,
      "loss": 0.6498,
      "num_input_tokens_seen": 8816432,
      "step": 6405
    },
    {
      "epoch": 0.20517252416618653,
      "grad_norm": 0.008313342928886414,
      "learning_rate": 0.2813948894315564,
      "loss": 0.6912,
      "num_input_tokens_seen": 8823568,
      "step": 6410
    },
    {
      "epoch": 0.205332565136675,
      "grad_norm": 0.006247850600630045,
      "learning_rate": 0.2813664651870677,
      "loss": 0.6303,
      "num_input_tokens_seen": 8830736,
      "step": 6415
    },
    {
      "epoch": 0.20549260610716344,
      "grad_norm": 0.01147557608783245,
      "learning_rate": 0.28133802068421926,
      "loss": 0.675,
      "num_input_tokens_seen": 8837824,
      "step": 6420
    },
    {
      "epoch": 0.20565264707765188,
      "grad_norm": 0.0170818530023098,
      "learning_rate": 0.28130955592739754,
      "loss": 0.7093,
      "num_input_tokens_seen": 8844432,
      "step": 6425
    },
    {
      "epoch": 0.20581268804814032,
      "grad_norm": 0.005286037921905518,
      "learning_rate": 0.2812810709209922,
      "loss": 0.645,
      "num_input_tokens_seen": 8851200,
      "step": 6430
    },
    {
      "epoch": 0.20597272901862876,
      "grad_norm": 0.013163374736905098,
      "learning_rate": 0.2812525656693959,
      "loss": 0.7401,
      "num_input_tokens_seen": 8857680,
      "step": 6435
    },
    {
      "epoch": 0.20613276998911723,
      "grad_norm": 0.009790965355932713,
      "learning_rate": 0.28122404017700453,
      "loss": 0.7238,
      "num_input_tokens_seen": 8864896,
      "step": 6440
    },
    {
      "epoch": 0.20629281095960567,
      "grad_norm": 0.007599156815558672,
      "learning_rate": 0.2811954944482171,
      "loss": 0.6135,
      "num_input_tokens_seen": 8871632,
      "step": 6445
    },
    {
      "epoch": 0.2064528519300941,
      "grad_norm": 0.005401734262704849,
      "learning_rate": 0.2811669284874358,
      "loss": 0.7978,
      "num_input_tokens_seen": 8878848,
      "step": 6450
    },
    {
      "epoch": 0.20661289290058255,
      "grad_norm": 0.005290291737765074,
      "learning_rate": 0.2811383422990657,
      "loss": 0.628,
      "num_input_tokens_seen": 8885872,
      "step": 6455
    },
    {
      "epoch": 0.20677293387107099,
      "grad_norm": 0.00904136523604393,
      "learning_rate": 0.2811097358875152,
      "loss": 0.8501,
      "num_input_tokens_seen": 8892480,
      "step": 6460
    },
    {
      "epoch": 0.20693297484155945,
      "grad_norm": 0.008955871686339378,
      "learning_rate": 0.2810811092571959,
      "loss": 0.7596,
      "num_input_tokens_seen": 8899616,
      "step": 6465
    },
    {
      "epoch": 0.2070930158120479,
      "grad_norm": 0.009583374485373497,
      "learning_rate": 0.28105246241252224,
      "loss": 0.692,
      "num_input_tokens_seen": 8906304,
      "step": 6470
    },
    {
      "epoch": 0.20725305678253633,
      "grad_norm": 0.013226358219981194,
      "learning_rate": 0.28102379535791194,
      "loss": 0.8674,
      "num_input_tokens_seen": 8912896,
      "step": 6475
    },
    {
      "epoch": 0.20741309775302477,
      "grad_norm": 0.006102576851844788,
      "learning_rate": 0.2809951080977859,
      "loss": 0.7749,
      "num_input_tokens_seen": 8919664,
      "step": 6480
    },
    {
      "epoch": 0.2075731387235132,
      "grad_norm": 0.0070810867473483086,
      "learning_rate": 0.28096640063656797,
      "loss": 0.6552,
      "num_input_tokens_seen": 8926464,
      "step": 6485
    },
    {
      "epoch": 0.20773317969400165,
      "grad_norm": 0.006936549674719572,
      "learning_rate": 0.2809376729786852,
      "loss": 0.843,
      "num_input_tokens_seen": 8933472,
      "step": 6490
    },
    {
      "epoch": 0.20789322066449012,
      "grad_norm": 0.008248903788626194,
      "learning_rate": 0.28090892512856785,
      "loss": 0.914,
      "num_input_tokens_seen": 8940544,
      "step": 6495
    },
    {
      "epoch": 0.20805326163497856,
      "grad_norm": 0.006772045977413654,
      "learning_rate": 0.2808801570906491,
      "loss": 0.6638,
      "num_input_tokens_seen": 8947392,
      "step": 6500
    },
    {
      "epoch": 0.208213302605467,
      "grad_norm": 0.008250614628195763,
      "learning_rate": 0.2808513688693654,
      "loss": 0.8282,
      "num_input_tokens_seen": 8954400,
      "step": 6505
    },
    {
      "epoch": 0.20837334357595544,
      "grad_norm": 0.008624937385320663,
      "learning_rate": 0.28082256046915627,
      "loss": 0.7845,
      "num_input_tokens_seen": 8960864,
      "step": 6510
    },
    {
      "epoch": 0.20853338454644388,
      "grad_norm": 0.01630803942680359,
      "learning_rate": 0.28079373189446427,
      "loss": 0.8471,
      "num_input_tokens_seen": 8967520,
      "step": 6515
    },
    {
      "epoch": 0.20869342551693235,
      "grad_norm": 0.010842268355190754,
      "learning_rate": 0.28076488314973513,
      "loss": 0.6881,
      "num_input_tokens_seen": 8974256,
      "step": 6520
    },
    {
      "epoch": 0.20885346648742079,
      "grad_norm": 0.006630707532167435,
      "learning_rate": 0.28073601423941774,
      "loss": 0.7613,
      "num_input_tokens_seen": 8980976,
      "step": 6525
    },
    {
      "epoch": 0.20901350745790923,
      "grad_norm": 0.007007596082985401,
      "learning_rate": 0.28070712516796403,
      "loss": 0.82,
      "num_input_tokens_seen": 8987680,
      "step": 6530
    },
    {
      "epoch": 0.20917354842839767,
      "grad_norm": 0.006246758159250021,
      "learning_rate": 0.28067821593982906,
      "loss": 0.6878,
      "num_input_tokens_seen": 8994384,
      "step": 6535
    },
    {
      "epoch": 0.2093335893988861,
      "grad_norm": 0.005485184490680695,
      "learning_rate": 0.28064928655947097,
      "loss": 0.6479,
      "num_input_tokens_seen": 9001360,
      "step": 6540
    },
    {
      "epoch": 0.20949363036937457,
      "grad_norm": 0.006268834229558706,
      "learning_rate": 0.28062033703135103,
      "loss": 0.8611,
      "num_input_tokens_seen": 9008704,
      "step": 6545
    },
    {
      "epoch": 0.209653671339863,
      "grad_norm": 0.005345207639038563,
      "learning_rate": 0.2805913673599337,
      "loss": 0.5553,
      "num_input_tokens_seen": 9015712,
      "step": 6550
    },
    {
      "epoch": 0.20981371231035145,
      "grad_norm": 0.009858453646302223,
      "learning_rate": 0.2805623775496864,
      "loss": 0.9331,
      "num_input_tokens_seen": 9022736,
      "step": 6555
    },
    {
      "epoch": 0.2099737532808399,
      "grad_norm": 0.010561661794781685,
      "learning_rate": 0.2805333676050797,
      "loss": 0.9516,
      "num_input_tokens_seen": 9029648,
      "step": 6560
    },
    {
      "epoch": 0.21013379425132833,
      "grad_norm": 0.005508619826287031,
      "learning_rate": 0.2805043375305873,
      "loss": 0.7954,
      "num_input_tokens_seen": 9036672,
      "step": 6565
    },
    {
      "epoch": 0.21029383522181677,
      "grad_norm": 0.007363644428551197,
      "learning_rate": 0.2804752873306861,
      "loss": 0.5244,
      "num_input_tokens_seen": 9043456,
      "step": 6570
    },
    {
      "epoch": 0.21045387619230524,
      "grad_norm": 0.005350339226424694,
      "learning_rate": 0.2804462170098559,
      "loss": 0.6038,
      "num_input_tokens_seen": 9050320,
      "step": 6575
    },
    {
      "epoch": 0.21061391716279368,
      "grad_norm": 0.009279624558985233,
      "learning_rate": 0.2804171265725797,
      "loss": 0.6401,
      "num_input_tokens_seen": 9056976,
      "step": 6580
    },
    {
      "epoch": 0.21077395813328212,
      "grad_norm": 0.006327273324131966,
      "learning_rate": 0.28038801602334373,
      "loss": 0.8486,
      "num_input_tokens_seen": 9063504,
      "step": 6585
    },
    {
      "epoch": 0.21093399910377056,
      "grad_norm": 0.009774543344974518,
      "learning_rate": 0.28035888536663717,
      "loss": 0.7694,
      "num_input_tokens_seen": 9070240,
      "step": 6590
    },
    {
      "epoch": 0.211094040074259,
      "grad_norm": 0.015831872820854187,
      "learning_rate": 0.2803297346069522,
      "loss": 0.7909,
      "num_input_tokens_seen": 9076880,
      "step": 6595
    },
    {
      "epoch": 0.21125408104474747,
      "grad_norm": 0.0069837854243814945,
      "learning_rate": 0.28030056374878437,
      "loss": 0.8651,
      "num_input_tokens_seen": 9084016,
      "step": 6600
    },
    {
      "epoch": 0.21125408104474747,
      "eval_loss": 0.7741808295249939,
      "eval_runtime": 332.3994,
      "eval_samples_per_second": 41.775,
      "eval_steps_per_second": 20.888,
      "num_input_tokens_seen": 9084016,
      "step": 6600
    },
    {
      "epoch": 0.2114141220152359,
      "grad_norm": 0.013852512463927269,
      "learning_rate": 0.2802713727966321,
      "loss": 0.881,
      "num_input_tokens_seen": 9091120,
      "step": 6605
    },
    {
      "epoch": 0.21157416298572435,
      "grad_norm": 0.009764902293682098,
      "learning_rate": 0.28024216175499717,
      "loss": 0.7142,
      "num_input_tokens_seen": 9097552,
      "step": 6610
    },
    {
      "epoch": 0.21173420395621279,
      "grad_norm": 0.010285402648150921,
      "learning_rate": 0.2802129306283841,
      "loss": 0.7843,
      "num_input_tokens_seen": 9104416,
      "step": 6615
    },
    {
      "epoch": 0.21189424492670123,
      "grad_norm": 0.011627767235040665,
      "learning_rate": 0.28018367942130074,
      "loss": 0.635,
      "num_input_tokens_seen": 9111008,
      "step": 6620
    },
    {
      "epoch": 0.2120542858971897,
      "grad_norm": 0.06358083337545395,
      "learning_rate": 0.28015440813825804,
      "loss": 0.91,
      "num_input_tokens_seen": 9117664,
      "step": 6625
    },
    {
      "epoch": 0.21221432686767813,
      "grad_norm": 0.010422845371067524,
      "learning_rate": 0.28012511678377006,
      "loss": 1.0566,
      "num_input_tokens_seen": 9124400,
      "step": 6630
    },
    {
      "epoch": 0.21237436783816657,
      "grad_norm": 0.031110476702451706,
      "learning_rate": 0.28009580536235373,
      "loss": 0.7122,
      "num_input_tokens_seen": 9131024,
      "step": 6635
    },
    {
      "epoch": 0.212534408808655,
      "grad_norm": 0.007440597750246525,
      "learning_rate": 0.28006647387852934,
      "loss": 0.7357,
      "num_input_tokens_seen": 9138064,
      "step": 6640
    },
    {
      "epoch": 0.21269444977914345,
      "grad_norm": 0.008198872208595276,
      "learning_rate": 0.28003712233682015,
      "loss": 0.7533,
      "num_input_tokens_seen": 9144976,
      "step": 6645
    },
    {
      "epoch": 0.21285449074963192,
      "grad_norm": 0.007891638204455376,
      "learning_rate": 0.2800077507417526,
      "loss": 0.7709,
      "num_input_tokens_seen": 9151664,
      "step": 6650
    },
    {
      "epoch": 0.21301453172012036,
      "grad_norm": 0.004610520321875811,
      "learning_rate": 0.2799783590978561,
      "loss": 0.7177,
      "num_input_tokens_seen": 9158784,
      "step": 6655
    },
    {
      "epoch": 0.2131745726906088,
      "grad_norm": 0.010614047758281231,
      "learning_rate": 0.2799489474096632,
      "loss": 0.7256,
      "num_input_tokens_seen": 9165664,
      "step": 6660
    },
    {
      "epoch": 0.21333461366109724,
      "grad_norm": 0.01005402859300375,
      "learning_rate": 0.27991951568170953,
      "loss": 0.6459,
      "num_input_tokens_seen": 9172096,
      "step": 6665
    },
    {
      "epoch": 0.21349465463158568,
      "grad_norm": 0.014971456490457058,
      "learning_rate": 0.2798900639185339,
      "loss": 0.7185,
      "num_input_tokens_seen": 9178528,
      "step": 6670
    },
    {
      "epoch": 0.21365469560207412,
      "grad_norm": 0.00803318154066801,
      "learning_rate": 0.2798605921246781,
      "loss": 0.8084,
      "num_input_tokens_seen": 9185872,
      "step": 6675
    },
    {
      "epoch": 0.21381473657256259,
      "grad_norm": 0.011306422762572765,
      "learning_rate": 0.2798311003046871,
      "loss": 0.7774,
      "num_input_tokens_seen": 9193232,
      "step": 6680
    },
    {
      "epoch": 0.21397477754305103,
      "grad_norm": 0.00720314821228385,
      "learning_rate": 0.2798015884631089,
      "loss": 0.7915,
      "num_input_tokens_seen": 9199808,
      "step": 6685
    },
    {
      "epoch": 0.21413481851353947,
      "grad_norm": 0.009711635299026966,
      "learning_rate": 0.27977205660449445,
      "loss": 0.8177,
      "num_input_tokens_seen": 9206672,
      "step": 6690
    },
    {
      "epoch": 0.2142948594840279,
      "grad_norm": 0.00998744461685419,
      "learning_rate": 0.2797425047333981,
      "loss": 0.7035,
      "num_input_tokens_seen": 9213600,
      "step": 6695
    },
    {
      "epoch": 0.21445490045451635,
      "grad_norm": 0.00629368657246232,
      "learning_rate": 0.27971293285437715,
      "loss": 0.7604,
      "num_input_tokens_seen": 9220352,
      "step": 6700
    },
    {
      "epoch": 0.2146149414250048,
      "grad_norm": 0.007616323884576559,
      "learning_rate": 0.2796833409719918,
      "loss": 0.7822,
      "num_input_tokens_seen": 9227328,
      "step": 6705
    },
    {
      "epoch": 0.21477498239549325,
      "grad_norm": 0.009631659835577011,
      "learning_rate": 0.27965372909080566,
      "loss": 0.8125,
      "num_input_tokens_seen": 9234304,
      "step": 6710
    },
    {
      "epoch": 0.2149350233659817,
      "grad_norm": 0.006187789607793093,
      "learning_rate": 0.27962409721538506,
      "loss": 0.6975,
      "num_input_tokens_seen": 9241136,
      "step": 6715
    },
    {
      "epoch": 0.21509506433647013,
      "grad_norm": 0.005578339099884033,
      "learning_rate": 0.27959444535029976,
      "loss": 0.6735,
      "num_input_tokens_seen": 9248576,
      "step": 6720
    },
    {
      "epoch": 0.21525510530695857,
      "grad_norm": 0.008327590301632881,
      "learning_rate": 0.27956477350012243,
      "loss": 0.7574,
      "num_input_tokens_seen": 9255168,
      "step": 6725
    },
    {
      "epoch": 0.21541514627744704,
      "grad_norm": 0.008377056568861008,
      "learning_rate": 0.27953508166942875,
      "loss": 0.741,
      "num_input_tokens_seen": 9261904,
      "step": 6730
    },
    {
      "epoch": 0.21557518724793548,
      "grad_norm": 0.005650956649333239,
      "learning_rate": 0.27950536986279767,
      "loss": 0.6324,
      "num_input_tokens_seen": 9268816,
      "step": 6735
    },
    {
      "epoch": 0.21573522821842392,
      "grad_norm": 0.008611026220023632,
      "learning_rate": 0.2794756380848111,
      "loss": 0.7047,
      "num_input_tokens_seen": 9275408,
      "step": 6740
    },
    {
      "epoch": 0.21589526918891236,
      "grad_norm": 0.018567921593785286,
      "learning_rate": 0.279445886340054,
      "loss": 0.8973,
      "num_input_tokens_seen": 9282272,
      "step": 6745
    },
    {
      "epoch": 0.2160553101594008,
      "grad_norm": 0.010167974978685379,
      "learning_rate": 0.27941611463311455,
      "loss": 0.7507,
      "num_input_tokens_seen": 9288560,
      "step": 6750
    },
    {
      "epoch": 0.21621535112988924,
      "grad_norm": 0.005401963833719492,
      "learning_rate": 0.2793863229685839,
      "loss": 0.5362,
      "num_input_tokens_seen": 9295488,
      "step": 6755
    },
    {
      "epoch": 0.2163753921003777,
      "grad_norm": 0.007051957305520773,
      "learning_rate": 0.27935651135105627,
      "loss": 0.9404,
      "num_input_tokens_seen": 9302352,
      "step": 6760
    },
    {
      "epoch": 0.21653543307086615,
      "grad_norm": 0.004601518157869577,
      "learning_rate": 0.279326679785129,
      "loss": 0.5765,
      "num_input_tokens_seen": 9309232,
      "step": 6765
    },
    {
      "epoch": 0.21669547404135459,
      "grad_norm": 0.013526241295039654,
      "learning_rate": 0.2792968282754024,
      "loss": 0.7005,
      "num_input_tokens_seen": 9315792,
      "step": 6770
    },
    {
      "epoch": 0.21685551501184303,
      "grad_norm": 0.007720015477389097,
      "learning_rate": 0.2792669568264801,
      "loss": 0.8223,
      "num_input_tokens_seen": 9322976,
      "step": 6775
    },
    {
      "epoch": 0.21701555598233147,
      "grad_norm": 0.008720126003026962,
      "learning_rate": 0.27923706544296856,
      "loss": 0.7957,
      "num_input_tokens_seen": 9330080,
      "step": 6780
    },
    {
      "epoch": 0.21717559695281993,
      "grad_norm": 0.011334408074617386,
      "learning_rate": 0.2792071541294775,
      "loss": 0.634,
      "num_input_tokens_seen": 9336848,
      "step": 6785
    },
    {
      "epoch": 0.21733563792330837,
      "grad_norm": 0.008365643210709095,
      "learning_rate": 0.27917722289061947,
      "loss": 0.7528,
      "num_input_tokens_seen": 9343632,
      "step": 6790
    },
    {
      "epoch": 0.2174956788937968,
      "grad_norm": 0.00997112412005663,
      "learning_rate": 0.27914727173101034,
      "loss": 0.6964,
      "num_input_tokens_seen": 9350400,
      "step": 6795
    },
    {
      "epoch": 0.21765571986428525,
      "grad_norm": 0.00799122266471386,
      "learning_rate": 0.279117300655269,
      "loss": 0.8679,
      "num_input_tokens_seen": 9357456,
      "step": 6800
    },
    {
      "epoch": 0.21765571986428525,
      "eval_loss": 0.7769171595573425,
      "eval_runtime": 331.948,
      "eval_samples_per_second": 41.832,
      "eval_steps_per_second": 20.916,
      "num_input_tokens_seen": 9357456,
      "step": 6800
    },
    {
      "epoch": 0.2178157608347737,
      "grad_norm": 0.011308832094073296,
      "learning_rate": 0.2790873096680173,
      "loss": 0.8191,
      "num_input_tokens_seen": 9363968,
      "step": 6805
    },
    {
      "epoch": 0.21797580180526216,
      "grad_norm": 0.009764011017978191,
      "learning_rate": 0.2790572987738802,
      "loss": 0.8093,
      "num_input_tokens_seen": 9370848,
      "step": 6810
    },
    {
      "epoch": 0.2181358427757506,
      "grad_norm": 0.006482793018221855,
      "learning_rate": 0.27902726797748584,
      "loss": 0.7747,
      "num_input_tokens_seen": 9377584,
      "step": 6815
    },
    {
      "epoch": 0.21829588374623904,
      "grad_norm": 0.0068987575359642506,
      "learning_rate": 0.2789972172834652,
      "loss": 0.7918,
      "num_input_tokens_seen": 9384416,
      "step": 6820
    },
    {
      "epoch": 0.21845592471672748,
      "grad_norm": 0.005598864052444696,
      "learning_rate": 0.2789671466964527,
      "loss": 0.7988,
      "num_input_tokens_seen": 9391072,
      "step": 6825
    },
    {
      "epoch": 0.21861596568721592,
      "grad_norm": 0.00891961995512247,
      "learning_rate": 0.2789370562210854,
      "loss": 0.7798,
      "num_input_tokens_seen": 9397808,
      "step": 6830
    },
    {
      "epoch": 0.21877600665770439,
      "grad_norm": 0.010939392261207104,
      "learning_rate": 0.27890694586200376,
      "loss": 0.8235,
      "num_input_tokens_seen": 9404960,
      "step": 6835
    },
    {
      "epoch": 0.21893604762819283,
      "grad_norm": 0.012928230687975883,
      "learning_rate": 0.2788768156238511,
      "loss": 0.7731,
      "num_input_tokens_seen": 9411584,
      "step": 6840
    },
    {
      "epoch": 0.21909608859868127,
      "grad_norm": 0.012965746223926544,
      "learning_rate": 0.27884666551127385,
      "loss": 0.7947,
      "num_input_tokens_seen": 9418368,
      "step": 6845
    },
    {
      "epoch": 0.2192561295691697,
      "grad_norm": 0.012642547488212585,
      "learning_rate": 0.2788164955289217,
      "loss": 0.8617,
      "num_input_tokens_seen": 9425376,
      "step": 6850
    },
    {
      "epoch": 0.21941617053965815,
      "grad_norm": 0.010647843591868877,
      "learning_rate": 0.27878630568144697,
      "loss": 0.7923,
      "num_input_tokens_seen": 9432608,
      "step": 6855
    },
    {
      "epoch": 0.21957621151014659,
      "grad_norm": 0.009303695522248745,
      "learning_rate": 0.2787560959735056,
      "loss": 0.8981,
      "num_input_tokens_seen": 9439328,
      "step": 6860
    },
    {
      "epoch": 0.21973625248063505,
      "grad_norm": 0.007296333089470863,
      "learning_rate": 0.27872586640975616,
      "loss": 0.7377,
      "num_input_tokens_seen": 9445664,
      "step": 6865
    },
    {
      "epoch": 0.2198962934511235,
      "grad_norm": 0.00532235624268651,
      "learning_rate": 0.27869561699486045,
      "loss": 0.8745,
      "num_input_tokens_seen": 9452688,
      "step": 6870
    },
    {
      "epoch": 0.22005633442161193,
      "grad_norm": 0.004808558616787195,
      "learning_rate": 0.2786653477334833,
      "loss": 0.7528,
      "num_input_tokens_seen": 9459664,
      "step": 6875
    },
    {
      "epoch": 0.22021637539210037,
      "grad_norm": 0.006912432610988617,
      "learning_rate": 0.2786350586302926,
      "loss": 0.7131,
      "num_input_tokens_seen": 9466688,
      "step": 6880
    },
    {
      "epoch": 0.2203764163625888,
      "grad_norm": 0.009063828736543655,
      "learning_rate": 0.27860474968995935,
      "loss": 0.7862,
      "num_input_tokens_seen": 9473616,
      "step": 6885
    },
    {
      "epoch": 0.22053645733307728,
      "grad_norm": 0.008398594334721565,
      "learning_rate": 0.27857442091715756,
      "loss": 0.7612,
      "num_input_tokens_seen": 9480400,
      "step": 6890
    },
    {
      "epoch": 0.22069649830356572,
      "grad_norm": 0.01031077653169632,
      "learning_rate": 0.27854407231656425,
      "loss": 0.8987,
      "num_input_tokens_seen": 9487456,
      "step": 6895
    },
    {
      "epoch": 0.22085653927405416,
      "grad_norm": 0.008153977803885937,
      "learning_rate": 0.2785137038928596,
      "loss": 0.8152,
      "num_input_tokens_seen": 9494016,
      "step": 6900
    },
    {
      "epoch": 0.2210165802445426,
      "grad_norm": 0.011446037329733372,
      "learning_rate": 0.27848331565072687,
      "loss": 0.7649,
      "num_input_tokens_seen": 9501232,
      "step": 6905
    },
    {
      "epoch": 0.22117662121503104,
      "grad_norm": 0.0052955299615859985,
      "learning_rate": 0.27845290759485225,
      "loss": 0.6741,
      "num_input_tokens_seen": 9507904,
      "step": 6910
    },
    {
      "epoch": 0.2213366621855195,
      "grad_norm": 0.007167611736804247,
      "learning_rate": 0.278422479729925,
      "loss": 0.8459,
      "num_input_tokens_seen": 9514736,
      "step": 6915
    },
    {
      "epoch": 0.22149670315600795,
      "grad_norm": 0.008100121282041073,
      "learning_rate": 0.2783920320606375,
      "loss": 0.8146,
      "num_input_tokens_seen": 9521312,
      "step": 6920
    },
    {
      "epoch": 0.22165674412649639,
      "grad_norm": 0.007301321253180504,
      "learning_rate": 0.2783615645916852,
      "loss": 0.8542,
      "num_input_tokens_seen": 9528176,
      "step": 6925
    },
    {
      "epoch": 0.22181678509698483,
      "grad_norm": 0.005757525563240051,
      "learning_rate": 0.2783310773277666,
      "loss": 0.7262,
      "num_input_tokens_seen": 9535040,
      "step": 6930
    },
    {
      "epoch": 0.22197682606747327,
      "grad_norm": 0.005359482020139694,
      "learning_rate": 0.2783005702735831,
      "loss": 0.8872,
      "num_input_tokens_seen": 9542240,
      "step": 6935
    },
    {
      "epoch": 0.2221368670379617,
      "grad_norm": 0.014628958888351917,
      "learning_rate": 0.2782700434338394,
      "loss": 0.8515,
      "num_input_tokens_seen": 9549360,
      "step": 6940
    },
    {
      "epoch": 0.22229690800845017,
      "grad_norm": 0.007337990682572126,
      "learning_rate": 0.278239496813243,
      "loss": 0.5819,
      "num_input_tokens_seen": 9556288,
      "step": 6945
    },
    {
      "epoch": 0.2224569489789386,
      "grad_norm": 0.007038144860416651,
      "learning_rate": 0.27820893041650463,
      "loss": 0.7008,
      "num_input_tokens_seen": 9563040,
      "step": 6950
    },
    {
      "epoch": 0.22261698994942705,
      "grad_norm": 0.009728698991239071,
      "learning_rate": 0.27817834424833804,
      "loss": 0.8665,
      "num_input_tokens_seen": 9569552,
      "step": 6955
    },
    {
      "epoch": 0.2227770309199155,
      "grad_norm": 0.008236253634095192,
      "learning_rate": 0.27814773831345996,
      "loss": 0.7967,
      "num_input_tokens_seen": 9576176,
      "step": 6960
    },
    {
      "epoch": 0.22293707189040393,
      "grad_norm": 0.004923753906041384,
      "learning_rate": 0.2781171126165902,
      "loss": 0.6819,
      "num_input_tokens_seen": 9583408,
      "step": 6965
    },
    {
      "epoch": 0.2230971128608924,
      "grad_norm": 0.005723756738007069,
      "learning_rate": 0.2780864671624517,
      "loss": 0.6979,
      "num_input_tokens_seen": 9590224,
      "step": 6970
    },
    {
      "epoch": 0.22325715383138084,
      "grad_norm": 0.006962136831134558,
      "learning_rate": 0.27805580195577034,
      "loss": 0.7947,
      "num_input_tokens_seen": 9596816,
      "step": 6975
    },
    {
      "epoch": 0.22341719480186928,
      "grad_norm": 0.006069603376090527,
      "learning_rate": 0.2780251170012751,
      "loss": 0.6684,
      "num_input_tokens_seen": 9603232,
      "step": 6980
    },
    {
      "epoch": 0.22357723577235772,
      "grad_norm": 0.007685716263949871,
      "learning_rate": 0.27799441230369787,
      "loss": 0.6448,
      "num_input_tokens_seen": 9609920,
      "step": 6985
    },
    {
      "epoch": 0.22373727674284616,
      "grad_norm": 0.005671888589859009,
      "learning_rate": 0.27796368786777387,
      "loss": 0.6441,
      "num_input_tokens_seen": 9616720,
      "step": 6990
    },
    {
      "epoch": 0.22389731771333463,
      "grad_norm": 0.010294433683156967,
      "learning_rate": 0.277932943698241,
      "loss": 0.9355,
      "num_input_tokens_seen": 9623712,
      "step": 6995
    },
    {
      "epoch": 0.22405735868382307,
      "grad_norm": 0.008738717064261436,
      "learning_rate": 0.2779021797998406,
      "loss": 0.7892,
      "num_input_tokens_seen": 9630608,
      "step": 7000
    },
    {
      "epoch": 0.22405735868382307,
      "eval_loss": 0.7837414145469666,
      "eval_runtime": 332.0743,
      "eval_samples_per_second": 41.816,
      "eval_steps_per_second": 20.908,
      "num_input_tokens_seen": 9630608,
      "step": 7000
    },
    {
      "epoch": 0.2242173996543115,
      "grad_norm": 0.007734429556876421,
      "learning_rate": 0.2778713961773167,
      "loss": 0.7882,
      "num_input_tokens_seen": 9637248,
      "step": 7005
    },
    {
      "epoch": 0.22437744062479995,
      "grad_norm": 0.01042780838906765,
      "learning_rate": 0.2778405928354166,
      "loss": 0.7184,
      "num_input_tokens_seen": 9644064,
      "step": 7010
    },
    {
      "epoch": 0.22453748159528839,
      "grad_norm": 0.011777355335652828,
      "learning_rate": 0.27780976977889055,
      "loss": 0.7003,
      "num_input_tokens_seen": 9651424,
      "step": 7015
    },
    {
      "epoch": 0.22469752256577685,
      "grad_norm": 0.010437076911330223,
      "learning_rate": 0.27777892701249185,
      "loss": 0.8519,
      "num_input_tokens_seen": 9658272,
      "step": 7020
    },
    {
      "epoch": 0.2248575635362653,
      "grad_norm": 0.013256513513624668,
      "learning_rate": 0.2777480645409768,
      "loss": 0.7598,
      "num_input_tokens_seen": 9665792,
      "step": 7025
    },
    {
      "epoch": 0.22501760450675373,
      "grad_norm": 0.030314290896058083,
      "learning_rate": 0.27771718236910486,
      "loss": 0.7915,
      "num_input_tokens_seen": 9672640,
      "step": 7030
    },
    {
      "epoch": 0.22517764547724217,
      "grad_norm": 0.02735884115099907,
      "learning_rate": 0.27768628050163835,
      "loss": 0.7398,
      "num_input_tokens_seen": 9679568,
      "step": 7035
    },
    {
      "epoch": 0.2253376864477306,
      "grad_norm": 0.011243275366723537,
      "learning_rate": 0.2776553589433428,
      "loss": 0.8616,
      "num_input_tokens_seen": 9686800,
      "step": 7040
    },
    {
      "epoch": 0.22549772741821905,
      "grad_norm": 0.00683556217700243,
      "learning_rate": 0.27762441769898666,
      "loss": 0.9094,
      "num_input_tokens_seen": 9693984,
      "step": 7045
    },
    {
      "epoch": 0.22565776838870752,
      "grad_norm": 0.08169539272785187,
      "learning_rate": 0.2775934567733415,
      "loss": 0.7711,
      "num_input_tokens_seen": 9700560,
      "step": 7050
    },
    {
      "epoch": 0.22581780935919596,
      "grad_norm": 0.007507712114602327,
      "learning_rate": 0.2775624761711819,
      "loss": 0.707,
      "num_input_tokens_seen": 9707392,
      "step": 7055
    },
    {
      "epoch": 0.2259778503296844,
      "grad_norm": 0.00901772640645504,
      "learning_rate": 0.2775314758972854,
      "loss": 0.8663,
      "num_input_tokens_seen": 9714480,
      "step": 7060
    },
    {
      "epoch": 0.22613789130017284,
      "grad_norm": 0.00513352919369936,
      "learning_rate": 0.2775004559564327,
      "loss": 0.5869,
      "num_input_tokens_seen": 9721616,
      "step": 7065
    },
    {
      "epoch": 0.22629793227066128,
      "grad_norm": 0.007109996862709522,
      "learning_rate": 0.2774694163534073,
      "loss": 0.5899,
      "num_input_tokens_seen": 9728656,
      "step": 7070
    },
    {
      "epoch": 0.22645797324114975,
      "grad_norm": 0.019340965896844864,
      "learning_rate": 0.27743835709299614,
      "loss": 0.927,
      "num_input_tokens_seen": 9735184,
      "step": 7075
    },
    {
      "epoch": 0.22661801421163819,
      "grad_norm": 0.01125310454517603,
      "learning_rate": 0.2774072781799888,
      "loss": 0.6952,
      "num_input_tokens_seen": 9742112,
      "step": 7080
    },
    {
      "epoch": 0.22677805518212663,
      "grad_norm": 0.009324902668595314,
      "learning_rate": 0.27737617961917804,
      "loss": 0.7105,
      "num_input_tokens_seen": 9748880,
      "step": 7085
    },
    {
      "epoch": 0.22693809615261507,
      "grad_norm": 0.009816372767090797,
      "learning_rate": 0.27734506141535964,
      "loss": 0.6138,
      "num_input_tokens_seen": 9755824,
      "step": 7090
    },
    {
      "epoch": 0.2270981371231035,
      "grad_norm": 0.010193655267357826,
      "learning_rate": 0.2773139235733325,
      "loss": 0.7879,
      "num_input_tokens_seen": 9762416,
      "step": 7095
    },
    {
      "epoch": 0.22725817809359197,
      "grad_norm": 0.010665523819625378,
      "learning_rate": 0.2772827660978984,
      "loss": 0.791,
      "num_input_tokens_seen": 9769680,
      "step": 7100
    },
    {
      "epoch": 0.2274182190640804,
      "grad_norm": 0.008093771524727345,
      "learning_rate": 0.27725158899386226,
      "loss": 0.5249,
      "num_input_tokens_seen": 9776192,
      "step": 7105
    },
    {
      "epoch": 0.22757826003456885,
      "grad_norm": 0.007577726151794195,
      "learning_rate": 0.27722039226603196,
      "loss": 0.5836,
      "num_input_tokens_seen": 9783136,
      "step": 7110
    },
    {
      "epoch": 0.2277383010050573,
      "grad_norm": 0.008275214582681656,
      "learning_rate": 0.2771891759192184,
      "loss": 0.8052,
      "num_input_tokens_seen": 9789952,
      "step": 7115
    },
    {
      "epoch": 0.22789834197554573,
      "grad_norm": 0.012214920483529568,
      "learning_rate": 0.2771579399582355,
      "loss": 0.5308,
      "num_input_tokens_seen": 9796784,
      "step": 7120
    },
    {
      "epoch": 0.22805838294603417,
      "grad_norm": 0.009941665455698967,
      "learning_rate": 0.2771266843879004,
      "loss": 0.7578,
      "num_input_tokens_seen": 9803792,
      "step": 7125
    },
    {
      "epoch": 0.22821842391652264,
      "grad_norm": 0.012696162797510624,
      "learning_rate": 0.2770954092130329,
      "loss": 0.8589,
      "num_input_tokens_seen": 9810320,
      "step": 7130
    },
    {
      "epoch": 0.22837846488701108,
      "grad_norm": 0.06983236968517303,
      "learning_rate": 0.27706411443845613,
      "loss": 0.804,
      "num_input_tokens_seen": 9816960,
      "step": 7135
    },
    {
      "epoch": 0.22853850585749952,
      "grad_norm": 0.012736797332763672,
      "learning_rate": 0.27703280006899617,
      "loss": 0.9289,
      "num_input_tokens_seen": 9823968,
      "step": 7140
    },
    {
      "epoch": 0.22869854682798796,
      "grad_norm": 0.008412552066147327,
      "learning_rate": 0.277001466109482,
      "loss": 0.9538,
      "num_input_tokens_seen": 9830752,
      "step": 7145
    },
    {
      "epoch": 0.2288585877984764,
      "grad_norm": 0.010802467353641987,
      "learning_rate": 0.2769701125647458,
      "loss": 0.9381,
      "num_input_tokens_seen": 9837488,
      "step": 7150
    },
    {
      "epoch": 0.22901862876896487,
      "grad_norm": 0.008104544132947922,
      "learning_rate": 0.27693873943962266,
      "loss": 0.7584,
      "num_input_tokens_seen": 9845216,
      "step": 7155
    },
    {
      "epoch": 0.2291786697394533,
      "grad_norm": 0.010556623339653015,
      "learning_rate": 0.2769073467389506,
      "loss": 0.777,
      "num_input_tokens_seen": 9852000,
      "step": 7160
    },
    {
      "epoch": 0.22933871070994175,
      "grad_norm": 0.007131753023713827,
      "learning_rate": 0.2768759344675709,
      "loss": 0.7007,
      "num_input_tokens_seen": 9858608,
      "step": 7165
    },
    {
      "epoch": 0.22949875168043019,
      "grad_norm": 0.014550565741956234,
      "learning_rate": 0.27684450263032767,
      "loss": 0.9143,
      "num_input_tokens_seen": 9865392,
      "step": 7170
    },
    {
      "epoch": 0.22965879265091863,
      "grad_norm": 0.012071533128619194,
      "learning_rate": 0.2768130512320682,
      "loss": 0.6435,
      "num_input_tokens_seen": 9872352,
      "step": 7175
    },
    {
      "epoch": 0.2298188336214071,
      "grad_norm": 0.010502328164875507,
      "learning_rate": 0.27678158027764244,
      "loss": 0.6929,
      "num_input_tokens_seen": 9879408,
      "step": 7180
    },
    {
      "epoch": 0.22997887459189553,
      "grad_norm": 0.0062859500758349895,
      "learning_rate": 0.27675008977190385,
      "loss": 1.3535,
      "num_input_tokens_seen": 9886400,
      "step": 7185
    },
    {
      "epoch": 0.23013891556238397,
      "grad_norm": 0.008582514710724354,
      "learning_rate": 0.2767185797197086,
      "loss": 1.0389,
      "num_input_tokens_seen": 9893376,
      "step": 7190
    },
    {
      "epoch": 0.2302989565328724,
      "grad_norm": 0.022453399375081062,
      "learning_rate": 0.2766870501259159,
      "loss": 0.792,
      "num_input_tokens_seen": 9900400,
      "step": 7195
    },
    {
      "epoch": 0.23045899750336085,
      "grad_norm": 0.009674412198364735,
      "learning_rate": 0.276655500995388,
      "loss": 0.7311,
      "num_input_tokens_seen": 9907888,
      "step": 7200
    },
    {
      "epoch": 0.23045899750336085,
      "eval_loss": 0.7955318689346313,
      "eval_runtime": 332.1508,
      "eval_samples_per_second": 41.806,
      "eval_steps_per_second": 20.903,
      "num_input_tokens_seen": 9907888,
      "step": 7200
    },
    {
      "epoch": 0.2306190384738493,
      "grad_norm": 0.03207160532474518,
      "learning_rate": 0.27662393233299015,
      "loss": 0.963,
      "num_input_tokens_seen": 9915168,
      "step": 7205
    },
    {
      "epoch": 0.23077907944433776,
      "grad_norm": 0.00704196747392416,
      "learning_rate": 0.27659234414359074,
      "loss": 0.6749,
      "num_input_tokens_seen": 9922144,
      "step": 7210
    },
    {
      "epoch": 0.2309391204148262,
      "grad_norm": 0.008951089344918728,
      "learning_rate": 0.27656073643206097,
      "loss": 0.5818,
      "num_input_tokens_seen": 9928976,
      "step": 7215
    },
    {
      "epoch": 0.23109916138531464,
      "grad_norm": 0.011486553587019444,
      "learning_rate": 0.27652910920327517,
      "loss": 0.6282,
      "num_input_tokens_seen": 9935520,
      "step": 7220
    },
    {
      "epoch": 0.23125920235580308,
      "grad_norm": 0.0182874146848917,
      "learning_rate": 0.2764974624621107,
      "loss": 0.9695,
      "num_input_tokens_seen": 9942320,
      "step": 7225
    },
    {
      "epoch": 0.23141924332629152,
      "grad_norm": 0.007773166988044977,
      "learning_rate": 0.2764657962134479,
      "loss": 0.6918,
      "num_input_tokens_seen": 9949360,
      "step": 7230
    },
    {
      "epoch": 0.23157928429677999,
      "grad_norm": 0.008006321266293526,
      "learning_rate": 0.27643411046217,
      "loss": 0.7287,
      "num_input_tokens_seen": 9956144,
      "step": 7235
    },
    {
      "epoch": 0.23173932526726843,
      "grad_norm": 0.008509339764714241,
      "learning_rate": 0.27640240521316334,
      "loss": 0.9197,
      "num_input_tokens_seen": 9963120,
      "step": 7240
    },
    {
      "epoch": 0.23189936623775687,
      "grad_norm": 0.009460530243813992,
      "learning_rate": 0.2763706804713174,
      "loss": 0.642,
      "num_input_tokens_seen": 9969808,
      "step": 7245
    },
    {
      "epoch": 0.2320594072082453,
      "grad_norm": 0.007745206356048584,
      "learning_rate": 0.2763389362415245,
      "loss": 0.5924,
      "num_input_tokens_seen": 9976736,
      "step": 7250
    },
    {
      "epoch": 0.23221944817873375,
      "grad_norm": 0.01019737496972084,
      "learning_rate": 0.27630717252867987,
      "loss": 0.8007,
      "num_input_tokens_seen": 9983616,
      "step": 7255
    },
    {
      "epoch": 0.2323794891492222,
      "grad_norm": 0.007022456265985966,
      "learning_rate": 0.276275389337682,
      "loss": 0.8248,
      "num_input_tokens_seen": 9990448,
      "step": 7260
    },
    {
      "epoch": 0.23253953011971065,
      "grad_norm": 0.013794797472655773,
      "learning_rate": 0.2762435866734322,
      "loss": 0.9212,
      "num_input_tokens_seen": 9997664,
      "step": 7265
    },
    {
      "epoch": 0.2326995710901991,
      "grad_norm": 0.009427715092897415,
      "learning_rate": 0.27621176454083485,
      "loss": 0.703,
      "num_input_tokens_seen": 10004896,
      "step": 7270
    },
    {
      "epoch": 0.23285961206068753,
      "grad_norm": 0.00739242322742939,
      "learning_rate": 0.2761799229447973,
      "loss": 0.8556,
      "num_input_tokens_seen": 10011776,
      "step": 7275
    },
    {
      "epoch": 0.23301965303117597,
      "grad_norm": 0.0073457201942801476,
      "learning_rate": 0.27614806189023006,
      "loss": 0.6876,
      "num_input_tokens_seen": 10019136,
      "step": 7280
    },
    {
      "epoch": 0.23317969400166444,
      "grad_norm": 0.010150252841413021,
      "learning_rate": 0.27611618138204636,
      "loss": 0.7756,
      "num_input_tokens_seen": 10026016,
      "step": 7285
    },
    {
      "epoch": 0.23333973497215288,
      "grad_norm": 0.005310993641614914,
      "learning_rate": 0.2760842814251626,
      "loss": 0.8352,
      "num_input_tokens_seen": 10033056,
      "step": 7290
    },
    {
      "epoch": 0.23349977594264132,
      "grad_norm": 0.0139920087531209,
      "learning_rate": 0.2760523620244982,
      "loss": 0.9269,
      "num_input_tokens_seen": 10039920,
      "step": 7295
    },
    {
      "epoch": 0.23365981691312976,
      "grad_norm": 0.018063241615891457,
      "learning_rate": 0.27602042318497544,
      "loss": 0.8429,
      "num_input_tokens_seen": 10046384,
      "step": 7300
    },
    {
      "epoch": 0.2338198578836182,
      "grad_norm": 0.010112784802913666,
      "learning_rate": 0.2759884649115198,
      "loss": 0.5928,
      "num_input_tokens_seen": 10053296,
      "step": 7305
    },
    {
      "epoch": 0.23397989885410664,
      "grad_norm": 0.007347604259848595,
      "learning_rate": 0.2759564872090596,
      "loss": 0.8691,
      "num_input_tokens_seen": 10060096,
      "step": 7310
    },
    {
      "epoch": 0.2341399398245951,
      "grad_norm": 0.006619023624807596,
      "learning_rate": 0.2759244900825262,
      "loss": 0.7399,
      "num_input_tokens_seen": 10067424,
      "step": 7315
    },
    {
      "epoch": 0.23429998079508355,
      "grad_norm": 0.010516642592847347,
      "learning_rate": 0.2758924735368539,
      "loss": 0.68,
      "num_input_tokens_seen": 10074160,
      "step": 7320
    },
    {
      "epoch": 0.23446002176557199,
      "grad_norm": 0.013970642350614071,
      "learning_rate": 0.27586043757698014,
      "loss": 0.8379,
      "num_input_tokens_seen": 10080736,
      "step": 7325
    },
    {
      "epoch": 0.23462006273606043,
      "grad_norm": 0.020693693310022354,
      "learning_rate": 0.27582838220784534,
      "loss": 0.988,
      "num_input_tokens_seen": 10087696,
      "step": 7330
    },
    {
      "epoch": 0.23478010370654886,
      "grad_norm": 0.007229202426970005,
      "learning_rate": 0.27579630743439265,
      "loss": 0.6359,
      "num_input_tokens_seen": 10094400,
      "step": 7335
    },
    {
      "epoch": 0.23494014467703733,
      "grad_norm": 0.005924350582063198,
      "learning_rate": 0.2757642132615686,
      "loss": 0.7417,
      "num_input_tokens_seen": 10100864,
      "step": 7340
    },
    {
      "epoch": 0.23510018564752577,
      "grad_norm": 0.013905544765293598,
      "learning_rate": 0.2757320996943223,
      "loss": 0.9481,
      "num_input_tokens_seen": 10107520,
      "step": 7345
    },
    {
      "epoch": 0.2352602266180142,
      "grad_norm": 0.006060462910681963,
      "learning_rate": 0.2756999667376062,
      "loss": 0.8249,
      "num_input_tokens_seen": 10114288,
      "step": 7350
    },
    {
      "epoch": 0.23542026758850265,
      "grad_norm": 0.010960781946778297,
      "learning_rate": 0.2756678143963756,
      "loss": 0.8469,
      "num_input_tokens_seen": 10120752,
      "step": 7355
    },
    {
      "epoch": 0.2355803085589911,
      "grad_norm": 0.007246183231472969,
      "learning_rate": 0.2756356426755888,
      "loss": 0.6908,
      "num_input_tokens_seen": 10127616,
      "step": 7360
    },
    {
      "epoch": 0.23574034952947956,
      "grad_norm": 0.014863626100122929,
      "learning_rate": 0.27560345158020705,
      "loss": 0.9277,
      "num_input_tokens_seen": 10134480,
      "step": 7365
    },
    {
      "epoch": 0.235900390499968,
      "grad_norm": 0.007890014909207821,
      "learning_rate": 0.27557124111519465,
      "loss": 0.7359,
      "num_input_tokens_seen": 10141248,
      "step": 7370
    },
    {
      "epoch": 0.23606043147045644,
      "grad_norm": 0.007797855883836746,
      "learning_rate": 0.27553901128551883,
      "loss": 0.8488,
      "num_input_tokens_seen": 10148016,
      "step": 7375
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 0.007585299666970968,
      "learning_rate": 0.2755067620961498,
      "loss": 0.863,
      "num_input_tokens_seen": 10154688,
      "step": 7380
    },
    {
      "epoch": 0.23638051341143332,
      "grad_norm": 0.005341827403753996,
      "learning_rate": 0.27547449355206094,
      "loss": 0.7723,
      "num_input_tokens_seen": 10161232,
      "step": 7385
    },
    {
      "epoch": 0.23654055438192176,
      "grad_norm": 0.011808124370872974,
      "learning_rate": 0.2754422056582283,
      "loss": 0.816,
      "num_input_tokens_seen": 10167648,
      "step": 7390
    },
    {
      "epoch": 0.23670059535241023,
      "grad_norm": 0.005632084794342518,
      "learning_rate": 0.27540989841963115,
      "loss": 0.7993,
      "num_input_tokens_seen": 10174912,
      "step": 7395
    },
    {
      "epoch": 0.23686063632289867,
      "grad_norm": 0.007231582887470722,
      "learning_rate": 0.27537757184125167,
      "loss": 0.7695,
      "num_input_tokens_seen": 10182048,
      "step": 7400
    },
    {
      "epoch": 0.23686063632289867,
      "eval_loss": 0.772653341293335,
      "eval_runtime": 332.3462,
      "eval_samples_per_second": 41.782,
      "eval_steps_per_second": 20.891,
      "num_input_tokens_seen": 10182048,
      "step": 7400
    },
    {
      "epoch": 0.2370206772933871,
      "grad_norm": 0.004683514591306448,
      "learning_rate": 0.275345225928075,
      "loss": 0.7532,
      "num_input_tokens_seen": 10188544,
      "step": 7405
    },
    {
      "epoch": 0.23718071826387555,
      "grad_norm": 0.006657206919044256,
      "learning_rate": 0.2753128606850893,
      "loss": 0.9044,
      "num_input_tokens_seen": 10195824,
      "step": 7410
    },
    {
      "epoch": 0.23734075923436398,
      "grad_norm": 0.006297136656939983,
      "learning_rate": 0.2752804761172858,
      "loss": 0.7126,
      "num_input_tokens_seen": 10202672,
      "step": 7415
    },
    {
      "epoch": 0.23750080020485245,
      "grad_norm": 0.007595492992550135,
      "learning_rate": 0.27524807222965836,
      "loss": 0.7717,
      "num_input_tokens_seen": 10209728,
      "step": 7420
    },
    {
      "epoch": 0.2376608411753409,
      "grad_norm": 0.0073567042127251625,
      "learning_rate": 0.27521564902720436,
      "loss": 0.9444,
      "num_input_tokens_seen": 10216480,
      "step": 7425
    },
    {
      "epoch": 0.23782088214582933,
      "grad_norm": 0.0070734950713813305,
      "learning_rate": 0.2751832065149236,
      "loss": 0.6185,
      "num_input_tokens_seen": 10223104,
      "step": 7430
    },
    {
      "epoch": 0.23798092311631777,
      "grad_norm": 0.006891967263072729,
      "learning_rate": 0.2751507446978193,
      "loss": 0.7187,
      "num_input_tokens_seen": 10230336,
      "step": 7435
    },
    {
      "epoch": 0.2381409640868062,
      "grad_norm": 0.008101019077003002,
      "learning_rate": 0.2751182635808974,
      "loss": 0.6879,
      "num_input_tokens_seen": 10237152,
      "step": 7440
    },
    {
      "epoch": 0.23830100505729468,
      "grad_norm": 0.006348316092044115,
      "learning_rate": 0.27508576316916694,
      "loss": 0.8292,
      "num_input_tokens_seen": 10243824,
      "step": 7445
    },
    {
      "epoch": 0.23846104602778312,
      "grad_norm": 0.012420416809618473,
      "learning_rate": 0.2750532434676399,
      "loss": 0.8014,
      "num_input_tokens_seen": 10250400,
      "step": 7450
    },
    {
      "epoch": 0.23862108699827156,
      "grad_norm": 0.008541727438569069,
      "learning_rate": 0.27502070448133115,
      "loss": 0.7132,
      "num_input_tokens_seen": 10257088,
      "step": 7455
    },
    {
      "epoch": 0.23878112796876,
      "grad_norm": 0.00590846361592412,
      "learning_rate": 0.2749881462152587,
      "loss": 0.7496,
      "num_input_tokens_seen": 10264336,
      "step": 7460
    },
    {
      "epoch": 0.23894116893924844,
      "grad_norm": 0.005430371500551701,
      "learning_rate": 0.2749555686744434,
      "loss": 0.7269,
      "num_input_tokens_seen": 10271200,
      "step": 7465
    },
    {
      "epoch": 0.2391012099097369,
      "grad_norm": 0.007444627583026886,
      "learning_rate": 0.2749229718639091,
      "loss": 0.7874,
      "num_input_tokens_seen": 10278336,
      "step": 7470
    },
    {
      "epoch": 0.23926125088022535,
      "grad_norm": 0.008960473351180553,
      "learning_rate": 0.27489035578868265,
      "loss": 0.7489,
      "num_input_tokens_seen": 10285136,
      "step": 7475
    },
    {
      "epoch": 0.23942129185071379,
      "grad_norm": 0.004636173136532307,
      "learning_rate": 0.2748577204537939,
      "loss": 0.8766,
      "num_input_tokens_seen": 10292032,
      "step": 7480
    },
    {
      "epoch": 0.23958133282120223,
      "grad_norm": 0.005053180269896984,
      "learning_rate": 0.2748250658642756,
      "loss": 0.8136,
      "num_input_tokens_seen": 10299008,
      "step": 7485
    },
    {
      "epoch": 0.23974137379169067,
      "grad_norm": 0.007385753560811281,
      "learning_rate": 0.2747923920251634,
      "loss": 0.5495,
      "num_input_tokens_seen": 10305872,
      "step": 7490
    },
    {
      "epoch": 0.2399014147621791,
      "grad_norm": 0.007255032658576965,
      "learning_rate": 0.27475969894149627,
      "loss": 0.5923,
      "num_input_tokens_seen": 10312560,
      "step": 7495
    },
    {
      "epoch": 0.24006145573266757,
      "grad_norm": 0.009573272429406643,
      "learning_rate": 0.2747269866183156,
      "loss": 0.84,
      "num_input_tokens_seen": 10319712,
      "step": 7500
    },
    {
      "epoch": 0.240221496703156,
      "grad_norm": 0.006856395862996578,
      "learning_rate": 0.27469425506066625,
      "loss": 0.8216,
      "num_input_tokens_seen": 10326640,
      "step": 7505
    },
    {
      "epoch": 0.24038153767364445,
      "grad_norm": 0.005337701179087162,
      "learning_rate": 0.27466150427359576,
      "loss": 0.7121,
      "num_input_tokens_seen": 10333552,
      "step": 7510
    },
    {
      "epoch": 0.2405415786441329,
      "grad_norm": 0.008706532418727875,
      "learning_rate": 0.2746287342621547,
      "loss": 0.8527,
      "num_input_tokens_seen": 10340368,
      "step": 7515
    },
    {
      "epoch": 0.24070161961462133,
      "grad_norm": 0.005509525537490845,
      "learning_rate": 0.2745959450313966,
      "loss": 0.8084,
      "num_input_tokens_seen": 10347376,
      "step": 7520
    },
    {
      "epoch": 0.2408616605851098,
      "grad_norm": 0.006590875796973705,
      "learning_rate": 0.27456313658637804,
      "loss": 0.5905,
      "num_input_tokens_seen": 10353936,
      "step": 7525
    },
    {
      "epoch": 0.24102170155559824,
      "grad_norm": 0.01200808398425579,
      "learning_rate": 0.27453030893215846,
      "loss": 0.6627,
      "num_input_tokens_seen": 10360880,
      "step": 7530
    },
    {
      "epoch": 0.24118174252608668,
      "grad_norm": 0.0057974206283688545,
      "learning_rate": 0.2744974620738003,
      "loss": 0.8546,
      "num_input_tokens_seen": 10368064,
      "step": 7535
    },
    {
      "epoch": 0.24134178349657512,
      "grad_norm": 0.010570148937404156,
      "learning_rate": 0.27446459601636897,
      "loss": 0.8359,
      "num_input_tokens_seen": 10375328,
      "step": 7540
    },
    {
      "epoch": 0.24150182446706356,
      "grad_norm": 0.004944505635648966,
      "learning_rate": 0.2744317107649328,
      "loss": 0.5654,
      "num_input_tokens_seen": 10382352,
      "step": 7545
    },
    {
      "epoch": 0.24166186543755203,
      "grad_norm": 0.006361719220876694,
      "learning_rate": 0.2743988063245631,
      "loss": 0.7854,
      "num_input_tokens_seen": 10389392,
      "step": 7550
    },
    {
      "epoch": 0.24182190640804047,
      "grad_norm": 0.00645810179412365,
      "learning_rate": 0.2743658827003342,
      "loss": 0.6988,
      "num_input_tokens_seen": 10396448,
      "step": 7555
    },
    {
      "epoch": 0.2419819473785289,
      "grad_norm": 0.0079046580940485,
      "learning_rate": 0.27433293989732327,
      "loss": 0.924,
      "num_input_tokens_seen": 10403488,
      "step": 7560
    },
    {
      "epoch": 0.24214198834901735,
      "grad_norm": 0.005674941930919886,
      "learning_rate": 0.27429997792061056,
      "loss": 0.6853,
      "num_input_tokens_seen": 10410144,
      "step": 7565
    },
    {
      "epoch": 0.24230202931950579,
      "grad_norm": 0.009239474311470985,
      "learning_rate": 0.27426699677527927,
      "loss": 0.7821,
      "num_input_tokens_seen": 10416768,
      "step": 7570
    },
    {
      "epoch": 0.24246207028999422,
      "grad_norm": 0.004975178278982639,
      "learning_rate": 0.2742339964664154,
      "loss": 0.6583,
      "num_input_tokens_seen": 10423536,
      "step": 7575
    },
    {
      "epoch": 0.2426221112604827,
      "grad_norm": 0.007226089481264353,
      "learning_rate": 0.274200976999108,
      "loss": 0.7514,
      "num_input_tokens_seen": 10430784,
      "step": 7580
    },
    {
      "epoch": 0.24278215223097113,
      "grad_norm": 0.006373984273523092,
      "learning_rate": 0.27416793837844916,
      "loss": 0.6135,
      "num_input_tokens_seen": 10437904,
      "step": 7585
    },
    {
      "epoch": 0.24294219320145957,
      "grad_norm": 0.004584962502121925,
      "learning_rate": 0.27413488060953384,
      "loss": 0.6793,
      "num_input_tokens_seen": 10444560,
      "step": 7590
    },
    {
      "epoch": 0.243102234171948,
      "grad_norm": 0.005412232130765915,
      "learning_rate": 0.27410180369745996,
      "loss": 0.8033,
      "num_input_tokens_seen": 10451840,
      "step": 7595
    },
    {
      "epoch": 0.24326227514243645,
      "grad_norm": 0.008798685856163502,
      "learning_rate": 0.27406870764732844,
      "loss": 0.9169,
      "num_input_tokens_seen": 10458544,
      "step": 7600
    },
    {
      "epoch": 0.24326227514243645,
      "eval_loss": 0.7676618099212646,
      "eval_runtime": 332.22,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 10458544,
      "step": 7600
    },
    {
      "epoch": 0.24342231611292492,
      "grad_norm": 0.005437752697616816,
      "learning_rate": 0.27403559246424297,
      "loss": 0.792,
      "num_input_tokens_seen": 10465648,
      "step": 7605
    },
    {
      "epoch": 0.24358235708341336,
      "grad_norm": 0.004953858442604542,
      "learning_rate": 0.2740024581533105,
      "loss": 0.7834,
      "num_input_tokens_seen": 10472624,
      "step": 7610
    },
    {
      "epoch": 0.2437423980539018,
      "grad_norm": 0.010466689243912697,
      "learning_rate": 0.2739693047196406,
      "loss": 0.7402,
      "num_input_tokens_seen": 10479808,
      "step": 7615
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.00739706726744771,
      "learning_rate": 0.27393613216834606,
      "loss": 0.6894,
      "num_input_tokens_seen": 10486560,
      "step": 7620
    },
    {
      "epoch": 0.24406247999487868,
      "grad_norm": 0.007644067984074354,
      "learning_rate": 0.2739029405045424,
      "loss": 0.8185,
      "num_input_tokens_seen": 10493920,
      "step": 7625
    },
    {
      "epoch": 0.24422252096536715,
      "grad_norm": 0.007339237257838249,
      "learning_rate": 0.2738697297333483,
      "loss": 0.7088,
      "num_input_tokens_seen": 10500640,
      "step": 7630
    },
    {
      "epoch": 0.24438256193585559,
      "grad_norm": 0.007830987684428692,
      "learning_rate": 0.2738364998598852,
      "loss": 0.7041,
      "num_input_tokens_seen": 10507312,
      "step": 7635
    },
    {
      "epoch": 0.24454260290634403,
      "grad_norm": 0.004492427688091993,
      "learning_rate": 0.27380325088927765,
      "loss": 0.825,
      "num_input_tokens_seen": 10514304,
      "step": 7640
    },
    {
      "epoch": 0.24470264387683247,
      "grad_norm": 0.007802753709256649,
      "learning_rate": 0.27376998282665294,
      "loss": 0.7627,
      "num_input_tokens_seen": 10521600,
      "step": 7645
    },
    {
      "epoch": 0.2448626848473209,
      "grad_norm": 0.010345598682761192,
      "learning_rate": 0.27373669567714154,
      "loss": 0.878,
      "num_input_tokens_seen": 10528544,
      "step": 7650
    },
    {
      "epoch": 0.24502272581780937,
      "grad_norm": 0.006318162195384502,
      "learning_rate": 0.27370338944587663,
      "loss": 0.7643,
      "num_input_tokens_seen": 10535408,
      "step": 7655
    },
    {
      "epoch": 0.2451827667882978,
      "grad_norm": 0.004368324764072895,
      "learning_rate": 0.27367006413799455,
      "loss": 0.7442,
      "num_input_tokens_seen": 10542384,
      "step": 7660
    },
    {
      "epoch": 0.24534280775878625,
      "grad_norm": 0.005129069555550814,
      "learning_rate": 0.2736367197586345,
      "loss": 0.6739,
      "num_input_tokens_seen": 10549328,
      "step": 7665
    },
    {
      "epoch": 0.2455028487292747,
      "grad_norm": 0.00723248440772295,
      "learning_rate": 0.2736033563129385,
      "loss": 0.8343,
      "num_input_tokens_seen": 10556048,
      "step": 7670
    },
    {
      "epoch": 0.24566288969976313,
      "grad_norm": 0.007049359381198883,
      "learning_rate": 0.27356997380605164,
      "loss": 0.7087,
      "num_input_tokens_seen": 10563264,
      "step": 7675
    },
    {
      "epoch": 0.24582293067025157,
      "grad_norm": 0.005818269215524197,
      "learning_rate": 0.27353657224312194,
      "loss": 0.742,
      "num_input_tokens_seen": 10569952,
      "step": 7680
    },
    {
      "epoch": 0.24598297164074004,
      "grad_norm": 0.0064355311915278435,
      "learning_rate": 0.2735031516293004,
      "loss": 0.8386,
      "num_input_tokens_seen": 10576416,
      "step": 7685
    },
    {
      "epoch": 0.24614301261122848,
      "grad_norm": 0.013234546408057213,
      "learning_rate": 0.2734697119697408,
      "loss": 0.7699,
      "num_input_tokens_seen": 10583216,
      "step": 7690
    },
    {
      "epoch": 0.24630305358171692,
      "grad_norm": 0.011294964700937271,
      "learning_rate": 0.27343625326959997,
      "loss": 0.6284,
      "num_input_tokens_seen": 10590064,
      "step": 7695
    },
    {
      "epoch": 0.24646309455220536,
      "grad_norm": 0.0066700708121061325,
      "learning_rate": 0.27340277553403775,
      "loss": 0.8238,
      "num_input_tokens_seen": 10597152,
      "step": 7700
    },
    {
      "epoch": 0.2466231355226938,
      "grad_norm": 0.01957801729440689,
      "learning_rate": 0.2733692787682167,
      "loss": 0.7856,
      "num_input_tokens_seen": 10604176,
      "step": 7705
    },
    {
      "epoch": 0.24678317649318227,
      "grad_norm": 0.010543019510805607,
      "learning_rate": 0.27333576297730255,
      "loss": 0.8758,
      "num_input_tokens_seen": 10611600,
      "step": 7710
    },
    {
      "epoch": 0.2469432174636707,
      "grad_norm": 0.0063520558178424835,
      "learning_rate": 0.2733022281664638,
      "loss": 0.7914,
      "num_input_tokens_seen": 10618384,
      "step": 7715
    },
    {
      "epoch": 0.24710325843415915,
      "grad_norm": 0.00998634472489357,
      "learning_rate": 0.273268674340872,
      "loss": 0.7356,
      "num_input_tokens_seen": 10625200,
      "step": 7720
    },
    {
      "epoch": 0.24726329940464759,
      "grad_norm": 0.008114517666399479,
      "learning_rate": 0.27323510150570146,
      "loss": 0.9398,
      "num_input_tokens_seen": 10631920,
      "step": 7725
    },
    {
      "epoch": 0.24742334037513602,
      "grad_norm": 0.01084892824292183,
      "learning_rate": 0.27320150966612966,
      "loss": 0.7571,
      "num_input_tokens_seen": 10638608,
      "step": 7730
    },
    {
      "epoch": 0.2475833813456245,
      "grad_norm": 0.009015893563628197,
      "learning_rate": 0.2731678988273368,
      "loss": 0.7742,
      "num_input_tokens_seen": 10645568,
      "step": 7735
    },
    {
      "epoch": 0.24774342231611293,
      "grad_norm": 0.0054274084977805614,
      "learning_rate": 0.27313426899450605,
      "loss": 0.6002,
      "num_input_tokens_seen": 10653104,
      "step": 7740
    },
    {
      "epoch": 0.24790346328660137,
      "grad_norm": 0.005199416074901819,
      "learning_rate": 0.27310062017282366,
      "loss": 0.5985,
      "num_input_tokens_seen": 10660000,
      "step": 7745
    },
    {
      "epoch": 0.2480635042570898,
      "grad_norm": 0.0060507506132125854,
      "learning_rate": 0.2730669523674787,
      "loss": 0.8209,
      "num_input_tokens_seen": 10666880,
      "step": 7750
    },
    {
      "epoch": 0.24822354522757825,
      "grad_norm": 0.009252738207578659,
      "learning_rate": 0.2730332655836631,
      "loss": 0.886,
      "num_input_tokens_seen": 10674416,
      "step": 7755
    },
    {
      "epoch": 0.2483835861980667,
      "grad_norm": 0.008753868751227856,
      "learning_rate": 0.2729995598265718,
      "loss": 0.7639,
      "num_input_tokens_seen": 10681152,
      "step": 7760
    },
    {
      "epoch": 0.24854362716855516,
      "grad_norm": 0.0111595643684268,
      "learning_rate": 0.2729658351014027,
      "loss": 0.6882,
      "num_input_tokens_seen": 10687616,
      "step": 7765
    },
    {
      "epoch": 0.2487036681390436,
      "grad_norm": 0.004730585031211376,
      "learning_rate": 0.27293209141335656,
      "loss": 0.6187,
      "num_input_tokens_seen": 10694672,
      "step": 7770
    },
    {
      "epoch": 0.24886370910953204,
      "grad_norm": 0.005610242951661348,
      "learning_rate": 0.27289832876763703,
      "loss": 0.8205,
      "num_input_tokens_seen": 10702240,
      "step": 7775
    },
    {
      "epoch": 0.24902375008002048,
      "grad_norm": 0.0064122360199689865,
      "learning_rate": 0.27286454716945074,
      "loss": 0.6498,
      "num_input_tokens_seen": 10708752,
      "step": 7780
    },
    {
      "epoch": 0.24918379105050892,
      "grad_norm": 0.006481515243649483,
      "learning_rate": 0.27283074662400725,
      "loss": 0.7962,
      "num_input_tokens_seen": 10715648,
      "step": 7785
    },
    {
      "epoch": 0.24934383202099739,
      "grad_norm": 0.005934955086559057,
      "learning_rate": 0.2727969271365191,
      "loss": 0.6999,
      "num_input_tokens_seen": 10722624,
      "step": 7790
    },
    {
      "epoch": 0.24950387299148583,
      "grad_norm": 0.008214727975428104,
      "learning_rate": 0.2727630887122016,
      "loss": 1.0006,
      "num_input_tokens_seen": 10729264,
      "step": 7795
    },
    {
      "epoch": 0.24966391396197427,
      "grad_norm": 0.00487761665135622,
      "learning_rate": 0.27272923135627314,
      "loss": 0.7497,
      "num_input_tokens_seen": 10736144,
      "step": 7800
    },
    {
      "epoch": 0.24966391396197427,
      "eval_loss": 0.7552260756492615,
      "eval_runtime": 332.1477,
      "eval_samples_per_second": 41.807,
      "eval_steps_per_second": 20.903,
      "num_input_tokens_seen": 10736144,
      "step": 7800
    },
    {
      "epoch": 0.2498239549324627,
      "grad_norm": 0.006295541767030954,
      "learning_rate": 0.2726953550739548,
      "loss": 0.7086,
      "num_input_tokens_seen": 10742688,
      "step": 7805
    },
    {
      "epoch": 0.24998399590295114,
      "grad_norm": 0.004062939900904894,
      "learning_rate": 0.27266145987047086,
      "loss": 0.7774,
      "num_input_tokens_seen": 10749680,
      "step": 7810
    },
    {
      "epoch": 0.2501440368734396,
      "grad_norm": 0.0033149418886750937,
      "learning_rate": 0.27262754575104836,
      "loss": 0.5979,
      "num_input_tokens_seen": 10756800,
      "step": 7815
    },
    {
      "epoch": 0.25030407784392805,
      "grad_norm": 0.004983740393072367,
      "learning_rate": 0.27259361272091726,
      "loss": 0.567,
      "num_input_tokens_seen": 10763552,
      "step": 7820
    },
    {
      "epoch": 0.2504641188144165,
      "grad_norm": 0.004722462501376867,
      "learning_rate": 0.27255966078531046,
      "loss": 0.6816,
      "num_input_tokens_seen": 10769984,
      "step": 7825
    },
    {
      "epoch": 0.25062415978490493,
      "grad_norm": 0.003324538003653288,
      "learning_rate": 0.2725256899494638,
      "loss": 0.8192,
      "num_input_tokens_seen": 10776880,
      "step": 7830
    },
    {
      "epoch": 0.25078420075539337,
      "grad_norm": 0.00789639726281166,
      "learning_rate": 0.272491700218616,
      "loss": 0.762,
      "num_input_tokens_seen": 10783616,
      "step": 7835
    },
    {
      "epoch": 0.2509442417258818,
      "grad_norm": 0.0068956343457102776,
      "learning_rate": 0.27245769159800876,
      "loss": 0.6883,
      "num_input_tokens_seen": 10790624,
      "step": 7840
    },
    {
      "epoch": 0.25110428269637025,
      "grad_norm": 0.006863664370030165,
      "learning_rate": 0.2724236640928865,
      "loss": 0.9303,
      "num_input_tokens_seen": 10797552,
      "step": 7845
    },
    {
      "epoch": 0.2512643236668587,
      "grad_norm": 0.009022410959005356,
      "learning_rate": 0.27238961770849673,
      "loss": 0.8083,
      "num_input_tokens_seen": 10804576,
      "step": 7850
    },
    {
      "epoch": 0.2514243646373472,
      "grad_norm": 0.006785233970731497,
      "learning_rate": 0.27235555245008997,
      "loss": 0.7486,
      "num_input_tokens_seen": 10811840,
      "step": 7855
    },
    {
      "epoch": 0.2515844056078356,
      "grad_norm": 0.006466659717261791,
      "learning_rate": 0.2723214683229193,
      "loss": 0.7472,
      "num_input_tokens_seen": 10818800,
      "step": 7860
    },
    {
      "epoch": 0.25174444657832407,
      "grad_norm": 0.007322905585169792,
      "learning_rate": 0.27228736533224107,
      "loss": 0.8281,
      "num_input_tokens_seen": 10825584,
      "step": 7865
    },
    {
      "epoch": 0.2519044875488125,
      "grad_norm": 0.005264654755592346,
      "learning_rate": 0.27225324348331437,
      "loss": 0.611,
      "num_input_tokens_seen": 10832560,
      "step": 7870
    },
    {
      "epoch": 0.25206452851930095,
      "grad_norm": 0.006572661455720663,
      "learning_rate": 0.27221910278140116,
      "loss": 0.5598,
      "num_input_tokens_seen": 10839264,
      "step": 7875
    },
    {
      "epoch": 0.2522245694897894,
      "grad_norm": 0.009618516080081463,
      "learning_rate": 0.2721849432317664,
      "loss": 0.9592,
      "num_input_tokens_seen": 10846128,
      "step": 7880
    },
    {
      "epoch": 0.2523846104602778,
      "grad_norm": 0.019066818058490753,
      "learning_rate": 0.2721507648396779,
      "loss": 0.9153,
      "num_input_tokens_seen": 10853376,
      "step": 7885
    },
    {
      "epoch": 0.25254465143076626,
      "grad_norm": 0.005978269036859274,
      "learning_rate": 0.27211656761040653,
      "loss": 0.8477,
      "num_input_tokens_seen": 10860384,
      "step": 7890
    },
    {
      "epoch": 0.2527046924012547,
      "grad_norm": 0.005046115722507238,
      "learning_rate": 0.2720823515492257,
      "loss": 0.7802,
      "num_input_tokens_seen": 10867264,
      "step": 7895
    },
    {
      "epoch": 0.25286473337174314,
      "grad_norm": 0.011514746583998203,
      "learning_rate": 0.27204811666141215,
      "loss": 0.7389,
      "num_input_tokens_seen": 10874848,
      "step": 7900
    },
    {
      "epoch": 0.2530247743422316,
      "grad_norm": 0.00549840135499835,
      "learning_rate": 0.2720138629522452,
      "loss": 0.5191,
      "num_input_tokens_seen": 10881344,
      "step": 7905
    },
    {
      "epoch": 0.2531848153127201,
      "grad_norm": 0.005632615182548761,
      "learning_rate": 0.2719795904270073,
      "loss": 0.8143,
      "num_input_tokens_seen": 10888080,
      "step": 7910
    },
    {
      "epoch": 0.2533448562832085,
      "grad_norm": 0.013170948252081871,
      "learning_rate": 0.2719452990909837,
      "loss": 0.5765,
      "num_input_tokens_seen": 10895072,
      "step": 7915
    },
    {
      "epoch": 0.25350489725369696,
      "grad_norm": 0.012306636199355125,
      "learning_rate": 0.2719109889494625,
      "loss": 0.7707,
      "num_input_tokens_seen": 10902128,
      "step": 7920
    },
    {
      "epoch": 0.2536649382241854,
      "grad_norm": 0.01849871687591076,
      "learning_rate": 0.27187666000773475,
      "loss": 0.6989,
      "num_input_tokens_seen": 10909104,
      "step": 7925
    },
    {
      "epoch": 0.25382497919467384,
      "grad_norm": 0.011490707285702229,
      "learning_rate": 0.2718423122710944,
      "loss": 0.8061,
      "num_input_tokens_seen": 10915776,
      "step": 7930
    },
    {
      "epoch": 0.2539850201651623,
      "grad_norm": 0.010130563750863075,
      "learning_rate": 0.2718079457448384,
      "loss": 0.7618,
      "num_input_tokens_seen": 10922336,
      "step": 7935
    },
    {
      "epoch": 0.2541450611356507,
      "grad_norm": 0.006429702043533325,
      "learning_rate": 0.27177356043426637,
      "loss": 0.64,
      "num_input_tokens_seen": 10928752,
      "step": 7940
    },
    {
      "epoch": 0.25430510210613916,
      "grad_norm": 0.014860779978334904,
      "learning_rate": 0.27173915634468104,
      "loss": 0.8282,
      "num_input_tokens_seen": 10935120,
      "step": 7945
    },
    {
      "epoch": 0.2544651430766276,
      "grad_norm": 0.008686023764312267,
      "learning_rate": 0.27170473348138796,
      "loss": 0.8114,
      "num_input_tokens_seen": 10942448,
      "step": 7950
    },
    {
      "epoch": 0.25462518404711604,
      "grad_norm": 0.013568894006311893,
      "learning_rate": 0.27167029184969554,
      "loss": 0.8803,
      "num_input_tokens_seen": 10949392,
      "step": 7955
    },
    {
      "epoch": 0.25478522501760453,
      "grad_norm": 0.0069716195575892925,
      "learning_rate": 0.27163583145491504,
      "loss": 0.7341,
      "num_input_tokens_seen": 10956144,
      "step": 7960
    },
    {
      "epoch": 0.254945265988093,
      "grad_norm": 0.013991416431963444,
      "learning_rate": 0.2716013523023608,
      "loss": 0.8646,
      "num_input_tokens_seen": 10962736,
      "step": 7965
    },
    {
      "epoch": 0.2551053069585814,
      "grad_norm": 0.008697553537786007,
      "learning_rate": 0.27156685439734995,
      "loss": 0.6599,
      "num_input_tokens_seen": 10969488,
      "step": 7970
    },
    {
      "epoch": 0.25526534792906985,
      "grad_norm": 0.009524884633719921,
      "learning_rate": 0.2715323377452024,
      "loss": 0.7316,
      "num_input_tokens_seen": 10976416,
      "step": 7975
    },
    {
      "epoch": 0.2554253888995583,
      "grad_norm": 0.007359870709478855,
      "learning_rate": 0.2714978023512411,
      "loss": 0.9063,
      "num_input_tokens_seen": 10983328,
      "step": 7980
    },
    {
      "epoch": 0.25558542987004673,
      "grad_norm": 0.020115960389375687,
      "learning_rate": 0.2714632482207918,
      "loss": 0.6819,
      "num_input_tokens_seen": 10990112,
      "step": 7985
    },
    {
      "epoch": 0.25574547084053517,
      "grad_norm": 0.013960708864033222,
      "learning_rate": 0.2714286753591833,
      "loss": 0.8484,
      "num_input_tokens_seen": 10996928,
      "step": 7990
    },
    {
      "epoch": 0.2559055118110236,
      "grad_norm": 0.008697359822690487,
      "learning_rate": 0.27139408377174706,
      "loss": 0.7156,
      "num_input_tokens_seen": 11003632,
      "step": 7995
    },
    {
      "epoch": 0.25606555278151205,
      "grad_norm": 0.010988409630954266,
      "learning_rate": 0.27135947346381756,
      "loss": 0.8553,
      "num_input_tokens_seen": 11010512,
      "step": 8000
    },
    {
      "epoch": 0.25606555278151205,
      "eval_loss": 0.7170922756195068,
      "eval_runtime": 332.0013,
      "eval_samples_per_second": 41.825,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 11010512,
      "step": 8000
    },
    {
      "epoch": 0.2562255937520005,
      "grad_norm": 0.012522922828793526,
      "learning_rate": 0.2713248444407322,
      "loss": 0.5741,
      "num_input_tokens_seen": 11017616,
      "step": 8005
    },
    {
      "epoch": 0.25638563472248893,
      "grad_norm": 0.006475528702139854,
      "learning_rate": 0.27129019670783106,
      "loss": 0.6241,
      "num_input_tokens_seen": 11024288,
      "step": 8010
    },
    {
      "epoch": 0.2565456756929774,
      "grad_norm": 0.006927291862666607,
      "learning_rate": 0.27125553027045746,
      "loss": 0.6727,
      "num_input_tokens_seen": 11031024,
      "step": 8015
    },
    {
      "epoch": 0.25670571666346587,
      "grad_norm": 0.014967563562095165,
      "learning_rate": 0.2712208451339572,
      "loss": 0.7778,
      "num_input_tokens_seen": 11037952,
      "step": 8020
    },
    {
      "epoch": 0.2568657576339543,
      "grad_norm": 0.011385800316929817,
      "learning_rate": 0.27118614130367935,
      "loss": 0.7087,
      "num_input_tokens_seen": 11044640,
      "step": 8025
    },
    {
      "epoch": 0.25702579860444275,
      "grad_norm": 0.012016605585813522,
      "learning_rate": 0.2711514187849756,
      "loss": 0.6513,
      "num_input_tokens_seen": 11051376,
      "step": 8030
    },
    {
      "epoch": 0.2571858395749312,
      "grad_norm": 0.009891977533698082,
      "learning_rate": 0.27111667758320057,
      "loss": 0.778,
      "num_input_tokens_seen": 11058496,
      "step": 8035
    },
    {
      "epoch": 0.2573458805454196,
      "grad_norm": 0.00793931819498539,
      "learning_rate": 0.27108191770371176,
      "loss": 0.6169,
      "num_input_tokens_seen": 11065264,
      "step": 8040
    },
    {
      "epoch": 0.25750592151590807,
      "grad_norm": 0.006909654010087252,
      "learning_rate": 0.2710471391518697,
      "loss": 0.7221,
      "num_input_tokens_seen": 11072288,
      "step": 8045
    },
    {
      "epoch": 0.2576659624863965,
      "grad_norm": 0.011014196090400219,
      "learning_rate": 0.2710123419330375,
      "loss": 1.0908,
      "num_input_tokens_seen": 11079264,
      "step": 8050
    },
    {
      "epoch": 0.25782600345688494,
      "grad_norm": 0.005583024583756924,
      "learning_rate": 0.2709775260525816,
      "loss": 0.7704,
      "num_input_tokens_seen": 11086080,
      "step": 8055
    },
    {
      "epoch": 0.2579860444273734,
      "grad_norm": 0.007234042044728994,
      "learning_rate": 0.27094269151587075,
      "loss": 0.7418,
      "num_input_tokens_seen": 11093264,
      "step": 8060
    },
    {
      "epoch": 0.2581460853978619,
      "grad_norm": 0.005714014172554016,
      "learning_rate": 0.27090783832827703,
      "loss": 0.8278,
      "num_input_tokens_seen": 11100080,
      "step": 8065
    },
    {
      "epoch": 0.2583061263683503,
      "grad_norm": 0.005615189205855131,
      "learning_rate": 0.2708729664951753,
      "loss": 0.7449,
      "num_input_tokens_seen": 11107248,
      "step": 8070
    },
    {
      "epoch": 0.25846616733883876,
      "grad_norm": 0.006605070549994707,
      "learning_rate": 0.27083807602194304,
      "loss": 0.6294,
      "num_input_tokens_seen": 11114112,
      "step": 8075
    },
    {
      "epoch": 0.2586262083093272,
      "grad_norm": 0.0118270143866539,
      "learning_rate": 0.270803166913961,
      "loss": 0.7249,
      "num_input_tokens_seen": 11120976,
      "step": 8080
    },
    {
      "epoch": 0.25878624927981564,
      "grad_norm": 0.009405777789652348,
      "learning_rate": 0.27076823917661247,
      "loss": 0.7329,
      "num_input_tokens_seen": 11127680,
      "step": 8085
    },
    {
      "epoch": 0.2589462902503041,
      "grad_norm": 0.008124049752950668,
      "learning_rate": 0.2707332928152838,
      "loss": 0.7013,
      "num_input_tokens_seen": 11134736,
      "step": 8090
    },
    {
      "epoch": 0.2591063312207925,
      "grad_norm": 0.008762523531913757,
      "learning_rate": 0.2706983278353641,
      "loss": 0.5989,
      "num_input_tokens_seen": 11141440,
      "step": 8095
    },
    {
      "epoch": 0.25926637219128096,
      "grad_norm": 0.007345728576183319,
      "learning_rate": 0.27066334424224553,
      "loss": 0.559,
      "num_input_tokens_seen": 11147968,
      "step": 8100
    },
    {
      "epoch": 0.2594264131617694,
      "grad_norm": 0.010720835998654366,
      "learning_rate": 0.27062834204132297,
      "loss": 0.5011,
      "num_input_tokens_seen": 11154880,
      "step": 8105
    },
    {
      "epoch": 0.25958645413225784,
      "grad_norm": 0.02126622386276722,
      "learning_rate": 0.27059332123799407,
      "loss": 0.8311,
      "num_input_tokens_seen": 11161504,
      "step": 8110
    },
    {
      "epoch": 0.2597464951027463,
      "grad_norm": 0.0116244126111269,
      "learning_rate": 0.27055828183765956,
      "loss": 0.6787,
      "num_input_tokens_seen": 11168912,
      "step": 8115
    },
    {
      "epoch": 0.2599065360732348,
      "grad_norm": 0.010658898390829563,
      "learning_rate": 0.270523223845723,
      "loss": 0.6894,
      "num_input_tokens_seen": 11175744,
      "step": 8120
    },
    {
      "epoch": 0.2600665770437232,
      "grad_norm": 0.018593663349747658,
      "learning_rate": 0.2704881472675907,
      "loss": 0.7327,
      "num_input_tokens_seen": 11182464,
      "step": 8125
    },
    {
      "epoch": 0.26022661801421165,
      "grad_norm": 0.025547513738274574,
      "learning_rate": 0.270453052108672,
      "loss": 0.6309,
      "num_input_tokens_seen": 11189536,
      "step": 8130
    },
    {
      "epoch": 0.2603866589847001,
      "grad_norm": 0.005909955594688654,
      "learning_rate": 0.2704179383743789,
      "loss": 0.873,
      "num_input_tokens_seen": 11196672,
      "step": 8135
    },
    {
      "epoch": 0.26054669995518853,
      "grad_norm": 0.008847068063914776,
      "learning_rate": 0.27038280607012644,
      "loss": 0.7538,
      "num_input_tokens_seen": 11203248,
      "step": 8140
    },
    {
      "epoch": 0.26070674092567697,
      "grad_norm": 0.013415513560175896,
      "learning_rate": 0.27034765520133247,
      "loss": 0.7369,
      "num_input_tokens_seen": 11210352,
      "step": 8145
    },
    {
      "epoch": 0.2608667818961654,
      "grad_norm": 0.006500020157545805,
      "learning_rate": 0.2703124857734177,
      "loss": 0.567,
      "num_input_tokens_seen": 11217248,
      "step": 8150
    },
    {
      "epoch": 0.26102682286665385,
      "grad_norm": 0.011540479958057404,
      "learning_rate": 0.27027729779180565,
      "loss": 0.5507,
      "num_input_tokens_seen": 11223792,
      "step": 8155
    },
    {
      "epoch": 0.2611868638371423,
      "grad_norm": 0.009609749540686607,
      "learning_rate": 0.27024209126192283,
      "loss": 0.6529,
      "num_input_tokens_seen": 11230544,
      "step": 8160
    },
    {
      "epoch": 0.26134690480763073,
      "grad_norm": 0.011730894446372986,
      "learning_rate": 0.2702068661891984,
      "loss": 0.7174,
      "num_input_tokens_seen": 11236976,
      "step": 8165
    },
    {
      "epoch": 0.26150694577811917,
      "grad_norm": 0.012627842836081982,
      "learning_rate": 0.2701716225790647,
      "loss": 0.6876,
      "num_input_tokens_seen": 11243936,
      "step": 8170
    },
    {
      "epoch": 0.26166698674860767,
      "grad_norm": 0.011240657418966293,
      "learning_rate": 0.27013636043695655,
      "loss": 0.7423,
      "num_input_tokens_seen": 11250576,
      "step": 8175
    },
    {
      "epoch": 0.2618270277190961,
      "grad_norm": 0.007822379469871521,
      "learning_rate": 0.27010107976831194,
      "loss": 0.5209,
      "num_input_tokens_seen": 11257008,
      "step": 8180
    },
    {
      "epoch": 0.26198706868958455,
      "grad_norm": 0.011538907885551453,
      "learning_rate": 0.2700657805785715,
      "loss": 0.5714,
      "num_input_tokens_seen": 11263952,
      "step": 8185
    },
    {
      "epoch": 0.262147109660073,
      "grad_norm": 0.00645653298124671,
      "learning_rate": 0.2700304628731789,
      "loss": 0.6895,
      "num_input_tokens_seen": 11271056,
      "step": 8190
    },
    {
      "epoch": 0.2623071506305614,
      "grad_norm": 0.007929829880595207,
      "learning_rate": 0.26999512665758046,
      "loss": 0.7259,
      "num_input_tokens_seen": 11277456,
      "step": 8195
    },
    {
      "epoch": 0.26246719160104987,
      "grad_norm": 0.010756120085716248,
      "learning_rate": 0.2699597719372256,
      "loss": 0.6293,
      "num_input_tokens_seen": 11284128,
      "step": 8200
    },
    {
      "epoch": 0.26246719160104987,
      "eval_loss": 0.6950517296791077,
      "eval_runtime": 332.0889,
      "eval_samples_per_second": 41.814,
      "eval_steps_per_second": 20.907,
      "num_input_tokens_seen": 11284128,
      "step": 8200
    },
    {
      "epoch": 0.2626272325715383,
      "grad_norm": 0.011558943428099155,
      "learning_rate": 0.26992439871756635,
      "loss": 0.7285,
      "num_input_tokens_seen": 11290816,
      "step": 8205
    },
    {
      "epoch": 0.26278727354202674,
      "grad_norm": 0.007842072285711765,
      "learning_rate": 0.2698890070040578,
      "loss": 0.7092,
      "num_input_tokens_seen": 11297120,
      "step": 8210
    },
    {
      "epoch": 0.2629473145125152,
      "grad_norm": 0.011464515700936317,
      "learning_rate": 0.2698535968021577,
      "loss": 0.6885,
      "num_input_tokens_seen": 11304192,
      "step": 8215
    },
    {
      "epoch": 0.2631073554830036,
      "grad_norm": 0.007150901015847921,
      "learning_rate": 0.26981816811732684,
      "loss": 0.7019,
      "num_input_tokens_seen": 11310736,
      "step": 8220
    },
    {
      "epoch": 0.2632673964534921,
      "grad_norm": 0.009371697902679443,
      "learning_rate": 0.26978272095502875,
      "loss": 0.6954,
      "num_input_tokens_seen": 11317520,
      "step": 8225
    },
    {
      "epoch": 0.26342743742398056,
      "grad_norm": 0.008449207060039043,
      "learning_rate": 0.26974725532072974,
      "loss": 0.8574,
      "num_input_tokens_seen": 11324544,
      "step": 8230
    },
    {
      "epoch": 0.263587478394469,
      "grad_norm": 0.009743397124111652,
      "learning_rate": 0.26971177121989914,
      "loss": 0.7761,
      "num_input_tokens_seen": 11331296,
      "step": 8235
    },
    {
      "epoch": 0.26374751936495744,
      "grad_norm": 0.006077510304749012,
      "learning_rate": 0.2696762686580091,
      "loss": 0.7592,
      "num_input_tokens_seen": 11337920,
      "step": 8240
    },
    {
      "epoch": 0.2639075603354459,
      "grad_norm": 0.007273699622601271,
      "learning_rate": 0.26964074764053436,
      "loss": 0.5456,
      "num_input_tokens_seen": 11345072,
      "step": 8245
    },
    {
      "epoch": 0.2640676013059343,
      "grad_norm": 0.0072770873084664345,
      "learning_rate": 0.2696052081729529,
      "loss": 0.6146,
      "num_input_tokens_seen": 11351856,
      "step": 8250
    },
    {
      "epoch": 0.26422764227642276,
      "grad_norm": 0.009554846212267876,
      "learning_rate": 0.2695696502607453,
      "loss": 0.7692,
      "num_input_tokens_seen": 11359088,
      "step": 8255
    },
    {
      "epoch": 0.2643876832469112,
      "grad_norm": 0.005392825230956078,
      "learning_rate": 0.26953407390939504,
      "loss": 0.7005,
      "num_input_tokens_seen": 11365680,
      "step": 8260
    },
    {
      "epoch": 0.26454772421739964,
      "grad_norm": 0.009507431648671627,
      "learning_rate": 0.26949847912438835,
      "loss": 0.4819,
      "num_input_tokens_seen": 11372576,
      "step": 8265
    },
    {
      "epoch": 0.2647077651878881,
      "grad_norm": 0.0039152647368609905,
      "learning_rate": 0.26946286591121454,
      "loss": 0.6989,
      "num_input_tokens_seen": 11379120,
      "step": 8270
    },
    {
      "epoch": 0.2648678061583765,
      "grad_norm": 0.009904664941132069,
      "learning_rate": 0.2694272342753655,
      "loss": 0.8709,
      "num_input_tokens_seen": 11385888,
      "step": 8275
    },
    {
      "epoch": 0.265027847128865,
      "grad_norm": 0.006711646448820829,
      "learning_rate": 0.26939158422233617,
      "loss": 0.5267,
      "num_input_tokens_seen": 11393152,
      "step": 8280
    },
    {
      "epoch": 0.26518788809935345,
      "grad_norm": 0.007665413431823254,
      "learning_rate": 0.26935591575762413,
      "loss": 0.6924,
      "num_input_tokens_seen": 11399584,
      "step": 8285
    },
    {
      "epoch": 0.2653479290698419,
      "grad_norm": 0.01243884488940239,
      "learning_rate": 0.26932022888672996,
      "loss": 0.7244,
      "num_input_tokens_seen": 11406384,
      "step": 8290
    },
    {
      "epoch": 0.26550797004033033,
      "grad_norm": 0.005931397434324026,
      "learning_rate": 0.26928452361515703,
      "loss": 0.7987,
      "num_input_tokens_seen": 11413232,
      "step": 8295
    },
    {
      "epoch": 0.26566801101081877,
      "grad_norm": 0.006602325011044741,
      "learning_rate": 0.26924879994841155,
      "loss": 0.6587,
      "num_input_tokens_seen": 11420560,
      "step": 8300
    },
    {
      "epoch": 0.2658280519813072,
      "grad_norm": 0.0083188246935606,
      "learning_rate": 0.2692130578920025,
      "loss": 0.5984,
      "num_input_tokens_seen": 11427024,
      "step": 8305
    },
    {
      "epoch": 0.26598809295179565,
      "grad_norm": 0.009259200654923916,
      "learning_rate": 0.26917729745144187,
      "loss": 0.8283,
      "num_input_tokens_seen": 11433696,
      "step": 8310
    },
    {
      "epoch": 0.2661481339222841,
      "grad_norm": 0.006635971833020449,
      "learning_rate": 0.2691415186322443,
      "loss": 0.5221,
      "num_input_tokens_seen": 11440352,
      "step": 8315
    },
    {
      "epoch": 0.26630817489277253,
      "grad_norm": 0.009373366832733154,
      "learning_rate": 0.2691057214399273,
      "loss": 0.7803,
      "num_input_tokens_seen": 11446912,
      "step": 8320
    },
    {
      "epoch": 0.26646821586326097,
      "grad_norm": 0.009493043646216393,
      "learning_rate": 0.2690699058800113,
      "loss": 0.7414,
      "num_input_tokens_seen": 11453872,
      "step": 8325
    },
    {
      "epoch": 0.26662825683374947,
      "grad_norm": 0.006813171319663525,
      "learning_rate": 0.2690340719580194,
      "loss": 0.5984,
      "num_input_tokens_seen": 11460832,
      "step": 8330
    },
    {
      "epoch": 0.2667882978042379,
      "grad_norm": 0.006854232866317034,
      "learning_rate": 0.2689982196794778,
      "loss": 0.689,
      "num_input_tokens_seen": 11467904,
      "step": 8335
    },
    {
      "epoch": 0.26694833877472635,
      "grad_norm": 0.009286942891776562,
      "learning_rate": 0.2689623490499153,
      "loss": 0.5858,
      "num_input_tokens_seen": 11474560,
      "step": 8340
    },
    {
      "epoch": 0.2671083797452148,
      "grad_norm": 0.008495469577610493,
      "learning_rate": 0.2689264600748636,
      "loss": 0.4803,
      "num_input_tokens_seen": 11481264,
      "step": 8345
    },
    {
      "epoch": 0.2672684207157032,
      "grad_norm": 0.006974439136683941,
      "learning_rate": 0.26889055275985724,
      "loss": 0.6158,
      "num_input_tokens_seen": 11489360,
      "step": 8350
    },
    {
      "epoch": 0.26742846168619167,
      "grad_norm": 0.011327465996146202,
      "learning_rate": 0.2688546271104335,
      "loss": 0.86,
      "num_input_tokens_seen": 11496112,
      "step": 8355
    },
    {
      "epoch": 0.2675885026566801,
      "grad_norm": 0.016003059223294258,
      "learning_rate": 0.26881868313213275,
      "loss": 0.891,
      "num_input_tokens_seen": 11503088,
      "step": 8360
    },
    {
      "epoch": 0.26774854362716854,
      "grad_norm": 0.004570802673697472,
      "learning_rate": 0.2687827208304978,
      "loss": 0.7133,
      "num_input_tokens_seen": 11509632,
      "step": 8365
    },
    {
      "epoch": 0.267908584597657,
      "grad_norm": 0.005690330173820257,
      "learning_rate": 0.26874674021107464,
      "loss": 0.6383,
      "num_input_tokens_seen": 11516160,
      "step": 8370
    },
    {
      "epoch": 0.2680686255681454,
      "grad_norm": 0.006123042665421963,
      "learning_rate": 0.2687107412794118,
      "loss": 0.6144,
      "num_input_tokens_seen": 11522720,
      "step": 8375
    },
    {
      "epoch": 0.26822866653863386,
      "grad_norm": 0.005340490024536848,
      "learning_rate": 0.26867472404106096,
      "loss": 0.5255,
      "num_input_tokens_seen": 11529696,
      "step": 8380
    },
    {
      "epoch": 0.26838870750912236,
      "grad_norm": 0.007437787018716335,
      "learning_rate": 0.26863868850157624,
      "loss": 0.5865,
      "num_input_tokens_seen": 11536304,
      "step": 8385
    },
    {
      "epoch": 0.2685487484796108,
      "grad_norm": 0.009700038470327854,
      "learning_rate": 0.26860263466651485,
      "loss": 0.4207,
      "num_input_tokens_seen": 11543392,
      "step": 8390
    },
    {
      "epoch": 0.26870878945009924,
      "grad_norm": 0.03670084476470947,
      "learning_rate": 0.26856656254143674,
      "loss": 0.7319,
      "num_input_tokens_seen": 11550112,
      "step": 8395
    },
    {
      "epoch": 0.2688688304205877,
      "grad_norm": 0.014733667485415936,
      "learning_rate": 0.2685304721319047,
      "loss": 0.6334,
      "num_input_tokens_seen": 11556816,
      "step": 8400
    },
    {
      "epoch": 0.2688688304205877,
      "eval_loss": 0.6879321336746216,
      "eval_runtime": 332.7248,
      "eval_samples_per_second": 41.734,
      "eval_steps_per_second": 20.867,
      "num_input_tokens_seen": 11556816,
      "step": 8400
    },
    {
      "epoch": 0.2690288713910761,
      "grad_norm": 0.011301757767796516,
      "learning_rate": 0.2684943634434843,
      "loss": 0.4954,
      "num_input_tokens_seen": 11563776,
      "step": 8405
    },
    {
      "epoch": 0.26918891236156456,
      "grad_norm": 0.009746763855218887,
      "learning_rate": 0.268458236481744,
      "loss": 0.5978,
      "num_input_tokens_seen": 11570752,
      "step": 8410
    },
    {
      "epoch": 0.269348953332053,
      "grad_norm": 0.009461709298193455,
      "learning_rate": 0.2684220912522549,
      "loss": 0.6485,
      "num_input_tokens_seen": 11577392,
      "step": 8415
    },
    {
      "epoch": 0.26950899430254144,
      "grad_norm": 0.005236112512648106,
      "learning_rate": 0.2683859277605913,
      "loss": 0.5652,
      "num_input_tokens_seen": 11583760,
      "step": 8420
    },
    {
      "epoch": 0.2696690352730299,
      "grad_norm": 0.006198915187269449,
      "learning_rate": 0.2683497460123298,
      "loss": 0.6609,
      "num_input_tokens_seen": 11590304,
      "step": 8425
    },
    {
      "epoch": 0.2698290762435183,
      "grad_norm": 0.0065734004601836205,
      "learning_rate": 0.26831354601305013,
      "loss": 0.6706,
      "num_input_tokens_seen": 11597232,
      "step": 8430
    },
    {
      "epoch": 0.2699891172140068,
      "grad_norm": 0.01644887402653694,
      "learning_rate": 0.26827732776833496,
      "loss": 0.7211,
      "num_input_tokens_seen": 11604192,
      "step": 8435
    },
    {
      "epoch": 0.27014915818449525,
      "grad_norm": 0.005048952531069517,
      "learning_rate": 0.26824109128376944,
      "loss": 0.664,
      "num_input_tokens_seen": 11611152,
      "step": 8440
    },
    {
      "epoch": 0.2703091991549837,
      "grad_norm": 0.004889735020697117,
      "learning_rate": 0.2682048365649417,
      "loss": 0.7053,
      "num_input_tokens_seen": 11618640,
      "step": 8445
    },
    {
      "epoch": 0.27046924012547213,
      "grad_norm": 0.00884926412254572,
      "learning_rate": 0.2681685636174428,
      "loss": 0.6269,
      "num_input_tokens_seen": 11625520,
      "step": 8450
    },
    {
      "epoch": 0.27062928109596057,
      "grad_norm": 0.005334661342203617,
      "learning_rate": 0.2681322724468663,
      "loss": 0.5405,
      "num_input_tokens_seen": 11632224,
      "step": 8455
    },
    {
      "epoch": 0.270789322066449,
      "grad_norm": 0.007501024752855301,
      "learning_rate": 0.2680959630588089,
      "loss": 0.7242,
      "num_input_tokens_seen": 11638736,
      "step": 8460
    },
    {
      "epoch": 0.27094936303693745,
      "grad_norm": 0.008555927313864231,
      "learning_rate": 0.26805963545886985,
      "loss": 0.5797,
      "num_input_tokens_seen": 11645440,
      "step": 8465
    },
    {
      "epoch": 0.2711094040074259,
      "grad_norm": 0.010778768919408321,
      "learning_rate": 0.26802328965265143,
      "loss": 0.4869,
      "num_input_tokens_seen": 11651840,
      "step": 8470
    },
    {
      "epoch": 0.27126944497791433,
      "grad_norm": 0.006680928170681,
      "learning_rate": 0.26798692564575854,
      "loss": 0.7156,
      "num_input_tokens_seen": 11658848,
      "step": 8475
    },
    {
      "epoch": 0.27142948594840277,
      "grad_norm": 0.006544272415339947,
      "learning_rate": 0.26795054344379904,
      "loss": 0.7321,
      "num_input_tokens_seen": 11665728,
      "step": 8480
    },
    {
      "epoch": 0.2715895269188912,
      "grad_norm": 0.007015740033239126,
      "learning_rate": 0.2679141430523835,
      "loss": 0.7529,
      "num_input_tokens_seen": 11672352,
      "step": 8485
    },
    {
      "epoch": 0.2717495678893797,
      "grad_norm": 0.00876945722848177,
      "learning_rate": 0.2678777244771252,
      "loss": 0.5501,
      "num_input_tokens_seen": 11678912,
      "step": 8490
    },
    {
      "epoch": 0.27190960885986815,
      "grad_norm": 0.008769613690674305,
      "learning_rate": 0.2678412877236405,
      "loss": 0.7159,
      "num_input_tokens_seen": 11685408,
      "step": 8495
    },
    {
      "epoch": 0.2720696498303566,
      "grad_norm": 0.01159744057804346,
      "learning_rate": 0.2678048327975484,
      "loss": 0.6351,
      "num_input_tokens_seen": 11692336,
      "step": 8500
    },
    {
      "epoch": 0.272229690800845,
      "grad_norm": 0.006748630665242672,
      "learning_rate": 0.2677683597044706,
      "loss": 0.5155,
      "num_input_tokens_seen": 11699536,
      "step": 8505
    },
    {
      "epoch": 0.27238973177133347,
      "grad_norm": 0.0036971857771277428,
      "learning_rate": 0.2677318684500318,
      "loss": 0.5234,
      "num_input_tokens_seen": 11706816,
      "step": 8510
    },
    {
      "epoch": 0.2725497727418219,
      "grad_norm": 0.012051122263073921,
      "learning_rate": 0.2676953590398593,
      "loss": 0.7026,
      "num_input_tokens_seen": 11714224,
      "step": 8515
    },
    {
      "epoch": 0.27270981371231034,
      "grad_norm": 0.01676187478005886,
      "learning_rate": 0.2676588314795834,
      "loss": 0.6641,
      "num_input_tokens_seen": 11721024,
      "step": 8520
    },
    {
      "epoch": 0.2728698546827988,
      "grad_norm": 0.0078828614205122,
      "learning_rate": 0.26762228577483715,
      "loss": 0.5688,
      "num_input_tokens_seen": 11727872,
      "step": 8525
    },
    {
      "epoch": 0.2730298956532872,
      "grad_norm": 0.006649298127740622,
      "learning_rate": 0.2675857219312563,
      "loss": 0.719,
      "num_input_tokens_seen": 11734688,
      "step": 8530
    },
    {
      "epoch": 0.27318993662377566,
      "grad_norm": 0.008280339650809765,
      "learning_rate": 0.2675491399544794,
      "loss": 0.5513,
      "num_input_tokens_seen": 11741392,
      "step": 8535
    },
    {
      "epoch": 0.2733499775942641,
      "grad_norm": 0.006033977959305048,
      "learning_rate": 0.2675125398501479,
      "loss": 0.516,
      "num_input_tokens_seen": 11748000,
      "step": 8540
    },
    {
      "epoch": 0.2735100185647526,
      "grad_norm": 0.010564738884568214,
      "learning_rate": 0.26747592162390604,
      "loss": 0.673,
      "num_input_tokens_seen": 11754896,
      "step": 8545
    },
    {
      "epoch": 0.27367005953524104,
      "grad_norm": 0.009208228439092636,
      "learning_rate": 0.26743928528140076,
      "loss": 0.7809,
      "num_input_tokens_seen": 11761952,
      "step": 8550
    },
    {
      "epoch": 0.2738301005057295,
      "grad_norm": 0.006581317633390427,
      "learning_rate": 0.26740263082828186,
      "loss": 0.5294,
      "num_input_tokens_seen": 11768688,
      "step": 8555
    },
    {
      "epoch": 0.2739901414762179,
      "grad_norm": 0.011490979231894016,
      "learning_rate": 0.2673659582702019,
      "loss": 0.792,
      "num_input_tokens_seen": 11775632,
      "step": 8560
    },
    {
      "epoch": 0.27415018244670636,
      "grad_norm": 0.004846623167395592,
      "learning_rate": 0.2673292676128163,
      "loss": 0.6038,
      "num_input_tokens_seen": 11782464,
      "step": 8565
    },
    {
      "epoch": 0.2743102234171948,
      "grad_norm": 0.01139429584145546,
      "learning_rate": 0.2672925588617831,
      "loss": 0.9336,
      "num_input_tokens_seen": 11789264,
      "step": 8570
    },
    {
      "epoch": 0.27447026438768324,
      "grad_norm": 0.006463234778493643,
      "learning_rate": 0.2672558320227634,
      "loss": 0.6252,
      "num_input_tokens_seen": 11795680,
      "step": 8575
    },
    {
      "epoch": 0.2746303053581717,
      "grad_norm": 0.008846582844853401,
      "learning_rate": 0.2672190871014209,
      "loss": 0.6536,
      "num_input_tokens_seen": 11802064,
      "step": 8580
    },
    {
      "epoch": 0.2747903463286601,
      "grad_norm": 0.008103794418275356,
      "learning_rate": 0.267182324103422,
      "loss": 0.719,
      "num_input_tokens_seen": 11808720,
      "step": 8585
    },
    {
      "epoch": 0.27495038729914856,
      "grad_norm": 0.010140721686184406,
      "learning_rate": 0.2671455430344362,
      "loss": 0.4495,
      "num_input_tokens_seen": 11815536,
      "step": 8590
    },
    {
      "epoch": 0.27511042826963705,
      "grad_norm": 0.008127265609800816,
      "learning_rate": 0.2671087439001355,
      "loss": 0.6814,
      "num_input_tokens_seen": 11822080,
      "step": 8595
    },
    {
      "epoch": 0.2752704692401255,
      "grad_norm": 0.015258848667144775,
      "learning_rate": 0.2670719267061948,
      "loss": 0.8279,
      "num_input_tokens_seen": 11828816,
      "step": 8600
    },
    {
      "epoch": 0.2752704692401255,
      "eval_loss": 0.7054733633995056,
      "eval_runtime": 331.6324,
      "eval_samples_per_second": 41.872,
      "eval_steps_per_second": 20.936,
      "num_input_tokens_seen": 11828816,
      "step": 8600
    },
    {
      "epoch": 0.27543051021061393,
      "grad_norm": 0.015467997640371323,
      "learning_rate": 0.2670350914582918,
      "loss": 0.7706,
      "num_input_tokens_seen": 11835328,
      "step": 8605
    },
    {
      "epoch": 0.2755905511811024,
      "grad_norm": 0.008065232075750828,
      "learning_rate": 0.26699823816210694,
      "loss": 0.4673,
      "num_input_tokens_seen": 11842192,
      "step": 8610
    },
    {
      "epoch": 0.2757505921515908,
      "grad_norm": 0.025897838175296783,
      "learning_rate": 0.26696136682332344,
      "loss": 0.8017,
      "num_input_tokens_seen": 11849296,
      "step": 8615
    },
    {
      "epoch": 0.27591063312207925,
      "grad_norm": 0.01021454855799675,
      "learning_rate": 0.2669244774476274,
      "loss": 0.7464,
      "num_input_tokens_seen": 11856048,
      "step": 8620
    },
    {
      "epoch": 0.2760706740925677,
      "grad_norm": 0.010139137506484985,
      "learning_rate": 0.2668875700407075,
      "loss": 0.8089,
      "num_input_tokens_seen": 11862848,
      "step": 8625
    },
    {
      "epoch": 0.27623071506305613,
      "grad_norm": 0.021038269624114037,
      "learning_rate": 0.26685064460825547,
      "loss": 0.6841,
      "num_input_tokens_seen": 11869760,
      "step": 8630
    },
    {
      "epoch": 0.27639075603354457,
      "grad_norm": 0.019782651215791702,
      "learning_rate": 0.26681370115596553,
      "loss": 0.7227,
      "num_input_tokens_seen": 11876576,
      "step": 8635
    },
    {
      "epoch": 0.276550797004033,
      "grad_norm": 0.00738466577604413,
      "learning_rate": 0.26677673968953497,
      "loss": 0.6574,
      "num_input_tokens_seen": 11883104,
      "step": 8640
    },
    {
      "epoch": 0.27671083797452145,
      "grad_norm": 0.015635160729289055,
      "learning_rate": 0.2667397602146636,
      "loss": 0.5939,
      "num_input_tokens_seen": 11889872,
      "step": 8645
    },
    {
      "epoch": 0.27687087894500995,
      "grad_norm": 0.01902231015264988,
      "learning_rate": 0.2667027627370542,
      "loss": 0.7467,
      "num_input_tokens_seen": 11896784,
      "step": 8650
    },
    {
      "epoch": 0.2770309199154984,
      "grad_norm": 0.005624221637845039,
      "learning_rate": 0.26666574726241216,
      "loss": 0.7033,
      "num_input_tokens_seen": 11903264,
      "step": 8655
    },
    {
      "epoch": 0.2771909608859868,
      "grad_norm": 0.007232906296849251,
      "learning_rate": 0.2666287137964458,
      "loss": 0.7602,
      "num_input_tokens_seen": 11910464,
      "step": 8660
    },
    {
      "epoch": 0.27735100185647527,
      "grad_norm": 0.008390610106289387,
      "learning_rate": 0.26659166234486614,
      "loss": 0.4146,
      "num_input_tokens_seen": 11917152,
      "step": 8665
    },
    {
      "epoch": 0.2775110428269637,
      "grad_norm": 0.008842998184263706,
      "learning_rate": 0.2665545929133869,
      "loss": 0.6579,
      "num_input_tokens_seen": 11924240,
      "step": 8670
    },
    {
      "epoch": 0.27767108379745215,
      "grad_norm": 0.013377797789871693,
      "learning_rate": 0.2665175055077248,
      "loss": 0.6647,
      "num_input_tokens_seen": 11931376,
      "step": 8675
    },
    {
      "epoch": 0.2778311247679406,
      "grad_norm": 0.010320200584828854,
      "learning_rate": 0.2664804001335991,
      "loss": 0.9332,
      "num_input_tokens_seen": 11938352,
      "step": 8680
    },
    {
      "epoch": 0.277991165738429,
      "grad_norm": 0.006507988553494215,
      "learning_rate": 0.26644327679673185,
      "loss": 0.7795,
      "num_input_tokens_seen": 11945504,
      "step": 8685
    },
    {
      "epoch": 0.27815120670891746,
      "grad_norm": 0.009655999951064587,
      "learning_rate": 0.26640613550284803,
      "loss": 0.6045,
      "num_input_tokens_seen": 11952224,
      "step": 8690
    },
    {
      "epoch": 0.2783112476794059,
      "grad_norm": 0.015135655179619789,
      "learning_rate": 0.26636897625767525,
      "loss": 0.749,
      "num_input_tokens_seen": 11959104,
      "step": 8695
    },
    {
      "epoch": 0.2784712886498944,
      "grad_norm": 0.009232353419065475,
      "learning_rate": 0.266331799066944,
      "loss": 0.7155,
      "num_input_tokens_seen": 11965568,
      "step": 8700
    },
    {
      "epoch": 0.27863132962038284,
      "grad_norm": 0.011117063462734222,
      "learning_rate": 0.2662946039363874,
      "loss": 0.7497,
      "num_input_tokens_seen": 11972336,
      "step": 8705
    },
    {
      "epoch": 0.2787913705908713,
      "grad_norm": 0.006511324550956488,
      "learning_rate": 0.2662573908717414,
      "loss": 0.6663,
      "num_input_tokens_seen": 11979392,
      "step": 8710
    },
    {
      "epoch": 0.2789514115613597,
      "grad_norm": 0.015053525567054749,
      "learning_rate": 0.2662201598787447,
      "loss": 0.7616,
      "num_input_tokens_seen": 11986528,
      "step": 8715
    },
    {
      "epoch": 0.27911145253184816,
      "grad_norm": 0.015125627629458904,
      "learning_rate": 0.2661829109631389,
      "loss": 0.7649,
      "num_input_tokens_seen": 11993344,
      "step": 8720
    },
    {
      "epoch": 0.2792714935023366,
      "grad_norm": 0.00962990801781416,
      "learning_rate": 0.26614564413066816,
      "loss": 0.6814,
      "num_input_tokens_seen": 12000400,
      "step": 8725
    },
    {
      "epoch": 0.27943153447282504,
      "grad_norm": 0.011324038729071617,
      "learning_rate": 0.2661083593870795,
      "loss": 0.689,
      "num_input_tokens_seen": 12007120,
      "step": 8730
    },
    {
      "epoch": 0.2795915754433135,
      "grad_norm": 0.010671140626072884,
      "learning_rate": 0.26607105673812276,
      "loss": 0.6404,
      "num_input_tokens_seen": 12013760,
      "step": 8735
    },
    {
      "epoch": 0.2797516164138019,
      "grad_norm": 0.014262614771723747,
      "learning_rate": 0.2660337361895504,
      "loss": 0.7426,
      "num_input_tokens_seen": 12020512,
      "step": 8740
    },
    {
      "epoch": 0.27991165738429036,
      "grad_norm": 0.007003130856901407,
      "learning_rate": 0.26599639774711775,
      "loss": 0.4693,
      "num_input_tokens_seen": 12027392,
      "step": 8745
    },
    {
      "epoch": 0.2800716983547788,
      "grad_norm": 0.008341572247445583,
      "learning_rate": 0.2659590414165829,
      "loss": 0.7865,
      "num_input_tokens_seen": 12034320,
      "step": 8750
    },
    {
      "epoch": 0.2802317393252673,
      "grad_norm": 0.010071524418890476,
      "learning_rate": 0.2659216672037066,
      "loss": 0.4765,
      "num_input_tokens_seen": 12041520,
      "step": 8755
    },
    {
      "epoch": 0.28039178029575573,
      "grad_norm": 0.0052046459168195724,
      "learning_rate": 0.26588427511425244,
      "loss": 0.7381,
      "num_input_tokens_seen": 12048400,
      "step": 8760
    },
    {
      "epoch": 0.2805518212662442,
      "grad_norm": 0.006870572455227375,
      "learning_rate": 0.26584686515398676,
      "loss": 0.4691,
      "num_input_tokens_seen": 12055072,
      "step": 8765
    },
    {
      "epoch": 0.2807118622367326,
      "grad_norm": 0.005315572489053011,
      "learning_rate": 0.2658094373286787,
      "loss": 0.6467,
      "num_input_tokens_seen": 12061920,
      "step": 8770
    },
    {
      "epoch": 0.28087190320722105,
      "grad_norm": 0.005988674238324165,
      "learning_rate": 0.2657719916441,
      "loss": 0.6067,
      "num_input_tokens_seen": 12068976,
      "step": 8775
    },
    {
      "epoch": 0.2810319441777095,
      "grad_norm": 0.00887224543839693,
      "learning_rate": 0.2657345281060253,
      "loss": 0.5321,
      "num_input_tokens_seen": 12076112,
      "step": 8780
    },
    {
      "epoch": 0.28119198514819793,
      "grad_norm": 0.014161346480250359,
      "learning_rate": 0.26569704672023203,
      "loss": 0.585,
      "num_input_tokens_seen": 12082640,
      "step": 8785
    },
    {
      "epoch": 0.28135202611868637,
      "grad_norm": 0.009067794308066368,
      "learning_rate": 0.26565954749250015,
      "loss": 0.7459,
      "num_input_tokens_seen": 12090016,
      "step": 8790
    },
    {
      "epoch": 0.2815120670891748,
      "grad_norm": 0.008635942824184895,
      "learning_rate": 0.2656220304286126,
      "loss": 0.7097,
      "num_input_tokens_seen": 12097280,
      "step": 8795
    },
    {
      "epoch": 0.28167210805966325,
      "grad_norm": 0.004396345466375351,
      "learning_rate": 0.265584495534355,
      "loss": 0.6041,
      "num_input_tokens_seen": 12104176,
      "step": 8800
    },
    {
      "epoch": 0.28167210805966325,
      "eval_loss": 0.6626179218292236,
      "eval_runtime": 331.769,
      "eval_samples_per_second": 41.854,
      "eval_steps_per_second": 20.927,
      "num_input_tokens_seen": 12104176,
      "step": 8800
    },
    {
      "epoch": 0.28183214903015175,
      "grad_norm": 0.004717938136309385,
      "learning_rate": 0.2655469428155156,
      "loss": 0.605,
      "num_input_tokens_seen": 12110960,
      "step": 8805
    },
    {
      "epoch": 0.2819921900006402,
      "grad_norm": 0.006328154355287552,
      "learning_rate": 0.2655093722778856,
      "loss": 0.7025,
      "num_input_tokens_seen": 12117776,
      "step": 8810
    },
    {
      "epoch": 0.2821522309711286,
      "grad_norm": 0.014908558689057827,
      "learning_rate": 0.2654717839272588,
      "loss": 0.5633,
      "num_input_tokens_seen": 12124480,
      "step": 8815
    },
    {
      "epoch": 0.28231227194161707,
      "grad_norm": 0.009608385153114796,
      "learning_rate": 0.2654341777694318,
      "loss": 0.7648,
      "num_input_tokens_seen": 12131568,
      "step": 8820
    },
    {
      "epoch": 0.2824723129121055,
      "grad_norm": 0.008448909968137741,
      "learning_rate": 0.265396553810204,
      "loss": 0.4998,
      "num_input_tokens_seen": 12138896,
      "step": 8825
    },
    {
      "epoch": 0.28263235388259395,
      "grad_norm": 0.007727531250566244,
      "learning_rate": 0.26535891205537737,
      "loss": 0.6209,
      "num_input_tokens_seen": 12145920,
      "step": 8830
    },
    {
      "epoch": 0.2827923948530824,
      "grad_norm": 0.011930359527468681,
      "learning_rate": 0.26532125251075683,
      "loss": 0.7325,
      "num_input_tokens_seen": 12153008,
      "step": 8835
    },
    {
      "epoch": 0.2829524358235708,
      "grad_norm": 0.00807550735771656,
      "learning_rate": 0.26528357518214996,
      "loss": 0.6639,
      "num_input_tokens_seen": 12159792,
      "step": 8840
    },
    {
      "epoch": 0.28311247679405926,
      "grad_norm": 0.008678814396262169,
      "learning_rate": 0.26524588007536704,
      "loss": 0.5538,
      "num_input_tokens_seen": 12166624,
      "step": 8845
    },
    {
      "epoch": 0.2832725177645477,
      "grad_norm": 0.022035863250494003,
      "learning_rate": 0.26520816719622115,
      "loss": 0.6972,
      "num_input_tokens_seen": 12173248,
      "step": 8850
    },
    {
      "epoch": 0.28343255873503614,
      "grad_norm": 0.0075431084260344505,
      "learning_rate": 0.2651704365505281,
      "loss": 0.6382,
      "num_input_tokens_seen": 12179952,
      "step": 8855
    },
    {
      "epoch": 0.28359259970552464,
      "grad_norm": 0.011307477951049805,
      "learning_rate": 0.26513268814410634,
      "loss": 0.7589,
      "num_input_tokens_seen": 12187216,
      "step": 8860
    },
    {
      "epoch": 0.2837526406760131,
      "grad_norm": 0.010990317910909653,
      "learning_rate": 0.2650949219827773,
      "loss": 0.6558,
      "num_input_tokens_seen": 12193712,
      "step": 8865
    },
    {
      "epoch": 0.2839126816465015,
      "grad_norm": 0.012537635862827301,
      "learning_rate": 0.26505713807236486,
      "loss": 0.6284,
      "num_input_tokens_seen": 12200400,
      "step": 8870
    },
    {
      "epoch": 0.28407272261698996,
      "grad_norm": 0.007650337181985378,
      "learning_rate": 0.26501933641869585,
      "loss": 0.6971,
      "num_input_tokens_seen": 12207280,
      "step": 8875
    },
    {
      "epoch": 0.2842327635874784,
      "grad_norm": 0.006919007748365402,
      "learning_rate": 0.26498151702759976,
      "loss": 0.7183,
      "num_input_tokens_seen": 12214064,
      "step": 8880
    },
    {
      "epoch": 0.28439280455796684,
      "grad_norm": 0.007221536710858345,
      "learning_rate": 0.2649436799049088,
      "loss": 0.5765,
      "num_input_tokens_seen": 12221104,
      "step": 8885
    },
    {
      "epoch": 0.2845528455284553,
      "grad_norm": 0.009833513759076595,
      "learning_rate": 0.2649058250564579,
      "loss": 0.6196,
      "num_input_tokens_seen": 12227968,
      "step": 8890
    },
    {
      "epoch": 0.2847128864989437,
      "grad_norm": 0.00798302236944437,
      "learning_rate": 0.26486795248808476,
      "loss": 0.6968,
      "num_input_tokens_seen": 12234752,
      "step": 8895
    },
    {
      "epoch": 0.28487292746943216,
      "grad_norm": 0.01457277126610279,
      "learning_rate": 0.2648300622056298,
      "loss": 0.6474,
      "num_input_tokens_seen": 12241696,
      "step": 8900
    },
    {
      "epoch": 0.2850329684399206,
      "grad_norm": 0.007219317369163036,
      "learning_rate": 0.2647921542149363,
      "loss": 0.8306,
      "num_input_tokens_seen": 12249024,
      "step": 8905
    },
    {
      "epoch": 0.28519300941040904,
      "grad_norm": 0.01632436364889145,
      "learning_rate": 0.26475422852185,
      "loss": 0.8065,
      "num_input_tokens_seen": 12255792,
      "step": 8910
    },
    {
      "epoch": 0.28535305038089753,
      "grad_norm": 0.008250514976680279,
      "learning_rate": 0.2647162851322196,
      "loss": 0.6111,
      "num_input_tokens_seen": 12262720,
      "step": 8915
    },
    {
      "epoch": 0.285513091351386,
      "grad_norm": 0.009566945023834705,
      "learning_rate": 0.2646783240518964,
      "loss": 0.683,
      "num_input_tokens_seen": 12269568,
      "step": 8920
    },
    {
      "epoch": 0.2856731323218744,
      "grad_norm": 0.009926403872668743,
      "learning_rate": 0.26464034528673447,
      "loss": 0.541,
      "num_input_tokens_seen": 12276288,
      "step": 8925
    },
    {
      "epoch": 0.28583317329236285,
      "grad_norm": 0.006901978515088558,
      "learning_rate": 0.26460234884259065,
      "loss": 0.633,
      "num_input_tokens_seen": 12282800,
      "step": 8930
    },
    {
      "epoch": 0.2859932142628513,
      "grad_norm": 0.006817108951508999,
      "learning_rate": 0.2645643347253245,
      "loss": 0.5245,
      "num_input_tokens_seen": 12289840,
      "step": 8935
    },
    {
      "epoch": 0.28615325523333973,
      "grad_norm": 0.022324129939079285,
      "learning_rate": 0.2645263029407982,
      "loss": 0.7609,
      "num_input_tokens_seen": 12296432,
      "step": 8940
    },
    {
      "epoch": 0.28631329620382817,
      "grad_norm": 0.006341546308249235,
      "learning_rate": 0.2644882534948767,
      "loss": 0.7236,
      "num_input_tokens_seen": 12303248,
      "step": 8945
    },
    {
      "epoch": 0.2864733371743166,
      "grad_norm": 0.00869255606085062,
      "learning_rate": 0.2644501863934278,
      "loss": 0.6601,
      "num_input_tokens_seen": 12310400,
      "step": 8950
    },
    {
      "epoch": 0.28663337814480505,
      "grad_norm": 0.0071814050897955894,
      "learning_rate": 0.26441210164232193,
      "loss": 0.5105,
      "num_input_tokens_seen": 12317120,
      "step": 8955
    },
    {
      "epoch": 0.2867934191152935,
      "grad_norm": 0.010935398750007153,
      "learning_rate": 0.26437399924743216,
      "loss": 0.4855,
      "num_input_tokens_seen": 12323920,
      "step": 8960
    },
    {
      "epoch": 0.286953460085782,
      "grad_norm": 0.012808280996978283,
      "learning_rate": 0.26433587921463436,
      "loss": 0.6452,
      "num_input_tokens_seen": 12330720,
      "step": 8965
    },
    {
      "epoch": 0.2871135010562704,
      "grad_norm": 0.012849031016230583,
      "learning_rate": 0.2642977415498072,
      "loss": 0.5804,
      "num_input_tokens_seen": 12337888,
      "step": 8970
    },
    {
      "epoch": 0.28727354202675887,
      "grad_norm": 0.009677700698375702,
      "learning_rate": 0.26425958625883195,
      "loss": 0.6609,
      "num_input_tokens_seen": 12344592,
      "step": 8975
    },
    {
      "epoch": 0.2874335829972473,
      "grad_norm": 0.01410185731947422,
      "learning_rate": 0.2642214133475926,
      "loss": 0.6689,
      "num_input_tokens_seen": 12351568,
      "step": 8980
    },
    {
      "epoch": 0.28759362396773575,
      "grad_norm": 0.010511890053749084,
      "learning_rate": 0.26418322282197587,
      "loss": 0.6817,
      "num_input_tokens_seen": 12358064,
      "step": 8985
    },
    {
      "epoch": 0.2877536649382242,
      "grad_norm": 0.008195876143872738,
      "learning_rate": 0.2641450146878714,
      "loss": 0.707,
      "num_input_tokens_seen": 12364928,
      "step": 8990
    },
    {
      "epoch": 0.2879137059087126,
      "grad_norm": 0.007839297875761986,
      "learning_rate": 0.26410678895117107,
      "loss": 0.5582,
      "num_input_tokens_seen": 12372064,
      "step": 8995
    },
    {
      "epoch": 0.28807374687920106,
      "grad_norm": 0.010732516646385193,
      "learning_rate": 0.26406854561777,
      "loss": 0.6143,
      "num_input_tokens_seen": 12378784,
      "step": 9000
    },
    {
      "epoch": 0.28807374687920106,
      "eval_loss": 0.6639882326126099,
      "eval_runtime": 331.7809,
      "eval_samples_per_second": 41.853,
      "eval_steps_per_second": 20.926,
      "num_input_tokens_seen": 12378784,
      "step": 9000
    },
    {
      "epoch": 0.2882337878496895,
      "grad_norm": 0.009385496377944946,
      "learning_rate": 0.26403028469356576,
      "loss": 0.8614,
      "num_input_tokens_seen": 12385984,
      "step": 9005
    },
    {
      "epoch": 0.28839382882017794,
      "grad_norm": 0.010620811954140663,
      "learning_rate": 0.2639920061844585,
      "loss": 0.7046,
      "num_input_tokens_seen": 12392912,
      "step": 9010
    },
    {
      "epoch": 0.2885538697906664,
      "grad_norm": 0.007230508606880903,
      "learning_rate": 0.2639537100963515,
      "loss": 0.6356,
      "num_input_tokens_seen": 12400304,
      "step": 9015
    },
    {
      "epoch": 0.2887139107611549,
      "grad_norm": 0.007094886153936386,
      "learning_rate": 0.26391539643515033,
      "loss": 0.7468,
      "num_input_tokens_seen": 12407088,
      "step": 9020
    },
    {
      "epoch": 0.2888739517316433,
      "grad_norm": 0.006844061426818371,
      "learning_rate": 0.26387706520676346,
      "loss": 0.673,
      "num_input_tokens_seen": 12414336,
      "step": 9025
    },
    {
      "epoch": 0.28903399270213176,
      "grad_norm": 0.007956981658935547,
      "learning_rate": 0.26383871641710205,
      "loss": 0.7803,
      "num_input_tokens_seen": 12421056,
      "step": 9030
    },
    {
      "epoch": 0.2891940336726202,
      "grad_norm": 0.003932608757168055,
      "learning_rate": 0.26380035007208,
      "loss": 0.421,
      "num_input_tokens_seen": 12427984,
      "step": 9035
    },
    {
      "epoch": 0.28935407464310864,
      "grad_norm": 0.016090942546725273,
      "learning_rate": 0.26376196617761394,
      "loss": 0.765,
      "num_input_tokens_seen": 12434848,
      "step": 9040
    },
    {
      "epoch": 0.2895141156135971,
      "grad_norm": 0.007160667330026627,
      "learning_rate": 0.263723564739623,
      "loss": 0.55,
      "num_input_tokens_seen": 12441664,
      "step": 9045
    },
    {
      "epoch": 0.2896741565840855,
      "grad_norm": 0.012244940735399723,
      "learning_rate": 0.2636851457640293,
      "loss": 0.6022,
      "num_input_tokens_seen": 12448512,
      "step": 9050
    },
    {
      "epoch": 0.28983419755457396,
      "grad_norm": 0.005717778578400612,
      "learning_rate": 0.26364670925675737,
      "loss": 0.6832,
      "num_input_tokens_seen": 12455584,
      "step": 9055
    },
    {
      "epoch": 0.2899942385250624,
      "grad_norm": 0.012269129045307636,
      "learning_rate": 0.2636082552237347,
      "loss": 0.4917,
      "num_input_tokens_seen": 12462176,
      "step": 9060
    },
    {
      "epoch": 0.29015427949555084,
      "grad_norm": 0.007636896800249815,
      "learning_rate": 0.26356978367089146,
      "loss": 0.7788,
      "num_input_tokens_seen": 12469024,
      "step": 9065
    },
    {
      "epoch": 0.29031432046603933,
      "grad_norm": 0.007585485465824604,
      "learning_rate": 0.26353129460416036,
      "loss": 0.6982,
      "num_input_tokens_seen": 12476144,
      "step": 9070
    },
    {
      "epoch": 0.2904743614365278,
      "grad_norm": 0.008058891631662846,
      "learning_rate": 0.2634927880294769,
      "loss": 0.567,
      "num_input_tokens_seen": 12483072,
      "step": 9075
    },
    {
      "epoch": 0.2906344024070162,
      "grad_norm": 0.007067918311804533,
      "learning_rate": 0.26345426395277927,
      "loss": 0.6903,
      "num_input_tokens_seen": 12489776,
      "step": 9080
    },
    {
      "epoch": 0.29079444337750465,
      "grad_norm": 0.03445572406053543,
      "learning_rate": 0.2634157223800084,
      "loss": 0.7098,
      "num_input_tokens_seen": 12496832,
      "step": 9085
    },
    {
      "epoch": 0.2909544843479931,
      "grad_norm": 0.0112479068338871,
      "learning_rate": 0.26337716331710787,
      "loss": 0.5376,
      "num_input_tokens_seen": 12503680,
      "step": 9090
    },
    {
      "epoch": 0.29111452531848153,
      "grad_norm": 0.00636708689853549,
      "learning_rate": 0.2633385867700239,
      "loss": 0.6364,
      "num_input_tokens_seen": 12510352,
      "step": 9095
    },
    {
      "epoch": 0.29127456628896997,
      "grad_norm": 0.008000856265425682,
      "learning_rate": 0.2632999927447056,
      "loss": 0.6392,
      "num_input_tokens_seen": 12516944,
      "step": 9100
    },
    {
      "epoch": 0.2914346072594584,
      "grad_norm": 0.006039201747626066,
      "learning_rate": 0.2632613812471046,
      "loss": 0.5866,
      "num_input_tokens_seen": 12523792,
      "step": 9105
    },
    {
      "epoch": 0.29159464822994685,
      "grad_norm": 0.011403429321944714,
      "learning_rate": 0.2632227522831753,
      "loss": 0.6931,
      "num_input_tokens_seen": 12530944,
      "step": 9110
    },
    {
      "epoch": 0.2917546892004353,
      "grad_norm": 0.00636592973023653,
      "learning_rate": 0.26318410585887475,
      "loss": 0.6955,
      "num_input_tokens_seen": 12538064,
      "step": 9115
    },
    {
      "epoch": 0.29191473017092373,
      "grad_norm": 0.009267051704227924,
      "learning_rate": 0.2631454419801627,
      "loss": 0.5287,
      "num_input_tokens_seen": 12544800,
      "step": 9120
    },
    {
      "epoch": 0.2920747711414122,
      "grad_norm": 0.012518778443336487,
      "learning_rate": 0.2631067606530016,
      "loss": 0.7788,
      "num_input_tokens_seen": 12551552,
      "step": 9125
    },
    {
      "epoch": 0.29223481211190067,
      "grad_norm": 0.008870705030858517,
      "learning_rate": 0.2630680618833567,
      "loss": 0.7514,
      "num_input_tokens_seen": 12558608,
      "step": 9130
    },
    {
      "epoch": 0.2923948530823891,
      "grad_norm": 0.009904327802360058,
      "learning_rate": 0.26302934567719566,
      "loss": 0.8347,
      "num_input_tokens_seen": 12565552,
      "step": 9135
    },
    {
      "epoch": 0.29255489405287755,
      "grad_norm": 0.007372736930847168,
      "learning_rate": 0.2629906120404892,
      "loss": 0.7675,
      "num_input_tokens_seen": 12572560,
      "step": 9140
    },
    {
      "epoch": 0.292714935023366,
      "grad_norm": 0.010312055237591267,
      "learning_rate": 0.26295186097921036,
      "loss": 0.7418,
      "num_input_tokens_seen": 12579216,
      "step": 9145
    },
    {
      "epoch": 0.2928749759938544,
      "grad_norm": 0.005850558169186115,
      "learning_rate": 0.2629130924993351,
      "loss": 0.8681,
      "num_input_tokens_seen": 12586080,
      "step": 9150
    },
    {
      "epoch": 0.29303501696434286,
      "grad_norm": 0.011422601528465748,
      "learning_rate": 0.2628743066068421,
      "loss": 0.8216,
      "num_input_tokens_seen": 12593136,
      "step": 9155
    },
    {
      "epoch": 0.2931950579348313,
      "grad_norm": 0.009871602058410645,
      "learning_rate": 0.26283550330771244,
      "loss": 0.7026,
      "num_input_tokens_seen": 12600144,
      "step": 9160
    },
    {
      "epoch": 0.29335509890531974,
      "grad_norm": 0.009169838391244411,
      "learning_rate": 0.2627966826079303,
      "loss": 0.8255,
      "num_input_tokens_seen": 12606896,
      "step": 9165
    },
    {
      "epoch": 0.2935151398758082,
      "grad_norm": 0.006591183599084616,
      "learning_rate": 0.26275784451348216,
      "loss": 0.6572,
      "num_input_tokens_seen": 12613824,
      "step": 9170
    },
    {
      "epoch": 0.2936751808462967,
      "grad_norm": 0.006687216926366091,
      "learning_rate": 0.2627189890303574,
      "loss": 0.5805,
      "num_input_tokens_seen": 12620448,
      "step": 9175
    },
    {
      "epoch": 0.2938352218167851,
      "grad_norm": 0.007849445566534996,
      "learning_rate": 0.262680116164548,
      "loss": 0.6244,
      "num_input_tokens_seen": 12627232,
      "step": 9180
    },
    {
      "epoch": 0.29399526278727356,
      "grad_norm": 0.010281077586114407,
      "learning_rate": 0.2626412259220487,
      "loss": 0.6138,
      "num_input_tokens_seen": 12633792,
      "step": 9185
    },
    {
      "epoch": 0.294155303757762,
      "grad_norm": 0.010387124493718147,
      "learning_rate": 0.2626023183088568,
      "loss": 0.7687,
      "num_input_tokens_seen": 12640384,
      "step": 9190
    },
    {
      "epoch": 0.29431534472825044,
      "grad_norm": 0.00725965341553092,
      "learning_rate": 0.26256339333097234,
      "loss": 0.5591,
      "num_input_tokens_seen": 12647392,
      "step": 9195
    },
    {
      "epoch": 0.2944753856987389,
      "grad_norm": 0.009639243595302105,
      "learning_rate": 0.2625244509943981,
      "loss": 0.8597,
      "num_input_tokens_seen": 12654368,
      "step": 9200
    },
    {
      "epoch": 0.2944753856987389,
      "eval_loss": 0.6496866345405579,
      "eval_runtime": 332.3519,
      "eval_samples_per_second": 41.781,
      "eval_steps_per_second": 20.891,
      "num_input_tokens_seen": 12654368,
      "step": 9200
    },
    {
      "epoch": 0.2946354266692273,
      "grad_norm": 0.006671532057225704,
      "learning_rate": 0.2624854913051395,
      "loss": 0.6491,
      "num_input_tokens_seen": 12661088,
      "step": 9205
    },
    {
      "epoch": 0.29479546763971576,
      "grad_norm": 0.017344221472740173,
      "learning_rate": 0.26244651426920446,
      "loss": 0.6376,
      "num_input_tokens_seen": 12667888,
      "step": 9210
    },
    {
      "epoch": 0.2949555086102042,
      "grad_norm": 0.008152629248797894,
      "learning_rate": 0.26240751989260386,
      "loss": 0.6428,
      "num_input_tokens_seen": 12674832,
      "step": 9215
    },
    {
      "epoch": 0.29511554958069264,
      "grad_norm": 0.008266136050224304,
      "learning_rate": 0.2623685081813511,
      "loss": 0.5738,
      "num_input_tokens_seen": 12681792,
      "step": 9220
    },
    {
      "epoch": 0.2952755905511811,
      "grad_norm": 0.011235547251999378,
      "learning_rate": 0.2623294791414623,
      "loss": 0.6038,
      "num_input_tokens_seen": 12688544,
      "step": 9225
    },
    {
      "epoch": 0.2954356315216696,
      "grad_norm": 0.007345035206526518,
      "learning_rate": 0.26229043277895614,
      "loss": 0.4518,
      "num_input_tokens_seen": 12695312,
      "step": 9230
    },
    {
      "epoch": 0.295595672492158,
      "grad_norm": 0.00807307381182909,
      "learning_rate": 0.2622513690998542,
      "loss": 0.5511,
      "num_input_tokens_seen": 12701872,
      "step": 9235
    },
    {
      "epoch": 0.29575571346264645,
      "grad_norm": 0.005714510567486286,
      "learning_rate": 0.26221228811018044,
      "loss": 0.7358,
      "num_input_tokens_seen": 12708976,
      "step": 9240
    },
    {
      "epoch": 0.2959157544331349,
      "grad_norm": 0.009866871871054173,
      "learning_rate": 0.2621731898159617,
      "loss": 0.6272,
      "num_input_tokens_seen": 12715872,
      "step": 9245
    },
    {
      "epoch": 0.29607579540362333,
      "grad_norm": 0.006026131566613913,
      "learning_rate": 0.26213407422322743,
      "loss": 0.5691,
      "num_input_tokens_seen": 12723008,
      "step": 9250
    },
    {
      "epoch": 0.29623583637411177,
      "grad_norm": 0.006954412441700697,
      "learning_rate": 0.2620949413380098,
      "loss": 0.7482,
      "num_input_tokens_seen": 12729360,
      "step": 9255
    },
    {
      "epoch": 0.2963958773446002,
      "grad_norm": 0.005403397139161825,
      "learning_rate": 0.26205579116634353,
      "loss": 0.5075,
      "num_input_tokens_seen": 12736432,
      "step": 9260
    },
    {
      "epoch": 0.29655591831508865,
      "grad_norm": 0.0065576764754951,
      "learning_rate": 0.26201662371426604,
      "loss": 0.5071,
      "num_input_tokens_seen": 12742880,
      "step": 9265
    },
    {
      "epoch": 0.2967159592855771,
      "grad_norm": 0.007043034769594669,
      "learning_rate": 0.2619774389878175,
      "loss": 0.5453,
      "num_input_tokens_seen": 12749904,
      "step": 9270
    },
    {
      "epoch": 0.29687600025606553,
      "grad_norm": 0.006890338845551014,
      "learning_rate": 0.2619382369930407,
      "loss": 0.7608,
      "num_input_tokens_seen": 12756656,
      "step": 9275
    },
    {
      "epoch": 0.29703604122655397,
      "grad_norm": 0.00915397796779871,
      "learning_rate": 0.261899017735981,
      "loss": 0.6913,
      "num_input_tokens_seen": 12763216,
      "step": 9280
    },
    {
      "epoch": 0.29719608219704247,
      "grad_norm": 0.008495728485286236,
      "learning_rate": 0.2618597812226866,
      "loss": 0.7489,
      "num_input_tokens_seen": 12770256,
      "step": 9285
    },
    {
      "epoch": 0.2973561231675309,
      "grad_norm": 0.0037311904598027468,
      "learning_rate": 0.2618205274592082,
      "loss": 0.6665,
      "num_input_tokens_seen": 12776784,
      "step": 9290
    },
    {
      "epoch": 0.29751616413801935,
      "grad_norm": 0.008253427222371101,
      "learning_rate": 0.2617812564515992,
      "loss": 0.6377,
      "num_input_tokens_seen": 12783264,
      "step": 9295
    },
    {
      "epoch": 0.2976762051085078,
      "grad_norm": 0.010174769908189774,
      "learning_rate": 0.2617419682059158,
      "loss": 0.7443,
      "num_input_tokens_seen": 12790272,
      "step": 9300
    },
    {
      "epoch": 0.2978362460789962,
      "grad_norm": 0.010402833111584187,
      "learning_rate": 0.26170266272821663,
      "loss": 0.9262,
      "num_input_tokens_seen": 12797168,
      "step": 9305
    },
    {
      "epoch": 0.29799628704948466,
      "grad_norm": 0.00532355485484004,
      "learning_rate": 0.26166334002456315,
      "loss": 0.5507,
      "num_input_tokens_seen": 12803664,
      "step": 9310
    },
    {
      "epoch": 0.2981563280199731,
      "grad_norm": 0.004938360303640366,
      "learning_rate": 0.2616240001010194,
      "loss": 0.7496,
      "num_input_tokens_seen": 12810864,
      "step": 9315
    },
    {
      "epoch": 0.29831636899046154,
      "grad_norm": 0.007940770126879215,
      "learning_rate": 0.26158464296365197,
      "loss": 0.7059,
      "num_input_tokens_seen": 12817440,
      "step": 9320
    },
    {
      "epoch": 0.29847640996095,
      "grad_norm": 0.006776353809982538,
      "learning_rate": 0.2615452686185304,
      "loss": 0.5583,
      "num_input_tokens_seen": 12824096,
      "step": 9325
    },
    {
      "epoch": 0.2986364509314384,
      "grad_norm": 0.006312118377536535,
      "learning_rate": 0.26150587707172673,
      "loss": 0.6001,
      "num_input_tokens_seen": 12831088,
      "step": 9330
    },
    {
      "epoch": 0.2987964919019269,
      "grad_norm": 0.02032378688454628,
      "learning_rate": 0.2614664683293154,
      "loss": 0.7568,
      "num_input_tokens_seen": 12838176,
      "step": 9335
    },
    {
      "epoch": 0.29895653287241536,
      "grad_norm": 0.011339906603097916,
      "learning_rate": 0.26142704239737397,
      "loss": 0.7843,
      "num_input_tokens_seen": 12844800,
      "step": 9340
    },
    {
      "epoch": 0.2991165738429038,
      "grad_norm": 0.006106159184128046,
      "learning_rate": 0.26138759928198235,
      "loss": 0.7904,
      "num_input_tokens_seen": 12851696,
      "step": 9345
    },
    {
      "epoch": 0.29927661481339224,
      "grad_norm": 0.006344963796436787,
      "learning_rate": 0.26134813898922304,
      "loss": 0.4492,
      "num_input_tokens_seen": 12858224,
      "step": 9350
    },
    {
      "epoch": 0.2994366557838807,
      "grad_norm": 0.007833428680896759,
      "learning_rate": 0.26130866152518145,
      "loss": 0.4952,
      "num_input_tokens_seen": 12864896,
      "step": 9355
    },
    {
      "epoch": 0.2995966967543691,
      "grad_norm": 0.007490812335163355,
      "learning_rate": 0.2612691668959455,
      "loss": 0.5568,
      "num_input_tokens_seen": 12871456,
      "step": 9360
    },
    {
      "epoch": 0.29975673772485756,
      "grad_norm": 0.01156582124531269,
      "learning_rate": 0.2612296551076057,
      "loss": 0.7058,
      "num_input_tokens_seen": 12878144,
      "step": 9365
    },
    {
      "epoch": 0.299916778695346,
      "grad_norm": 0.004205952864140272,
      "learning_rate": 0.26119012616625525,
      "loss": 0.6315,
      "num_input_tokens_seen": 12885296,
      "step": 9370
    },
    {
      "epoch": 0.30007681966583444,
      "grad_norm": 0.011271764524281025,
      "learning_rate": 0.26115058007799,
      "loss": 0.7078,
      "num_input_tokens_seen": 12892160,
      "step": 9375
    },
    {
      "epoch": 0.3002368606363229,
      "grad_norm": 0.0058518061414361,
      "learning_rate": 0.26111101684890864,
      "loss": 0.4751,
      "num_input_tokens_seen": 12899520,
      "step": 9380
    },
    {
      "epoch": 0.3003969016068113,
      "grad_norm": 0.004784691147506237,
      "learning_rate": 0.26107143648511205,
      "loss": 0.4318,
      "num_input_tokens_seen": 12906752,
      "step": 9385
    },
    {
      "epoch": 0.3005569425772998,
      "grad_norm": 0.011727767065167427,
      "learning_rate": 0.2610318389927042,
      "loss": 0.5815,
      "num_input_tokens_seen": 12913632,
      "step": 9390
    },
    {
      "epoch": 0.30071698354778825,
      "grad_norm": 0.009229347109794617,
      "learning_rate": 0.26099222437779146,
      "loss": 0.5117,
      "num_input_tokens_seen": 12920528,
      "step": 9395
    },
    {
      "epoch": 0.3008770245182767,
      "grad_norm": 0.013130825012922287,
      "learning_rate": 0.26095259264648285,
      "loss": 0.6403,
      "num_input_tokens_seen": 12927088,
      "step": 9400
    },
    {
      "epoch": 0.3008770245182767,
      "eval_loss": 0.6553226709365845,
      "eval_runtime": 331.8569,
      "eval_samples_per_second": 41.843,
      "eval_steps_per_second": 20.922,
      "num_input_tokens_seen": 12927088,
      "step": 9400
    },
    {
      "epoch": 0.30103706548876513,
      "grad_norm": 0.007025111932307482,
      "learning_rate": 0.2609129438048902,
      "loss": 0.5073,
      "num_input_tokens_seen": 12933744,
      "step": 9405
    },
    {
      "epoch": 0.30119710645925357,
      "grad_norm": 0.007470875512808561,
      "learning_rate": 0.2608732778591278,
      "loss": 0.6955,
      "num_input_tokens_seen": 12940384,
      "step": 9410
    },
    {
      "epoch": 0.301357147429742,
      "grad_norm": 0.010535404086112976,
      "learning_rate": 0.2608335948153126,
      "loss": 0.6335,
      "num_input_tokens_seen": 12947584,
      "step": 9415
    },
    {
      "epoch": 0.30151718840023045,
      "grad_norm": 0.012496751733124256,
      "learning_rate": 0.26079389467956426,
      "loss": 0.6135,
      "num_input_tokens_seen": 12954384,
      "step": 9420
    },
    {
      "epoch": 0.3016772293707189,
      "grad_norm": 0.007619759067893028,
      "learning_rate": 0.26075417745800505,
      "loss": 0.4747,
      "num_input_tokens_seen": 12961328,
      "step": 9425
    },
    {
      "epoch": 0.30183727034120733,
      "grad_norm": 0.010600415989756584,
      "learning_rate": 0.26071444315675985,
      "loss": 0.72,
      "num_input_tokens_seen": 12968320,
      "step": 9430
    },
    {
      "epoch": 0.30199731131169577,
      "grad_norm": 0.007758009247481823,
      "learning_rate": 0.2606746917819562,
      "loss": 0.5445,
      "num_input_tokens_seen": 12974688,
      "step": 9435
    },
    {
      "epoch": 0.30215735228218427,
      "grad_norm": 0.008234265260398388,
      "learning_rate": 0.2606349233397242,
      "loss": 0.6299,
      "num_input_tokens_seen": 12981568,
      "step": 9440
    },
    {
      "epoch": 0.3023173932526727,
      "grad_norm": 0.0061437520198524,
      "learning_rate": 0.26059513783619676,
      "loss": 0.5017,
      "num_input_tokens_seen": 12988448,
      "step": 9445
    },
    {
      "epoch": 0.30247743422316115,
      "grad_norm": 0.008831211365759373,
      "learning_rate": 0.26055533527750924,
      "loss": 0.7482,
      "num_input_tokens_seen": 12995376,
      "step": 9450
    },
    {
      "epoch": 0.3026374751936496,
      "grad_norm": 0.009500599466264248,
      "learning_rate": 0.26051551566979964,
      "loss": 0.4675,
      "num_input_tokens_seen": 13002320,
      "step": 9455
    },
    {
      "epoch": 0.302797516164138,
      "grad_norm": 0.012512067332863808,
      "learning_rate": 0.26047567901920876,
      "loss": 0.5991,
      "num_input_tokens_seen": 13008880,
      "step": 9460
    },
    {
      "epoch": 0.30295755713462647,
      "grad_norm": 0.005864270962774754,
      "learning_rate": 0.2604358253318798,
      "loss": 0.6783,
      "num_input_tokens_seen": 13016000,
      "step": 9465
    },
    {
      "epoch": 0.3031175981051149,
      "grad_norm": 0.008900439366698265,
      "learning_rate": 0.26039595461395876,
      "loss": 0.6742,
      "num_input_tokens_seen": 13022944,
      "step": 9470
    },
    {
      "epoch": 0.30327763907560334,
      "grad_norm": 0.014293044805526733,
      "learning_rate": 0.26035606687159424,
      "loss": 0.8346,
      "num_input_tokens_seen": 13029776,
      "step": 9475
    },
    {
      "epoch": 0.3034376800460918,
      "grad_norm": 0.007364095188677311,
      "learning_rate": 0.26031616211093733,
      "loss": 0.6719,
      "num_input_tokens_seen": 13036608,
      "step": 9480
    },
    {
      "epoch": 0.3035977210165802,
      "grad_norm": 0.007874422706663609,
      "learning_rate": 0.26027624033814195,
      "loss": 0.566,
      "num_input_tokens_seen": 13043344,
      "step": 9485
    },
    {
      "epoch": 0.30375776198706866,
      "grad_norm": 0.004871288314461708,
      "learning_rate": 0.2602363015593645,
      "loss": 0.5438,
      "num_input_tokens_seen": 13050496,
      "step": 9490
    },
    {
      "epoch": 0.30391780295755716,
      "grad_norm": 0.007800579071044922,
      "learning_rate": 0.26019634578076395,
      "loss": 0.6455,
      "num_input_tokens_seen": 13057968,
      "step": 9495
    },
    {
      "epoch": 0.3040778439280456,
      "grad_norm": 0.006921196822077036,
      "learning_rate": 0.26015637300850214,
      "loss": 0.5965,
      "num_input_tokens_seen": 13064672,
      "step": 9500
    },
    {
      "epoch": 0.30423788489853404,
      "grad_norm": 0.004646347835659981,
      "learning_rate": 0.26011638324874325,
      "loss": 0.5852,
      "num_input_tokens_seen": 13071088,
      "step": 9505
    },
    {
      "epoch": 0.3043979258690225,
      "grad_norm": 0.007194059435278177,
      "learning_rate": 0.2600763765076543,
      "loss": 0.7399,
      "num_input_tokens_seen": 13077632,
      "step": 9510
    },
    {
      "epoch": 0.3045579668395109,
      "grad_norm": 0.010671357624232769,
      "learning_rate": 0.2600363527914048,
      "loss": 0.5439,
      "num_input_tokens_seen": 13084048,
      "step": 9515
    },
    {
      "epoch": 0.30471800780999936,
      "grad_norm": 0.006575864739716053,
      "learning_rate": 0.25999631210616686,
      "loss": 0.6315,
      "num_input_tokens_seen": 13090912,
      "step": 9520
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 0.009178603067994118,
      "learning_rate": 0.25995625445811527,
      "loss": 0.7597,
      "num_input_tokens_seen": 13097872,
      "step": 9525
    },
    {
      "epoch": 0.30503808975097624,
      "grad_norm": 0.005516283214092255,
      "learning_rate": 0.2599161798534275,
      "loss": 0.5134,
      "num_input_tokens_seen": 13104944,
      "step": 9530
    },
    {
      "epoch": 0.3051981307214647,
      "grad_norm": 0.009142724797129631,
      "learning_rate": 0.25987608829828346,
      "loss": 0.6461,
      "num_input_tokens_seen": 13111264,
      "step": 9535
    },
    {
      "epoch": 0.3053581716919531,
      "grad_norm": 0.011842687614262104,
      "learning_rate": 0.25983597979886586,
      "loss": 0.5029,
      "num_input_tokens_seen": 13117936,
      "step": 9540
    },
    {
      "epoch": 0.3055182126624416,
      "grad_norm": 0.007930971682071686,
      "learning_rate": 0.2597958543613599,
      "loss": 0.6868,
      "num_input_tokens_seen": 13124736,
      "step": 9545
    },
    {
      "epoch": 0.30567825363293005,
      "grad_norm": 0.007129996083676815,
      "learning_rate": 0.25975571199195335,
      "loss": 0.4428,
      "num_input_tokens_seen": 13131520,
      "step": 9550
    },
    {
      "epoch": 0.3058382946034185,
      "grad_norm": 0.012857728637754917,
      "learning_rate": 0.25971555269683677,
      "loss": 0.5687,
      "num_input_tokens_seen": 13138336,
      "step": 9555
    },
    {
      "epoch": 0.30599833557390693,
      "grad_norm": 0.021578313782811165,
      "learning_rate": 0.25967537648220324,
      "loss": 0.8178,
      "num_input_tokens_seen": 13145360,
      "step": 9560
    },
    {
      "epoch": 0.30615837654439537,
      "grad_norm": 0.006078542210161686,
      "learning_rate": 0.2596351833542483,
      "loss": 0.5367,
      "num_input_tokens_seen": 13152448,
      "step": 9565
    },
    {
      "epoch": 0.3063184175148838,
      "grad_norm": 0.00612006988376379,
      "learning_rate": 0.25959497331917036,
      "loss": 0.5451,
      "num_input_tokens_seen": 13159280,
      "step": 9570
    },
    {
      "epoch": 0.30647845848537225,
      "grad_norm": 0.011780543252825737,
      "learning_rate": 0.2595547463831703,
      "loss": 0.8536,
      "num_input_tokens_seen": 13165616,
      "step": 9575
    },
    {
      "epoch": 0.3066384994558607,
      "grad_norm": 0.01077545527368784,
      "learning_rate": 0.25951450255245156,
      "loss": 0.7894,
      "num_input_tokens_seen": 13172016,
      "step": 9580
    },
    {
      "epoch": 0.30679854042634913,
      "grad_norm": 0.004775175359100103,
      "learning_rate": 0.2594742418332203,
      "loss": 0.7685,
      "num_input_tokens_seen": 13178848,
      "step": 9585
    },
    {
      "epoch": 0.30695858139683757,
      "grad_norm": 0.004219943657517433,
      "learning_rate": 0.2594339642316852,
      "loss": 0.6391,
      "num_input_tokens_seen": 13185600,
      "step": 9590
    },
    {
      "epoch": 0.307118622367326,
      "grad_norm": 0.009908239357173443,
      "learning_rate": 0.2593936697540576,
      "loss": 0.5674,
      "num_input_tokens_seen": 13192720,
      "step": 9595
    },
    {
      "epoch": 0.3072786633378145,
      "grad_norm": 0.012665347196161747,
      "learning_rate": 0.2593533584065514,
      "loss": 0.6487,
      "num_input_tokens_seen": 13199552,
      "step": 9600
    },
    {
      "epoch": 0.3072786633378145,
      "eval_loss": 0.6483891010284424,
      "eval_runtime": 332.6514,
      "eval_samples_per_second": 41.743,
      "eval_steps_per_second": 20.872,
      "num_input_tokens_seen": 13199552,
      "step": 9600
    },
    {
      "epoch": 0.30743870430830295,
      "grad_norm": 0.009091125801205635,
      "learning_rate": 0.2593130301953831,
      "loss": 0.5491,
      "num_input_tokens_seen": 13206208,
      "step": 9605
    },
    {
      "epoch": 0.3075987452787914,
      "grad_norm": 0.009263678453862667,
      "learning_rate": 0.2592726851267718,
      "loss": 0.6752,
      "num_input_tokens_seen": 13213296,
      "step": 9610
    },
    {
      "epoch": 0.3077587862492798,
      "grad_norm": 0.014823497273027897,
      "learning_rate": 0.2592323232069393,
      "loss": 0.5916,
      "num_input_tokens_seen": 13219920,
      "step": 9615
    },
    {
      "epoch": 0.30791882721976827,
      "grad_norm": 0.008301733061671257,
      "learning_rate": 0.25919194444210986,
      "loss": 0.7011,
      "num_input_tokens_seen": 13226816,
      "step": 9620
    },
    {
      "epoch": 0.3080788681902567,
      "grad_norm": 0.005598084069788456,
      "learning_rate": 0.2591515488385103,
      "loss": 0.6117,
      "num_input_tokens_seen": 13233664,
      "step": 9625
    },
    {
      "epoch": 0.30823890916074514,
      "grad_norm": 0.014659088104963303,
      "learning_rate": 0.2591111364023704,
      "loss": 0.6543,
      "num_input_tokens_seen": 13240160,
      "step": 9630
    },
    {
      "epoch": 0.3083989501312336,
      "grad_norm": 0.00903757568448782,
      "learning_rate": 0.259070707139922,
      "loss": 0.6199,
      "num_input_tokens_seen": 13247344,
      "step": 9635
    },
    {
      "epoch": 0.308558991101722,
      "grad_norm": 0.008966974914073944,
      "learning_rate": 0.25903026105739985,
      "loss": 0.5094,
      "num_input_tokens_seen": 13254448,
      "step": 9640
    },
    {
      "epoch": 0.30871903207221046,
      "grad_norm": 0.0057980893179774284,
      "learning_rate": 0.2589897981610413,
      "loss": 0.4736,
      "num_input_tokens_seen": 13261216,
      "step": 9645
    },
    {
      "epoch": 0.3088790730426989,
      "grad_norm": 0.013411333784461021,
      "learning_rate": 0.2589493184570863,
      "loss": 0.6499,
      "num_input_tokens_seen": 13268272,
      "step": 9650
    },
    {
      "epoch": 0.3090391140131874,
      "grad_norm": 0.011493904516100883,
      "learning_rate": 0.25890882195177717,
      "loss": 0.4193,
      "num_input_tokens_seen": 13275296,
      "step": 9655
    },
    {
      "epoch": 0.30919915498367584,
      "grad_norm": 0.01519360113888979,
      "learning_rate": 0.25886830865135907,
      "loss": 0.6616,
      "num_input_tokens_seen": 13282240,
      "step": 9660
    },
    {
      "epoch": 0.3093591959541643,
      "grad_norm": 0.007929768413305283,
      "learning_rate": 0.25882777856207967,
      "loss": 0.8337,
      "num_input_tokens_seen": 13289808,
      "step": 9665
    },
    {
      "epoch": 0.3095192369246527,
      "grad_norm": 0.014123025350272655,
      "learning_rate": 0.2587872316901892,
      "loss": 0.62,
      "num_input_tokens_seen": 13296800,
      "step": 9670
    },
    {
      "epoch": 0.30967927789514116,
      "grad_norm": 0.006511315703392029,
      "learning_rate": 0.25874666804194046,
      "loss": 0.5723,
      "num_input_tokens_seen": 13303792,
      "step": 9675
    },
    {
      "epoch": 0.3098393188656296,
      "grad_norm": 0.007617831230163574,
      "learning_rate": 0.258706087623589,
      "loss": 0.6187,
      "num_input_tokens_seen": 13310368,
      "step": 9680
    },
    {
      "epoch": 0.30999935983611804,
      "grad_norm": 0.007815046235918999,
      "learning_rate": 0.25866549044139264,
      "loss": 0.5352,
      "num_input_tokens_seen": 13316928,
      "step": 9685
    },
    {
      "epoch": 0.3101594008066065,
      "grad_norm": 0.008382990956306458,
      "learning_rate": 0.25862487650161214,
      "loss": 0.8462,
      "num_input_tokens_seen": 13323840,
      "step": 9690
    },
    {
      "epoch": 0.3103194417770949,
      "grad_norm": 0.011508084833621979,
      "learning_rate": 0.2585842458105106,
      "loss": 0.5716,
      "num_input_tokens_seen": 13330704,
      "step": 9695
    },
    {
      "epoch": 0.31047948274758336,
      "grad_norm": 0.010409809648990631,
      "learning_rate": 0.2585435983743538,
      "loss": 0.6767,
      "num_input_tokens_seen": 13337472,
      "step": 9700
    },
    {
      "epoch": 0.31063952371807185,
      "grad_norm": 0.0038585823494940996,
      "learning_rate": 0.2585029341994101,
      "loss": 0.4972,
      "num_input_tokens_seen": 13344320,
      "step": 9705
    },
    {
      "epoch": 0.3107995646885603,
      "grad_norm": 0.009472579695284367,
      "learning_rate": 0.2584622532919504,
      "loss": 0.6702,
      "num_input_tokens_seen": 13351072,
      "step": 9710
    },
    {
      "epoch": 0.31095960565904873,
      "grad_norm": 0.008474478498101234,
      "learning_rate": 0.2584215556582482,
      "loss": 0.724,
      "num_input_tokens_seen": 13358016,
      "step": 9715
    },
    {
      "epoch": 0.31111964662953717,
      "grad_norm": 0.010508954524993896,
      "learning_rate": 0.25838084130457967,
      "loss": 0.4825,
      "num_input_tokens_seen": 13364704,
      "step": 9720
    },
    {
      "epoch": 0.3112796876000256,
      "grad_norm": 0.00853115227073431,
      "learning_rate": 0.2583401102372234,
      "loss": 0.6712,
      "num_input_tokens_seen": 13371456,
      "step": 9725
    },
    {
      "epoch": 0.31143972857051405,
      "grad_norm": 0.005021542776376009,
      "learning_rate": 0.2582993624624606,
      "loss": 0.6213,
      "num_input_tokens_seen": 13378672,
      "step": 9730
    },
    {
      "epoch": 0.3115997695410025,
      "grad_norm": 0.00876977201551199,
      "learning_rate": 0.25825859798657513,
      "loss": 0.6657,
      "num_input_tokens_seen": 13385584,
      "step": 9735
    },
    {
      "epoch": 0.31175981051149093,
      "grad_norm": 0.00777070876210928,
      "learning_rate": 0.25821781681585343,
      "loss": 0.5874,
      "num_input_tokens_seen": 13392608,
      "step": 9740
    },
    {
      "epoch": 0.31191985148197937,
      "grad_norm": 0.008551366627216339,
      "learning_rate": 0.2581770189565844,
      "loss": 0.7019,
      "num_input_tokens_seen": 13399424,
      "step": 9745
    },
    {
      "epoch": 0.3120798924524678,
      "grad_norm": 0.00620243139564991,
      "learning_rate": 0.25813620441505963,
      "loss": 0.5295,
      "num_input_tokens_seen": 13405952,
      "step": 9750
    },
    {
      "epoch": 0.31223993342295625,
      "grad_norm": 0.007175668608397245,
      "learning_rate": 0.2580953731975732,
      "loss": 0.6389,
      "num_input_tokens_seen": 13412560,
      "step": 9755
    },
    {
      "epoch": 0.31239997439344475,
      "grad_norm": 0.007262269966304302,
      "learning_rate": 0.2580545253104218,
      "loss": 0.6578,
      "num_input_tokens_seen": 13419600,
      "step": 9760
    },
    {
      "epoch": 0.3125600153639332,
      "grad_norm": 0.007814034819602966,
      "learning_rate": 0.2580136607599047,
      "loss": 0.5848,
      "num_input_tokens_seen": 13426560,
      "step": 9765
    },
    {
      "epoch": 0.3127200563344216,
      "grad_norm": 0.006538683082908392,
      "learning_rate": 0.2579727795523238,
      "loss": 0.6023,
      "num_input_tokens_seen": 13433392,
      "step": 9770
    },
    {
      "epoch": 0.31288009730491007,
      "grad_norm": 0.0066025941632688046,
      "learning_rate": 0.25793188169398334,
      "loss": 0.5645,
      "num_input_tokens_seen": 13440352,
      "step": 9775
    },
    {
      "epoch": 0.3130401382753985,
      "grad_norm": 0.013621504418551922,
      "learning_rate": 0.25789096719119037,
      "loss": 0.8335,
      "num_input_tokens_seen": 13447008,
      "step": 9780
    },
    {
      "epoch": 0.31320017924588694,
      "grad_norm": 0.008214623667299747,
      "learning_rate": 0.2578500360502544,
      "loss": 0.7027,
      "num_input_tokens_seen": 13453856,
      "step": 9785
    },
    {
      "epoch": 0.3133602202163754,
      "grad_norm": 0.008625369518995285,
      "learning_rate": 0.2578090882774876,
      "loss": 0.5898,
      "num_input_tokens_seen": 13460720,
      "step": 9790
    },
    {
      "epoch": 0.3135202611868638,
      "grad_norm": 0.010528559796512127,
      "learning_rate": 0.25776812387920456,
      "loss": 0.7336,
      "num_input_tokens_seen": 13467280,
      "step": 9795
    },
    {
      "epoch": 0.31368030215735226,
      "grad_norm": 0.008844278752803802,
      "learning_rate": 0.2577271428617225,
      "loss": 0.5817,
      "num_input_tokens_seen": 13473952,
      "step": 9800
    },
    {
      "epoch": 0.31368030215735226,
      "eval_loss": 0.6354929208755493,
      "eval_runtime": 332.5478,
      "eval_samples_per_second": 41.756,
      "eval_steps_per_second": 20.878,
      "num_input_tokens_seen": 13473952,
      "step": 9800
    },
    {
      "epoch": 0.3138403431278407,
      "grad_norm": 0.008222613483667374,
      "learning_rate": 0.25768614523136124,
      "loss": 0.7121,
      "num_input_tokens_seen": 13481168,
      "step": 9805
    },
    {
      "epoch": 0.3140003840983292,
      "grad_norm": 0.011119064874947071,
      "learning_rate": 0.25764513099444314,
      "loss": 0.4701,
      "num_input_tokens_seen": 13488240,
      "step": 9810
    },
    {
      "epoch": 0.31416042506881764,
      "grad_norm": 0.007410811260342598,
      "learning_rate": 0.25760410015729307,
      "loss": 0.6223,
      "num_input_tokens_seen": 13495072,
      "step": 9815
    },
    {
      "epoch": 0.3143204660393061,
      "grad_norm": 0.005890044383704662,
      "learning_rate": 0.2575630527262385,
      "loss": 0.4571,
      "num_input_tokens_seen": 13501760,
      "step": 9820
    },
    {
      "epoch": 0.3144805070097945,
      "grad_norm": 0.008691269904375076,
      "learning_rate": 0.25752198870760945,
      "loss": 0.7421,
      "num_input_tokens_seen": 13508304,
      "step": 9825
    },
    {
      "epoch": 0.31464054798028296,
      "grad_norm": 0.008985619060695171,
      "learning_rate": 0.2574809081077386,
      "loss": 0.5895,
      "num_input_tokens_seen": 13514816,
      "step": 9830
    },
    {
      "epoch": 0.3148005889507714,
      "grad_norm": 0.00778541574254632,
      "learning_rate": 0.257439810932961,
      "loss": 0.5884,
      "num_input_tokens_seen": 13521472,
      "step": 9835
    },
    {
      "epoch": 0.31496062992125984,
      "grad_norm": 0.005007551517337561,
      "learning_rate": 0.2573986971896144,
      "loss": 0.4822,
      "num_input_tokens_seen": 13528032,
      "step": 9840
    },
    {
      "epoch": 0.3151206708917483,
      "grad_norm": 0.007291983813047409,
      "learning_rate": 0.257357566884039,
      "loss": 0.7999,
      "num_input_tokens_seen": 13534800,
      "step": 9845
    },
    {
      "epoch": 0.3152807118622367,
      "grad_norm": 0.004570115823298693,
      "learning_rate": 0.25731642002257765,
      "loss": 0.4983,
      "num_input_tokens_seen": 13541360,
      "step": 9850
    },
    {
      "epoch": 0.31544075283272516,
      "grad_norm": 0.009505018591880798,
      "learning_rate": 0.25727525661157574,
      "loss": 0.6652,
      "num_input_tokens_seen": 13548880,
      "step": 9855
    },
    {
      "epoch": 0.3156007938032136,
      "grad_norm": 0.012118065729737282,
      "learning_rate": 0.2572340766573811,
      "loss": 0.5963,
      "num_input_tokens_seen": 13556080,
      "step": 9860
    },
    {
      "epoch": 0.3157608347737021,
      "grad_norm": 0.008503141812980175,
      "learning_rate": 0.25719288016634434,
      "loss": 0.7484,
      "num_input_tokens_seen": 13562992,
      "step": 9865
    },
    {
      "epoch": 0.31592087574419053,
      "grad_norm": 0.007541250437498093,
      "learning_rate": 0.25715166714481835,
      "loss": 0.7611,
      "num_input_tokens_seen": 13569984,
      "step": 9870
    },
    {
      "epoch": 0.31608091671467897,
      "grad_norm": 0.00936080887913704,
      "learning_rate": 0.2571104375991587,
      "loss": 0.813,
      "num_input_tokens_seen": 13576784,
      "step": 9875
    },
    {
      "epoch": 0.3162409576851674,
      "grad_norm": 0.005093112122267485,
      "learning_rate": 0.2570691915357236,
      "loss": 0.5395,
      "num_input_tokens_seen": 13583712,
      "step": 9880
    },
    {
      "epoch": 0.31640099865565585,
      "grad_norm": 0.008055702783167362,
      "learning_rate": 0.2570279289608736,
      "loss": 0.7021,
      "num_input_tokens_seen": 13590656,
      "step": 9885
    },
    {
      "epoch": 0.3165610396261443,
      "grad_norm": 0.012293040752410889,
      "learning_rate": 0.256986649880972,
      "loss": 0.5664,
      "num_input_tokens_seen": 13598000,
      "step": 9890
    },
    {
      "epoch": 0.31672108059663273,
      "grad_norm": 0.004459221847355366,
      "learning_rate": 0.25694535430238447,
      "loss": 0.4812,
      "num_input_tokens_seen": 13605456,
      "step": 9895
    },
    {
      "epoch": 0.31688112156712117,
      "grad_norm": 0.006278933957219124,
      "learning_rate": 0.25690404223147933,
      "loss": 0.4849,
      "num_input_tokens_seen": 13611968,
      "step": 9900
    },
    {
      "epoch": 0.3170411625376096,
      "grad_norm": 0.006977429147809744,
      "learning_rate": 0.2568627136746275,
      "loss": 0.5608,
      "num_input_tokens_seen": 13618864,
      "step": 9905
    },
    {
      "epoch": 0.31720120350809805,
      "grad_norm": 0.008683331310749054,
      "learning_rate": 0.25682136863820226,
      "loss": 0.6842,
      "num_input_tokens_seen": 13625680,
      "step": 9910
    },
    {
      "epoch": 0.3173612444785865,
      "grad_norm": 0.008243151940405369,
      "learning_rate": 0.25678000712857957,
      "loss": 0.5559,
      "num_input_tokens_seen": 13632928,
      "step": 9915
    },
    {
      "epoch": 0.317521285449075,
      "grad_norm": 0.005419593304395676,
      "learning_rate": 0.2567386291521379,
      "loss": 0.508,
      "num_input_tokens_seen": 13640512,
      "step": 9920
    },
    {
      "epoch": 0.3176813264195634,
      "grad_norm": 0.006983200553804636,
      "learning_rate": 0.2566972347152583,
      "loss": 0.5724,
      "num_input_tokens_seen": 13647376,
      "step": 9925
    },
    {
      "epoch": 0.31784136739005187,
      "grad_norm": 0.006551362108439207,
      "learning_rate": 0.2566558238243242,
      "loss": 0.5403,
      "num_input_tokens_seen": 13654352,
      "step": 9930
    },
    {
      "epoch": 0.3180014083605403,
      "grad_norm": 0.00721052335575223,
      "learning_rate": 0.25661439648572176,
      "loss": 0.4821,
      "num_input_tokens_seen": 13661424,
      "step": 9935
    },
    {
      "epoch": 0.31816144933102875,
      "grad_norm": 0.009668475948274136,
      "learning_rate": 0.25657295270583963,
      "loss": 0.6156,
      "num_input_tokens_seen": 13667984,
      "step": 9940
    },
    {
      "epoch": 0.3183214903015172,
      "grad_norm": 0.006838646717369556,
      "learning_rate": 0.25653149249106894,
      "loss": 0.5725,
      "num_input_tokens_seen": 13674608,
      "step": 9945
    },
    {
      "epoch": 0.3184815312720056,
      "grad_norm": 0.006915644742548466,
      "learning_rate": 0.25649001584780323,
      "loss": 0.6236,
      "num_input_tokens_seen": 13681536,
      "step": 9950
    },
    {
      "epoch": 0.31864157224249406,
      "grad_norm": 0.006315354723483324,
      "learning_rate": 0.2564485227824389,
      "loss": 0.739,
      "num_input_tokens_seen": 13688576,
      "step": 9955
    },
    {
      "epoch": 0.3188016132129825,
      "grad_norm": 0.010083982720971107,
      "learning_rate": 0.25640701330137466,
      "loss": 0.6315,
      "num_input_tokens_seen": 13695488,
      "step": 9960
    },
    {
      "epoch": 0.31896165418347094,
      "grad_norm": 0.01721445471048355,
      "learning_rate": 0.2563654874110117,
      "loss": 0.6613,
      "num_input_tokens_seen": 13702352,
      "step": 9965
    },
    {
      "epoch": 0.31912169515395944,
      "grad_norm": 0.009850578382611275,
      "learning_rate": 0.256323945117754,
      "loss": 0.6541,
      "num_input_tokens_seen": 13709456,
      "step": 9970
    },
    {
      "epoch": 0.3192817361244479,
      "grad_norm": 0.010721787810325623,
      "learning_rate": 0.2562823864280078,
      "loss": 0.6763,
      "num_input_tokens_seen": 13716256,
      "step": 9975
    },
    {
      "epoch": 0.3194417770949363,
      "grad_norm": 0.006811526603996754,
      "learning_rate": 0.25624081134818194,
      "loss": 0.5283,
      "num_input_tokens_seen": 13723296,
      "step": 9980
    },
    {
      "epoch": 0.31960181806542476,
      "grad_norm": 0.009035428054630756,
      "learning_rate": 0.2561992198846879,
      "loss": 0.7061,
      "num_input_tokens_seen": 13729808,
      "step": 9985
    },
    {
      "epoch": 0.3197618590359132,
      "grad_norm": 0.00625214260071516,
      "learning_rate": 0.25615761204393955,
      "loss": 0.825,
      "num_input_tokens_seen": 13736800,
      "step": 9990
    },
    {
      "epoch": 0.31992190000640164,
      "grad_norm": 0.0032553395722061396,
      "learning_rate": 0.2561159878323534,
      "loss": 0.5203,
      "num_input_tokens_seen": 13743792,
      "step": 9995
    },
    {
      "epoch": 0.3200819409768901,
      "grad_norm": 0.006508545484393835,
      "learning_rate": 0.2560743472563483,
      "loss": 0.6924,
      "num_input_tokens_seen": 13750288,
      "step": 10000
    },
    {
      "epoch": 0.3200819409768901,
      "eval_loss": 0.6558617949485779,
      "eval_runtime": 332.2164,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 13750288,
      "step": 10000
    },
    {
      "epoch": 0.3202419819473785,
      "grad_norm": 0.006248945370316505,
      "learning_rate": 0.25603269032234593,
      "loss": 0.5811,
      "num_input_tokens_seen": 13757184,
      "step": 10005
    },
    {
      "epoch": 0.32040202291786696,
      "grad_norm": 0.006959843914955854,
      "learning_rate": 0.2559910170367702,
      "loss": 0.6176,
      "num_input_tokens_seen": 13763920,
      "step": 10010
    },
    {
      "epoch": 0.3205620638883554,
      "grad_norm": 0.006639881059527397,
      "learning_rate": 0.2559493274060477,
      "loss": 0.5784,
      "num_input_tokens_seen": 13771056,
      "step": 10015
    },
    {
      "epoch": 0.32072210485884384,
      "grad_norm": 0.016249485313892365,
      "learning_rate": 0.2559076214366074,
      "loss": 0.4259,
      "num_input_tokens_seen": 13778112,
      "step": 10020
    },
    {
      "epoch": 0.32088214582933233,
      "grad_norm": 0.009472782723605633,
      "learning_rate": 0.25586589913488106,
      "loss": 0.6076,
      "num_input_tokens_seen": 13785184,
      "step": 10025
    },
    {
      "epoch": 0.3210421867998208,
      "grad_norm": 0.00900257844477892,
      "learning_rate": 0.2558241605073026,
      "loss": 0.7609,
      "num_input_tokens_seen": 13791968,
      "step": 10030
    },
    {
      "epoch": 0.3212022277703092,
      "grad_norm": 0.007221254054456949,
      "learning_rate": 0.25578240556030873,
      "loss": 0.6354,
      "num_input_tokens_seen": 13799552,
      "step": 10035
    },
    {
      "epoch": 0.32136226874079765,
      "grad_norm": 0.0047965575940907,
      "learning_rate": 0.2557406343003386,
      "loss": 0.4894,
      "num_input_tokens_seen": 13806224,
      "step": 10040
    },
    {
      "epoch": 0.3215223097112861,
      "grad_norm": 0.008966987021267414,
      "learning_rate": 0.25569884673383375,
      "loss": 0.6588,
      "num_input_tokens_seen": 13813664,
      "step": 10045
    },
    {
      "epoch": 0.32168235068177453,
      "grad_norm": 0.007435140665620565,
      "learning_rate": 0.25565704286723856,
      "loss": 0.6268,
      "num_input_tokens_seen": 13820256,
      "step": 10050
    },
    {
      "epoch": 0.32184239165226297,
      "grad_norm": 0.012483682483434677,
      "learning_rate": 0.25561522270699955,
      "loss": 0.6501,
      "num_input_tokens_seen": 13826736,
      "step": 10055
    },
    {
      "epoch": 0.3220024326227514,
      "grad_norm": 0.0058286068961024284,
      "learning_rate": 0.25557338625956594,
      "loss": 0.6692,
      "num_input_tokens_seen": 13833440,
      "step": 10060
    },
    {
      "epoch": 0.32216247359323985,
      "grad_norm": 0.009110119193792343,
      "learning_rate": 0.25553153353138947,
      "loss": 0.6391,
      "num_input_tokens_seen": 13840288,
      "step": 10065
    },
    {
      "epoch": 0.3223225145637283,
      "grad_norm": 0.01128315832465887,
      "learning_rate": 0.2554896645289243,
      "loss": 0.5927,
      "num_input_tokens_seen": 13847328,
      "step": 10070
    },
    {
      "epoch": 0.3224825555342168,
      "grad_norm": 0.007110651582479477,
      "learning_rate": 0.2554477792586272,
      "loss": 0.7958,
      "num_input_tokens_seen": 13853952,
      "step": 10075
    },
    {
      "epoch": 0.3226425965047052,
      "grad_norm": 0.0036052283830940723,
      "learning_rate": 0.25540587772695744,
      "loss": 0.4951,
      "num_input_tokens_seen": 13860816,
      "step": 10080
    },
    {
      "epoch": 0.32280263747519367,
      "grad_norm": 0.005467899143695831,
      "learning_rate": 0.2553639599403767,
      "loss": 0.6402,
      "num_input_tokens_seen": 13867440,
      "step": 10085
    },
    {
      "epoch": 0.3229626784456821,
      "grad_norm": 0.008776753209531307,
      "learning_rate": 0.2553220259053493,
      "loss": 0.8625,
      "num_input_tokens_seen": 13874016,
      "step": 10090
    },
    {
      "epoch": 0.32312271941617055,
      "grad_norm": 0.006992165464907885,
      "learning_rate": 0.2552800756283419,
      "loss": 0.7771,
      "num_input_tokens_seen": 13880768,
      "step": 10095
    },
    {
      "epoch": 0.323282760386659,
      "grad_norm": 0.007477920968085527,
      "learning_rate": 0.25523810911582373,
      "loss": 0.7222,
      "num_input_tokens_seen": 13887984,
      "step": 10100
    },
    {
      "epoch": 0.3234428013571474,
      "grad_norm": 0.0029435434844344854,
      "learning_rate": 0.25519612637426675,
      "loss": 0.5622,
      "num_input_tokens_seen": 13894656,
      "step": 10105
    },
    {
      "epoch": 0.32360284232763586,
      "grad_norm": 0.004549854900687933,
      "learning_rate": 0.25515412741014504,
      "loss": 0.6709,
      "num_input_tokens_seen": 13901968,
      "step": 10110
    },
    {
      "epoch": 0.3237628832981243,
      "grad_norm": 0.007830461487174034,
      "learning_rate": 0.2551121122299355,
      "loss": 0.5954,
      "num_input_tokens_seen": 13908896,
      "step": 10115
    },
    {
      "epoch": 0.32392292426861274,
      "grad_norm": 0.010749570094048977,
      "learning_rate": 0.2550700808401173,
      "loss": 0.7563,
      "num_input_tokens_seen": 13915504,
      "step": 10120
    },
    {
      "epoch": 0.3240829652391012,
      "grad_norm": 0.005418696906417608,
      "learning_rate": 0.2550280332471722,
      "loss": 0.5028,
      "num_input_tokens_seen": 13922240,
      "step": 10125
    },
    {
      "epoch": 0.3242430062095897,
      "grad_norm": 0.008235369808971882,
      "learning_rate": 0.2549859694575845,
      "loss": 0.7436,
      "num_input_tokens_seen": 13929312,
      "step": 10130
    },
    {
      "epoch": 0.3244030471800781,
      "grad_norm": 0.004924540873616934,
      "learning_rate": 0.254943889477841,
      "loss": 0.6989,
      "num_input_tokens_seen": 13935872,
      "step": 10135
    },
    {
      "epoch": 0.32456308815056656,
      "grad_norm": 0.006124699488282204,
      "learning_rate": 0.25490179331443097,
      "loss": 0.5982,
      "num_input_tokens_seen": 13942928,
      "step": 10140
    },
    {
      "epoch": 0.324723129121055,
      "grad_norm": 0.005419937428086996,
      "learning_rate": 0.25485968097384615,
      "loss": 0.6654,
      "num_input_tokens_seen": 13949680,
      "step": 10145
    },
    {
      "epoch": 0.32488317009154344,
      "grad_norm": 0.0030400168616324663,
      "learning_rate": 0.25481755246258075,
      "loss": 0.5145,
      "num_input_tokens_seen": 13956208,
      "step": 10150
    },
    {
      "epoch": 0.3250432110620319,
      "grad_norm": 0.008398096077144146,
      "learning_rate": 0.2547754077871315,
      "loss": 0.459,
      "num_input_tokens_seen": 13962992,
      "step": 10155
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 0.009333460591733456,
      "learning_rate": 0.25473324695399774,
      "loss": 0.5085,
      "num_input_tokens_seen": 13969680,
      "step": 10160
    },
    {
      "epoch": 0.32536329300300876,
      "grad_norm": 0.0067320819944143295,
      "learning_rate": 0.25469106996968105,
      "loss": 0.6618,
      "num_input_tokens_seen": 13976592,
      "step": 10165
    },
    {
      "epoch": 0.3255233339734972,
      "grad_norm": 0.01088318694382906,
      "learning_rate": 0.2546488768406858,
      "loss": 0.6713,
      "num_input_tokens_seen": 13983568,
      "step": 10170
    },
    {
      "epoch": 0.32568337494398564,
      "grad_norm": 0.005121106281876564,
      "learning_rate": 0.25460666757351863,
      "loss": 0.7225,
      "num_input_tokens_seen": 13990928,
      "step": 10175
    },
    {
      "epoch": 0.32584341591447413,
      "grad_norm": 0.005411363672465086,
      "learning_rate": 0.25456444217468877,
      "loss": 0.5372,
      "num_input_tokens_seen": 13997776,
      "step": 10180
    },
    {
      "epoch": 0.3260034568849626,
      "grad_norm": 0.003512304276227951,
      "learning_rate": 0.25452220065070785,
      "loss": 0.646,
      "num_input_tokens_seen": 14004464,
      "step": 10185
    },
    {
      "epoch": 0.326163497855451,
      "grad_norm": 0.007218777667731047,
      "learning_rate": 0.2544799430080901,
      "loss": 0.8858,
      "num_input_tokens_seen": 14011520,
      "step": 10190
    },
    {
      "epoch": 0.32632353882593945,
      "grad_norm": 0.003718978026881814,
      "learning_rate": 0.2544376692533522,
      "loss": 0.5219,
      "num_input_tokens_seen": 14018464,
      "step": 10195
    },
    {
      "epoch": 0.3264835797964279,
      "grad_norm": 0.004668921232223511,
      "learning_rate": 0.2543953793930132,
      "loss": 0.5349,
      "num_input_tokens_seen": 14025248,
      "step": 10200
    },
    {
      "epoch": 0.3264835797964279,
      "eval_loss": 0.6239352226257324,
      "eval_runtime": 332.401,
      "eval_samples_per_second": 41.775,
      "eval_steps_per_second": 20.887,
      "num_input_tokens_seen": 14025248,
      "step": 10200
    },
    {
      "epoch": 0.32664362076691633,
      "grad_norm": 0.006586578208953142,
      "learning_rate": 0.2543530734335948,
      "loss": 0.6399,
      "num_input_tokens_seen": 14032096,
      "step": 10205
    },
    {
      "epoch": 0.32680366173740477,
      "grad_norm": 0.011735272593796253,
      "learning_rate": 0.2543107513816211,
      "loss": 0.8908,
      "num_input_tokens_seen": 14038944,
      "step": 10210
    },
    {
      "epoch": 0.3269637027078932,
      "grad_norm": 0.005525809712707996,
      "learning_rate": 0.25426841324361865,
      "loss": 0.509,
      "num_input_tokens_seen": 14045920,
      "step": 10215
    },
    {
      "epoch": 0.32712374367838165,
      "grad_norm": 0.006308292038738728,
      "learning_rate": 0.2542260590261166,
      "loss": 0.4489,
      "num_input_tokens_seen": 14052992,
      "step": 10220
    },
    {
      "epoch": 0.3272837846488701,
      "grad_norm": 0.008176153525710106,
      "learning_rate": 0.2541836887356465,
      "loss": 0.7614,
      "num_input_tokens_seen": 14059632,
      "step": 10225
    },
    {
      "epoch": 0.32744382561935853,
      "grad_norm": 0.014002848416566849,
      "learning_rate": 0.2541413023787423,
      "loss": 0.7002,
      "num_input_tokens_seen": 14066080,
      "step": 10230
    },
    {
      "epoch": 0.327603866589847,
      "grad_norm": 0.009851115755736828,
      "learning_rate": 0.2540988999619405,
      "loss": 0.7133,
      "num_input_tokens_seen": 14073024,
      "step": 10235
    },
    {
      "epoch": 0.32776390756033547,
      "grad_norm": 0.00421861931681633,
      "learning_rate": 0.25405648149178023,
      "loss": 0.5286,
      "num_input_tokens_seen": 14080304,
      "step": 10240
    },
    {
      "epoch": 0.3279239485308239,
      "grad_norm": 0.008479919284582138,
      "learning_rate": 0.2540140469748028,
      "loss": 0.5952,
      "num_input_tokens_seen": 14087104,
      "step": 10245
    },
    {
      "epoch": 0.32808398950131235,
      "grad_norm": 0.005121774040162563,
      "learning_rate": 0.25397159641755224,
      "loss": 0.5323,
      "num_input_tokens_seen": 14093440,
      "step": 10250
    },
    {
      "epoch": 0.3282440304718008,
      "grad_norm": 0.004282086621969938,
      "learning_rate": 0.2539291298265749,
      "loss": 0.5837,
      "num_input_tokens_seen": 14100720,
      "step": 10255
    },
    {
      "epoch": 0.3284040714422892,
      "grad_norm": 0.005475756712257862,
      "learning_rate": 0.2538866472084197,
      "loss": 0.8058,
      "num_input_tokens_seen": 14107552,
      "step": 10260
    },
    {
      "epoch": 0.32856411241277766,
      "grad_norm": 0.008579809218645096,
      "learning_rate": 0.25384414856963794,
      "loss": 0.5717,
      "num_input_tokens_seen": 14114352,
      "step": 10265
    },
    {
      "epoch": 0.3287241533832661,
      "grad_norm": 0.009772374294698238,
      "learning_rate": 0.25380163391678356,
      "loss": 0.8449,
      "num_input_tokens_seen": 14121376,
      "step": 10270
    },
    {
      "epoch": 0.32888419435375454,
      "grad_norm": 0.009105863980948925,
      "learning_rate": 0.2537591032564127,
      "loss": 0.4822,
      "num_input_tokens_seen": 14127792,
      "step": 10275
    },
    {
      "epoch": 0.329044235324243,
      "grad_norm": 0.0062294877134263515,
      "learning_rate": 0.25371655659508424,
      "loss": 0.5982,
      "num_input_tokens_seen": 14134864,
      "step": 10280
    },
    {
      "epoch": 0.3292042762947314,
      "grad_norm": 0.008662257343530655,
      "learning_rate": 0.25367399393935935,
      "loss": 0.6911,
      "num_input_tokens_seen": 14141856,
      "step": 10285
    },
    {
      "epoch": 0.3293643172652199,
      "grad_norm": 0.006269056815654039,
      "learning_rate": 0.25363141529580174,
      "loss": 0.6061,
      "num_input_tokens_seen": 14148992,
      "step": 10290
    },
    {
      "epoch": 0.32952435823570836,
      "grad_norm": 0.005542110651731491,
      "learning_rate": 0.2535888206709776,
      "loss": 0.7489,
      "num_input_tokens_seen": 14155856,
      "step": 10295
    },
    {
      "epoch": 0.3296843992061968,
      "grad_norm": 0.0048302593640983105,
      "learning_rate": 0.2535462100714555,
      "loss": 0.5936,
      "num_input_tokens_seen": 14162896,
      "step": 10300
    },
    {
      "epoch": 0.32984444017668524,
      "grad_norm": 0.011608595959842205,
      "learning_rate": 0.2535035835038066,
      "loss": 0.6191,
      "num_input_tokens_seen": 14169824,
      "step": 10305
    },
    {
      "epoch": 0.3300044811471737,
      "grad_norm": 0.002728360239416361,
      "learning_rate": 0.2534609409746044,
      "loss": 0.5613,
      "num_input_tokens_seen": 14176576,
      "step": 10310
    },
    {
      "epoch": 0.3301645221176621,
      "grad_norm": 0.004956952296197414,
      "learning_rate": 0.253418282490425,
      "loss": 0.6156,
      "num_input_tokens_seen": 14183728,
      "step": 10315
    },
    {
      "epoch": 0.33032456308815056,
      "grad_norm": 0.010443545877933502,
      "learning_rate": 0.2533756080578467,
      "loss": 0.6513,
      "num_input_tokens_seen": 14190272,
      "step": 10320
    },
    {
      "epoch": 0.330484604058639,
      "grad_norm": 0.00802115723490715,
      "learning_rate": 0.25333291768345056,
      "loss": 0.7424,
      "num_input_tokens_seen": 14196880,
      "step": 10325
    },
    {
      "epoch": 0.33064464502912744,
      "grad_norm": 0.006540502421557903,
      "learning_rate": 0.25329021137381996,
      "loss": 0.4816,
      "num_input_tokens_seen": 14204096,
      "step": 10330
    },
    {
      "epoch": 0.3308046859996159,
      "grad_norm": 0.008839964866638184,
      "learning_rate": 0.25324748913554074,
      "loss": 0.6177,
      "num_input_tokens_seen": 14210784,
      "step": 10335
    },
    {
      "epoch": 0.3309647269701044,
      "grad_norm": 0.007929418236017227,
      "learning_rate": 0.2532047509752013,
      "loss": 0.5576,
      "num_input_tokens_seen": 14217312,
      "step": 10340
    },
    {
      "epoch": 0.3311247679405928,
      "grad_norm": 0.006906774826347828,
      "learning_rate": 0.25316199689939217,
      "loss": 0.5738,
      "num_input_tokens_seen": 14223936,
      "step": 10345
    },
    {
      "epoch": 0.33128480891108125,
      "grad_norm": 0.009130747988820076,
      "learning_rate": 0.2531192269147068,
      "loss": 0.6296,
      "num_input_tokens_seen": 14231184,
      "step": 10350
    },
    {
      "epoch": 0.3314448498815697,
      "grad_norm": 0.007852130569517612,
      "learning_rate": 0.2530764410277407,
      "loss": 0.5259,
      "num_input_tokens_seen": 14238048,
      "step": 10355
    },
    {
      "epoch": 0.33160489085205813,
      "grad_norm": 0.0044796341098845005,
      "learning_rate": 0.25303363924509203,
      "loss": 0.6329,
      "num_input_tokens_seen": 14245040,
      "step": 10360
    },
    {
      "epoch": 0.33176493182254657,
      "grad_norm": 0.0038204099982976913,
      "learning_rate": 0.25299082157336145,
      "loss": 0.6008,
      "num_input_tokens_seen": 14251664,
      "step": 10365
    },
    {
      "epoch": 0.331924972793035,
      "grad_norm": 0.004838475026190281,
      "learning_rate": 0.2529479880191519,
      "loss": 0.5832,
      "num_input_tokens_seen": 14258752,
      "step": 10370
    },
    {
      "epoch": 0.33208501376352345,
      "grad_norm": 0.005843034014105797,
      "learning_rate": 0.2529051385890689,
      "loss": 0.8055,
      "num_input_tokens_seen": 14265584,
      "step": 10375
    },
    {
      "epoch": 0.3322450547340119,
      "grad_norm": 0.006795758847147226,
      "learning_rate": 0.2528622732897203,
      "loss": 0.7067,
      "num_input_tokens_seen": 14272144,
      "step": 10380
    },
    {
      "epoch": 0.33240509570450033,
      "grad_norm": 0.00872464757412672,
      "learning_rate": 0.25281939212771654,
      "loss": 0.6089,
      "num_input_tokens_seen": 14279056,
      "step": 10385
    },
    {
      "epoch": 0.33256513667498877,
      "grad_norm": 0.004256832879036665,
      "learning_rate": 0.2527764951096704,
      "loss": 0.7149,
      "num_input_tokens_seen": 14285856,
      "step": 10390
    },
    {
      "epoch": 0.33272517764547727,
      "grad_norm": 0.008487577550113201,
      "learning_rate": 0.2527335822421971,
      "loss": 0.6341,
      "num_input_tokens_seen": 14292928,
      "step": 10395
    },
    {
      "epoch": 0.3328852186159657,
      "grad_norm": 0.008169720880687237,
      "learning_rate": 0.25269065353191444,
      "loss": 0.556,
      "num_input_tokens_seen": 14300160,
      "step": 10400
    },
    {
      "epoch": 0.3328852186159657,
      "eval_loss": 0.641886830329895,
      "eval_runtime": 332.5545,
      "eval_samples_per_second": 41.756,
      "eval_steps_per_second": 20.878,
      "num_input_tokens_seen": 14300160,
      "step": 10400
    },
    {
      "epoch": 0.33304525958645415,
      "grad_norm": 0.006468845997005701,
      "learning_rate": 0.2526477089854425,
      "loss": 0.6006,
      "num_input_tokens_seen": 14306800,
      "step": 10405
    },
    {
      "epoch": 0.3332053005569426,
      "grad_norm": 0.00891763623803854,
      "learning_rate": 0.25260474860940385,
      "loss": 0.7594,
      "num_input_tokens_seen": 14313792,
      "step": 10410
    },
    {
      "epoch": 0.333365341527431,
      "grad_norm": 0.009246611967682838,
      "learning_rate": 0.2525617724104236,
      "loss": 0.6335,
      "num_input_tokens_seen": 14320992,
      "step": 10415
    },
    {
      "epoch": 0.33352538249791946,
      "grad_norm": 0.008574585430324078,
      "learning_rate": 0.25251878039512915,
      "loss": 0.6868,
      "num_input_tokens_seen": 14327872,
      "step": 10420
    },
    {
      "epoch": 0.3336854234684079,
      "grad_norm": 0.004732375964522362,
      "learning_rate": 0.25247577257015047,
      "loss": 0.6488,
      "num_input_tokens_seen": 14334624,
      "step": 10425
    },
    {
      "epoch": 0.33384546443889634,
      "grad_norm": 0.01126795168966055,
      "learning_rate": 0.2524327489421198,
      "loss": 0.866,
      "num_input_tokens_seen": 14341648,
      "step": 10430
    },
    {
      "epoch": 0.3340055054093848,
      "grad_norm": 0.010304995812475681,
      "learning_rate": 0.25238970951767203,
      "loss": 0.6472,
      "num_input_tokens_seen": 14348000,
      "step": 10435
    },
    {
      "epoch": 0.3341655463798732,
      "grad_norm": 0.012058787979185581,
      "learning_rate": 0.25234665430344433,
      "loss": 0.6335,
      "num_input_tokens_seen": 14355024,
      "step": 10440
    },
    {
      "epoch": 0.3343255873503617,
      "grad_norm": 0.006463197059929371,
      "learning_rate": 0.2523035833060764,
      "loss": 0.7273,
      "num_input_tokens_seen": 14361904,
      "step": 10445
    },
    {
      "epoch": 0.33448562832085016,
      "grad_norm": 0.0056928060948848724,
      "learning_rate": 0.2522604965322103,
      "loss": 0.4841,
      "num_input_tokens_seen": 14368752,
      "step": 10450
    },
    {
      "epoch": 0.3346456692913386,
      "grad_norm": 0.007446163333952427,
      "learning_rate": 0.25221739398849047,
      "loss": 0.4857,
      "num_input_tokens_seen": 14375552,
      "step": 10455
    },
    {
      "epoch": 0.33480571026182704,
      "grad_norm": 0.005199469160288572,
      "learning_rate": 0.252174275681564,
      "loss": 0.666,
      "num_input_tokens_seen": 14382304,
      "step": 10460
    },
    {
      "epoch": 0.3349657512323155,
      "grad_norm": 0.007069489918649197,
      "learning_rate": 0.2521311416180802,
      "loss": 0.7119,
      "num_input_tokens_seen": 14389392,
      "step": 10465
    },
    {
      "epoch": 0.3351257922028039,
      "grad_norm": 0.0050874738954007626,
      "learning_rate": 0.25208799180469094,
      "loss": 0.457,
      "num_input_tokens_seen": 14396640,
      "step": 10470
    },
    {
      "epoch": 0.33528583317329236,
      "grad_norm": 0.005531508941203356,
      "learning_rate": 0.2520448262480504,
      "loss": 0.5425,
      "num_input_tokens_seen": 14403632,
      "step": 10475
    },
    {
      "epoch": 0.3354458741437808,
      "grad_norm": 0.006286353804171085,
      "learning_rate": 0.25200164495481525,
      "loss": 0.6768,
      "num_input_tokens_seen": 14410496,
      "step": 10480
    },
    {
      "epoch": 0.33560591511426924,
      "grad_norm": 0.008634201250970364,
      "learning_rate": 0.25195844793164474,
      "loss": 0.6864,
      "num_input_tokens_seen": 14417904,
      "step": 10485
    },
    {
      "epoch": 0.3357659560847577,
      "grad_norm": 0.016588488593697548,
      "learning_rate": 0.2519152351852001,
      "loss": 0.5033,
      "num_input_tokens_seen": 14425216,
      "step": 10490
    },
    {
      "epoch": 0.3359259970552461,
      "grad_norm": 0.006853383034467697,
      "learning_rate": 0.25187200672214555,
      "loss": 0.7085,
      "num_input_tokens_seen": 14432048,
      "step": 10495
    },
    {
      "epoch": 0.3360860380257346,
      "grad_norm": 0.005664461757987738,
      "learning_rate": 0.2518287625491473,
      "loss": 0.6752,
      "num_input_tokens_seen": 14438832,
      "step": 10500
    },
    {
      "epoch": 0.33624607899622305,
      "grad_norm": 0.005770180840045214,
      "learning_rate": 0.25178550267287425,
      "loss": 0.5359,
      "num_input_tokens_seen": 14445792,
      "step": 10505
    },
    {
      "epoch": 0.3364061199667115,
      "grad_norm": 0.013353699818253517,
      "learning_rate": 0.2517422270999976,
      "loss": 0.7815,
      "num_input_tokens_seen": 14453200,
      "step": 10510
    },
    {
      "epoch": 0.33656616093719993,
      "grad_norm": 0.0055867754854261875,
      "learning_rate": 0.2516989358371909,
      "loss": 0.5037,
      "num_input_tokens_seen": 14459920,
      "step": 10515
    },
    {
      "epoch": 0.33672620190768837,
      "grad_norm": 0.009469469077885151,
      "learning_rate": 0.25165562889113025,
      "loss": 0.7865,
      "num_input_tokens_seen": 14466544,
      "step": 10520
    },
    {
      "epoch": 0.3368862428781768,
      "grad_norm": 0.004530004225671291,
      "learning_rate": 0.2516123062684942,
      "loss": 0.5106,
      "num_input_tokens_seen": 14473424,
      "step": 10525
    },
    {
      "epoch": 0.33704628384866525,
      "grad_norm": 0.004416310228407383,
      "learning_rate": 0.25156896797596356,
      "loss": 0.4488,
      "num_input_tokens_seen": 14480800,
      "step": 10530
    },
    {
      "epoch": 0.3372063248191537,
      "grad_norm": 0.005884842481464148,
      "learning_rate": 0.2515256140202216,
      "loss": 0.4953,
      "num_input_tokens_seen": 14488256,
      "step": 10535
    },
    {
      "epoch": 0.33736636578964213,
      "grad_norm": 0.009039074182510376,
      "learning_rate": 0.25148224440795425,
      "loss": 0.73,
      "num_input_tokens_seen": 14495040,
      "step": 10540
    },
    {
      "epoch": 0.33752640676013057,
      "grad_norm": 0.004908687435090542,
      "learning_rate": 0.2514388591458494,
      "loss": 0.45,
      "num_input_tokens_seen": 14501792,
      "step": 10545
    },
    {
      "epoch": 0.33768644773061907,
      "grad_norm": 0.008542473427951336,
      "learning_rate": 0.2513954582405977,
      "loss": 0.6336,
      "num_input_tokens_seen": 14508848,
      "step": 10550
    },
    {
      "epoch": 0.3378464887011075,
      "grad_norm": 0.019772090017795563,
      "learning_rate": 0.2513520416988922,
      "loss": 0.7513,
      "num_input_tokens_seen": 14515712,
      "step": 10555
    },
    {
      "epoch": 0.33800652967159595,
      "grad_norm": 0.008526772260665894,
      "learning_rate": 0.2513086095274281,
      "loss": 0.7167,
      "num_input_tokens_seen": 14522704,
      "step": 10560
    },
    {
      "epoch": 0.3381665706420844,
      "grad_norm": 0.012522893957793713,
      "learning_rate": 0.25126516173290336,
      "loss": 0.8353,
      "num_input_tokens_seen": 14529472,
      "step": 10565
    },
    {
      "epoch": 0.3383266116125728,
      "grad_norm": 0.011395756155252457,
      "learning_rate": 0.2512216983220181,
      "loss": 0.7757,
      "num_input_tokens_seen": 14536784,
      "step": 10570
    },
    {
      "epoch": 0.33848665258306126,
      "grad_norm": 0.0077110351994633675,
      "learning_rate": 0.25117821930147494,
      "loss": 0.7166,
      "num_input_tokens_seen": 14543856,
      "step": 10575
    },
    {
      "epoch": 0.3386466935535497,
      "grad_norm": 0.008118054829537868,
      "learning_rate": 0.2511347246779788,
      "loss": 0.5828,
      "num_input_tokens_seen": 14550992,
      "step": 10580
    },
    {
      "epoch": 0.33880673452403814,
      "grad_norm": 0.012855308130383492,
      "learning_rate": 0.25109121445823723,
      "loss": 0.7309,
      "num_input_tokens_seen": 14557584,
      "step": 10585
    },
    {
      "epoch": 0.3389667754945266,
      "grad_norm": 0.00609563710168004,
      "learning_rate": 0.25104768864896004,
      "loss": 0.5733,
      "num_input_tokens_seen": 14564640,
      "step": 10590
    },
    {
      "epoch": 0.339126816465015,
      "grad_norm": 0.0042165713384747505,
      "learning_rate": 0.2510041472568594,
      "loss": 0.5216,
      "num_input_tokens_seen": 14571232,
      "step": 10595
    },
    {
      "epoch": 0.33928685743550346,
      "grad_norm": 0.004522073082625866,
      "learning_rate": 0.25096059028864987,
      "loss": 0.5425,
      "num_input_tokens_seen": 14577760,
      "step": 10600
    },
    {
      "epoch": 0.33928685743550346,
      "eval_loss": 0.638480007648468,
      "eval_runtime": 332.1332,
      "eval_samples_per_second": 41.809,
      "eval_steps_per_second": 20.904,
      "num_input_tokens_seen": 14577760,
      "step": 10600
    },
    {
      "epoch": 0.33944689840599196,
      "grad_norm": 0.005898833274841309,
      "learning_rate": 0.25091701775104863,
      "loss": 0.5726,
      "num_input_tokens_seen": 14584336,
      "step": 10605
    },
    {
      "epoch": 0.3396069393764804,
      "grad_norm": 0.003326419275254011,
      "learning_rate": 0.250873429650775,
      "loss": 0.5054,
      "num_input_tokens_seen": 14591232,
      "step": 10610
    },
    {
      "epoch": 0.33976698034696884,
      "grad_norm": 0.007752681151032448,
      "learning_rate": 0.25082982599455095,
      "loss": 0.6057,
      "num_input_tokens_seen": 14598064,
      "step": 10615
    },
    {
      "epoch": 0.3399270213174573,
      "grad_norm": 0.00841052271425724,
      "learning_rate": 0.2507862067891006,
      "loss": 0.7359,
      "num_input_tokens_seen": 14604896,
      "step": 10620
    },
    {
      "epoch": 0.3400870622879457,
      "grad_norm": 0.005182433873414993,
      "learning_rate": 0.25074257204115064,
      "loss": 0.521,
      "num_input_tokens_seen": 14611904,
      "step": 10625
    },
    {
      "epoch": 0.34024710325843416,
      "grad_norm": 0.012467177584767342,
      "learning_rate": 0.25069892175742997,
      "loss": 0.5681,
      "num_input_tokens_seen": 14618544,
      "step": 10630
    },
    {
      "epoch": 0.3404071442289226,
      "grad_norm": 0.00529807573184371,
      "learning_rate": 0.25065525594467014,
      "loss": 0.5915,
      "num_input_tokens_seen": 14625296,
      "step": 10635
    },
    {
      "epoch": 0.34056718519941104,
      "grad_norm": 0.004706353414803743,
      "learning_rate": 0.2506115746096049,
      "loss": 0.6883,
      "num_input_tokens_seen": 14632528,
      "step": 10640
    },
    {
      "epoch": 0.3407272261698995,
      "grad_norm": 0.007157750893384218,
      "learning_rate": 0.25056787775897055,
      "loss": 0.3979,
      "num_input_tokens_seen": 14638976,
      "step": 10645
    },
    {
      "epoch": 0.3408872671403879,
      "grad_norm": 0.009632536210119724,
      "learning_rate": 0.2505241653995056,
      "loss": 0.748,
      "num_input_tokens_seen": 14645792,
      "step": 10650
    },
    {
      "epoch": 0.34104730811087636,
      "grad_norm": 0.008360541425645351,
      "learning_rate": 0.25048043753795113,
      "loss": 0.6489,
      "num_input_tokens_seen": 14653040,
      "step": 10655
    },
    {
      "epoch": 0.34120734908136485,
      "grad_norm": 0.009841225109994411,
      "learning_rate": 0.2504366941810504,
      "loss": 0.6671,
      "num_input_tokens_seen": 14659792,
      "step": 10660
    },
    {
      "epoch": 0.3413673900518533,
      "grad_norm": 0.004864535294473171,
      "learning_rate": 0.2503929353355493,
      "loss": 0.5857,
      "num_input_tokens_seen": 14666400,
      "step": 10665
    },
    {
      "epoch": 0.34152743102234173,
      "grad_norm": 0.00626362394541502,
      "learning_rate": 0.250349161008196,
      "loss": 0.5473,
      "num_input_tokens_seen": 14673360,
      "step": 10670
    },
    {
      "epoch": 0.34168747199283017,
      "grad_norm": 0.004988749977201223,
      "learning_rate": 0.2503053712057409,
      "loss": 0.6797,
      "num_input_tokens_seen": 14680192,
      "step": 10675
    },
    {
      "epoch": 0.3418475129633186,
      "grad_norm": 0.008814803324639797,
      "learning_rate": 0.25026156593493715,
      "loss": 0.6897,
      "num_input_tokens_seen": 14686912,
      "step": 10680
    },
    {
      "epoch": 0.34200755393380705,
      "grad_norm": 0.0077622802928090096,
      "learning_rate": 0.2502177452025399,
      "loss": 0.5153,
      "num_input_tokens_seen": 14693792,
      "step": 10685
    },
    {
      "epoch": 0.3421675949042955,
      "grad_norm": 0.005160643253475428,
      "learning_rate": 0.25017390901530695,
      "loss": 0.5554,
      "num_input_tokens_seen": 14700992,
      "step": 10690
    },
    {
      "epoch": 0.34232763587478393,
      "grad_norm": 0.007145180366933346,
      "learning_rate": 0.2501300573799984,
      "loss": 0.5374,
      "num_input_tokens_seen": 14707840,
      "step": 10695
    },
    {
      "epoch": 0.34248767684527237,
      "grad_norm": 0.009933768771588802,
      "learning_rate": 0.2500861903033766,
      "loss": 0.5609,
      "num_input_tokens_seen": 14714656,
      "step": 10700
    },
    {
      "epoch": 0.3426477178157608,
      "grad_norm": 0.007206473965197802,
      "learning_rate": 0.25004230779220654,
      "loss": 0.5177,
      "num_input_tokens_seen": 14721808,
      "step": 10705
    },
    {
      "epoch": 0.3428077587862493,
      "grad_norm": 0.009071281179785728,
      "learning_rate": 0.24999840985325542,
      "loss": 0.759,
      "num_input_tokens_seen": 14729008,
      "step": 10710
    },
    {
      "epoch": 0.34296779975673775,
      "grad_norm": 0.004333717282861471,
      "learning_rate": 0.24995449649329285,
      "loss": 0.5117,
      "num_input_tokens_seen": 14735952,
      "step": 10715
    },
    {
      "epoch": 0.3431278407272262,
      "grad_norm": 0.00926240161061287,
      "learning_rate": 0.2499105677190908,
      "loss": 0.8867,
      "num_input_tokens_seen": 14742736,
      "step": 10720
    },
    {
      "epoch": 0.3432878816977146,
      "grad_norm": 0.006993136368691921,
      "learning_rate": 0.24986662353742364,
      "loss": 0.6383,
      "num_input_tokens_seen": 14749120,
      "step": 10725
    },
    {
      "epoch": 0.34344792266820307,
      "grad_norm": 0.0071345968171954155,
      "learning_rate": 0.24982266395506814,
      "loss": 0.7477,
      "num_input_tokens_seen": 14756064,
      "step": 10730
    },
    {
      "epoch": 0.3436079636386915,
      "grad_norm": 0.006661946419626474,
      "learning_rate": 0.2497786889788034,
      "loss": 0.5531,
      "num_input_tokens_seen": 14763184,
      "step": 10735
    },
    {
      "epoch": 0.34376800460917994,
      "grad_norm": 0.005004459526389837,
      "learning_rate": 0.24973469861541095,
      "loss": 0.5735,
      "num_input_tokens_seen": 14769984,
      "step": 10740
    },
    {
      "epoch": 0.3439280455796684,
      "grad_norm": 0.008220951072871685,
      "learning_rate": 0.24969069287167456,
      "loss": 0.6148,
      "num_input_tokens_seen": 14777008,
      "step": 10745
    },
    {
      "epoch": 0.3440880865501568,
      "grad_norm": 0.005319297779351473,
      "learning_rate": 0.2496466717543806,
      "loss": 0.5835,
      "num_input_tokens_seen": 14783808,
      "step": 10750
    },
    {
      "epoch": 0.34424812752064526,
      "grad_norm": 0.005139423068612814,
      "learning_rate": 0.24960263527031762,
      "loss": 0.4419,
      "num_input_tokens_seen": 14791040,
      "step": 10755
    },
    {
      "epoch": 0.3444081684911337,
      "grad_norm": 0.005742767825722694,
      "learning_rate": 0.24955858342627657,
      "loss": 0.4671,
      "num_input_tokens_seen": 14797840,
      "step": 10760
    },
    {
      "epoch": 0.3445682094616222,
      "grad_norm": 0.013228274881839752,
      "learning_rate": 0.24951451622905083,
      "loss": 0.5436,
      "num_input_tokens_seen": 14804240,
      "step": 10765
    },
    {
      "epoch": 0.34472825043211064,
      "grad_norm": 0.007169139571487904,
      "learning_rate": 0.24947043368543612,
      "loss": 0.6328,
      "num_input_tokens_seen": 14810768,
      "step": 10770
    },
    {
      "epoch": 0.3448882914025991,
      "grad_norm": 0.006057925522327423,
      "learning_rate": 0.2494263358022305,
      "loss": 0.5414,
      "num_input_tokens_seen": 14817280,
      "step": 10775
    },
    {
      "epoch": 0.3450483323730875,
      "grad_norm": 0.007294554263353348,
      "learning_rate": 0.24938222258623444,
      "loss": 0.5179,
      "num_input_tokens_seen": 14824016,
      "step": 10780
    },
    {
      "epoch": 0.34520837334357596,
      "grad_norm": 0.00714082270860672,
      "learning_rate": 0.24933809404425075,
      "loss": 0.6083,
      "num_input_tokens_seen": 14830720,
      "step": 10785
    },
    {
      "epoch": 0.3453684143140644,
      "grad_norm": 0.005525049287825823,
      "learning_rate": 0.24929395018308453,
      "loss": 0.4939,
      "num_input_tokens_seen": 14837312,
      "step": 10790
    },
    {
      "epoch": 0.34552845528455284,
      "grad_norm": 0.0035751922987401485,
      "learning_rate": 0.24924979100954348,
      "loss": 0.4679,
      "num_input_tokens_seen": 14844016,
      "step": 10795
    },
    {
      "epoch": 0.3456884962550413,
      "grad_norm": 0.006915854290127754,
      "learning_rate": 0.24920561653043735,
      "loss": 0.6794,
      "num_input_tokens_seen": 14851280,
      "step": 10800
    },
    {
      "epoch": 0.3456884962550413,
      "eval_loss": 0.61854088306427,
      "eval_runtime": 331.992,
      "eval_samples_per_second": 41.826,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 14851280,
      "step": 10800
    },
    {
      "epoch": 0.3458485372255297,
      "grad_norm": 0.007988526485860348,
      "learning_rate": 0.24916142675257846,
      "loss": 0.5703,
      "num_input_tokens_seen": 14858480,
      "step": 10805
    },
    {
      "epoch": 0.34600857819601816,
      "grad_norm": 0.006838198285549879,
      "learning_rate": 0.24911722168278144,
      "loss": 0.7006,
      "num_input_tokens_seen": 14865456,
      "step": 10810
    },
    {
      "epoch": 0.34616861916650665,
      "grad_norm": 0.0071050976403057575,
      "learning_rate": 0.24907300132786328,
      "loss": 0.5428,
      "num_input_tokens_seen": 14872096,
      "step": 10815
    },
    {
      "epoch": 0.3463286601369951,
      "grad_norm": 0.0028193567413836718,
      "learning_rate": 0.24902876569464322,
      "loss": 0.433,
      "num_input_tokens_seen": 14878640,
      "step": 10820
    },
    {
      "epoch": 0.34648870110748353,
      "grad_norm": 0.0052422103472054005,
      "learning_rate": 0.24898451478994305,
      "loss": 0.6728,
      "num_input_tokens_seen": 14885216,
      "step": 10825
    },
    {
      "epoch": 0.34664874207797197,
      "grad_norm": 0.00505789602175355,
      "learning_rate": 0.2489402486205868,
      "loss": 0.6957,
      "num_input_tokens_seen": 14891936,
      "step": 10830
    },
    {
      "epoch": 0.3468087830484604,
      "grad_norm": 0.008208677172660828,
      "learning_rate": 0.24889596719340085,
      "loss": 0.7904,
      "num_input_tokens_seen": 14899088,
      "step": 10835
    },
    {
      "epoch": 0.34696882401894885,
      "grad_norm": 0.007302225101739168,
      "learning_rate": 0.24885167051521392,
      "loss": 0.6106,
      "num_input_tokens_seen": 14906032,
      "step": 10840
    },
    {
      "epoch": 0.3471288649894373,
      "grad_norm": 0.005141447763890028,
      "learning_rate": 0.24880735859285716,
      "loss": 0.6136,
      "num_input_tokens_seen": 14912752,
      "step": 10845
    },
    {
      "epoch": 0.34728890595992573,
      "grad_norm": 0.009195703081786633,
      "learning_rate": 0.24876303143316406,
      "loss": 0.7765,
      "num_input_tokens_seen": 14919776,
      "step": 10850
    },
    {
      "epoch": 0.34744894693041417,
      "grad_norm": 0.005492561962455511,
      "learning_rate": 0.24871868904297031,
      "loss": 0.588,
      "num_input_tokens_seen": 14926304,
      "step": 10855
    },
    {
      "epoch": 0.3476089879009026,
      "grad_norm": 0.007868201471865177,
      "learning_rate": 0.24867433142911416,
      "loss": 0.5305,
      "num_input_tokens_seen": 14933376,
      "step": 10860
    },
    {
      "epoch": 0.34776902887139105,
      "grad_norm": 0.006754644680768251,
      "learning_rate": 0.24862995859843612,
      "loss": 0.6518,
      "num_input_tokens_seen": 14939792,
      "step": 10865
    },
    {
      "epoch": 0.34792906984187955,
      "grad_norm": 0.0077536143362522125,
      "learning_rate": 0.24858557055777897,
      "loss": 0.5563,
      "num_input_tokens_seen": 14947104,
      "step": 10870
    },
    {
      "epoch": 0.348089110812368,
      "grad_norm": 0.0038851855788379908,
      "learning_rate": 0.24854116731398793,
      "loss": 0.6207,
      "num_input_tokens_seen": 14954336,
      "step": 10875
    },
    {
      "epoch": 0.3482491517828564,
      "grad_norm": 0.008198010735213757,
      "learning_rate": 0.24849674887391052,
      "loss": 0.7332,
      "num_input_tokens_seen": 14960928,
      "step": 10880
    },
    {
      "epoch": 0.34840919275334487,
      "grad_norm": 0.00897998921573162,
      "learning_rate": 0.2484523152443967,
      "loss": 0.4697,
      "num_input_tokens_seen": 14967824,
      "step": 10885
    },
    {
      "epoch": 0.3485692337238333,
      "grad_norm": 0.006450921297073364,
      "learning_rate": 0.24840786643229862,
      "loss": 0.7146,
      "num_input_tokens_seen": 14974928,
      "step": 10890
    },
    {
      "epoch": 0.34872927469432174,
      "grad_norm": 0.013932818546891212,
      "learning_rate": 0.2483634024444709,
      "loss": 0.6185,
      "num_input_tokens_seen": 14981696,
      "step": 10895
    },
    {
      "epoch": 0.3488893156648102,
      "grad_norm": 0.0029510518070310354,
      "learning_rate": 0.24831892328777033,
      "loss": 0.4709,
      "num_input_tokens_seen": 14988336,
      "step": 10900
    },
    {
      "epoch": 0.3490493566352986,
      "grad_norm": 0.008552859537303448,
      "learning_rate": 0.2482744289690563,
      "loss": 0.6501,
      "num_input_tokens_seen": 14995440,
      "step": 10905
    },
    {
      "epoch": 0.34920939760578706,
      "grad_norm": 0.0070005618035793304,
      "learning_rate": 0.2482299194951903,
      "loss": 0.5598,
      "num_input_tokens_seen": 15002048,
      "step": 10910
    },
    {
      "epoch": 0.3493694385762755,
      "grad_norm": 0.00638953223824501,
      "learning_rate": 0.2481853948730363,
      "loss": 0.5397,
      "num_input_tokens_seen": 15008960,
      "step": 10915
    },
    {
      "epoch": 0.349529479546764,
      "grad_norm": 0.00621002446860075,
      "learning_rate": 0.24814085510946052,
      "loss": 0.4077,
      "num_input_tokens_seen": 15015904,
      "step": 10920
    },
    {
      "epoch": 0.34968952051725244,
      "grad_norm": 0.010094067081809044,
      "learning_rate": 0.24809630021133158,
      "loss": 0.5901,
      "num_input_tokens_seen": 15022800,
      "step": 10925
    },
    {
      "epoch": 0.3498495614877409,
      "grad_norm": 0.010991927236318588,
      "learning_rate": 0.24805173018552037,
      "loss": 0.5964,
      "num_input_tokens_seen": 15029424,
      "step": 10930
    },
    {
      "epoch": 0.3500096024582293,
      "grad_norm": 0.007581927347928286,
      "learning_rate": 0.2480071450389002,
      "loss": 0.7717,
      "num_input_tokens_seen": 15036656,
      "step": 10935
    },
    {
      "epoch": 0.35016964342871776,
      "grad_norm": 0.005177496932446957,
      "learning_rate": 0.24796254477834662,
      "loss": 0.6524,
      "num_input_tokens_seen": 15043376,
      "step": 10940
    },
    {
      "epoch": 0.3503296843992062,
      "grad_norm": 0.00689172837883234,
      "learning_rate": 0.24791792941073754,
      "loss": 0.5864,
      "num_input_tokens_seen": 15050400,
      "step": 10945
    },
    {
      "epoch": 0.35048972536969464,
      "grad_norm": 0.008241884410381317,
      "learning_rate": 0.2478732989429533,
      "loss": 0.8122,
      "num_input_tokens_seen": 15057168,
      "step": 10950
    },
    {
      "epoch": 0.3506497663401831,
      "grad_norm": 0.008290293626487255,
      "learning_rate": 0.24782865338187632,
      "loss": 0.5753,
      "num_input_tokens_seen": 15063872,
      "step": 10955
    },
    {
      "epoch": 0.3508098073106715,
      "grad_norm": 0.010513114742934704,
      "learning_rate": 0.2477839927343916,
      "loss": 0.7271,
      "num_input_tokens_seen": 15070640,
      "step": 10960
    },
    {
      "epoch": 0.35096984828115996,
      "grad_norm": 0.005850061308592558,
      "learning_rate": 0.2477393170073864,
      "loss": 0.7344,
      "num_input_tokens_seen": 15077664,
      "step": 10965
    },
    {
      "epoch": 0.3511298892516484,
      "grad_norm": 0.006732120178639889,
      "learning_rate": 0.2476946262077503,
      "loss": 0.4482,
      "num_input_tokens_seen": 15084704,
      "step": 10970
    },
    {
      "epoch": 0.3512899302221369,
      "grad_norm": 0.005800638347864151,
      "learning_rate": 0.24764992034237507,
      "loss": 0.5574,
      "num_input_tokens_seen": 15091664,
      "step": 10975
    },
    {
      "epoch": 0.35144997119262533,
      "grad_norm": 0.01682785525918007,
      "learning_rate": 0.24760519941815498,
      "loss": 0.7048,
      "num_input_tokens_seen": 15098192,
      "step": 10980
    },
    {
      "epoch": 0.35161001216311377,
      "grad_norm": 0.007605898194015026,
      "learning_rate": 0.2475604634419866,
      "loss": 0.8154,
      "num_input_tokens_seen": 15104960,
      "step": 10985
    },
    {
      "epoch": 0.3517700531336022,
      "grad_norm": 0.009364730678498745,
      "learning_rate": 0.24751571242076872,
      "loss": 0.8721,
      "num_input_tokens_seen": 15111680,
      "step": 10990
    },
    {
      "epoch": 0.35193009410409065,
      "grad_norm": 0.012791736982762814,
      "learning_rate": 0.2474709463614025,
      "loss": 0.7708,
      "num_input_tokens_seen": 15118480,
      "step": 10995
    },
    {
      "epoch": 0.3520901350745791,
      "grad_norm": 0.005072949919849634,
      "learning_rate": 0.24742616527079145,
      "loss": 0.5348,
      "num_input_tokens_seen": 15125104,
      "step": 11000
    },
    {
      "epoch": 0.3520901350745791,
      "eval_loss": 0.6384974122047424,
      "eval_runtime": 332.0057,
      "eval_samples_per_second": 41.825,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 15125104,
      "step": 11000
    },
    {
      "epoch": 0.35225017604506753,
      "grad_norm": 0.007588368374854326,
      "learning_rate": 0.24738136915584139,
      "loss": 0.5576,
      "num_input_tokens_seen": 15131968,
      "step": 11005
    },
    {
      "epoch": 0.35241021701555597,
      "grad_norm": 0.004890263546258211,
      "learning_rate": 0.24733655802346047,
      "loss": 0.5864,
      "num_input_tokens_seen": 15138752,
      "step": 11010
    },
    {
      "epoch": 0.3525702579860444,
      "grad_norm": 0.005226462613791227,
      "learning_rate": 0.24729173188055906,
      "loss": 0.6347,
      "num_input_tokens_seen": 15145600,
      "step": 11015
    },
    {
      "epoch": 0.35273029895653285,
      "grad_norm": 0.0047945198602974415,
      "learning_rate": 0.24724689073404996,
      "loss": 0.4713,
      "num_input_tokens_seen": 15152560,
      "step": 11020
    },
    {
      "epoch": 0.3528903399270213,
      "grad_norm": 0.006955576594918966,
      "learning_rate": 0.24720203459084822,
      "loss": 0.671,
      "num_input_tokens_seen": 15159328,
      "step": 11025
    },
    {
      "epoch": 0.3530503808975098,
      "grad_norm": 0.005542691797018051,
      "learning_rate": 0.24715716345787123,
      "loss": 0.7412,
      "num_input_tokens_seen": 15166448,
      "step": 11030
    },
    {
      "epoch": 0.3532104218679982,
      "grad_norm": 0.004372313152998686,
      "learning_rate": 0.2471122773420387,
      "loss": 0.6367,
      "num_input_tokens_seen": 15173232,
      "step": 11035
    },
    {
      "epoch": 0.35337046283848667,
      "grad_norm": 0.007299772463738918,
      "learning_rate": 0.24706737625027259,
      "loss": 0.5548,
      "num_input_tokens_seen": 15179792,
      "step": 11040
    },
    {
      "epoch": 0.3535305038089751,
      "grad_norm": 0.006478786934167147,
      "learning_rate": 0.24702246018949725,
      "loss": 0.6071,
      "num_input_tokens_seen": 15186576,
      "step": 11045
    },
    {
      "epoch": 0.35369054477946354,
      "grad_norm": 0.008600303903222084,
      "learning_rate": 0.2469775291666393,
      "loss": 0.7882,
      "num_input_tokens_seen": 15193328,
      "step": 11050
    },
    {
      "epoch": 0.353850585749952,
      "grad_norm": 0.006271657533943653,
      "learning_rate": 0.24693258318862765,
      "loss": 0.6515,
      "num_input_tokens_seen": 15200208,
      "step": 11055
    },
    {
      "epoch": 0.3540106267204404,
      "grad_norm": 0.009205232374370098,
      "learning_rate": 0.2468876222623935,
      "loss": 0.6272,
      "num_input_tokens_seen": 15207040,
      "step": 11060
    },
    {
      "epoch": 0.35417066769092886,
      "grad_norm": 0.00325025524944067,
      "learning_rate": 0.2468426463948705,
      "loss": 0.6593,
      "num_input_tokens_seen": 15213984,
      "step": 11065
    },
    {
      "epoch": 0.3543307086614173,
      "grad_norm": 0.008172614499926567,
      "learning_rate": 0.24679765559299438,
      "loss": 0.5346,
      "num_input_tokens_seen": 15220768,
      "step": 11070
    },
    {
      "epoch": 0.35449074963190574,
      "grad_norm": 0.00579586997628212,
      "learning_rate": 0.24675264986370332,
      "loss": 0.6355,
      "num_input_tokens_seen": 15227104,
      "step": 11075
    },
    {
      "epoch": 0.35465079060239424,
      "grad_norm": 0.005949322134256363,
      "learning_rate": 0.2467076292139378,
      "loss": 0.5876,
      "num_input_tokens_seen": 15234432,
      "step": 11080
    },
    {
      "epoch": 0.3548108315728827,
      "grad_norm": 0.0021558478474617004,
      "learning_rate": 0.24666259365064055,
      "loss": 0.4733,
      "num_input_tokens_seen": 15241280,
      "step": 11085
    },
    {
      "epoch": 0.3549708725433711,
      "grad_norm": 0.008977525867521763,
      "learning_rate": 0.24661754318075663,
      "loss": 0.5217,
      "num_input_tokens_seen": 15248128,
      "step": 11090
    },
    {
      "epoch": 0.35513091351385956,
      "grad_norm": 0.0030334130860865116,
      "learning_rate": 0.2465724778112334,
      "loss": 0.493,
      "num_input_tokens_seen": 15254640,
      "step": 11095
    },
    {
      "epoch": 0.355290954484348,
      "grad_norm": 0.004835005383938551,
      "learning_rate": 0.24652739754902042,
      "loss": 0.6591,
      "num_input_tokens_seen": 15261488,
      "step": 11100
    },
    {
      "epoch": 0.35545099545483644,
      "grad_norm": 0.015489216893911362,
      "learning_rate": 0.24648230240106975,
      "loss": 0.4525,
      "num_input_tokens_seen": 15268240,
      "step": 11105
    },
    {
      "epoch": 0.3556110364253249,
      "grad_norm": 0.011386454105377197,
      "learning_rate": 0.2464371923743356,
      "loss": 0.7126,
      "num_input_tokens_seen": 15275120,
      "step": 11110
    },
    {
      "epoch": 0.3557710773958133,
      "grad_norm": 0.007850337773561478,
      "learning_rate": 0.24639206747577444,
      "loss": 0.7231,
      "num_input_tokens_seen": 15282656,
      "step": 11115
    },
    {
      "epoch": 0.35593111836630176,
      "grad_norm": 0.01306613814085722,
      "learning_rate": 0.24634692771234515,
      "loss": 0.673,
      "num_input_tokens_seen": 15289856,
      "step": 11120
    },
    {
      "epoch": 0.3560911593367902,
      "grad_norm": 0.00766422925516963,
      "learning_rate": 0.2463017730910088,
      "loss": 0.7386,
      "num_input_tokens_seen": 15296400,
      "step": 11125
    },
    {
      "epoch": 0.35625120030727864,
      "grad_norm": 0.004922203719615936,
      "learning_rate": 0.2462566036187289,
      "loss": 0.6873,
      "num_input_tokens_seen": 15303152,
      "step": 11130
    },
    {
      "epoch": 0.35641124127776713,
      "grad_norm": 0.007006898522377014,
      "learning_rate": 0.24621141930247106,
      "loss": 0.5548,
      "num_input_tokens_seen": 15309648,
      "step": 11135
    },
    {
      "epoch": 0.35657128224825557,
      "grad_norm": 0.010969094932079315,
      "learning_rate": 0.2461662201492033,
      "loss": 0.7063,
      "num_input_tokens_seen": 15316768,
      "step": 11140
    },
    {
      "epoch": 0.356731323218744,
      "grad_norm": 0.00552840530872345,
      "learning_rate": 0.24612100616589586,
      "loss": 0.6065,
      "num_input_tokens_seen": 15323856,
      "step": 11145
    },
    {
      "epoch": 0.35689136418923245,
      "grad_norm": 0.0077193258330225945,
      "learning_rate": 0.24607577735952135,
      "loss": 0.5147,
      "num_input_tokens_seen": 15330752,
      "step": 11150
    },
    {
      "epoch": 0.3570514051597209,
      "grad_norm": 0.009341519325971603,
      "learning_rate": 0.24603053373705464,
      "loss": 0.5202,
      "num_input_tokens_seen": 15337584,
      "step": 11155
    },
    {
      "epoch": 0.35721144613020933,
      "grad_norm": 0.007757590617984533,
      "learning_rate": 0.2459852753054728,
      "loss": 0.6451,
      "num_input_tokens_seen": 15344288,
      "step": 11160
    },
    {
      "epoch": 0.35737148710069777,
      "grad_norm": 0.004239576868712902,
      "learning_rate": 0.24594000207175526,
      "loss": 0.606,
      "num_input_tokens_seen": 15351280,
      "step": 11165
    },
    {
      "epoch": 0.3575315280711862,
      "grad_norm": 0.00990801677107811,
      "learning_rate": 0.2458947140428838,
      "loss": 0.5373,
      "num_input_tokens_seen": 15357696,
      "step": 11170
    },
    {
      "epoch": 0.35769156904167465,
      "grad_norm": 0.005283024162054062,
      "learning_rate": 0.24584941122584233,
      "loss": 0.56,
      "num_input_tokens_seen": 15364464,
      "step": 11175
    },
    {
      "epoch": 0.3578516100121631,
      "grad_norm": 0.008474182337522507,
      "learning_rate": 0.24580409362761713,
      "loss": 0.4669,
      "num_input_tokens_seen": 15371408,
      "step": 11180
    },
    {
      "epoch": 0.3580116509826516,
      "grad_norm": 0.006034253630787134,
      "learning_rate": 0.2457587612551967,
      "loss": 0.5929,
      "num_input_tokens_seen": 15378240,
      "step": 11185
    },
    {
      "epoch": 0.35817169195314,
      "grad_norm": 0.010033795610070229,
      "learning_rate": 0.24571341411557193,
      "loss": 0.6361,
      "num_input_tokens_seen": 15385056,
      "step": 11190
    },
    {
      "epoch": 0.35833173292362847,
      "grad_norm": 0.007233035750687122,
      "learning_rate": 0.2456680522157359,
      "loss": 0.6035,
      "num_input_tokens_seen": 15392048,
      "step": 11195
    },
    {
      "epoch": 0.3584917738941169,
      "grad_norm": 0.0040999120101332664,
      "learning_rate": 0.245622675562684,
      "loss": 0.6781,
      "num_input_tokens_seen": 15398624,
      "step": 11200
    },
    {
      "epoch": 0.3584917738941169,
      "eval_loss": 0.6158634424209595,
      "eval_runtime": 331.6879,
      "eval_samples_per_second": 41.865,
      "eval_steps_per_second": 20.932,
      "num_input_tokens_seen": 15398624,
      "step": 11200
    },
    {
      "epoch": 0.35865181486460534,
      "grad_norm": 0.008378767408430576,
      "learning_rate": 0.24557728416341384,
      "loss": 0.7182,
      "num_input_tokens_seen": 15405712,
      "step": 11205
    },
    {
      "epoch": 0.3588118558350938,
      "grad_norm": 0.0033062759321182966,
      "learning_rate": 0.24553187802492538,
      "loss": 0.4363,
      "num_input_tokens_seen": 15412512,
      "step": 11210
    },
    {
      "epoch": 0.3589718968055822,
      "grad_norm": 0.0040353573858737946,
      "learning_rate": 0.24548645715422074,
      "loss": 0.5581,
      "num_input_tokens_seen": 15419520,
      "step": 11215
    },
    {
      "epoch": 0.35913193777607066,
      "grad_norm": 0.005133124999701977,
      "learning_rate": 0.2454410215583045,
      "loss": 0.6441,
      "num_input_tokens_seen": 15426080,
      "step": 11220
    },
    {
      "epoch": 0.3592919787465591,
      "grad_norm": 0.006272599101066589,
      "learning_rate": 0.24539557124418332,
      "loss": 0.4766,
      "num_input_tokens_seen": 15432928,
      "step": 11225
    },
    {
      "epoch": 0.35945201971704754,
      "grad_norm": 0.010090595111250877,
      "learning_rate": 0.24535010621886624,
      "loss": 0.4677,
      "num_input_tokens_seen": 15439296,
      "step": 11230
    },
    {
      "epoch": 0.359612060687536,
      "grad_norm": 0.007169957738369703,
      "learning_rate": 0.2453046264893646,
      "loss": 0.7155,
      "num_input_tokens_seen": 15446576,
      "step": 11235
    },
    {
      "epoch": 0.3597721016580245,
      "grad_norm": 0.008679534308612347,
      "learning_rate": 0.24525913206269184,
      "loss": 0.5416,
      "num_input_tokens_seen": 15453280,
      "step": 11240
    },
    {
      "epoch": 0.3599321426285129,
      "grad_norm": 0.0051596155390143394,
      "learning_rate": 0.2452136229458638,
      "loss": 0.5547,
      "num_input_tokens_seen": 15460416,
      "step": 11245
    },
    {
      "epoch": 0.36009218359900136,
      "grad_norm": 0.005590378772467375,
      "learning_rate": 0.24516809914589857,
      "loss": 0.6467,
      "num_input_tokens_seen": 15467248,
      "step": 11250
    },
    {
      "epoch": 0.3602522245694898,
      "grad_norm": 0.0077466475777328014,
      "learning_rate": 0.2451225606698165,
      "loss": 0.6559,
      "num_input_tokens_seen": 15473856,
      "step": 11255
    },
    {
      "epoch": 0.36041226553997824,
      "grad_norm": 0.005644656252115965,
      "learning_rate": 0.2450770075246402,
      "loss": 0.4486,
      "num_input_tokens_seen": 15480608,
      "step": 11260
    },
    {
      "epoch": 0.3605723065104667,
      "grad_norm": 0.013499526306986809,
      "learning_rate": 0.24503143971739455,
      "loss": 0.6773,
      "num_input_tokens_seen": 15487920,
      "step": 11265
    },
    {
      "epoch": 0.3607323474809551,
      "grad_norm": 0.00374894542619586,
      "learning_rate": 0.24498585725510663,
      "loss": 0.6424,
      "num_input_tokens_seen": 15494864,
      "step": 11270
    },
    {
      "epoch": 0.36089238845144356,
      "grad_norm": 0.0030833540949970484,
      "learning_rate": 0.24494026014480583,
      "loss": 0.7027,
      "num_input_tokens_seen": 15501376,
      "step": 11275
    },
    {
      "epoch": 0.361052429421932,
      "grad_norm": 0.007237750571221113,
      "learning_rate": 0.24489464839352387,
      "loss": 0.5022,
      "num_input_tokens_seen": 15508160,
      "step": 11280
    },
    {
      "epoch": 0.36121247039242044,
      "grad_norm": 0.004505747929215431,
      "learning_rate": 0.2448490220082946,
      "loss": 0.6764,
      "num_input_tokens_seen": 15515088,
      "step": 11285
    },
    {
      "epoch": 0.3613725113629089,
      "grad_norm": 0.011605488136410713,
      "learning_rate": 0.24480338099615415,
      "loss": 0.5296,
      "num_input_tokens_seen": 15522288,
      "step": 11290
    },
    {
      "epoch": 0.3615325523333974,
      "grad_norm": 0.005954917054623365,
      "learning_rate": 0.244757725364141,
      "loss": 0.579,
      "num_input_tokens_seen": 15528976,
      "step": 11295
    },
    {
      "epoch": 0.3616925933038858,
      "grad_norm": 0.007190226577222347,
      "learning_rate": 0.24471205511929583,
      "loss": 0.7933,
      "num_input_tokens_seen": 15536096,
      "step": 11300
    },
    {
      "epoch": 0.36185263427437425,
      "grad_norm": 0.004166801925748587,
      "learning_rate": 0.24466637026866145,
      "loss": 0.598,
      "num_input_tokens_seen": 15542896,
      "step": 11305
    },
    {
      "epoch": 0.3620126752448627,
      "grad_norm": 0.0047438256442546844,
      "learning_rate": 0.2446206708192832,
      "loss": 0.4239,
      "num_input_tokens_seen": 15549808,
      "step": 11310
    },
    {
      "epoch": 0.36217271621535113,
      "grad_norm": 0.005950138904154301,
      "learning_rate": 0.2445749567782084,
      "loss": 0.7461,
      "num_input_tokens_seen": 15556688,
      "step": 11315
    },
    {
      "epoch": 0.36233275718583957,
      "grad_norm": 0.00401300610974431,
      "learning_rate": 0.2445292281524868,
      "loss": 0.7103,
      "num_input_tokens_seen": 15562992,
      "step": 11320
    },
    {
      "epoch": 0.362492798156328,
      "grad_norm": 0.005449684336781502,
      "learning_rate": 0.24448348494917022,
      "loss": 0.4769,
      "num_input_tokens_seen": 15569632,
      "step": 11325
    },
    {
      "epoch": 0.36265283912681645,
      "grad_norm": 0.0064971367828547955,
      "learning_rate": 0.24443772717531295,
      "loss": 0.5662,
      "num_input_tokens_seen": 15576368,
      "step": 11330
    },
    {
      "epoch": 0.3628128800973049,
      "grad_norm": 0.006698237732052803,
      "learning_rate": 0.24439195483797138,
      "loss": 0.7153,
      "num_input_tokens_seen": 15583136,
      "step": 11335
    },
    {
      "epoch": 0.36297292106779333,
      "grad_norm": 0.006808754988014698,
      "learning_rate": 0.24434616794420416,
      "loss": 0.4916,
      "num_input_tokens_seen": 15590112,
      "step": 11340
    },
    {
      "epoch": 0.3631329620382818,
      "grad_norm": 0.006572535261511803,
      "learning_rate": 0.24430036650107223,
      "loss": 0.8547,
      "num_input_tokens_seen": 15597216,
      "step": 11345
    },
    {
      "epoch": 0.36329300300877027,
      "grad_norm": 0.006025427486747503,
      "learning_rate": 0.2442545505156387,
      "loss": 0.481,
      "num_input_tokens_seen": 15604512,
      "step": 11350
    },
    {
      "epoch": 0.3634530439792587,
      "grad_norm": 0.005589632783085108,
      "learning_rate": 0.24420871999496904,
      "loss": 0.5019,
      "num_input_tokens_seen": 15611120,
      "step": 11355
    },
    {
      "epoch": 0.36361308494974715,
      "grad_norm": 0.006115617696195841,
      "learning_rate": 0.24416287494613084,
      "loss": 0.5775,
      "num_input_tokens_seen": 15618208,
      "step": 11360
    },
    {
      "epoch": 0.3637731259202356,
      "grad_norm": 0.009692149236798286,
      "learning_rate": 0.24411701537619399,
      "loss": 0.5901,
      "num_input_tokens_seen": 15624736,
      "step": 11365
    },
    {
      "epoch": 0.363933166890724,
      "grad_norm": 0.0041560581885278225,
      "learning_rate": 0.24407114129223062,
      "loss": 0.5601,
      "num_input_tokens_seen": 15631664,
      "step": 11370
    },
    {
      "epoch": 0.36409320786121246,
      "grad_norm": 0.010139300487935543,
      "learning_rate": 0.2440252527013151,
      "loss": 0.7406,
      "num_input_tokens_seen": 15638176,
      "step": 11375
    },
    {
      "epoch": 0.3642532488317009,
      "grad_norm": 0.00923470500856638,
      "learning_rate": 0.24397934961052403,
      "loss": 0.859,
      "num_input_tokens_seen": 15644816,
      "step": 11380
    },
    {
      "epoch": 0.36441328980218934,
      "grad_norm": 0.006790461018681526,
      "learning_rate": 0.24393343202693618,
      "loss": 0.5636,
      "num_input_tokens_seen": 15651504,
      "step": 11385
    },
    {
      "epoch": 0.3645733307726778,
      "grad_norm": 0.005213690921664238,
      "learning_rate": 0.2438874999576327,
      "loss": 0.6533,
      "num_input_tokens_seen": 15659168,
      "step": 11390
    },
    {
      "epoch": 0.3647333717431662,
      "grad_norm": 0.0062753125093877316,
      "learning_rate": 0.24384155340969688,
      "loss": 0.5867,
      "num_input_tokens_seen": 15665664,
      "step": 11395
    },
    {
      "epoch": 0.3648934127136547,
      "grad_norm": 0.008023789152503014,
      "learning_rate": 0.24379559239021423,
      "loss": 0.8133,
      "num_input_tokens_seen": 15672384,
      "step": 11400
    },
    {
      "epoch": 0.3648934127136547,
      "eval_loss": 0.6244256496429443,
      "eval_runtime": 332.2427,
      "eval_samples_per_second": 41.795,
      "eval_steps_per_second": 20.897,
      "num_input_tokens_seen": 15672384,
      "step": 11400
    },
    {
      "epoch": 0.36505345368414316,
      "grad_norm": 0.005522483494132757,
      "learning_rate": 0.2437496169062725,
      "loss": 0.58,
      "num_input_tokens_seen": 15679152,
      "step": 11405
    },
    {
      "epoch": 0.3652134946546316,
      "grad_norm": 0.005651562009006739,
      "learning_rate": 0.24370362696496176,
      "loss": 0.5374,
      "num_input_tokens_seen": 15686032,
      "step": 11410
    },
    {
      "epoch": 0.36537353562512004,
      "grad_norm": 0.007185099180787802,
      "learning_rate": 0.24365762257337417,
      "loss": 0.7261,
      "num_input_tokens_seen": 15693312,
      "step": 11415
    },
    {
      "epoch": 0.3655335765956085,
      "grad_norm": 0.01534960512071848,
      "learning_rate": 0.2436116037386042,
      "loss": 0.6618,
      "num_input_tokens_seen": 15700800,
      "step": 11420
    },
    {
      "epoch": 0.3656936175660969,
      "grad_norm": 0.0052008796483278275,
      "learning_rate": 0.24356557046774852,
      "loss": 0.5384,
      "num_input_tokens_seen": 15707328,
      "step": 11425
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 0.006301993038505316,
      "learning_rate": 0.24351952276790606,
      "loss": 0.6218,
      "num_input_tokens_seen": 15714016,
      "step": 11430
    },
    {
      "epoch": 0.3660136995070738,
      "grad_norm": 0.005365945864468813,
      "learning_rate": 0.24347346064617797,
      "loss": 0.5997,
      "num_input_tokens_seen": 15721088,
      "step": 11435
    },
    {
      "epoch": 0.36617374047756224,
      "grad_norm": 0.004532933700829744,
      "learning_rate": 0.24342738410966758,
      "loss": 0.5242,
      "num_input_tokens_seen": 15727552,
      "step": 11440
    },
    {
      "epoch": 0.3663337814480507,
      "grad_norm": 0.008052978664636612,
      "learning_rate": 0.24338129316548046,
      "loss": 0.5545,
      "num_input_tokens_seen": 15734288,
      "step": 11445
    },
    {
      "epoch": 0.3664938224185392,
      "grad_norm": 0.004088872577995062,
      "learning_rate": 0.24333518782072444,
      "loss": 0.4388,
      "num_input_tokens_seen": 15741136,
      "step": 11450
    },
    {
      "epoch": 0.3666538633890276,
      "grad_norm": 0.008056432008743286,
      "learning_rate": 0.24328906808250952,
      "loss": 0.5777,
      "num_input_tokens_seen": 15748288,
      "step": 11455
    },
    {
      "epoch": 0.36681390435951605,
      "grad_norm": 0.009476535022258759,
      "learning_rate": 0.243242933957948,
      "loss": 0.5907,
      "num_input_tokens_seen": 15755120,
      "step": 11460
    },
    {
      "epoch": 0.3669739453300045,
      "grad_norm": 0.009762156754732132,
      "learning_rate": 0.24319678545415427,
      "loss": 0.785,
      "num_input_tokens_seen": 15762032,
      "step": 11465
    },
    {
      "epoch": 0.36713398630049293,
      "grad_norm": 0.004983838647603989,
      "learning_rate": 0.24315062257824507,
      "loss": 0.6889,
      "num_input_tokens_seen": 15769296,
      "step": 11470
    },
    {
      "epoch": 0.36729402727098137,
      "grad_norm": 0.003826622385531664,
      "learning_rate": 0.24310444533733921,
      "loss": 0.6451,
      "num_input_tokens_seen": 15775872,
      "step": 11475
    },
    {
      "epoch": 0.3674540682414698,
      "grad_norm": 0.003850743407383561,
      "learning_rate": 0.2430582537385579,
      "loss": 0.7857,
      "num_input_tokens_seen": 15782720,
      "step": 11480
    },
    {
      "epoch": 0.36761410921195825,
      "grad_norm": 0.0037508425302803516,
      "learning_rate": 0.2430120477890244,
      "loss": 0.676,
      "num_input_tokens_seen": 15790080,
      "step": 11485
    },
    {
      "epoch": 0.3677741501824467,
      "grad_norm": 0.017738480120897293,
      "learning_rate": 0.24296582749586426,
      "loss": 0.7222,
      "num_input_tokens_seen": 15796992,
      "step": 11490
    },
    {
      "epoch": 0.36793419115293513,
      "grad_norm": 0.00494067557156086,
      "learning_rate": 0.24291959286620526,
      "loss": 0.6717,
      "num_input_tokens_seen": 15803728,
      "step": 11495
    },
    {
      "epoch": 0.36809423212342357,
      "grad_norm": 0.0037509314715862274,
      "learning_rate": 0.24287334390717738,
      "loss": 0.6118,
      "num_input_tokens_seen": 15810480,
      "step": 11500
    },
    {
      "epoch": 0.36825427309391207,
      "grad_norm": 0.011298328638076782,
      "learning_rate": 0.24282708062591268,
      "loss": 0.572,
      "num_input_tokens_seen": 15816960,
      "step": 11505
    },
    {
      "epoch": 0.3684143140644005,
      "grad_norm": 0.019873259589076042,
      "learning_rate": 0.24278080302954563,
      "loss": 0.7611,
      "num_input_tokens_seen": 15824032,
      "step": 11510
    },
    {
      "epoch": 0.36857435503488895,
      "grad_norm": 0.008034049533307552,
      "learning_rate": 0.24273451112521283,
      "loss": 0.7961,
      "num_input_tokens_seen": 15830896,
      "step": 11515
    },
    {
      "epoch": 0.3687343960053774,
      "grad_norm": 0.005625423975288868,
      "learning_rate": 0.242688204920053,
      "loss": 0.6401,
      "num_input_tokens_seen": 15837104,
      "step": 11520
    },
    {
      "epoch": 0.3688944369758658,
      "grad_norm": 0.005100298207253218,
      "learning_rate": 0.24264188442120715,
      "loss": 0.7131,
      "num_input_tokens_seen": 15843536,
      "step": 11525
    },
    {
      "epoch": 0.36905447794635426,
      "grad_norm": 0.004261687397956848,
      "learning_rate": 0.24259554963581853,
      "loss": 0.6447,
      "num_input_tokens_seen": 15849840,
      "step": 11530
    },
    {
      "epoch": 0.3692145189168427,
      "grad_norm": 0.006413452327251434,
      "learning_rate": 0.24254920057103257,
      "loss": 0.7197,
      "num_input_tokens_seen": 15856384,
      "step": 11535
    },
    {
      "epoch": 0.36937455988733114,
      "grad_norm": 0.006475765258073807,
      "learning_rate": 0.24250283723399685,
      "loss": 0.7535,
      "num_input_tokens_seen": 15863408,
      "step": 11540
    },
    {
      "epoch": 0.3695346008578196,
      "grad_norm": 0.0038498761132359505,
      "learning_rate": 0.24245645963186108,
      "loss": 0.4772,
      "num_input_tokens_seen": 15870160,
      "step": 11545
    },
    {
      "epoch": 0.369694641828308,
      "grad_norm": 0.006547307129949331,
      "learning_rate": 0.2424100677717774,
      "loss": 0.5069,
      "num_input_tokens_seen": 15877008,
      "step": 11550
    },
    {
      "epoch": 0.3698546827987965,
      "grad_norm": 0.003823584411293268,
      "learning_rate": 0.24236366166090004,
      "loss": 0.4931,
      "num_input_tokens_seen": 15883760,
      "step": 11555
    },
    {
      "epoch": 0.37001472376928496,
      "grad_norm": 0.004098910838365555,
      "learning_rate": 0.24231724130638527,
      "loss": 0.5211,
      "num_input_tokens_seen": 15890592,
      "step": 11560
    },
    {
      "epoch": 0.3701747647397734,
      "grad_norm": 0.005608234088867903,
      "learning_rate": 0.2422708067153917,
      "loss": 0.7263,
      "num_input_tokens_seen": 15897568,
      "step": 11565
    },
    {
      "epoch": 0.37033480571026184,
      "grad_norm": 0.0037263829726725817,
      "learning_rate": 0.24222435789508026,
      "loss": 0.4992,
      "num_input_tokens_seen": 15904384,
      "step": 11570
    },
    {
      "epoch": 0.3704948466807503,
      "grad_norm": 0.005699122790247202,
      "learning_rate": 0.24217789485261387,
      "loss": 0.4824,
      "num_input_tokens_seen": 15910960,
      "step": 11575
    },
    {
      "epoch": 0.3706548876512387,
      "grad_norm": 0.005160016473382711,
      "learning_rate": 0.2421314175951577,
      "loss": 0.5171,
      "num_input_tokens_seen": 15918256,
      "step": 11580
    },
    {
      "epoch": 0.37081492862172716,
      "grad_norm": 0.005507413763552904,
      "learning_rate": 0.2420849261298791,
      "loss": 0.5804,
      "num_input_tokens_seen": 15925024,
      "step": 11585
    },
    {
      "epoch": 0.3709749695922156,
      "grad_norm": 0.0129706971347332,
      "learning_rate": 0.24203842046394775,
      "loss": 0.7114,
      "num_input_tokens_seen": 15931824,
      "step": 11590
    },
    {
      "epoch": 0.37113501056270404,
      "grad_norm": 0.005483701825141907,
      "learning_rate": 0.24199190060453535,
      "loss": 0.5288,
      "num_input_tokens_seen": 15939104,
      "step": 11595
    },
    {
      "epoch": 0.3712950515331925,
      "grad_norm": 0.004393855109810829,
      "learning_rate": 0.2419453665588158,
      "loss": 0.5864,
      "num_input_tokens_seen": 15946384,
      "step": 11600
    },
    {
      "epoch": 0.3712950515331925,
      "eval_loss": 0.6170587539672852,
      "eval_runtime": 331.7536,
      "eval_samples_per_second": 41.856,
      "eval_steps_per_second": 20.928,
      "num_input_tokens_seen": 15946384,
      "step": 11600
    },
    {
      "epoch": 0.3714550925036809,
      "grad_norm": 0.002989707048982382,
      "learning_rate": 0.24189881833396523,
      "loss": 0.5951,
      "num_input_tokens_seen": 15953024,
      "step": 11605
    },
    {
      "epoch": 0.3716151334741694,
      "grad_norm": 0.01027976255863905,
      "learning_rate": 0.24185225593716203,
      "loss": 0.4953,
      "num_input_tokens_seen": 15959904,
      "step": 11610
    },
    {
      "epoch": 0.37177517444465785,
      "grad_norm": 0.008064164780080318,
      "learning_rate": 0.2418056793755867,
      "loss": 0.6465,
      "num_input_tokens_seen": 15967008,
      "step": 11615
    },
    {
      "epoch": 0.3719352154151463,
      "grad_norm": 0.003903935430571437,
      "learning_rate": 0.24175908865642187,
      "loss": 0.4883,
      "num_input_tokens_seen": 15973584,
      "step": 11620
    },
    {
      "epoch": 0.37209525638563473,
      "grad_norm": 0.005468128249049187,
      "learning_rate": 0.24171248378685248,
      "loss": 0.577,
      "num_input_tokens_seen": 15980400,
      "step": 11625
    },
    {
      "epoch": 0.37225529735612317,
      "grad_norm": 0.007422845810651779,
      "learning_rate": 0.24166586477406554,
      "loss": 0.7308,
      "num_input_tokens_seen": 15987136,
      "step": 11630
    },
    {
      "epoch": 0.3724153383266116,
      "grad_norm": 0.00503882672637701,
      "learning_rate": 0.24161923162525034,
      "loss": 0.6359,
      "num_input_tokens_seen": 15994064,
      "step": 11635
    },
    {
      "epoch": 0.37257537929710005,
      "grad_norm": 0.013464143499732018,
      "learning_rate": 0.2415725843475982,
      "loss": 0.6689,
      "num_input_tokens_seen": 16000752,
      "step": 11640
    },
    {
      "epoch": 0.3727354202675885,
      "grad_norm": 0.0071204230189323425,
      "learning_rate": 0.24152592294830286,
      "loss": 0.5135,
      "num_input_tokens_seen": 16007664,
      "step": 11645
    },
    {
      "epoch": 0.37289546123807693,
      "grad_norm": 0.00897814892232418,
      "learning_rate": 0.24147924743455995,
      "loss": 0.5227,
      "num_input_tokens_seen": 16014576,
      "step": 11650
    },
    {
      "epoch": 0.37305550220856537,
      "grad_norm": 0.007364772260189056,
      "learning_rate": 0.24143255781356754,
      "loss": 0.5604,
      "num_input_tokens_seen": 16021424,
      "step": 11655
    },
    {
      "epoch": 0.3732155431790538,
      "grad_norm": 0.004284643568098545,
      "learning_rate": 0.24138585409252566,
      "loss": 0.8842,
      "num_input_tokens_seen": 16028304,
      "step": 11660
    },
    {
      "epoch": 0.3733755841495423,
      "grad_norm": 0.0038875185418874025,
      "learning_rate": 0.24133913627863662,
      "loss": 0.5675,
      "num_input_tokens_seen": 16035360,
      "step": 11665
    },
    {
      "epoch": 0.37353562512003075,
      "grad_norm": 0.00730394059792161,
      "learning_rate": 0.241292404379105,
      "loss": 0.6425,
      "num_input_tokens_seen": 16041920,
      "step": 11670
    },
    {
      "epoch": 0.3736956660905192,
      "grad_norm": 0.006033908110111952,
      "learning_rate": 0.24124565840113735,
      "loss": 0.5554,
      "num_input_tokens_seen": 16049088,
      "step": 11675
    },
    {
      "epoch": 0.3738557070610076,
      "grad_norm": 0.01061287336051464,
      "learning_rate": 0.2411988983519425,
      "loss": 0.4658,
      "num_input_tokens_seen": 16055440,
      "step": 11680
    },
    {
      "epoch": 0.37401574803149606,
      "grad_norm": 0.007285709958523512,
      "learning_rate": 0.24115212423873145,
      "loss": 0.5564,
      "num_input_tokens_seen": 16062256,
      "step": 11685
    },
    {
      "epoch": 0.3741757890019845,
      "grad_norm": 0.009295105002820492,
      "learning_rate": 0.24110533606871737,
      "loss": 0.6241,
      "num_input_tokens_seen": 16068832,
      "step": 11690
    },
    {
      "epoch": 0.37433582997247294,
      "grad_norm": 0.007229440379887819,
      "learning_rate": 0.24105853384911552,
      "loss": 0.6669,
      "num_input_tokens_seen": 16075968,
      "step": 11695
    },
    {
      "epoch": 0.3744958709429614,
      "grad_norm": 0.006276176776736975,
      "learning_rate": 0.24101171758714346,
      "loss": 0.313,
      "num_input_tokens_seen": 16082896,
      "step": 11700
    },
    {
      "epoch": 0.3746559119134498,
      "grad_norm": 0.010929251089692116,
      "learning_rate": 0.24096488729002086,
      "loss": 0.5215,
      "num_input_tokens_seen": 16089584,
      "step": 11705
    },
    {
      "epoch": 0.37481595288393826,
      "grad_norm": 0.00813973881304264,
      "learning_rate": 0.24091804296496946,
      "loss": 0.7922,
      "num_input_tokens_seen": 16096608,
      "step": 11710
    },
    {
      "epoch": 0.37497599385442676,
      "grad_norm": 0.007638320792466402,
      "learning_rate": 0.2408711846192133,
      "loss": 0.5078,
      "num_input_tokens_seen": 16103424,
      "step": 11715
    },
    {
      "epoch": 0.3751360348249152,
      "grad_norm": 0.011018030345439911,
      "learning_rate": 0.24082431225997855,
      "loss": 0.5885,
      "num_input_tokens_seen": 16110240,
      "step": 11720
    },
    {
      "epoch": 0.37529607579540364,
      "grad_norm": 0.010088639333844185,
      "learning_rate": 0.24077742589449344,
      "loss": 0.646,
      "num_input_tokens_seen": 16117440,
      "step": 11725
    },
    {
      "epoch": 0.3754561167658921,
      "grad_norm": 0.007969040423631668,
      "learning_rate": 0.24073052552998844,
      "loss": 0.5629,
      "num_input_tokens_seen": 16124176,
      "step": 11730
    },
    {
      "epoch": 0.3756161577363805,
      "grad_norm": 0.006672032177448273,
      "learning_rate": 0.2406836111736963,
      "loss": 0.6109,
      "num_input_tokens_seen": 16131056,
      "step": 11735
    },
    {
      "epoch": 0.37577619870686896,
      "grad_norm": 0.007653433829545975,
      "learning_rate": 0.2406366828328517,
      "loss": 0.5504,
      "num_input_tokens_seen": 16137904,
      "step": 11740
    },
    {
      "epoch": 0.3759362396773574,
      "grad_norm": 0.005922604352235794,
      "learning_rate": 0.2405897405146915,
      "loss": 0.4605,
      "num_input_tokens_seen": 16144640,
      "step": 11745
    },
    {
      "epoch": 0.37609628064784584,
      "grad_norm": 0.008233539760112762,
      "learning_rate": 0.240542784226455,
      "loss": 0.6102,
      "num_input_tokens_seen": 16151408,
      "step": 11750
    },
    {
      "epoch": 0.3762563216183343,
      "grad_norm": 0.009757982566952705,
      "learning_rate": 0.24049581397538328,
      "loss": 0.6945,
      "num_input_tokens_seen": 16158640,
      "step": 11755
    },
    {
      "epoch": 0.3764163625888227,
      "grad_norm": 0.009462387301027775,
      "learning_rate": 0.24044882976871984,
      "loss": 0.6746,
      "num_input_tokens_seen": 16165808,
      "step": 11760
    },
    {
      "epoch": 0.37657640355931116,
      "grad_norm": 0.002806768985465169,
      "learning_rate": 0.2404018316137102,
      "loss": 0.6061,
      "num_input_tokens_seen": 16172480,
      "step": 11765
    },
    {
      "epoch": 0.37673644452979965,
      "grad_norm": 0.008621922694146633,
      "learning_rate": 0.24035481951760204,
      "loss": 0.6305,
      "num_input_tokens_seen": 16179152,
      "step": 11770
    },
    {
      "epoch": 0.3768964855002881,
      "grad_norm": 0.003324169898405671,
      "learning_rate": 0.2403077934876452,
      "loss": 0.5377,
      "num_input_tokens_seen": 16185696,
      "step": 11775
    },
    {
      "epoch": 0.37705652647077653,
      "grad_norm": 0.004109563305974007,
      "learning_rate": 0.2402607535310918,
      "loss": 0.4272,
      "num_input_tokens_seen": 16192672,
      "step": 11780
    },
    {
      "epoch": 0.37721656744126497,
      "grad_norm": 0.011993993073701859,
      "learning_rate": 0.2402136996551959,
      "loss": 0.7839,
      "num_input_tokens_seen": 16199712,
      "step": 11785
    },
    {
      "epoch": 0.3773766084117534,
      "grad_norm": 0.006536619737744331,
      "learning_rate": 0.24016663186721376,
      "loss": 0.5951,
      "num_input_tokens_seen": 16206464,
      "step": 11790
    },
    {
      "epoch": 0.37753664938224185,
      "grad_norm": 0.006806665565818548,
      "learning_rate": 0.24011955017440395,
      "loss": 0.5173,
      "num_input_tokens_seen": 16213376,
      "step": 11795
    },
    {
      "epoch": 0.3776966903527303,
      "grad_norm": 0.005951316095888615,
      "learning_rate": 0.24007245458402696,
      "loss": 0.3912,
      "num_input_tokens_seen": 16220112,
      "step": 11800
    },
    {
      "epoch": 0.3776966903527303,
      "eval_loss": 0.6201618909835815,
      "eval_runtime": 332.0633,
      "eval_samples_per_second": 41.817,
      "eval_steps_per_second": 20.909,
      "num_input_tokens_seen": 16220112,
      "step": 11800
    },
    {
      "epoch": 0.37785673132321873,
      "grad_norm": 0.01184309646487236,
      "learning_rate": 0.2400253451033456,
      "loss": 0.566,
      "num_input_tokens_seen": 16226912,
      "step": 11805
    },
    {
      "epoch": 0.37801677229370717,
      "grad_norm": 0.022304212674498558,
      "learning_rate": 0.23997822173962463,
      "loss": 0.664,
      "num_input_tokens_seen": 16234048,
      "step": 11810
    },
    {
      "epoch": 0.3781768132641956,
      "grad_norm": 0.006211261264979839,
      "learning_rate": 0.23993108450013118,
      "loss": 0.7092,
      "num_input_tokens_seen": 16240560,
      "step": 11815
    },
    {
      "epoch": 0.3783368542346841,
      "grad_norm": 0.004982175771147013,
      "learning_rate": 0.2398839333921343,
      "loss": 0.7233,
      "num_input_tokens_seen": 16247200,
      "step": 11820
    },
    {
      "epoch": 0.37849689520517255,
      "grad_norm": 0.0054908981546759605,
      "learning_rate": 0.23983676842290536,
      "loss": 0.5388,
      "num_input_tokens_seen": 16254080,
      "step": 11825
    },
    {
      "epoch": 0.378656936175661,
      "grad_norm": 0.006276071071624756,
      "learning_rate": 0.2397895895997178,
      "loss": 0.5825,
      "num_input_tokens_seen": 16260896,
      "step": 11830
    },
    {
      "epoch": 0.3788169771461494,
      "grad_norm": 0.006114635616540909,
      "learning_rate": 0.23974239692984714,
      "loss": 0.6868,
      "num_input_tokens_seen": 16267600,
      "step": 11835
    },
    {
      "epoch": 0.37897701811663786,
      "grad_norm": 0.0055734398774802685,
      "learning_rate": 0.2396951904205711,
      "loss": 0.6178,
      "num_input_tokens_seen": 16274672,
      "step": 11840
    },
    {
      "epoch": 0.3791370590871263,
      "grad_norm": 0.008015085011720657,
      "learning_rate": 0.23964797007916952,
      "loss": 0.4819,
      "num_input_tokens_seen": 16281760,
      "step": 11845
    },
    {
      "epoch": 0.37929710005761474,
      "grad_norm": 0.006969295907765627,
      "learning_rate": 0.23960073591292436,
      "loss": 0.6,
      "num_input_tokens_seen": 16288656,
      "step": 11850
    },
    {
      "epoch": 0.3794571410281032,
      "grad_norm": 0.005740500520914793,
      "learning_rate": 0.2395534879291197,
      "loss": 0.5176,
      "num_input_tokens_seen": 16295472,
      "step": 11855
    },
    {
      "epoch": 0.3796171819985916,
      "grad_norm": 0.004689689259976149,
      "learning_rate": 0.23950622613504186,
      "loss": 0.5743,
      "num_input_tokens_seen": 16302208,
      "step": 11860
    },
    {
      "epoch": 0.37977722296908006,
      "grad_norm": 0.007811552844941616,
      "learning_rate": 0.2394589505379791,
      "loss": 0.6314,
      "num_input_tokens_seen": 16308992,
      "step": 11865
    },
    {
      "epoch": 0.3799372639395685,
      "grad_norm": 0.005761782173067331,
      "learning_rate": 0.23941166114522197,
      "loss": 0.5266,
      "num_input_tokens_seen": 16315936,
      "step": 11870
    },
    {
      "epoch": 0.380097304910057,
      "grad_norm": 0.005854453891515732,
      "learning_rate": 0.23936435796406308,
      "loss": 0.6078,
      "num_input_tokens_seen": 16322352,
      "step": 11875
    },
    {
      "epoch": 0.38025734588054544,
      "grad_norm": 0.005885271355509758,
      "learning_rate": 0.23931704100179715,
      "loss": 0.7619,
      "num_input_tokens_seen": 16329248,
      "step": 11880
    },
    {
      "epoch": 0.3804173868510339,
      "grad_norm": 0.004895459860563278,
      "learning_rate": 0.2392697102657211,
      "loss": 0.5226,
      "num_input_tokens_seen": 16335984,
      "step": 11885
    },
    {
      "epoch": 0.3805774278215223,
      "grad_norm": 0.003580384189262986,
      "learning_rate": 0.23922236576313388,
      "loss": 0.5159,
      "num_input_tokens_seen": 16342864,
      "step": 11890
    },
    {
      "epoch": 0.38073746879201076,
      "grad_norm": 0.007401067763566971,
      "learning_rate": 0.2391750075013366,
      "loss": 0.6861,
      "num_input_tokens_seen": 16349904,
      "step": 11895
    },
    {
      "epoch": 0.3808975097624992,
      "grad_norm": 0.005428956355899572,
      "learning_rate": 0.2391276354876326,
      "loss": 0.4768,
      "num_input_tokens_seen": 16356256,
      "step": 11900
    },
    {
      "epoch": 0.38105755073298764,
      "grad_norm": 0.008215093985199928,
      "learning_rate": 0.23908024972932707,
      "loss": 0.5082,
      "num_input_tokens_seen": 16363424,
      "step": 11905
    },
    {
      "epoch": 0.3812175917034761,
      "grad_norm": 0.005099376197904348,
      "learning_rate": 0.2390328502337276,
      "loss": 0.6432,
      "num_input_tokens_seen": 16370784,
      "step": 11910
    },
    {
      "epoch": 0.3813776326739645,
      "grad_norm": 0.003868791041895747,
      "learning_rate": 0.23898543700814376,
      "loss": 0.6191,
      "num_input_tokens_seen": 16377952,
      "step": 11915
    },
    {
      "epoch": 0.38153767364445296,
      "grad_norm": 0.006912401877343655,
      "learning_rate": 0.2389380100598873,
      "loss": 0.3275,
      "num_input_tokens_seen": 16384640,
      "step": 11920
    },
    {
      "epoch": 0.38169771461494145,
      "grad_norm": 0.00411196518689394,
      "learning_rate": 0.23889056939627207,
      "loss": 0.5002,
      "num_input_tokens_seen": 16391312,
      "step": 11925
    },
    {
      "epoch": 0.3818577555854299,
      "grad_norm": 0.011513905599713326,
      "learning_rate": 0.23884311502461386,
      "loss": 0.6286,
      "num_input_tokens_seen": 16398176,
      "step": 11930
    },
    {
      "epoch": 0.38201779655591833,
      "grad_norm": 0.004355521872639656,
      "learning_rate": 0.23879564695223088,
      "loss": 0.5582,
      "num_input_tokens_seen": 16404608,
      "step": 11935
    },
    {
      "epoch": 0.38217783752640677,
      "grad_norm": 0.009151594713330269,
      "learning_rate": 0.23874816518644332,
      "loss": 0.6557,
      "num_input_tokens_seen": 16411696,
      "step": 11940
    },
    {
      "epoch": 0.3823378784968952,
      "grad_norm": 0.007119524758309126,
      "learning_rate": 0.23870066973457335,
      "loss": 0.5183,
      "num_input_tokens_seen": 16419184,
      "step": 11945
    },
    {
      "epoch": 0.38249791946738365,
      "grad_norm": 0.004672287032008171,
      "learning_rate": 0.23865316060394545,
      "loss": 0.5659,
      "num_input_tokens_seen": 16426208,
      "step": 11950
    },
    {
      "epoch": 0.3826579604378721,
      "grad_norm": 0.00503682903945446,
      "learning_rate": 0.2386056378018861,
      "loss": 0.5923,
      "num_input_tokens_seen": 16432832,
      "step": 11955
    },
    {
      "epoch": 0.38281800140836053,
      "grad_norm": 0.00661284988746047,
      "learning_rate": 0.2385581013357239,
      "loss": 0.703,
      "num_input_tokens_seen": 16439504,
      "step": 11960
    },
    {
      "epoch": 0.38297804237884897,
      "grad_norm": 0.008085505105555058,
      "learning_rate": 0.23851055121278958,
      "loss": 0.5308,
      "num_input_tokens_seen": 16446160,
      "step": 11965
    },
    {
      "epoch": 0.3831380833493374,
      "grad_norm": 0.00792680587619543,
      "learning_rate": 0.23846298744041594,
      "loss": 0.6212,
      "num_input_tokens_seen": 16453184,
      "step": 11970
    },
    {
      "epoch": 0.38329812431982585,
      "grad_norm": 0.006900610867887735,
      "learning_rate": 0.23841541002593802,
      "loss": 0.3436,
      "num_input_tokens_seen": 16459712,
      "step": 11975
    },
    {
      "epoch": 0.38345816529031435,
      "grad_norm": 0.012076684273779392,
      "learning_rate": 0.23836781897669276,
      "loss": 0.5953,
      "num_input_tokens_seen": 16466528,
      "step": 11980
    },
    {
      "epoch": 0.3836182062608028,
      "grad_norm": 0.009246500208973885,
      "learning_rate": 0.23832021430001926,
      "loss": 0.7308,
      "num_input_tokens_seen": 16473408,
      "step": 11985
    },
    {
      "epoch": 0.3837782472312912,
      "grad_norm": 0.004258433356881142,
      "learning_rate": 0.2382725960032588,
      "loss": 0.5035,
      "num_input_tokens_seen": 16480240,
      "step": 11990
    },
    {
      "epoch": 0.38393828820177966,
      "grad_norm": 0.007126282900571823,
      "learning_rate": 0.23822496409375482,
      "loss": 0.629,
      "num_input_tokens_seen": 16486880,
      "step": 11995
    },
    {
      "epoch": 0.3840983291722681,
      "grad_norm": 0.004332502372562885,
      "learning_rate": 0.2381773185788526,
      "loss": 0.6575,
      "num_input_tokens_seen": 16493920,
      "step": 12000
    },
    {
      "epoch": 0.3840983291722681,
      "eval_loss": 0.619225800037384,
      "eval_runtime": 331.9739,
      "eval_samples_per_second": 41.829,
      "eval_steps_per_second": 20.914,
      "num_input_tokens_seen": 16493920,
      "step": 12000
    },
    {
      "epoch": 0.38425837014275654,
      "grad_norm": 0.00479172682389617,
      "learning_rate": 0.2381296594658998,
      "loss": 0.5016,
      "num_input_tokens_seen": 16500720,
      "step": 12005
    },
    {
      "epoch": 0.384418411113245,
      "grad_norm": 0.028605621308088303,
      "learning_rate": 0.238081986762246,
      "loss": 0.7156,
      "num_input_tokens_seen": 16507504,
      "step": 12010
    },
    {
      "epoch": 0.3845784520837334,
      "grad_norm": 0.008796527981758118,
      "learning_rate": 0.23803430047524293,
      "loss": 0.5677,
      "num_input_tokens_seen": 16513920,
      "step": 12015
    },
    {
      "epoch": 0.38473849305422186,
      "grad_norm": 0.006764411460608244,
      "learning_rate": 0.23798660061224441,
      "loss": 0.6284,
      "num_input_tokens_seen": 16520912,
      "step": 12020
    },
    {
      "epoch": 0.3848985340247103,
      "grad_norm": 0.0059520951472222805,
      "learning_rate": 0.23793888718060632,
      "loss": 0.5112,
      "num_input_tokens_seen": 16528160,
      "step": 12025
    },
    {
      "epoch": 0.38505857499519874,
      "grad_norm": 0.00557298818603158,
      "learning_rate": 0.23789116018768675,
      "loss": 0.7185,
      "num_input_tokens_seen": 16534848,
      "step": 12030
    },
    {
      "epoch": 0.38521861596568724,
      "grad_norm": 0.005258277058601379,
      "learning_rate": 0.2378434196408458,
      "loss": 0.5072,
      "num_input_tokens_seen": 16541472,
      "step": 12035
    },
    {
      "epoch": 0.3853786569361757,
      "grad_norm": 0.004562930669635534,
      "learning_rate": 0.23779566554744563,
      "loss": 0.4711,
      "num_input_tokens_seen": 16548640,
      "step": 12040
    },
    {
      "epoch": 0.3855386979066641,
      "grad_norm": 0.007205691654235125,
      "learning_rate": 0.23774789791485051,
      "loss": 0.8741,
      "num_input_tokens_seen": 16555600,
      "step": 12045
    },
    {
      "epoch": 0.38569873887715256,
      "grad_norm": 0.007290413603186607,
      "learning_rate": 0.2377001167504268,
      "loss": 0.7385,
      "num_input_tokens_seen": 16562576,
      "step": 12050
    },
    {
      "epoch": 0.385858779847641,
      "grad_norm": 0.005758518818765879,
      "learning_rate": 0.23765232206154302,
      "loss": 0.4114,
      "num_input_tokens_seen": 16569552,
      "step": 12055
    },
    {
      "epoch": 0.38601882081812944,
      "grad_norm": 0.0036183991469442844,
      "learning_rate": 0.23760451385556966,
      "loss": 0.6167,
      "num_input_tokens_seen": 16576096,
      "step": 12060
    },
    {
      "epoch": 0.3861788617886179,
      "grad_norm": 0.0068656099028885365,
      "learning_rate": 0.23755669213987932,
      "loss": 0.5353,
      "num_input_tokens_seen": 16583152,
      "step": 12065
    },
    {
      "epoch": 0.3863389027591063,
      "grad_norm": 0.004406407941132784,
      "learning_rate": 0.23750885692184676,
      "loss": 0.6209,
      "num_input_tokens_seen": 16590272,
      "step": 12070
    },
    {
      "epoch": 0.38649894372959476,
      "grad_norm": 0.004251023754477501,
      "learning_rate": 0.23746100820884875,
      "loss": 0.5612,
      "num_input_tokens_seen": 16596752,
      "step": 12075
    },
    {
      "epoch": 0.3866589847000832,
      "grad_norm": 0.005303858779370785,
      "learning_rate": 0.23741314600826421,
      "loss": 0.3829,
      "num_input_tokens_seen": 16603616,
      "step": 12080
    },
    {
      "epoch": 0.3868190256705717,
      "grad_norm": 0.010618059895932674,
      "learning_rate": 0.23736527032747406,
      "loss": 0.5391,
      "num_input_tokens_seen": 16610256,
      "step": 12085
    },
    {
      "epoch": 0.38697906664106013,
      "grad_norm": 0.006532068829983473,
      "learning_rate": 0.23731738117386128,
      "loss": 0.6169,
      "num_input_tokens_seen": 16617120,
      "step": 12090
    },
    {
      "epoch": 0.38713910761154857,
      "grad_norm": 0.011559724807739258,
      "learning_rate": 0.237269478554811,
      "loss": 0.8254,
      "num_input_tokens_seen": 16623536,
      "step": 12095
    },
    {
      "epoch": 0.387299148582037,
      "grad_norm": 0.005416445899754763,
      "learning_rate": 0.23722156247771053,
      "loss": 0.5374,
      "num_input_tokens_seen": 16630032,
      "step": 12100
    },
    {
      "epoch": 0.38745918955252545,
      "grad_norm": 0.0098112178966403,
      "learning_rate": 0.23717363294994895,
      "loss": 0.8019,
      "num_input_tokens_seen": 16636976,
      "step": 12105
    },
    {
      "epoch": 0.3876192305230139,
      "grad_norm": 0.003934196662157774,
      "learning_rate": 0.2371256899789177,
      "loss": 0.5199,
      "num_input_tokens_seen": 16643824,
      "step": 12110
    },
    {
      "epoch": 0.38777927149350233,
      "grad_norm": 0.004911189433187246,
      "learning_rate": 0.23707773357201017,
      "loss": 0.4951,
      "num_input_tokens_seen": 16650400,
      "step": 12115
    },
    {
      "epoch": 0.38793931246399077,
      "grad_norm": 0.005412020720541477,
      "learning_rate": 0.2370297637366218,
      "loss": 0.5411,
      "num_input_tokens_seen": 16657056,
      "step": 12120
    },
    {
      "epoch": 0.3880993534344792,
      "grad_norm": 0.004848657175898552,
      "learning_rate": 0.23698178048015026,
      "loss": 0.5609,
      "num_input_tokens_seen": 16664240,
      "step": 12125
    },
    {
      "epoch": 0.38825939440496765,
      "grad_norm": 0.00314113637432456,
      "learning_rate": 0.236933783809995,
      "loss": 0.5183,
      "num_input_tokens_seen": 16670800,
      "step": 12130
    },
    {
      "epoch": 0.3884194353754561,
      "grad_norm": 0.005979869049042463,
      "learning_rate": 0.23688577373355785,
      "loss": 0.6536,
      "num_input_tokens_seen": 16677792,
      "step": 12135
    },
    {
      "epoch": 0.3885794763459446,
      "grad_norm": 0.007960785180330276,
      "learning_rate": 0.23683775025824247,
      "loss": 0.4362,
      "num_input_tokens_seen": 16684528,
      "step": 12140
    },
    {
      "epoch": 0.388739517316433,
      "grad_norm": 0.00840382743626833,
      "learning_rate": 0.2367897133914548,
      "loss": 0.5485,
      "num_input_tokens_seen": 16691376,
      "step": 12145
    },
    {
      "epoch": 0.38889955828692147,
      "grad_norm": 0.009089718572795391,
      "learning_rate": 0.2367416631406026,
      "loss": 0.6144,
      "num_input_tokens_seen": 16698400,
      "step": 12150
    },
    {
      "epoch": 0.3890595992574099,
      "grad_norm": 0.007935242727398872,
      "learning_rate": 0.23669359951309588,
      "loss": 0.57,
      "num_input_tokens_seen": 16705808,
      "step": 12155
    },
    {
      "epoch": 0.38921964022789834,
      "grad_norm": 0.006243469193577766,
      "learning_rate": 0.23664552251634666,
      "loss": 0.7246,
      "num_input_tokens_seen": 16713120,
      "step": 12160
    },
    {
      "epoch": 0.3893796811983868,
      "grad_norm": 0.006515325978398323,
      "learning_rate": 0.23659743215776907,
      "loss": 0.6939,
      "num_input_tokens_seen": 16719808,
      "step": 12165
    },
    {
      "epoch": 0.3895397221688752,
      "grad_norm": 0.006543295457959175,
      "learning_rate": 0.23654932844477908,
      "loss": 0.5656,
      "num_input_tokens_seen": 16726512,
      "step": 12170
    },
    {
      "epoch": 0.38969976313936366,
      "grad_norm": 0.007679723668843508,
      "learning_rate": 0.23650121138479507,
      "loss": 0.7633,
      "num_input_tokens_seen": 16733424,
      "step": 12175
    },
    {
      "epoch": 0.3898598041098521,
      "grad_norm": 0.002586660673841834,
      "learning_rate": 0.23645308098523724,
      "loss": 0.6467,
      "num_input_tokens_seen": 16740224,
      "step": 12180
    },
    {
      "epoch": 0.39001984508034054,
      "grad_norm": 0.005388784222304821,
      "learning_rate": 0.23640493725352785,
      "loss": 0.8036,
      "num_input_tokens_seen": 16747152,
      "step": 12185
    },
    {
      "epoch": 0.39017988605082904,
      "grad_norm": 0.008522900752723217,
      "learning_rate": 0.2363567801970913,
      "loss": 0.9059,
      "num_input_tokens_seen": 16757456,
      "step": 12190
    },
    {
      "epoch": 0.3903399270213175,
      "grad_norm": 0.0023352333810180426,
      "learning_rate": 0.236308609823354,
      "loss": 0.3315,
      "num_input_tokens_seen": 16764496,
      "step": 12195
    },
    {
      "epoch": 0.3904999679918059,
      "grad_norm": 0.004016270395368338,
      "learning_rate": 0.23626042613974452,
      "loss": 0.5405,
      "num_input_tokens_seen": 16771376,
      "step": 12200
    },
    {
      "epoch": 0.3904999679918059,
      "eval_loss": 0.5922185778617859,
      "eval_runtime": 331.7429,
      "eval_samples_per_second": 41.858,
      "eval_steps_per_second": 20.929,
      "num_input_tokens_seen": 16771376,
      "step": 12200
    },
    {
      "epoch": 0.39066000896229436,
      "grad_norm": 0.004531036596745253,
      "learning_rate": 0.23621222915369325,
      "loss": 0.5179,
      "num_input_tokens_seen": 16778048,
      "step": 12205
    },
    {
      "epoch": 0.3908200499327828,
      "grad_norm": 0.004970494192093611,
      "learning_rate": 0.23616401887263283,
      "loss": 0.5444,
      "num_input_tokens_seen": 16784864,
      "step": 12210
    },
    {
      "epoch": 0.39098009090327124,
      "grad_norm": 0.011316511780023575,
      "learning_rate": 0.23611579530399793,
      "loss": 0.5455,
      "num_input_tokens_seen": 16791760,
      "step": 12215
    },
    {
      "epoch": 0.3911401318737597,
      "grad_norm": 0.008788762614130974,
      "learning_rate": 0.23606755845522517,
      "loss": 0.4679,
      "num_input_tokens_seen": 16798752,
      "step": 12220
    },
    {
      "epoch": 0.3913001728442481,
      "grad_norm": 0.008752801455557346,
      "learning_rate": 0.23601930833375329,
      "loss": 0.6541,
      "num_input_tokens_seen": 16805648,
      "step": 12225
    },
    {
      "epoch": 0.39146021381473656,
      "grad_norm": 0.0039950180798769,
      "learning_rate": 0.23597104494702312,
      "loss": 0.6407,
      "num_input_tokens_seen": 16812624,
      "step": 12230
    },
    {
      "epoch": 0.391620254785225,
      "grad_norm": 0.005795512814074755,
      "learning_rate": 0.23592276830247744,
      "loss": 0.6205,
      "num_input_tokens_seen": 16819520,
      "step": 12235
    },
    {
      "epoch": 0.39178029575571344,
      "grad_norm": 0.004561323206871748,
      "learning_rate": 0.2358744784075611,
      "loss": 0.504,
      "num_input_tokens_seen": 16826224,
      "step": 12240
    },
    {
      "epoch": 0.39194033672620193,
      "grad_norm": 0.0065803383477032185,
      "learning_rate": 0.235826175269721,
      "loss": 0.6074,
      "num_input_tokens_seen": 16833152,
      "step": 12245
    },
    {
      "epoch": 0.39210037769669037,
      "grad_norm": 0.006138889119029045,
      "learning_rate": 0.23577785889640612,
      "loss": 0.4944,
      "num_input_tokens_seen": 16839904,
      "step": 12250
    },
    {
      "epoch": 0.3922604186671788,
      "grad_norm": 0.004482930060476065,
      "learning_rate": 0.23572952929506744,
      "loss": 0.5105,
      "num_input_tokens_seen": 16847008,
      "step": 12255
    },
    {
      "epoch": 0.39242045963766725,
      "grad_norm": 0.00708662299439311,
      "learning_rate": 0.23568118647315803,
      "loss": 0.6722,
      "num_input_tokens_seen": 16853888,
      "step": 12260
    },
    {
      "epoch": 0.3925805006081557,
      "grad_norm": 0.005137661937624216,
      "learning_rate": 0.23563283043813296,
      "loss": 0.5208,
      "num_input_tokens_seen": 16860720,
      "step": 12265
    },
    {
      "epoch": 0.39274054157864413,
      "grad_norm": 0.003679192392155528,
      "learning_rate": 0.23558446119744922,
      "loss": 0.4746,
      "num_input_tokens_seen": 16867520,
      "step": 12270
    },
    {
      "epoch": 0.39290058254913257,
      "grad_norm": 0.005753814242780209,
      "learning_rate": 0.23553607875856608,
      "loss": 0.7605,
      "num_input_tokens_seen": 16874032,
      "step": 12275
    },
    {
      "epoch": 0.393060623519621,
      "grad_norm": 0.006358230486512184,
      "learning_rate": 0.2354876831289447,
      "loss": 0.6077,
      "num_input_tokens_seen": 16881360,
      "step": 12280
    },
    {
      "epoch": 0.39322066449010945,
      "grad_norm": 0.005486936308443546,
      "learning_rate": 0.23543927431604827,
      "loss": 0.7424,
      "num_input_tokens_seen": 16888304,
      "step": 12285
    },
    {
      "epoch": 0.3933807054605979,
      "grad_norm": 0.005651196464896202,
      "learning_rate": 0.23539085232734203,
      "loss": 0.6104,
      "num_input_tokens_seen": 16895184,
      "step": 12290
    },
    {
      "epoch": 0.3935407464310864,
      "grad_norm": 0.006345326546579599,
      "learning_rate": 0.2353424171702933,
      "loss": 0.42,
      "num_input_tokens_seen": 16902240,
      "step": 12295
    },
    {
      "epoch": 0.3937007874015748,
      "grad_norm": 0.00696270726621151,
      "learning_rate": 0.23529396885237133,
      "loss": 0.697,
      "num_input_tokens_seen": 16909040,
      "step": 12300
    },
    {
      "epoch": 0.39386082837206327,
      "grad_norm": 0.003552954411134124,
      "learning_rate": 0.2352455073810475,
      "loss": 0.458,
      "num_input_tokens_seen": 16915936,
      "step": 12305
    },
    {
      "epoch": 0.3940208693425517,
      "grad_norm": 0.009308304637670517,
      "learning_rate": 0.23519703276379517,
      "loss": 0.539,
      "num_input_tokens_seen": 16923072,
      "step": 12310
    },
    {
      "epoch": 0.39418091031304014,
      "grad_norm": 0.005884869024157524,
      "learning_rate": 0.2351485450080897,
      "loss": 0.4282,
      "num_input_tokens_seen": 16929744,
      "step": 12315
    },
    {
      "epoch": 0.3943409512835286,
      "grad_norm": 0.006616002414375544,
      "learning_rate": 0.2351000441214086,
      "loss": 0.6116,
      "num_input_tokens_seen": 16936864,
      "step": 12320
    },
    {
      "epoch": 0.394500992254017,
      "grad_norm": 0.0164946299046278,
      "learning_rate": 0.23505153011123125,
      "loss": 0.4174,
      "num_input_tokens_seen": 16943680,
      "step": 12325
    },
    {
      "epoch": 0.39466103322450546,
      "grad_norm": 0.005225038155913353,
      "learning_rate": 0.23500300298503912,
      "loss": 0.6547,
      "num_input_tokens_seen": 16950288,
      "step": 12330
    },
    {
      "epoch": 0.3948210741949939,
      "grad_norm": 0.003623554017394781,
      "learning_rate": 0.23495446275031576,
      "loss": 0.5665,
      "num_input_tokens_seen": 16956880,
      "step": 12335
    },
    {
      "epoch": 0.39498111516548234,
      "grad_norm": 0.0074448054656386375,
      "learning_rate": 0.2349059094145466,
      "loss": 0.6666,
      "num_input_tokens_seen": 16963520,
      "step": 12340
    },
    {
      "epoch": 0.3951411561359708,
      "grad_norm": 0.005952558945864439,
      "learning_rate": 0.2348573429852192,
      "loss": 0.6116,
      "num_input_tokens_seen": 16970336,
      "step": 12345
    },
    {
      "epoch": 0.3953011971064593,
      "grad_norm": 0.006778406444936991,
      "learning_rate": 0.23480876346982313,
      "loss": 0.636,
      "num_input_tokens_seen": 16977696,
      "step": 12350
    },
    {
      "epoch": 0.3954612380769477,
      "grad_norm": 0.0054815891198813915,
      "learning_rate": 0.23476017087585,
      "loss": 0.6398,
      "num_input_tokens_seen": 16984784,
      "step": 12355
    },
    {
      "epoch": 0.39562127904743616,
      "grad_norm": 0.003552732989192009,
      "learning_rate": 0.23471156521079334,
      "loss": 0.7517,
      "num_input_tokens_seen": 16991488,
      "step": 12360
    },
    {
      "epoch": 0.3957813200179246,
      "grad_norm": 0.004101182334125042,
      "learning_rate": 0.23466294648214875,
      "loss": 0.6047,
      "num_input_tokens_seen": 16998080,
      "step": 12365
    },
    {
      "epoch": 0.39594136098841304,
      "grad_norm": 0.0025234746281057596,
      "learning_rate": 0.2346143146974139,
      "loss": 0.5007,
      "num_input_tokens_seen": 17004608,
      "step": 12370
    },
    {
      "epoch": 0.3961014019589015,
      "grad_norm": 0.00497986376285553,
      "learning_rate": 0.23456566986408836,
      "loss": 0.515,
      "num_input_tokens_seen": 17011408,
      "step": 12375
    },
    {
      "epoch": 0.3962614429293899,
      "grad_norm": 0.004492671694606543,
      "learning_rate": 0.23451701198967384,
      "loss": 0.6699,
      "num_input_tokens_seen": 17018352,
      "step": 12380
    },
    {
      "epoch": 0.39642148389987836,
      "grad_norm": 0.006397861056029797,
      "learning_rate": 0.23446834108167397,
      "loss": 0.592,
      "num_input_tokens_seen": 17025776,
      "step": 12385
    },
    {
      "epoch": 0.3965815248703668,
      "grad_norm": 0.007806683424860239,
      "learning_rate": 0.23441965714759438,
      "loss": 0.6265,
      "num_input_tokens_seen": 17032880,
      "step": 12390
    },
    {
      "epoch": 0.39674156584085524,
      "grad_norm": 0.0071825385093688965,
      "learning_rate": 0.23437096019494277,
      "loss": 0.7352,
      "num_input_tokens_seen": 17040016,
      "step": 12395
    },
    {
      "epoch": 0.3969016068113437,
      "grad_norm": 0.004968507215380669,
      "learning_rate": 0.23432225023122885,
      "loss": 0.7828,
      "num_input_tokens_seen": 17046656,
      "step": 12400
    },
    {
      "epoch": 0.3969016068113437,
      "eval_loss": 0.5893734693527222,
      "eval_runtime": 332.0746,
      "eval_samples_per_second": 41.816,
      "eval_steps_per_second": 20.908,
      "num_input_tokens_seen": 17046656,
      "step": 12400
    },
    {
      "epoch": 0.39706164778183217,
      "grad_norm": 0.0059188175946474075,
      "learning_rate": 0.23427352726396428,
      "loss": 0.631,
      "num_input_tokens_seen": 17053904,
      "step": 12405
    },
    {
      "epoch": 0.3972216887523206,
      "grad_norm": 0.004707273095846176,
      "learning_rate": 0.2342247913006628,
      "loss": 0.4844,
      "num_input_tokens_seen": 17060624,
      "step": 12410
    },
    {
      "epoch": 0.39738172972280905,
      "grad_norm": 0.005044525023549795,
      "learning_rate": 0.23417604234883999,
      "loss": 0.6424,
      "num_input_tokens_seen": 17067216,
      "step": 12415
    },
    {
      "epoch": 0.3975417706932975,
      "grad_norm": 0.004225213546305895,
      "learning_rate": 0.23412728041601363,
      "loss": 0.4712,
      "num_input_tokens_seen": 17073984,
      "step": 12420
    },
    {
      "epoch": 0.39770181166378593,
      "grad_norm": 0.0015195246087387204,
      "learning_rate": 0.23407850550970347,
      "loss": 0.455,
      "num_input_tokens_seen": 17080624,
      "step": 12425
    },
    {
      "epoch": 0.39786185263427437,
      "grad_norm": 0.006713912822306156,
      "learning_rate": 0.23402971763743116,
      "loss": 0.6054,
      "num_input_tokens_seen": 17087568,
      "step": 12430
    },
    {
      "epoch": 0.3980218936047628,
      "grad_norm": 0.0056214104406535625,
      "learning_rate": 0.23398091680672037,
      "loss": 0.6692,
      "num_input_tokens_seen": 17094352,
      "step": 12435
    },
    {
      "epoch": 0.39818193457525125,
      "grad_norm": 0.0055718026123940945,
      "learning_rate": 0.23393210302509687,
      "loss": 0.4113,
      "num_input_tokens_seen": 17101200,
      "step": 12440
    },
    {
      "epoch": 0.3983419755457397,
      "grad_norm": 0.005448836833238602,
      "learning_rate": 0.23388327630008832,
      "loss": 0.447,
      "num_input_tokens_seen": 17108336,
      "step": 12445
    },
    {
      "epoch": 0.39850201651622813,
      "grad_norm": 0.005267130210995674,
      "learning_rate": 0.23383443663922443,
      "loss": 0.6076,
      "num_input_tokens_seen": 17114624,
      "step": 12450
    },
    {
      "epoch": 0.3986620574867166,
      "grad_norm": 0.0044997152872383595,
      "learning_rate": 0.23378558405003685,
      "loss": 0.4627,
      "num_input_tokens_seen": 17121280,
      "step": 12455
    },
    {
      "epoch": 0.39882209845720507,
      "grad_norm": 0.0022903424687683582,
      "learning_rate": 0.2337367185400593,
      "loss": 0.434,
      "num_input_tokens_seen": 17128272,
      "step": 12460
    },
    {
      "epoch": 0.3989821394276935,
      "grad_norm": 0.004356305114924908,
      "learning_rate": 0.23368784011682747,
      "loss": 0.7685,
      "num_input_tokens_seen": 17135472,
      "step": 12465
    },
    {
      "epoch": 0.39914218039818194,
      "grad_norm": 0.0030097460839897394,
      "learning_rate": 0.23363894878787902,
      "loss": 0.4414,
      "num_input_tokens_seen": 17142624,
      "step": 12470
    },
    {
      "epoch": 0.3993022213686704,
      "grad_norm": 0.0041924649849534035,
      "learning_rate": 0.23359004456075352,
      "loss": 0.4174,
      "num_input_tokens_seen": 17149312,
      "step": 12475
    },
    {
      "epoch": 0.3994622623391588,
      "grad_norm": 0.004849342629313469,
      "learning_rate": 0.23354112744299277,
      "loss": 0.4897,
      "num_input_tokens_seen": 17155952,
      "step": 12480
    },
    {
      "epoch": 0.39962230330964726,
      "grad_norm": 0.0033637327142059803,
      "learning_rate": 0.2334921974421403,
      "loss": 0.5465,
      "num_input_tokens_seen": 17162960,
      "step": 12485
    },
    {
      "epoch": 0.3997823442801357,
      "grad_norm": 0.0073568252846598625,
      "learning_rate": 0.23344325456574178,
      "loss": 0.6212,
      "num_input_tokens_seen": 17170528,
      "step": 12490
    },
    {
      "epoch": 0.39994238525062414,
      "grad_norm": 0.007521119900047779,
      "learning_rate": 0.23339429882134477,
      "loss": 0.5113,
      "num_input_tokens_seen": 17177552,
      "step": 12495
    },
    {
      "epoch": 0.4001024262211126,
      "grad_norm": 0.005604064092040062,
      "learning_rate": 0.23334533021649884,
      "loss": 0.4723,
      "num_input_tokens_seen": 17184016,
      "step": 12500
    },
    {
      "epoch": 0.400262467191601,
      "grad_norm": 0.008113858290016651,
      "learning_rate": 0.23329634875875566,
      "loss": 0.4463,
      "num_input_tokens_seen": 17190752,
      "step": 12505
    },
    {
      "epoch": 0.4004225081620895,
      "grad_norm": 0.003990092780441046,
      "learning_rate": 0.23324735445566874,
      "loss": 0.4341,
      "num_input_tokens_seen": 17197552,
      "step": 12510
    },
    {
      "epoch": 0.40058254913257796,
      "grad_norm": 0.005804348737001419,
      "learning_rate": 0.2331983473147936,
      "loss": 0.6389,
      "num_input_tokens_seen": 17203840,
      "step": 12515
    },
    {
      "epoch": 0.4007425901030664,
      "grad_norm": 0.013071526773273945,
      "learning_rate": 0.23314932734368776,
      "loss": 0.5817,
      "num_input_tokens_seen": 17210784,
      "step": 12520
    },
    {
      "epoch": 0.40090263107355484,
      "grad_norm": 0.0069799222983419895,
      "learning_rate": 0.2331002945499107,
      "loss": 0.5454,
      "num_input_tokens_seen": 17217680,
      "step": 12525
    },
    {
      "epoch": 0.4010626720440433,
      "grad_norm": 0.004716984461992979,
      "learning_rate": 0.23305124894102397,
      "loss": 0.5587,
      "num_input_tokens_seen": 17224320,
      "step": 12530
    },
    {
      "epoch": 0.4012227130145317,
      "grad_norm": 0.006711691152304411,
      "learning_rate": 0.23300219052459092,
      "loss": 0.5372,
      "num_input_tokens_seen": 17230976,
      "step": 12535
    },
    {
      "epoch": 0.40138275398502016,
      "grad_norm": 0.004192539490759373,
      "learning_rate": 0.23295311930817708,
      "loss": 0.6489,
      "num_input_tokens_seen": 17237552,
      "step": 12540
    },
    {
      "epoch": 0.4015427949555086,
      "grad_norm": 0.004672365728765726,
      "learning_rate": 0.23290403529934972,
      "loss": 0.6641,
      "num_input_tokens_seen": 17243968,
      "step": 12545
    },
    {
      "epoch": 0.40170283592599704,
      "grad_norm": 0.006239278241991997,
      "learning_rate": 0.23285493850567832,
      "loss": 0.5304,
      "num_input_tokens_seen": 17251136,
      "step": 12550
    },
    {
      "epoch": 0.4018628768964855,
      "grad_norm": 0.002822363283485174,
      "learning_rate": 0.23280582893473414,
      "loss": 0.3494,
      "num_input_tokens_seen": 17258160,
      "step": 12555
    },
    {
      "epoch": 0.402022917866974,
      "grad_norm": 0.004855161998420954,
      "learning_rate": 0.2327567065940906,
      "loss": 0.5189,
      "num_input_tokens_seen": 17264784,
      "step": 12560
    },
    {
      "epoch": 0.4021829588374624,
      "grad_norm": 0.00303054042160511,
      "learning_rate": 0.23270757149132285,
      "loss": 0.4247,
      "num_input_tokens_seen": 17271584,
      "step": 12565
    },
    {
      "epoch": 0.40234299980795085,
      "grad_norm": 0.008394371718168259,
      "learning_rate": 0.23265842363400827,
      "loss": 0.5014,
      "num_input_tokens_seen": 17278368,
      "step": 12570
    },
    {
      "epoch": 0.4025030407784393,
      "grad_norm": 0.00856398232281208,
      "learning_rate": 0.23260926302972595,
      "loss": 0.6981,
      "num_input_tokens_seen": 17284912,
      "step": 12575
    },
    {
      "epoch": 0.40266308174892773,
      "grad_norm": 0.00416354276239872,
      "learning_rate": 0.2325600896860572,
      "loss": 0.5281,
      "num_input_tokens_seen": 17291536,
      "step": 12580
    },
    {
      "epoch": 0.40282312271941617,
      "grad_norm": 0.0054116398096084595,
      "learning_rate": 0.23251090361058505,
      "loss": 0.4861,
      "num_input_tokens_seen": 17298080,
      "step": 12585
    },
    {
      "epoch": 0.4029831636899046,
      "grad_norm": 0.006141660735011101,
      "learning_rate": 0.23246170481089476,
      "loss": 0.6027,
      "num_input_tokens_seen": 17304624,
      "step": 12590
    },
    {
      "epoch": 0.40314320466039305,
      "grad_norm": 0.00821454543620348,
      "learning_rate": 0.23241249329457317,
      "loss": 0.4453,
      "num_input_tokens_seen": 17311664,
      "step": 12595
    },
    {
      "epoch": 0.4033032456308815,
      "grad_norm": 0.005020621232688427,
      "learning_rate": 0.23236326906920957,
      "loss": 0.4658,
      "num_input_tokens_seen": 17318272,
      "step": 12600
    },
    {
      "epoch": 0.4033032456308815,
      "eval_loss": 0.5989137887954712,
      "eval_runtime": 332.2315,
      "eval_samples_per_second": 41.796,
      "eval_steps_per_second": 20.898,
      "num_input_tokens_seen": 17318272,
      "step": 12600
    },
    {
      "epoch": 0.40346328660136993,
      "grad_norm": 0.005066873040050268,
      "learning_rate": 0.2323140321423948,
      "loss": 0.5549,
      "num_input_tokens_seen": 17324960,
      "step": 12605
    },
    {
      "epoch": 0.40362332757185837,
      "grad_norm": 0.011133354157209396,
      "learning_rate": 0.23226478252172184,
      "loss": 0.6376,
      "num_input_tokens_seen": 17331728,
      "step": 12610
    },
    {
      "epoch": 0.40378336854234687,
      "grad_norm": 0.01456967182457447,
      "learning_rate": 0.23221552021478561,
      "loss": 0.7438,
      "num_input_tokens_seen": 17338320,
      "step": 12615
    },
    {
      "epoch": 0.4039434095128353,
      "grad_norm": 0.007483807858079672,
      "learning_rate": 0.232166245229183,
      "loss": 0.6828,
      "num_input_tokens_seen": 17344896,
      "step": 12620
    },
    {
      "epoch": 0.40410345048332375,
      "grad_norm": 0.020162830129265785,
      "learning_rate": 0.2321169575725128,
      "loss": 0.8536,
      "num_input_tokens_seen": 17352080,
      "step": 12625
    },
    {
      "epoch": 0.4042634914538122,
      "grad_norm": 0.014008191414177418,
      "learning_rate": 0.23206765725237577,
      "loss": 0.7656,
      "num_input_tokens_seen": 17359488,
      "step": 12630
    },
    {
      "epoch": 0.4044235324243006,
      "grad_norm": 0.022036060690879822,
      "learning_rate": 0.2320183442763747,
      "loss": 0.8255,
      "num_input_tokens_seen": 17366336,
      "step": 12635
    },
    {
      "epoch": 0.40458357339478906,
      "grad_norm": 0.023348931223154068,
      "learning_rate": 0.23196901865211422,
      "loss": 0.7823,
      "num_input_tokens_seen": 17372848,
      "step": 12640
    },
    {
      "epoch": 0.4047436143652775,
      "grad_norm": 0.01300831325352192,
      "learning_rate": 0.231919680387201,
      "loss": 0.8948,
      "num_input_tokens_seen": 17379712,
      "step": 12645
    },
    {
      "epoch": 0.40490365533576594,
      "grad_norm": 0.013411766849458218,
      "learning_rate": 0.23187032948924358,
      "loss": 0.6216,
      "num_input_tokens_seen": 17386576,
      "step": 12650
    },
    {
      "epoch": 0.4050636963062544,
      "grad_norm": 0.012571456842124462,
      "learning_rate": 0.23182096596585247,
      "loss": 0.6313,
      "num_input_tokens_seen": 17393200,
      "step": 12655
    },
    {
      "epoch": 0.4052237372767428,
      "grad_norm": 0.012577694840729237,
      "learning_rate": 0.23177158982464025,
      "loss": 0.5897,
      "num_input_tokens_seen": 17399824,
      "step": 12660
    },
    {
      "epoch": 0.4053837782472313,
      "grad_norm": 0.014219232834875584,
      "learning_rate": 0.23172220107322122,
      "loss": 0.6435,
      "num_input_tokens_seen": 17406512,
      "step": 12665
    },
    {
      "epoch": 0.40554381921771976,
      "grad_norm": 0.008430583402514458,
      "learning_rate": 0.23167279971921184,
      "loss": 0.5439,
      "num_input_tokens_seen": 17413248,
      "step": 12670
    },
    {
      "epoch": 0.4057038601882082,
      "grad_norm": 0.017287954688072205,
      "learning_rate": 0.23162338577023034,
      "loss": 1.036,
      "num_input_tokens_seen": 17420112,
      "step": 12675
    },
    {
      "epoch": 0.40586390115869664,
      "grad_norm": 0.012065430171787739,
      "learning_rate": 0.23157395923389704,
      "loss": 0.6417,
      "num_input_tokens_seen": 17427184,
      "step": 12680
    },
    {
      "epoch": 0.4060239421291851,
      "grad_norm": 0.0033788944128900766,
      "learning_rate": 0.2315245201178341,
      "loss": 0.7138,
      "num_input_tokens_seen": 17434032,
      "step": 12685
    },
    {
      "epoch": 0.4061839830996735,
      "grad_norm": 0.011108094826340675,
      "learning_rate": 0.23147506842966564,
      "loss": 0.6021,
      "num_input_tokens_seen": 17440544,
      "step": 12690
    },
    {
      "epoch": 0.40634402407016196,
      "grad_norm": 0.005938884802162647,
      "learning_rate": 0.23142560417701774,
      "loss": 0.4828,
      "num_input_tokens_seen": 17447616,
      "step": 12695
    },
    {
      "epoch": 0.4065040650406504,
      "grad_norm": 0.007756771519780159,
      "learning_rate": 0.23137612736751845,
      "loss": 0.5247,
      "num_input_tokens_seen": 17455008,
      "step": 12700
    },
    {
      "epoch": 0.40666410601113884,
      "grad_norm": 0.009034575894474983,
      "learning_rate": 0.23132663800879766,
      "loss": 0.585,
      "num_input_tokens_seen": 17461872,
      "step": 12705
    },
    {
      "epoch": 0.4068241469816273,
      "grad_norm": 0.0056383670307695866,
      "learning_rate": 0.2312771361084873,
      "loss": 0.6696,
      "num_input_tokens_seen": 17468656,
      "step": 12710
    },
    {
      "epoch": 0.4069841879521157,
      "grad_norm": 0.025653105229139328,
      "learning_rate": 0.23122762167422112,
      "loss": 0.6216,
      "num_input_tokens_seen": 17475552,
      "step": 12715
    },
    {
      "epoch": 0.4071442289226042,
      "grad_norm": 0.005987914279103279,
      "learning_rate": 0.23117809471363493,
      "loss": 0.6064,
      "num_input_tokens_seen": 17482256,
      "step": 12720
    },
    {
      "epoch": 0.40730426989309265,
      "grad_norm": 0.010094396770000458,
      "learning_rate": 0.23112855523436637,
      "loss": 0.778,
      "num_input_tokens_seen": 17489232,
      "step": 12725
    },
    {
      "epoch": 0.4074643108635811,
      "grad_norm": 0.00895518809556961,
      "learning_rate": 0.23107900324405511,
      "loss": 0.4983,
      "num_input_tokens_seen": 17495728,
      "step": 12730
    },
    {
      "epoch": 0.40762435183406953,
      "grad_norm": 0.0068906452506780624,
      "learning_rate": 0.2310294387503426,
      "loss": 0.7762,
      "num_input_tokens_seen": 17502640,
      "step": 12735
    },
    {
      "epoch": 0.40778439280455797,
      "grad_norm": 0.006685537286102772,
      "learning_rate": 0.23097986176087237,
      "loss": 0.5648,
      "num_input_tokens_seen": 17509536,
      "step": 12740
    },
    {
      "epoch": 0.4079444337750464,
      "grad_norm": 0.01340391393750906,
      "learning_rate": 0.23093027228328986,
      "loss": 0.6535,
      "num_input_tokens_seen": 17516352,
      "step": 12745
    },
    {
      "epoch": 0.40810447474553485,
      "grad_norm": 0.00581264728680253,
      "learning_rate": 0.23088067032524226,
      "loss": 0.5688,
      "num_input_tokens_seen": 17523344,
      "step": 12750
    },
    {
      "epoch": 0.4082645157160233,
      "grad_norm": 0.007318377029150724,
      "learning_rate": 0.23083105589437888,
      "loss": 0.7249,
      "num_input_tokens_seen": 17530512,
      "step": 12755
    },
    {
      "epoch": 0.40842455668651173,
      "grad_norm": 0.007085143122822046,
      "learning_rate": 0.23078142899835094,
      "loss": 0.5715,
      "num_input_tokens_seen": 17537136,
      "step": 12760
    },
    {
      "epoch": 0.40858459765700017,
      "grad_norm": 0.002841942012310028,
      "learning_rate": 0.23073178964481147,
      "loss": 0.4073,
      "num_input_tokens_seen": 17543888,
      "step": 12765
    },
    {
      "epoch": 0.4087446386274886,
      "grad_norm": 0.004410593304783106,
      "learning_rate": 0.2306821378414155,
      "loss": 0.4595,
      "num_input_tokens_seen": 17550896,
      "step": 12770
    },
    {
      "epoch": 0.4089046795979771,
      "grad_norm": 0.006388612557202578,
      "learning_rate": 0.2306324735958199,
      "loss": 0.6517,
      "num_input_tokens_seen": 17557520,
      "step": 12775
    },
    {
      "epoch": 0.40906472056846555,
      "grad_norm": 0.008769909851253033,
      "learning_rate": 0.23058279691568362,
      "loss": 0.6842,
      "num_input_tokens_seen": 17564352,
      "step": 12780
    },
    {
      "epoch": 0.409224761538954,
      "grad_norm": 0.004455515183508396,
      "learning_rate": 0.23053310780866745,
      "loss": 0.4707,
      "num_input_tokens_seen": 17571472,
      "step": 12785
    },
    {
      "epoch": 0.4093848025094424,
      "grad_norm": 0.007854712195694447,
      "learning_rate": 0.23048340628243397,
      "loss": 0.6779,
      "num_input_tokens_seen": 17578368,
      "step": 12790
    },
    {
      "epoch": 0.40954484347993086,
      "grad_norm": 0.01192496344447136,
      "learning_rate": 0.23043369234464783,
      "loss": 0.6017,
      "num_input_tokens_seen": 17585024,
      "step": 12795
    },
    {
      "epoch": 0.4097048844504193,
      "grad_norm": 0.019462458789348602,
      "learning_rate": 0.2303839660029755,
      "loss": 0.7755,
      "num_input_tokens_seen": 17591696,
      "step": 12800
    },
    {
      "epoch": 0.4097048844504193,
      "eval_loss": 0.6354094743728638,
      "eval_runtime": 332.3368,
      "eval_samples_per_second": 41.783,
      "eval_steps_per_second": 20.891,
      "num_input_tokens_seen": 17591696,
      "step": 12800
    },
    {
      "epoch": 0.40986492542090774,
      "grad_norm": 0.015240351669490337,
      "learning_rate": 0.23033422726508548,
      "loss": 0.8188,
      "num_input_tokens_seen": 17598736,
      "step": 12805
    },
    {
      "epoch": 0.4100249663913962,
      "grad_norm": 0.003489969065412879,
      "learning_rate": 0.23028447613864808,
      "loss": 0.6075,
      "num_input_tokens_seen": 17605088,
      "step": 12810
    },
    {
      "epoch": 0.4101850073618846,
      "grad_norm": 0.00427399342879653,
      "learning_rate": 0.2302347126313355,
      "loss": 0.5594,
      "num_input_tokens_seen": 17611920,
      "step": 12815
    },
    {
      "epoch": 0.41034504833237306,
      "grad_norm": 0.0031224400736391544,
      "learning_rate": 0.23018493675082197,
      "loss": 0.4643,
      "num_input_tokens_seen": 17618496,
      "step": 12820
    },
    {
      "epoch": 0.41050508930286156,
      "grad_norm": 0.006183386780321598,
      "learning_rate": 0.2301351485047835,
      "loss": 0.52,
      "num_input_tokens_seen": 17625472,
      "step": 12825
    },
    {
      "epoch": 0.41066513027335,
      "grad_norm": 0.006266348529607058,
      "learning_rate": 0.23008534790089813,
      "loss": 0.5036,
      "num_input_tokens_seen": 17632320,
      "step": 12830
    },
    {
      "epoch": 0.41082517124383844,
      "grad_norm": 0.005891724489629269,
      "learning_rate": 0.2300355349468457,
      "loss": 0.6167,
      "num_input_tokens_seen": 17639424,
      "step": 12835
    },
    {
      "epoch": 0.4109852122143269,
      "grad_norm": 0.003645229386165738,
      "learning_rate": 0.22998570965030793,
      "loss": 0.4035,
      "num_input_tokens_seen": 17646592,
      "step": 12840
    },
    {
      "epoch": 0.4111452531848153,
      "grad_norm": 0.005567802116274834,
      "learning_rate": 0.22993587201896862,
      "loss": 0.6171,
      "num_input_tokens_seen": 17653232,
      "step": 12845
    },
    {
      "epoch": 0.41130529415530376,
      "grad_norm": 0.006948922295123339,
      "learning_rate": 0.2298860220605133,
      "loss": 0.4492,
      "num_input_tokens_seen": 17660144,
      "step": 12850
    },
    {
      "epoch": 0.4114653351257922,
      "grad_norm": 0.007101916708052158,
      "learning_rate": 0.22983615978262942,
      "loss": 0.5591,
      "num_input_tokens_seen": 17666736,
      "step": 12855
    },
    {
      "epoch": 0.41162537609628064,
      "grad_norm": 0.005368309561163187,
      "learning_rate": 0.22978628519300648,
      "loss": 0.6419,
      "num_input_tokens_seen": 17673296,
      "step": 12860
    },
    {
      "epoch": 0.4117854170667691,
      "grad_norm": 0.004314608871936798,
      "learning_rate": 0.22973639829933568,
      "loss": 0.5193,
      "num_input_tokens_seen": 17680304,
      "step": 12865
    },
    {
      "epoch": 0.4119454580372575,
      "grad_norm": 0.0044587417505681515,
      "learning_rate": 0.22968649910931027,
      "loss": 0.6403,
      "num_input_tokens_seen": 17686912,
      "step": 12870
    },
    {
      "epoch": 0.41210549900774596,
      "grad_norm": 0.004285505507141352,
      "learning_rate": 0.22963658763062528,
      "loss": 0.5105,
      "num_input_tokens_seen": 17693728,
      "step": 12875
    },
    {
      "epoch": 0.41226553997823445,
      "grad_norm": 0.006817476823925972,
      "learning_rate": 0.22958666387097765,
      "loss": 0.4798,
      "num_input_tokens_seen": 17700752,
      "step": 12880
    },
    {
      "epoch": 0.4124255809487229,
      "grad_norm": 0.00875171273946762,
      "learning_rate": 0.22953672783806633,
      "loss": 0.5506,
      "num_input_tokens_seen": 17707168,
      "step": 12885
    },
    {
      "epoch": 0.41258562191921133,
      "grad_norm": 0.015178276225924492,
      "learning_rate": 0.22948677953959207,
      "loss": 0.5639,
      "num_input_tokens_seen": 17714080,
      "step": 12890
    },
    {
      "epoch": 0.41274566288969977,
      "grad_norm": 0.003522732062265277,
      "learning_rate": 0.2294368189832575,
      "loss": 0.4649,
      "num_input_tokens_seen": 17720704,
      "step": 12895
    },
    {
      "epoch": 0.4129057038601882,
      "grad_norm": 0.0048374575562775135,
      "learning_rate": 0.2293868461767672,
      "loss": 0.586,
      "num_input_tokens_seen": 17727392,
      "step": 12900
    },
    {
      "epoch": 0.41306574483067665,
      "grad_norm": 0.005805367603898048,
      "learning_rate": 0.22933686112782758,
      "loss": 0.5841,
      "num_input_tokens_seen": 17734448,
      "step": 12905
    },
    {
      "epoch": 0.4132257858011651,
      "grad_norm": 0.010035508312284946,
      "learning_rate": 0.22928686384414698,
      "loss": 0.4531,
      "num_input_tokens_seen": 17740944,
      "step": 12910
    },
    {
      "epoch": 0.41338582677165353,
      "grad_norm": 0.010325099341571331,
      "learning_rate": 0.22923685433343552,
      "loss": 0.6423,
      "num_input_tokens_seen": 17747296,
      "step": 12915
    },
    {
      "epoch": 0.41354586774214197,
      "grad_norm": 0.011817491613328457,
      "learning_rate": 0.22918683260340542,
      "loss": 0.4316,
      "num_input_tokens_seen": 17754544,
      "step": 12920
    },
    {
      "epoch": 0.4137059087126304,
      "grad_norm": 0.013911507092416286,
      "learning_rate": 0.2291367986617706,
      "loss": 0.6183,
      "num_input_tokens_seen": 17760864,
      "step": 12925
    },
    {
      "epoch": 0.4138659496831189,
      "grad_norm": 0.008028055541217327,
      "learning_rate": 0.22908675251624697,
      "loss": 0.497,
      "num_input_tokens_seen": 17768112,
      "step": 12930
    },
    {
      "epoch": 0.41402599065360735,
      "grad_norm": 0.010741157457232475,
      "learning_rate": 0.22903669417455216,
      "loss": 0.6868,
      "num_input_tokens_seen": 17775120,
      "step": 12935
    },
    {
      "epoch": 0.4141860316240958,
      "grad_norm": 0.011194010265171528,
      "learning_rate": 0.22898662364440592,
      "loss": 0.4614,
      "num_input_tokens_seen": 17782000,
      "step": 12940
    },
    {
      "epoch": 0.4143460725945842,
      "grad_norm": 0.009881057776510715,
      "learning_rate": 0.2289365409335297,
      "loss": 0.4836,
      "num_input_tokens_seen": 17789008,
      "step": 12945
    },
    {
      "epoch": 0.41450611356507266,
      "grad_norm": 0.0066117881797254086,
      "learning_rate": 0.2288864460496469,
      "loss": 0.5378,
      "num_input_tokens_seen": 17795744,
      "step": 12950
    },
    {
      "epoch": 0.4146661545355611,
      "grad_norm": 0.007205449976027012,
      "learning_rate": 0.22883633900048272,
      "loss": 0.5408,
      "num_input_tokens_seen": 17802672,
      "step": 12955
    },
    {
      "epoch": 0.41482619550604954,
      "grad_norm": 0.006756846327334642,
      "learning_rate": 0.2287862197937644,
      "loss": 0.5742,
      "num_input_tokens_seen": 17809696,
      "step": 12960
    },
    {
      "epoch": 0.414986236476538,
      "grad_norm": 0.003941135946661234,
      "learning_rate": 0.2287360884372209,
      "loss": 0.5786,
      "num_input_tokens_seen": 17816640,
      "step": 12965
    },
    {
      "epoch": 0.4151462774470264,
      "grad_norm": 0.005848877597600222,
      "learning_rate": 0.22868594493858307,
      "loss": 0.7479,
      "num_input_tokens_seen": 17823616,
      "step": 12970
    },
    {
      "epoch": 0.41530631841751486,
      "grad_norm": 0.0038459054194390774,
      "learning_rate": 0.2286357893055837,
      "loss": 0.5383,
      "num_input_tokens_seen": 17830464,
      "step": 12975
    },
    {
      "epoch": 0.4154663593880033,
      "grad_norm": 0.006860378663986921,
      "learning_rate": 0.22858562154595746,
      "loss": 0.4759,
      "num_input_tokens_seen": 17837184,
      "step": 12980
    },
    {
      "epoch": 0.4156264003584918,
      "grad_norm": 0.004117982462048531,
      "learning_rate": 0.22853544166744078,
      "loss": 0.662,
      "num_input_tokens_seen": 17843904,
      "step": 12985
    },
    {
      "epoch": 0.41578644132898024,
      "grad_norm": 0.00681807566434145,
      "learning_rate": 0.22848524967777206,
      "loss": 0.6416,
      "num_input_tokens_seen": 17850992,
      "step": 12990
    },
    {
      "epoch": 0.4159464822994687,
      "grad_norm": 0.006330044474452734,
      "learning_rate": 0.22843504558469152,
      "loss": 0.7152,
      "num_input_tokens_seen": 17857456,
      "step": 12995
    },
    {
      "epoch": 0.4161065232699571,
      "grad_norm": 0.004218107555061579,
      "learning_rate": 0.2283848293959413,
      "loss": 0.5444,
      "num_input_tokens_seen": 17864256,
      "step": 13000
    },
    {
      "epoch": 0.4161065232699571,
      "eval_loss": 0.6029016971588135,
      "eval_runtime": 332.3167,
      "eval_samples_per_second": 41.785,
      "eval_steps_per_second": 20.893,
      "num_input_tokens_seen": 17864256,
      "step": 13000
    },
    {
      "epoch": 0.41626656424044556,
      "grad_norm": 0.006063445005565882,
      "learning_rate": 0.22833460111926532,
      "loss": 0.6597,
      "num_input_tokens_seen": 17871376,
      "step": 13005
    },
    {
      "epoch": 0.416426605210934,
      "grad_norm": 0.004956502001732588,
      "learning_rate": 0.22828436076240946,
      "loss": 0.4772,
      "num_input_tokens_seen": 17878384,
      "step": 13010
    },
    {
      "epoch": 0.41658664618142244,
      "grad_norm": 0.006139935459941626,
      "learning_rate": 0.22823410833312135,
      "loss": 0.5713,
      "num_input_tokens_seen": 17885776,
      "step": 13015
    },
    {
      "epoch": 0.4167466871519109,
      "grad_norm": 0.0029282330069690943,
      "learning_rate": 0.2281838438391506,
      "loss": 0.3361,
      "num_input_tokens_seen": 17892560,
      "step": 13020
    },
    {
      "epoch": 0.4169067281223993,
      "grad_norm": 0.003340311348438263,
      "learning_rate": 0.22813356728824863,
      "loss": 0.4804,
      "num_input_tokens_seen": 17899056,
      "step": 13025
    },
    {
      "epoch": 0.41706676909288776,
      "grad_norm": 0.007910910062491894,
      "learning_rate": 0.2280832786881687,
      "loss": 0.7417,
      "num_input_tokens_seen": 17906016,
      "step": 13030
    },
    {
      "epoch": 0.4172268100633762,
      "grad_norm": 0.005927425343543291,
      "learning_rate": 0.22803297804666592,
      "loss": 0.423,
      "num_input_tokens_seen": 17913104,
      "step": 13035
    },
    {
      "epoch": 0.4173868510338647,
      "grad_norm": 0.0039216140285134315,
      "learning_rate": 0.22798266537149728,
      "loss": 0.6679,
      "num_input_tokens_seen": 17920240,
      "step": 13040
    },
    {
      "epoch": 0.41754689200435313,
      "grad_norm": 0.004467968363314867,
      "learning_rate": 0.22793234067042167,
      "loss": 0.5996,
      "num_input_tokens_seen": 17927232,
      "step": 13045
    },
    {
      "epoch": 0.41770693297484157,
      "grad_norm": 0.0045779915526509285,
      "learning_rate": 0.22788200395119979,
      "loss": 0.4712,
      "num_input_tokens_seen": 17933936,
      "step": 13050
    },
    {
      "epoch": 0.41786697394533,
      "grad_norm": 0.0031670848838984966,
      "learning_rate": 0.2278316552215942,
      "loss": 0.6034,
      "num_input_tokens_seen": 17941232,
      "step": 13055
    },
    {
      "epoch": 0.41802701491581845,
      "grad_norm": 0.007027822080999613,
      "learning_rate": 0.22778129448936918,
      "loss": 0.8749,
      "num_input_tokens_seen": 17948640,
      "step": 13060
    },
    {
      "epoch": 0.4181870558863069,
      "grad_norm": 0.006155027076601982,
      "learning_rate": 0.22773092176229118,
      "loss": 0.6261,
      "num_input_tokens_seen": 17955680,
      "step": 13065
    },
    {
      "epoch": 0.41834709685679533,
      "grad_norm": 0.006488954648375511,
      "learning_rate": 0.22768053704812816,
      "loss": 0.6216,
      "num_input_tokens_seen": 17962272,
      "step": 13070
    },
    {
      "epoch": 0.41850713782728377,
      "grad_norm": 0.004054652526974678,
      "learning_rate": 0.22763014035465018,
      "loss": 0.5178,
      "num_input_tokens_seen": 17968832,
      "step": 13075
    },
    {
      "epoch": 0.4186671787977722,
      "grad_norm": 0.006093171890825033,
      "learning_rate": 0.22757973168962892,
      "loss": 0.5796,
      "num_input_tokens_seen": 17975616,
      "step": 13080
    },
    {
      "epoch": 0.41882721976826065,
      "grad_norm": 0.0050099496729671955,
      "learning_rate": 0.22752931106083818,
      "loss": 0.7504,
      "num_input_tokens_seen": 17982272,
      "step": 13085
    },
    {
      "epoch": 0.41898726073874915,
      "grad_norm": 0.004684824962168932,
      "learning_rate": 0.22747887847605341,
      "loss": 0.4667,
      "num_input_tokens_seen": 17988944,
      "step": 13090
    },
    {
      "epoch": 0.4191473017092376,
      "grad_norm": 0.004371091723442078,
      "learning_rate": 0.22742843394305184,
      "loss": 0.6241,
      "num_input_tokens_seen": 17995408,
      "step": 13095
    },
    {
      "epoch": 0.419307342679726,
      "grad_norm": 0.006074781529605389,
      "learning_rate": 0.22737797746961272,
      "loss": 0.6233,
      "num_input_tokens_seen": 18002592,
      "step": 13100
    },
    {
      "epoch": 0.41946738365021446,
      "grad_norm": 0.0039291055873036385,
      "learning_rate": 0.22732750906351712,
      "loss": 0.5707,
      "num_input_tokens_seen": 18009392,
      "step": 13105
    },
    {
      "epoch": 0.4196274246207029,
      "grad_norm": 0.0035600343253463507,
      "learning_rate": 0.22727702873254785,
      "loss": 0.5779,
      "num_input_tokens_seen": 18016160,
      "step": 13110
    },
    {
      "epoch": 0.41978746559119134,
      "grad_norm": 0.0040109967812895775,
      "learning_rate": 0.22722653648448968,
      "loss": 0.7076,
      "num_input_tokens_seen": 18023056,
      "step": 13115
    },
    {
      "epoch": 0.4199475065616798,
      "grad_norm": 0.006021981593221426,
      "learning_rate": 0.22717603232712902,
      "loss": 0.5296,
      "num_input_tokens_seen": 18030048,
      "step": 13120
    },
    {
      "epoch": 0.4201075475321682,
      "grad_norm": 0.0038324252236634493,
      "learning_rate": 0.22712551626825436,
      "loss": 0.6631,
      "num_input_tokens_seen": 18036784,
      "step": 13125
    },
    {
      "epoch": 0.42026758850265666,
      "grad_norm": 0.005302208475768566,
      "learning_rate": 0.2270749883156559,
      "loss": 0.6449,
      "num_input_tokens_seen": 18043488,
      "step": 13130
    },
    {
      "epoch": 0.4204276294731451,
      "grad_norm": 0.005685826763510704,
      "learning_rate": 0.22702444847712563,
      "loss": 0.7258,
      "num_input_tokens_seen": 18050112,
      "step": 13135
    },
    {
      "epoch": 0.42058767044363354,
      "grad_norm": 0.007674261461943388,
      "learning_rate": 0.22697389676045743,
      "loss": 0.7693,
      "num_input_tokens_seen": 18056592,
      "step": 13140
    },
    {
      "epoch": 0.42074771141412204,
      "grad_norm": 0.0054158372804522514,
      "learning_rate": 0.22692333317344704,
      "loss": 0.6863,
      "num_input_tokens_seen": 18063136,
      "step": 13145
    },
    {
      "epoch": 0.4209077523846105,
      "grad_norm": 0.004015537444502115,
      "learning_rate": 0.22687275772389198,
      "loss": 0.7686,
      "num_input_tokens_seen": 18069616,
      "step": 13150
    },
    {
      "epoch": 0.4210677933550989,
      "grad_norm": 0.003927984740585089,
      "learning_rate": 0.22682217041959168,
      "loss": 0.6262,
      "num_input_tokens_seen": 18076416,
      "step": 13155
    },
    {
      "epoch": 0.42122783432558736,
      "grad_norm": 0.0048230914399027824,
      "learning_rate": 0.2267715712683473,
      "loss": 0.513,
      "num_input_tokens_seen": 18083360,
      "step": 13160
    },
    {
      "epoch": 0.4213878752960758,
      "grad_norm": 0.006092927884310484,
      "learning_rate": 0.22672096027796182,
      "loss": 0.509,
      "num_input_tokens_seen": 18090384,
      "step": 13165
    },
    {
      "epoch": 0.42154791626656424,
      "grad_norm": 0.01033793669193983,
      "learning_rate": 0.22667033745624016,
      "loss": 0.6462,
      "num_input_tokens_seen": 18097296,
      "step": 13170
    },
    {
      "epoch": 0.4217079572370527,
      "grad_norm": 0.0036106931511312723,
      "learning_rate": 0.22661970281098895,
      "loss": 0.8868,
      "num_input_tokens_seen": 18104096,
      "step": 13175
    },
    {
      "epoch": 0.4218679982075411,
      "grad_norm": 0.004866689443588257,
      "learning_rate": 0.22656905635001667,
      "loss": 0.4702,
      "num_input_tokens_seen": 18110976,
      "step": 13180
    },
    {
      "epoch": 0.42202803917802956,
      "grad_norm": 0.004318642430007458,
      "learning_rate": 0.2265183980811337,
      "loss": 0.6354,
      "num_input_tokens_seen": 18117968,
      "step": 13185
    },
    {
      "epoch": 0.422188080148518,
      "grad_norm": 0.005460456013679504,
      "learning_rate": 0.22646772801215218,
      "loss": 0.5821,
      "num_input_tokens_seen": 18124816,
      "step": 13190
    },
    {
      "epoch": 0.4223481211190065,
      "grad_norm": 0.004925435874611139,
      "learning_rate": 0.22641704615088598,
      "loss": 0.6042,
      "num_input_tokens_seen": 18131328,
      "step": 13195
    },
    {
      "epoch": 0.42250816208949493,
      "grad_norm": 0.008868376724421978,
      "learning_rate": 0.22636635250515103,
      "loss": 0.5456,
      "num_input_tokens_seen": 18137984,
      "step": 13200
    },
    {
      "epoch": 0.42250816208949493,
      "eval_loss": 0.5985685586929321,
      "eval_runtime": 331.8991,
      "eval_samples_per_second": 41.838,
      "eval_steps_per_second": 20.919,
      "num_input_tokens_seen": 18137984,
      "step": 13200
    },
    {
      "epoch": 0.42266820305998337,
      "grad_norm": 0.0029450298752635717,
      "learning_rate": 0.2263156470827648,
      "loss": 0.5728,
      "num_input_tokens_seen": 18145664,
      "step": 13205
    },
    {
      "epoch": 0.4228282440304718,
      "grad_norm": 0.005908298306167126,
      "learning_rate": 0.22626492989154678,
      "loss": 0.5196,
      "num_input_tokens_seen": 18152224,
      "step": 13210
    },
    {
      "epoch": 0.42298828500096025,
      "grad_norm": 0.005065548699349165,
      "learning_rate": 0.22621420093931813,
      "loss": 0.468,
      "num_input_tokens_seen": 18159264,
      "step": 13215
    },
    {
      "epoch": 0.4231483259714487,
      "grad_norm": 0.004356635268777609,
      "learning_rate": 0.22616346023390194,
      "loss": 0.6335,
      "num_input_tokens_seen": 18166128,
      "step": 13220
    },
    {
      "epoch": 0.42330836694193713,
      "grad_norm": 0.004693270660936832,
      "learning_rate": 0.22611270778312306,
      "loss": 0.8742,
      "num_input_tokens_seen": 18173040,
      "step": 13225
    },
    {
      "epoch": 0.42346840791242557,
      "grad_norm": 0.008835924789309502,
      "learning_rate": 0.2260619435948081,
      "loss": 0.6422,
      "num_input_tokens_seen": 18180192,
      "step": 13230
    },
    {
      "epoch": 0.423628448882914,
      "grad_norm": 0.005762007553130388,
      "learning_rate": 0.22601116767678567,
      "loss": 0.7649,
      "num_input_tokens_seen": 18186912,
      "step": 13235
    },
    {
      "epoch": 0.42378848985340245,
      "grad_norm": 0.006792389787733555,
      "learning_rate": 0.2259603800368859,
      "loss": 0.6256,
      "num_input_tokens_seen": 18194032,
      "step": 13240
    },
    {
      "epoch": 0.4239485308238909,
      "grad_norm": 0.009027284570038319,
      "learning_rate": 0.22590958068294098,
      "loss": 0.6868,
      "num_input_tokens_seen": 18200672,
      "step": 13245
    },
    {
      "epoch": 0.4241085717943794,
      "grad_norm": 0.009260551072657108,
      "learning_rate": 0.22585876962278478,
      "loss": 0.7734,
      "num_input_tokens_seen": 18207904,
      "step": 13250
    },
    {
      "epoch": 0.4242686127648678,
      "grad_norm": 0.00798783265054226,
      "learning_rate": 0.22580794686425298,
      "loss": 0.585,
      "num_input_tokens_seen": 18215216,
      "step": 13255
    },
    {
      "epoch": 0.42442865373535626,
      "grad_norm": 0.004771496169269085,
      "learning_rate": 0.22575711241518312,
      "loss": 0.749,
      "num_input_tokens_seen": 18222864,
      "step": 13260
    },
    {
      "epoch": 0.4245886947058447,
      "grad_norm": 0.003605863079428673,
      "learning_rate": 0.22570626628341453,
      "loss": 0.5735,
      "num_input_tokens_seen": 18229536,
      "step": 13265
    },
    {
      "epoch": 0.42474873567633314,
      "grad_norm": 0.004388139583170414,
      "learning_rate": 0.22565540847678828,
      "loss": 0.6346,
      "num_input_tokens_seen": 18236480,
      "step": 13270
    },
    {
      "epoch": 0.4249087766468216,
      "grad_norm": 0.0054145678877830505,
      "learning_rate": 0.2256045390031473,
      "loss": 0.5657,
      "num_input_tokens_seen": 18243248,
      "step": 13275
    },
    {
      "epoch": 0.42506881761731,
      "grad_norm": 0.004634513054043055,
      "learning_rate": 0.22555365787033627,
      "loss": 0.429,
      "num_input_tokens_seen": 18249952,
      "step": 13280
    },
    {
      "epoch": 0.42522885858779846,
      "grad_norm": 0.004524956922978163,
      "learning_rate": 0.22550276508620173,
      "loss": 0.4044,
      "num_input_tokens_seen": 18256784,
      "step": 13285
    },
    {
      "epoch": 0.4253888995582869,
      "grad_norm": 0.007391577120870352,
      "learning_rate": 0.22545186065859202,
      "loss": 0.4132,
      "num_input_tokens_seen": 18263936,
      "step": 13290
    },
    {
      "epoch": 0.42554894052877534,
      "grad_norm": 0.0042267004027962685,
      "learning_rate": 0.2254009445953572,
      "loss": 0.6569,
      "num_input_tokens_seen": 18270944,
      "step": 13295
    },
    {
      "epoch": 0.42570898149926384,
      "grad_norm": 0.006759474985301495,
      "learning_rate": 0.22535001690434917,
      "loss": 0.5546,
      "num_input_tokens_seen": 18278208,
      "step": 13300
    },
    {
      "epoch": 0.4258690224697523,
      "grad_norm": 0.006746519356966019,
      "learning_rate": 0.22529907759342163,
      "loss": 0.6244,
      "num_input_tokens_seen": 18284976,
      "step": 13305
    },
    {
      "epoch": 0.4260290634402407,
      "grad_norm": 0.008213583379983902,
      "learning_rate": 0.22524812667043007,
      "loss": 0.8309,
      "num_input_tokens_seen": 18291744,
      "step": 13310
    },
    {
      "epoch": 0.42618910441072916,
      "grad_norm": 0.0070467758923769,
      "learning_rate": 0.22519716414323177,
      "loss": 0.6032,
      "num_input_tokens_seen": 18298208,
      "step": 13315
    },
    {
      "epoch": 0.4263491453812176,
      "grad_norm": 0.005913734436035156,
      "learning_rate": 0.22514619001968567,
      "loss": 0.668,
      "num_input_tokens_seen": 18305248,
      "step": 13320
    },
    {
      "epoch": 0.42650918635170604,
      "grad_norm": 0.0058713387697935104,
      "learning_rate": 0.2250952043076528,
      "loss": 0.5155,
      "num_input_tokens_seen": 18311760,
      "step": 13325
    },
    {
      "epoch": 0.4266692273221945,
      "grad_norm": 0.005503051448613405,
      "learning_rate": 0.2250442070149957,
      "loss": 0.448,
      "num_input_tokens_seen": 18318768,
      "step": 13330
    },
    {
      "epoch": 0.4268292682926829,
      "grad_norm": 0.0057563078589737415,
      "learning_rate": 0.22499319814957885,
      "loss": 0.6445,
      "num_input_tokens_seen": 18325616,
      "step": 13335
    },
    {
      "epoch": 0.42698930926317136,
      "grad_norm": 0.007764182984828949,
      "learning_rate": 0.2249421777192684,
      "loss": 0.5641,
      "num_input_tokens_seen": 18332256,
      "step": 13340
    },
    {
      "epoch": 0.4271493502336598,
      "grad_norm": 0.0036127441562712193,
      "learning_rate": 0.22489114573193236,
      "loss": 0.6152,
      "num_input_tokens_seen": 18338832,
      "step": 13345
    },
    {
      "epoch": 0.42730939120414824,
      "grad_norm": 0.0035977179650217295,
      "learning_rate": 0.2248401021954405,
      "loss": 0.4051,
      "num_input_tokens_seen": 18345456,
      "step": 13350
    },
    {
      "epoch": 0.42746943217463673,
      "grad_norm": 0.0055062780156731606,
      "learning_rate": 0.22478904711766443,
      "loss": 0.695,
      "num_input_tokens_seen": 18352048,
      "step": 13355
    },
    {
      "epoch": 0.42762947314512517,
      "grad_norm": 0.005429331213235855,
      "learning_rate": 0.22473798050647734,
      "loss": 0.4613,
      "num_input_tokens_seen": 18358976,
      "step": 13360
    },
    {
      "epoch": 0.4277895141156136,
      "grad_norm": 0.0031101801432669163,
      "learning_rate": 0.22468690236975453,
      "loss": 0.4726,
      "num_input_tokens_seen": 18365776,
      "step": 13365
    },
    {
      "epoch": 0.42794955508610205,
      "grad_norm": 0.009458278305828571,
      "learning_rate": 0.22463581271537272,
      "loss": 0.4769,
      "num_input_tokens_seen": 18372528,
      "step": 13370
    },
    {
      "epoch": 0.4281095960565905,
      "grad_norm": 0.006879318505525589,
      "learning_rate": 0.22458471155121076,
      "loss": 0.6867,
      "num_input_tokens_seen": 18379104,
      "step": 13375
    },
    {
      "epoch": 0.42826963702707893,
      "grad_norm": 0.00510614225640893,
      "learning_rate": 0.2245335988851489,
      "loss": 0.4087,
      "num_input_tokens_seen": 18385824,
      "step": 13380
    },
    {
      "epoch": 0.42842967799756737,
      "grad_norm": 0.006685869302600622,
      "learning_rate": 0.2244824747250695,
      "loss": 0.6306,
      "num_input_tokens_seen": 18392368,
      "step": 13385
    },
    {
      "epoch": 0.4285897189680558,
      "grad_norm": 0.0036886499729007483,
      "learning_rate": 0.22443133907885646,
      "loss": 0.5257,
      "num_input_tokens_seen": 18399584,
      "step": 13390
    },
    {
      "epoch": 0.42874975993854425,
      "grad_norm": 0.007558690384030342,
      "learning_rate": 0.22438019195439557,
      "loss": 0.5649,
      "num_input_tokens_seen": 18406480,
      "step": 13395
    },
    {
      "epoch": 0.4289098009090327,
      "grad_norm": 0.010483070276677608,
      "learning_rate": 0.22432903335957435,
      "loss": 0.7185,
      "num_input_tokens_seen": 18413504,
      "step": 13400
    },
    {
      "epoch": 0.4289098009090327,
      "eval_loss": 0.5926409959793091,
      "eval_runtime": 332.0265,
      "eval_samples_per_second": 41.822,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 18413504,
      "step": 13400
    },
    {
      "epoch": 0.42906984187952113,
      "grad_norm": 0.005341228563338518,
      "learning_rate": 0.22427786330228214,
      "loss": 0.4189,
      "num_input_tokens_seen": 18420640,
      "step": 13405
    },
    {
      "epoch": 0.4292298828500096,
      "grad_norm": 0.0058737886138260365,
      "learning_rate": 0.22422668179040997,
      "loss": 0.5302,
      "num_input_tokens_seen": 18427360,
      "step": 13410
    },
    {
      "epoch": 0.42938992382049807,
      "grad_norm": 0.014879675582051277,
      "learning_rate": 0.2241754888318507,
      "loss": 0.7825,
      "num_input_tokens_seen": 18434304,
      "step": 13415
    },
    {
      "epoch": 0.4295499647909865,
      "grad_norm": 0.005887698382139206,
      "learning_rate": 0.22412428443449886,
      "loss": 0.4317,
      "num_input_tokens_seen": 18441440,
      "step": 13420
    },
    {
      "epoch": 0.42971000576147494,
      "grad_norm": 0.004614986479282379,
      "learning_rate": 0.22407306860625087,
      "loss": 0.4002,
      "num_input_tokens_seen": 18448144,
      "step": 13425
    },
    {
      "epoch": 0.4298700467319634,
      "grad_norm": 0.007149344775825739,
      "learning_rate": 0.22402184135500483,
      "loss": 0.6472,
      "num_input_tokens_seen": 18455472,
      "step": 13430
    },
    {
      "epoch": 0.4300300877024518,
      "grad_norm": 0.004416583571583033,
      "learning_rate": 0.22397060268866067,
      "loss": 0.5815,
      "num_input_tokens_seen": 18462592,
      "step": 13435
    },
    {
      "epoch": 0.43019012867294026,
      "grad_norm": 0.004641219042241573,
      "learning_rate": 0.22391935261511994,
      "loss": 0.452,
      "num_input_tokens_seen": 18469728,
      "step": 13440
    },
    {
      "epoch": 0.4303501696434287,
      "grad_norm": 0.006309022661298513,
      "learning_rate": 0.22386809114228615,
      "loss": 0.5629,
      "num_input_tokens_seen": 18476512,
      "step": 13445
    },
    {
      "epoch": 0.43051021061391714,
      "grad_norm": 0.00415460579097271,
      "learning_rate": 0.22381681827806446,
      "loss": 0.4775,
      "num_input_tokens_seen": 18483424,
      "step": 13450
    },
    {
      "epoch": 0.4306702515844056,
      "grad_norm": 0.0062982686795294285,
      "learning_rate": 0.22376553403036173,
      "loss": 0.5099,
      "num_input_tokens_seen": 18490272,
      "step": 13455
    },
    {
      "epoch": 0.4308302925548941,
      "grad_norm": 0.00454499339684844,
      "learning_rate": 0.22371423840708662,
      "loss": 0.5348,
      "num_input_tokens_seen": 18497056,
      "step": 13460
    },
    {
      "epoch": 0.4309903335253825,
      "grad_norm": 0.006036759819835424,
      "learning_rate": 0.22366293141614962,
      "loss": 0.5549,
      "num_input_tokens_seen": 18503744,
      "step": 13465
    },
    {
      "epoch": 0.43115037449587096,
      "grad_norm": 0.010432951152324677,
      "learning_rate": 0.22361161306546287,
      "loss": 0.5562,
      "num_input_tokens_seen": 18510544,
      "step": 13470
    },
    {
      "epoch": 0.4313104154663594,
      "grad_norm": 0.0071235401555895805,
      "learning_rate": 0.22356028336294037,
      "loss": 0.7635,
      "num_input_tokens_seen": 18517920,
      "step": 13475
    },
    {
      "epoch": 0.43147045643684784,
      "grad_norm": 0.00252223270945251,
      "learning_rate": 0.2235089423164977,
      "loss": 0.4345,
      "num_input_tokens_seen": 18525024,
      "step": 13480
    },
    {
      "epoch": 0.4316304974073363,
      "grad_norm": 0.006844705902040005,
      "learning_rate": 0.22345758993405243,
      "loss": 0.4679,
      "num_input_tokens_seen": 18532416,
      "step": 13485
    },
    {
      "epoch": 0.4317905383778247,
      "grad_norm": 0.005129039287567139,
      "learning_rate": 0.2234062262235236,
      "loss": 0.5949,
      "num_input_tokens_seen": 18539392,
      "step": 13490
    },
    {
      "epoch": 0.43195057934831316,
      "grad_norm": 0.002969620516523719,
      "learning_rate": 0.22335485119283222,
      "loss": 0.6243,
      "num_input_tokens_seen": 18546640,
      "step": 13495
    },
    {
      "epoch": 0.4321106203188016,
      "grad_norm": 0.002871645847335458,
      "learning_rate": 0.22330346484990093,
      "loss": 0.4633,
      "num_input_tokens_seen": 18553248,
      "step": 13500
    },
    {
      "epoch": 0.43227066128929004,
      "grad_norm": 0.0057830726727843285,
      "learning_rate": 0.22325206720265425,
      "loss": 0.6634,
      "num_input_tokens_seen": 18560048,
      "step": 13505
    },
    {
      "epoch": 0.4324307022597785,
      "grad_norm": 0.004786062985658646,
      "learning_rate": 0.2232006582590182,
      "loss": 0.4413,
      "num_input_tokens_seen": 18566688,
      "step": 13510
    },
    {
      "epoch": 0.43259074323026697,
      "grad_norm": 0.005868189036846161,
      "learning_rate": 0.22314923802692077,
      "loss": 0.5455,
      "num_input_tokens_seen": 18573520,
      "step": 13515
    },
    {
      "epoch": 0.4327507842007554,
      "grad_norm": 0.00915228296071291,
      "learning_rate": 0.22309780651429156,
      "loss": 0.3999,
      "num_input_tokens_seen": 18580576,
      "step": 13520
    },
    {
      "epoch": 0.43291082517124385,
      "grad_norm": 0.0060744863003492355,
      "learning_rate": 0.22304636372906203,
      "loss": 0.6737,
      "num_input_tokens_seen": 18587504,
      "step": 13525
    },
    {
      "epoch": 0.4330708661417323,
      "grad_norm": 0.00306342588737607,
      "learning_rate": 0.22299490967916522,
      "loss": 0.6538,
      "num_input_tokens_seen": 18594432,
      "step": 13530
    },
    {
      "epoch": 0.43323090711222073,
      "grad_norm": 0.004801356699317694,
      "learning_rate": 0.22294344437253602,
      "loss": 0.4012,
      "num_input_tokens_seen": 18601120,
      "step": 13535
    },
    {
      "epoch": 0.43339094808270917,
      "grad_norm": 0.006448380649089813,
      "learning_rate": 0.22289196781711101,
      "loss": 0.6925,
      "num_input_tokens_seen": 18608208,
      "step": 13540
    },
    {
      "epoch": 0.4335509890531976,
      "grad_norm": 0.0047388034872710705,
      "learning_rate": 0.2228404800208286,
      "loss": 0.5267,
      "num_input_tokens_seen": 18615456,
      "step": 13545
    },
    {
      "epoch": 0.43371103002368605,
      "grad_norm": 0.008843871764838696,
      "learning_rate": 0.22278898099162875,
      "loss": 0.6637,
      "num_input_tokens_seen": 18622432,
      "step": 13550
    },
    {
      "epoch": 0.4338710709941745,
      "grad_norm": 0.006665028166025877,
      "learning_rate": 0.22273747073745337,
      "loss": 0.6277,
      "num_input_tokens_seen": 18629584,
      "step": 13555
    },
    {
      "epoch": 0.43403111196466293,
      "grad_norm": 0.012428068555891514,
      "learning_rate": 0.22268594926624588,
      "loss": 0.5997,
      "num_input_tokens_seen": 18636416,
      "step": 13560
    },
    {
      "epoch": 0.4341911529351514,
      "grad_norm": 0.013269086368381977,
      "learning_rate": 0.22263441658595162,
      "loss": 0.5732,
      "num_input_tokens_seen": 18642800,
      "step": 13565
    },
    {
      "epoch": 0.43435119390563987,
      "grad_norm": 0.0059837172739207745,
      "learning_rate": 0.2225828727045175,
      "loss": 0.4073,
      "num_input_tokens_seen": 18649648,
      "step": 13570
    },
    {
      "epoch": 0.4345112348761283,
      "grad_norm": 0.002927779918536544,
      "learning_rate": 0.22253131762989228,
      "loss": 0.4001,
      "num_input_tokens_seen": 18656496,
      "step": 13575
    },
    {
      "epoch": 0.43467127584661674,
      "grad_norm": 0.008931845426559448,
      "learning_rate": 0.2224797513700264,
      "loss": 0.4513,
      "num_input_tokens_seen": 18663392,
      "step": 13580
    },
    {
      "epoch": 0.4348313168171052,
      "grad_norm": 0.0040091341361403465,
      "learning_rate": 0.22242817393287204,
      "loss": 0.8003,
      "num_input_tokens_seen": 18670304,
      "step": 13585
    },
    {
      "epoch": 0.4349913577875936,
      "grad_norm": 0.0026587904430925846,
      "learning_rate": 0.22237658532638305,
      "loss": 0.6122,
      "num_input_tokens_seen": 18676976,
      "step": 13590
    },
    {
      "epoch": 0.43515139875808206,
      "grad_norm": 0.005222569219768047,
      "learning_rate": 0.22232498555851513,
      "loss": 0.4928,
      "num_input_tokens_seen": 18683600,
      "step": 13595
    },
    {
      "epoch": 0.4353114397285705,
      "grad_norm": 0.004378434270620346,
      "learning_rate": 0.22227337463722546,
      "loss": 0.6642,
      "num_input_tokens_seen": 18690528,
      "step": 13600
    },
    {
      "epoch": 0.4353114397285705,
      "eval_loss": 0.581483006477356,
      "eval_runtime": 332.0285,
      "eval_samples_per_second": 41.822,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 18690528,
      "step": 13600
    },
    {
      "epoch": 0.43547148069905894,
      "grad_norm": 0.0058560436591506,
      "learning_rate": 0.2222217525704732,
      "loss": 0.6092,
      "num_input_tokens_seen": 18697200,
      "step": 13605
    },
    {
      "epoch": 0.4356315216695474,
      "grad_norm": 0.010464085265994072,
      "learning_rate": 0.22217011936621908,
      "loss": 0.5508,
      "num_input_tokens_seen": 18703552,
      "step": 13610
    },
    {
      "epoch": 0.4357915626400358,
      "grad_norm": 0.005592316389083862,
      "learning_rate": 0.22211847503242566,
      "loss": 0.5269,
      "num_input_tokens_seen": 18710416,
      "step": 13615
    },
    {
      "epoch": 0.4359516036105243,
      "grad_norm": 0.004251630045473576,
      "learning_rate": 0.22206681957705704,
      "loss": 0.6148,
      "num_input_tokens_seen": 18717328,
      "step": 13620
    },
    {
      "epoch": 0.43611164458101276,
      "grad_norm": 0.004781563300639391,
      "learning_rate": 0.2220151530080792,
      "loss": 0.7209,
      "num_input_tokens_seen": 18724080,
      "step": 13625
    },
    {
      "epoch": 0.4362716855515012,
      "grad_norm": 0.006745137739926577,
      "learning_rate": 0.2219634753334598,
      "loss": 0.596,
      "num_input_tokens_seen": 18730608,
      "step": 13630
    },
    {
      "epoch": 0.43643172652198964,
      "grad_norm": 0.005043960642069578,
      "learning_rate": 0.22191178656116817,
      "loss": 0.4789,
      "num_input_tokens_seen": 18737904,
      "step": 13635
    },
    {
      "epoch": 0.4365917674924781,
      "grad_norm": 0.0095107676461339,
      "learning_rate": 0.2218600866991753,
      "loss": 0.7875,
      "num_input_tokens_seen": 18744944,
      "step": 13640
    },
    {
      "epoch": 0.4367518084629665,
      "grad_norm": 0.005194756202399731,
      "learning_rate": 0.221808375755454,
      "loss": 0.7807,
      "num_input_tokens_seen": 18752384,
      "step": 13645
    },
    {
      "epoch": 0.43691184943345496,
      "grad_norm": 0.0064672911539673805,
      "learning_rate": 0.22175665373797881,
      "loss": 0.672,
      "num_input_tokens_seen": 18759600,
      "step": 13650
    },
    {
      "epoch": 0.4370718904039434,
      "grad_norm": 0.0060839541256427765,
      "learning_rate": 0.22170492065472583,
      "loss": 0.8479,
      "num_input_tokens_seen": 18766496,
      "step": 13655
    },
    {
      "epoch": 0.43723193137443184,
      "grad_norm": 0.0036357876379042864,
      "learning_rate": 0.221653176513673,
      "loss": 0.4027,
      "num_input_tokens_seen": 18773216,
      "step": 13660
    },
    {
      "epoch": 0.4373919723449203,
      "grad_norm": 0.005700281355530024,
      "learning_rate": 0.2216014213227999,
      "loss": 0.4494,
      "num_input_tokens_seen": 18780176,
      "step": 13665
    },
    {
      "epoch": 0.43755201331540877,
      "grad_norm": 0.005402098409831524,
      "learning_rate": 0.22154965509008784,
      "loss": 0.5642,
      "num_input_tokens_seen": 18787376,
      "step": 13670
    },
    {
      "epoch": 0.4377120542858972,
      "grad_norm": 0.0035929912701249123,
      "learning_rate": 0.2214978778235198,
      "loss": 0.4484,
      "num_input_tokens_seen": 18794176,
      "step": 13675
    },
    {
      "epoch": 0.43787209525638565,
      "grad_norm": 0.0038729095831513405,
      "learning_rate": 0.2214460895310805,
      "loss": 0.4147,
      "num_input_tokens_seen": 18801248,
      "step": 13680
    },
    {
      "epoch": 0.4380321362268741,
      "grad_norm": 0.007093150168657303,
      "learning_rate": 0.22139429022075635,
      "loss": 0.7745,
      "num_input_tokens_seen": 18808032,
      "step": 13685
    },
    {
      "epoch": 0.43819217719736253,
      "grad_norm": 0.009297777898609638,
      "learning_rate": 0.22134247990053546,
      "loss": 0.7525,
      "num_input_tokens_seen": 18815120,
      "step": 13690
    },
    {
      "epoch": 0.43835221816785097,
      "grad_norm": 0.007002192549407482,
      "learning_rate": 0.2212906585784076,
      "loss": 0.5846,
      "num_input_tokens_seen": 18821936,
      "step": 13695
    },
    {
      "epoch": 0.4385122591383394,
      "grad_norm": 0.004588388837873936,
      "learning_rate": 0.22123882626236432,
      "loss": 0.7243,
      "num_input_tokens_seen": 18828768,
      "step": 13700
    },
    {
      "epoch": 0.43867230010882785,
      "grad_norm": 0.0049143903888762,
      "learning_rate": 0.2211869829603988,
      "loss": 0.7049,
      "num_input_tokens_seen": 18835696,
      "step": 13705
    },
    {
      "epoch": 0.4388323410793163,
      "grad_norm": 0.006719362922012806,
      "learning_rate": 0.22113512868050592,
      "loss": 0.6289,
      "num_input_tokens_seen": 18843088,
      "step": 13710
    },
    {
      "epoch": 0.43899238204980473,
      "grad_norm": 0.0070996894501149654,
      "learning_rate": 0.2210832634306822,
      "loss": 0.6303,
      "num_input_tokens_seen": 18849840,
      "step": 13715
    },
    {
      "epoch": 0.43915242302029317,
      "grad_norm": 0.0036435085348784924,
      "learning_rate": 0.22103138721892598,
      "loss": 0.5562,
      "num_input_tokens_seen": 18856816,
      "step": 13720
    },
    {
      "epoch": 0.43931246399078167,
      "grad_norm": 0.0033662396017462015,
      "learning_rate": 0.22097950005323724,
      "loss": 0.6353,
      "num_input_tokens_seen": 18863616,
      "step": 13725
    },
    {
      "epoch": 0.4394725049612701,
      "grad_norm": 0.0036736358888447285,
      "learning_rate": 0.22092760194161762,
      "loss": 0.5642,
      "num_input_tokens_seen": 18870560,
      "step": 13730
    },
    {
      "epoch": 0.43963254593175854,
      "grad_norm": 0.006254172418266535,
      "learning_rate": 0.2208756928920704,
      "loss": 0.6061,
      "num_input_tokens_seen": 18877344,
      "step": 13735
    },
    {
      "epoch": 0.439792586902247,
      "grad_norm": 0.004430749453604221,
      "learning_rate": 0.22082377291260072,
      "loss": 0.678,
      "num_input_tokens_seen": 18884368,
      "step": 13740
    },
    {
      "epoch": 0.4399526278727354,
      "grad_norm": 0.008711040019989014,
      "learning_rate": 0.2207718420112152,
      "loss": 0.5463,
      "num_input_tokens_seen": 18891200,
      "step": 13745
    },
    {
      "epoch": 0.44011266884322386,
      "grad_norm": 0.009512564167380333,
      "learning_rate": 0.22071990019592228,
      "loss": 0.6025,
      "num_input_tokens_seen": 18897744,
      "step": 13750
    },
    {
      "epoch": 0.4402727098137123,
      "grad_norm": 0.006895586382597685,
      "learning_rate": 0.22066794747473198,
      "loss": 0.6114,
      "num_input_tokens_seen": 18904336,
      "step": 13755
    },
    {
      "epoch": 0.44043275078420074,
      "grad_norm": 0.005069756414741278,
      "learning_rate": 0.2206159838556562,
      "loss": 0.4612,
      "num_input_tokens_seen": 18911376,
      "step": 13760
    },
    {
      "epoch": 0.4405927917546892,
      "grad_norm": 0.004345351830124855,
      "learning_rate": 0.2205640093467082,
      "loss": 0.6592,
      "num_input_tokens_seen": 18918288,
      "step": 13765
    },
    {
      "epoch": 0.4407528327251776,
      "grad_norm": 0.0029107253067195415,
      "learning_rate": 0.22051202395590322,
      "loss": 0.5048,
      "num_input_tokens_seen": 18925392,
      "step": 13770
    },
    {
      "epoch": 0.44091287369566606,
      "grad_norm": 0.004983037244528532,
      "learning_rate": 0.22046002769125808,
      "loss": 0.6259,
      "num_input_tokens_seen": 18932144,
      "step": 13775
    },
    {
      "epoch": 0.44107291466615456,
      "grad_norm": 0.004501597955822945,
      "learning_rate": 0.2204080205607912,
      "loss": 0.4942,
      "num_input_tokens_seen": 18938912,
      "step": 13780
    },
    {
      "epoch": 0.441232955636643,
      "grad_norm": 0.004495460074394941,
      "learning_rate": 0.22035600257252272,
      "loss": 0.5636,
      "num_input_tokens_seen": 18945792,
      "step": 13785
    },
    {
      "epoch": 0.44139299660713144,
      "grad_norm": 0.005447358824312687,
      "learning_rate": 0.2203039737344745,
      "loss": 0.5162,
      "num_input_tokens_seen": 18952768,
      "step": 13790
    },
    {
      "epoch": 0.4415530375776199,
      "grad_norm": 0.007323459256440401,
      "learning_rate": 0.22025193405467003,
      "loss": 0.5741,
      "num_input_tokens_seen": 18959504,
      "step": 13795
    },
    {
      "epoch": 0.4417130785481083,
      "grad_norm": 0.01664169691503048,
      "learning_rate": 0.2201998835411345,
      "loss": 0.6459,
      "num_input_tokens_seen": 18966352,
      "step": 13800
    },
    {
      "epoch": 0.4417130785481083,
      "eval_loss": 0.5842155814170837,
      "eval_runtime": 332.0178,
      "eval_samples_per_second": 41.823,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 18966352,
      "step": 13800
    },
    {
      "epoch": 0.44187311951859676,
      "grad_norm": 0.00635602418333292,
      "learning_rate": 0.22014782220189474,
      "loss": 0.5851,
      "num_input_tokens_seen": 18973856,
      "step": 13805
    },
    {
      "epoch": 0.4420331604890852,
      "grad_norm": 0.00446308171376586,
      "learning_rate": 0.2200957500449793,
      "loss": 0.4812,
      "num_input_tokens_seen": 18980752,
      "step": 13810
    },
    {
      "epoch": 0.44219320145957364,
      "grad_norm": 0.006565245334059,
      "learning_rate": 0.22004366707841827,
      "loss": 0.6152,
      "num_input_tokens_seen": 18987712,
      "step": 13815
    },
    {
      "epoch": 0.4423532424300621,
      "grad_norm": 0.0040622130036354065,
      "learning_rate": 0.21999157331024358,
      "loss": 0.505,
      "num_input_tokens_seen": 18994704,
      "step": 13820
    },
    {
      "epoch": 0.4425132834005505,
      "grad_norm": 0.005523551721125841,
      "learning_rate": 0.21993946874848871,
      "loss": 0.6667,
      "num_input_tokens_seen": 19001344,
      "step": 13825
    },
    {
      "epoch": 0.442673324371039,
      "grad_norm": 0.008351782336831093,
      "learning_rate": 0.2198873534011888,
      "loss": 0.6622,
      "num_input_tokens_seen": 19008256,
      "step": 13830
    },
    {
      "epoch": 0.44283336534152745,
      "grad_norm": 0.004423895385116339,
      "learning_rate": 0.2198352272763808,
      "loss": 0.5713,
      "num_input_tokens_seen": 19015104,
      "step": 13835
    },
    {
      "epoch": 0.4429934063120159,
      "grad_norm": 0.004070617258548737,
      "learning_rate": 0.2197830903821031,
      "loss": 0.5896,
      "num_input_tokens_seen": 19022128,
      "step": 13840
    },
    {
      "epoch": 0.44315344728250433,
      "grad_norm": 0.004073071759194136,
      "learning_rate": 0.21973094272639598,
      "loss": 0.602,
      "num_input_tokens_seen": 19028960,
      "step": 13845
    },
    {
      "epoch": 0.44331348825299277,
      "grad_norm": 0.0026866665575653315,
      "learning_rate": 0.21967878431730117,
      "loss": 0.5498,
      "num_input_tokens_seen": 19035968,
      "step": 13850
    },
    {
      "epoch": 0.4434735292234812,
      "grad_norm": 0.003637227462604642,
      "learning_rate": 0.21962661516286217,
      "loss": 0.5086,
      "num_input_tokens_seen": 19042832,
      "step": 13855
    },
    {
      "epoch": 0.44363357019396965,
      "grad_norm": 0.004376082681119442,
      "learning_rate": 0.21957443527112414,
      "loss": 0.5375,
      "num_input_tokens_seen": 19049616,
      "step": 13860
    },
    {
      "epoch": 0.4437936111644581,
      "grad_norm": 0.0088508864864707,
      "learning_rate": 0.21952224465013384,
      "loss": 0.7491,
      "num_input_tokens_seen": 19056400,
      "step": 13865
    },
    {
      "epoch": 0.44395365213494653,
      "grad_norm": 0.00328950397670269,
      "learning_rate": 0.21947004330793976,
      "loss": 0.6119,
      "num_input_tokens_seen": 19062944,
      "step": 13870
    },
    {
      "epoch": 0.44411369310543497,
      "grad_norm": 0.0030088829807937145,
      "learning_rate": 0.21941783125259198,
      "loss": 0.5212,
      "num_input_tokens_seen": 19069584,
      "step": 13875
    },
    {
      "epoch": 0.4442737340759234,
      "grad_norm": 0.006782303564250469,
      "learning_rate": 0.21936560849214226,
      "loss": 0.5846,
      "num_input_tokens_seen": 19076576,
      "step": 13880
    },
    {
      "epoch": 0.4444337750464119,
      "grad_norm": 0.006142520345747471,
      "learning_rate": 0.21931337503464404,
      "loss": 0.5754,
      "num_input_tokens_seen": 19083552,
      "step": 13885
    },
    {
      "epoch": 0.44459381601690035,
      "grad_norm": 0.004215502180159092,
      "learning_rate": 0.21926113088815233,
      "loss": 0.6863,
      "num_input_tokens_seen": 19090352,
      "step": 13890
    },
    {
      "epoch": 0.4447538569873888,
      "grad_norm": 0.0035383147187530994,
      "learning_rate": 0.2192088760607238,
      "loss": 0.5415,
      "num_input_tokens_seen": 19097088,
      "step": 13895
    },
    {
      "epoch": 0.4449138979578772,
      "grad_norm": 0.005089843180030584,
      "learning_rate": 0.2191566105604169,
      "loss": 0.4539,
      "num_input_tokens_seen": 19103824,
      "step": 13900
    },
    {
      "epoch": 0.44507393892836566,
      "grad_norm": 0.006053334102034569,
      "learning_rate": 0.21910433439529153,
      "loss": 0.6409,
      "num_input_tokens_seen": 19110432,
      "step": 13905
    },
    {
      "epoch": 0.4452339798988541,
      "grad_norm": 0.004694880452007055,
      "learning_rate": 0.2190520475734094,
      "loss": 0.7138,
      "num_input_tokens_seen": 19117216,
      "step": 13910
    },
    {
      "epoch": 0.44539402086934254,
      "grad_norm": 0.0021966344211250544,
      "learning_rate": 0.2189997501028338,
      "loss": 0.6687,
      "num_input_tokens_seen": 19124160,
      "step": 13915
    },
    {
      "epoch": 0.445554061839831,
      "grad_norm": 0.004226788878440857,
      "learning_rate": 0.2189474419916296,
      "loss": 0.6276,
      "num_input_tokens_seen": 19131104,
      "step": 13920
    },
    {
      "epoch": 0.4457141028103194,
      "grad_norm": 0.005559123121201992,
      "learning_rate": 0.21889512324786342,
      "loss": 0.6561,
      "num_input_tokens_seen": 19138144,
      "step": 13925
    },
    {
      "epoch": 0.44587414378080786,
      "grad_norm": 0.0031376623082906008,
      "learning_rate": 0.21884279387960345,
      "loss": 0.6241,
      "num_input_tokens_seen": 19145328,
      "step": 13930
    },
    {
      "epoch": 0.44603418475129636,
      "grad_norm": 0.0045607248321175575,
      "learning_rate": 0.2187904538949195,
      "loss": 0.4378,
      "num_input_tokens_seen": 19152144,
      "step": 13935
    },
    {
      "epoch": 0.4461942257217848,
      "grad_norm": 0.0028711019549518824,
      "learning_rate": 0.2187381033018831,
      "loss": 0.547,
      "num_input_tokens_seen": 19159360,
      "step": 13940
    },
    {
      "epoch": 0.44635426669227324,
      "grad_norm": 0.004800685681402683,
      "learning_rate": 0.2186857421085673,
      "loss": 0.4866,
      "num_input_tokens_seen": 19165888,
      "step": 13945
    },
    {
      "epoch": 0.4465143076627617,
      "grad_norm": 0.0036908374167978764,
      "learning_rate": 0.21863337032304697,
      "loss": 0.5607,
      "num_input_tokens_seen": 19172752,
      "step": 13950
    },
    {
      "epoch": 0.4466743486332501,
      "grad_norm": 0.006221655290573835,
      "learning_rate": 0.21858098795339845,
      "loss": 0.5426,
      "num_input_tokens_seen": 19179936,
      "step": 13955
    },
    {
      "epoch": 0.44683438960373856,
      "grad_norm": 0.007657644338905811,
      "learning_rate": 0.21852859500769975,
      "loss": 0.7046,
      "num_input_tokens_seen": 19187056,
      "step": 13960
    },
    {
      "epoch": 0.446994430574227,
      "grad_norm": 0.003335784887894988,
      "learning_rate": 0.21847619149403044,
      "loss": 0.5656,
      "num_input_tokens_seen": 19193696,
      "step": 13965
    },
    {
      "epoch": 0.44715447154471544,
      "grad_norm": 0.0029588951729238033,
      "learning_rate": 0.21842377742047195,
      "loss": 0.5534,
      "num_input_tokens_seen": 19200720,
      "step": 13970
    },
    {
      "epoch": 0.4473145125152039,
      "grad_norm": 0.0025768959894776344,
      "learning_rate": 0.21837135279510705,
      "loss": 0.4085,
      "num_input_tokens_seen": 19207696,
      "step": 13975
    },
    {
      "epoch": 0.4474745534856923,
      "grad_norm": 0.004774769768118858,
      "learning_rate": 0.21831891762602038,
      "loss": 0.5314,
      "num_input_tokens_seen": 19214816,
      "step": 13980
    },
    {
      "epoch": 0.44763459445618076,
      "grad_norm": 0.006812114268541336,
      "learning_rate": 0.21826647192129806,
      "loss": 0.5845,
      "num_input_tokens_seen": 19221616,
      "step": 13985
    },
    {
      "epoch": 0.44779463542666925,
      "grad_norm": 0.0065218559466302395,
      "learning_rate": 0.21821401568902787,
      "loss": 0.5581,
      "num_input_tokens_seen": 19228304,
      "step": 13990
    },
    {
      "epoch": 0.4479546763971577,
      "grad_norm": 0.003645852906629443,
      "learning_rate": 0.21816154893729925,
      "loss": 0.5514,
      "num_input_tokens_seen": 19235120,
      "step": 13995
    },
    {
      "epoch": 0.44811471736764613,
      "grad_norm": 0.003850736888125539,
      "learning_rate": 0.2181090716742032,
      "loss": 0.5672,
      "num_input_tokens_seen": 19242160,
      "step": 14000
    },
    {
      "epoch": 0.44811471736764613,
      "eval_loss": 0.5743791460990906,
      "eval_runtime": 331.8238,
      "eval_samples_per_second": 41.848,
      "eval_steps_per_second": 20.924,
      "num_input_tokens_seen": 19242160,
      "step": 14000
    },
    {
      "epoch": 0.44827475833813457,
      "grad_norm": 0.0034400122240185738,
      "learning_rate": 0.21805658390783236,
      "loss": 0.4273,
      "num_input_tokens_seen": 19249008,
      "step": 14005
    },
    {
      "epoch": 0.448434799308623,
      "grad_norm": 0.00405319407582283,
      "learning_rate": 0.21800408564628107,
      "loss": 0.5361,
      "num_input_tokens_seen": 19255968,
      "step": 14010
    },
    {
      "epoch": 0.44859484027911145,
      "grad_norm": 0.009222842752933502,
      "learning_rate": 0.21795157689764516,
      "loss": 0.4829,
      "num_input_tokens_seen": 19262736,
      "step": 14015
    },
    {
      "epoch": 0.4487548812495999,
      "grad_norm": 0.0070220353081822395,
      "learning_rate": 0.21789905767002216,
      "loss": 0.6106,
      "num_input_tokens_seen": 19269616,
      "step": 14020
    },
    {
      "epoch": 0.44891492222008833,
      "grad_norm": 0.0036691308487206697,
      "learning_rate": 0.2178465279715112,
      "loss": 0.368,
      "num_input_tokens_seen": 19276320,
      "step": 14025
    },
    {
      "epoch": 0.44907496319057677,
      "grad_norm": 0.004997137933969498,
      "learning_rate": 0.21779398781021303,
      "loss": 0.557,
      "num_input_tokens_seen": 19282976,
      "step": 14030
    },
    {
      "epoch": 0.4492350041610652,
      "grad_norm": 0.009348716586828232,
      "learning_rate": 0.21774143719422998,
      "loss": 0.4247,
      "num_input_tokens_seen": 19289840,
      "step": 14035
    },
    {
      "epoch": 0.4493950451315537,
      "grad_norm": 0.009410431608557701,
      "learning_rate": 0.21768887613166601,
      "loss": 0.6133,
      "num_input_tokens_seen": 19297104,
      "step": 14040
    },
    {
      "epoch": 0.44955508610204215,
      "grad_norm": 0.005836624652147293,
      "learning_rate": 0.2176363046306267,
      "loss": 0.3666,
      "num_input_tokens_seen": 19303936,
      "step": 14045
    },
    {
      "epoch": 0.4497151270725306,
      "grad_norm": 0.008383831940591335,
      "learning_rate": 0.21758372269921925,
      "loss": 0.4931,
      "num_input_tokens_seen": 19310640,
      "step": 14050
    },
    {
      "epoch": 0.449875168043019,
      "grad_norm": 0.006667809095233679,
      "learning_rate": 0.21753113034555244,
      "loss": 0.6065,
      "num_input_tokens_seen": 19317488,
      "step": 14055
    },
    {
      "epoch": 0.45003520901350746,
      "grad_norm": 0.0045168776996433735,
      "learning_rate": 0.2174785275777367,
      "loss": 0.4989,
      "num_input_tokens_seen": 19324224,
      "step": 14060
    },
    {
      "epoch": 0.4501952499839959,
      "grad_norm": 0.006152520887553692,
      "learning_rate": 0.21742591440388404,
      "loss": 0.662,
      "num_input_tokens_seen": 19331216,
      "step": 14065
    },
    {
      "epoch": 0.45035529095448434,
      "grad_norm": 0.008301906287670135,
      "learning_rate": 0.21737329083210802,
      "loss": 0.6712,
      "num_input_tokens_seen": 19338528,
      "step": 14070
    },
    {
      "epoch": 0.4505153319249728,
      "grad_norm": 0.009036107920110226,
      "learning_rate": 0.2173206568705239,
      "loss": 0.5635,
      "num_input_tokens_seen": 19345328,
      "step": 14075
    },
    {
      "epoch": 0.4506753728954612,
      "grad_norm": 0.008538501337170601,
      "learning_rate": 0.2172680125272485,
      "loss": 0.6598,
      "num_input_tokens_seen": 19352080,
      "step": 14080
    },
    {
      "epoch": 0.45083541386594966,
      "grad_norm": 0.006040395237505436,
      "learning_rate": 0.2172153578104002,
      "loss": 0.4741,
      "num_input_tokens_seen": 19358512,
      "step": 14085
    },
    {
      "epoch": 0.4509954548364381,
      "grad_norm": 0.003841026686131954,
      "learning_rate": 0.21716269272809902,
      "loss": 0.48,
      "num_input_tokens_seen": 19365744,
      "step": 14090
    },
    {
      "epoch": 0.4511554958069266,
      "grad_norm": 0.003198122838512063,
      "learning_rate": 0.21711001728846666,
      "loss": 0.4632,
      "num_input_tokens_seen": 19372432,
      "step": 14095
    },
    {
      "epoch": 0.45131553677741504,
      "grad_norm": 0.004490072373300791,
      "learning_rate": 0.21705733149962628,
      "loss": 0.6097,
      "num_input_tokens_seen": 19379664,
      "step": 14100
    },
    {
      "epoch": 0.4514755777479035,
      "grad_norm": 0.0034536917228251696,
      "learning_rate": 0.21700463536970263,
      "loss": 0.4715,
      "num_input_tokens_seen": 19386992,
      "step": 14105
    },
    {
      "epoch": 0.4516356187183919,
      "grad_norm": 0.006663980428129435,
      "learning_rate": 0.21695192890682222,
      "loss": 0.5306,
      "num_input_tokens_seen": 19393856,
      "step": 14110
    },
    {
      "epoch": 0.45179565968888036,
      "grad_norm": 0.0048205237835645676,
      "learning_rate": 0.21689921211911298,
      "loss": 0.6083,
      "num_input_tokens_seen": 19400656,
      "step": 14115
    },
    {
      "epoch": 0.4519557006593688,
      "grad_norm": 0.002306311158463359,
      "learning_rate": 0.21684648501470452,
      "loss": 0.4528,
      "num_input_tokens_seen": 19407216,
      "step": 14120
    },
    {
      "epoch": 0.45211574162985724,
      "grad_norm": 0.004022337030619383,
      "learning_rate": 0.216793747601728,
      "loss": 0.5969,
      "num_input_tokens_seen": 19414384,
      "step": 14125
    },
    {
      "epoch": 0.4522757826003457,
      "grad_norm": 0.0032270292285829782,
      "learning_rate": 0.21674099988831627,
      "loss": 0.4706,
      "num_input_tokens_seen": 19421216,
      "step": 14130
    },
    {
      "epoch": 0.4524358235708341,
      "grad_norm": 0.009366092272102833,
      "learning_rate": 0.21668824188260363,
      "loss": 0.5175,
      "num_input_tokens_seen": 19428080,
      "step": 14135
    },
    {
      "epoch": 0.45259586454132256,
      "grad_norm": 0.004702613223344088,
      "learning_rate": 0.21663547359272606,
      "loss": 0.6244,
      "num_input_tokens_seen": 19435312,
      "step": 14140
    },
    {
      "epoch": 0.452755905511811,
      "grad_norm": 0.005211134906858206,
      "learning_rate": 0.216582695026821,
      "loss": 0.6386,
      "num_input_tokens_seen": 19442176,
      "step": 14145
    },
    {
      "epoch": 0.4529159464822995,
      "grad_norm": 0.004565619397908449,
      "learning_rate": 0.21652990619302767,
      "loss": 0.678,
      "num_input_tokens_seen": 19449584,
      "step": 14150
    },
    {
      "epoch": 0.45307598745278793,
      "grad_norm": 0.005151933990418911,
      "learning_rate": 0.21647710709948673,
      "loss": 0.5801,
      "num_input_tokens_seen": 19456560,
      "step": 14155
    },
    {
      "epoch": 0.45323602842327637,
      "grad_norm": 0.0031435545533895493,
      "learning_rate": 0.2164242977543405,
      "loss": 0.5094,
      "num_input_tokens_seen": 19463168,
      "step": 14160
    },
    {
      "epoch": 0.4533960693937648,
      "grad_norm": 0.0071044196374714375,
      "learning_rate": 0.21637147816573277,
      "loss": 0.6286,
      "num_input_tokens_seen": 19469728,
      "step": 14165
    },
    {
      "epoch": 0.45355611036425325,
      "grad_norm": 0.005444220267236233,
      "learning_rate": 0.21631864834180908,
      "loss": 0.603,
      "num_input_tokens_seen": 19476384,
      "step": 14170
    },
    {
      "epoch": 0.4537161513347417,
      "grad_norm": 0.002925675129517913,
      "learning_rate": 0.21626580829071637,
      "loss": 0.5162,
      "num_input_tokens_seen": 19483248,
      "step": 14175
    },
    {
      "epoch": 0.45387619230523013,
      "grad_norm": 0.0036091716028749943,
      "learning_rate": 0.21621295802060328,
      "loss": 0.575,
      "num_input_tokens_seen": 19490176,
      "step": 14180
    },
    {
      "epoch": 0.45403623327571857,
      "grad_norm": 0.0046430351212620735,
      "learning_rate": 0.21616009753961996,
      "loss": 0.4527,
      "num_input_tokens_seen": 19497104,
      "step": 14185
    },
    {
      "epoch": 0.454196274246207,
      "grad_norm": 0.006073473021388054,
      "learning_rate": 0.2161072268559182,
      "loss": 0.5756,
      "num_input_tokens_seen": 19504416,
      "step": 14190
    },
    {
      "epoch": 0.45435631521669545,
      "grad_norm": 0.003987867385149002,
      "learning_rate": 0.21605434597765133,
      "loss": 0.5501,
      "num_input_tokens_seen": 19511712,
      "step": 14195
    },
    {
      "epoch": 0.45451635618718395,
      "grad_norm": 0.004904948640614748,
      "learning_rate": 0.21600145491297418,
      "loss": 0.5367,
      "num_input_tokens_seen": 19518832,
      "step": 14200
    },
    {
      "epoch": 0.45451635618718395,
      "eval_loss": 0.5750772953033447,
      "eval_runtime": 332.0423,
      "eval_samples_per_second": 41.82,
      "eval_steps_per_second": 20.91,
      "num_input_tokens_seen": 19518832,
      "step": 14200
    },
    {
      "epoch": 0.4546763971576724,
      "grad_norm": 0.008376479148864746,
      "learning_rate": 0.21594855367004326,
      "loss": 0.512,
      "num_input_tokens_seen": 19525488,
      "step": 14205
    },
    {
      "epoch": 0.4548364381281608,
      "grad_norm": 0.003971856087446213,
      "learning_rate": 0.21589564225701663,
      "loss": 0.5207,
      "num_input_tokens_seen": 19532224,
      "step": 14210
    },
    {
      "epoch": 0.45499647909864926,
      "grad_norm": 0.0022163009271025658,
      "learning_rate": 0.21584272068205385,
      "loss": 0.3571,
      "num_input_tokens_seen": 19538992,
      "step": 14215
    },
    {
      "epoch": 0.4551565200691377,
      "grad_norm": 0.008147991262376308,
      "learning_rate": 0.2157897889533161,
      "loss": 0.5931,
      "num_input_tokens_seen": 19546048,
      "step": 14220
    },
    {
      "epoch": 0.45531656103962614,
      "grad_norm": 0.003819553880020976,
      "learning_rate": 0.21573684707896612,
      "loss": 0.5512,
      "num_input_tokens_seen": 19553008,
      "step": 14225
    },
    {
      "epoch": 0.4554766020101146,
      "grad_norm": 0.0023833010345697403,
      "learning_rate": 0.21568389506716826,
      "loss": 0.4322,
      "num_input_tokens_seen": 19560160,
      "step": 14230
    },
    {
      "epoch": 0.455636642980603,
      "grad_norm": 0.011556286364793777,
      "learning_rate": 0.21563093292608831,
      "loss": 0.7183,
      "num_input_tokens_seen": 19567056,
      "step": 14235
    },
    {
      "epoch": 0.45579668395109146,
      "grad_norm": 0.003922497853636742,
      "learning_rate": 0.21557796066389376,
      "loss": 0.7215,
      "num_input_tokens_seen": 19574128,
      "step": 14240
    },
    {
      "epoch": 0.4559567249215799,
      "grad_norm": 0.006760784890502691,
      "learning_rate": 0.21552497828875353,
      "loss": 0.4353,
      "num_input_tokens_seen": 19581520,
      "step": 14245
    },
    {
      "epoch": 0.45611676589206834,
      "grad_norm": 0.006081345025449991,
      "learning_rate": 0.21547198580883828,
      "loss": 0.5245,
      "num_input_tokens_seen": 19588432,
      "step": 14250
    },
    {
      "epoch": 0.45627680686255684,
      "grad_norm": 0.010255822911858559,
      "learning_rate": 0.21541898323232,
      "loss": 0.7414,
      "num_input_tokens_seen": 19595328,
      "step": 14255
    },
    {
      "epoch": 0.4564368478330453,
      "grad_norm": 0.004524414427578449,
      "learning_rate": 0.2153659705673724,
      "loss": 0.5766,
      "num_input_tokens_seen": 19601824,
      "step": 14260
    },
    {
      "epoch": 0.4565968888035337,
      "grad_norm": 0.003551561152562499,
      "learning_rate": 0.2153129478221707,
      "loss": 0.5497,
      "num_input_tokens_seen": 19608976,
      "step": 14265
    },
    {
      "epoch": 0.45675692977402216,
      "grad_norm": 0.007406030315905809,
      "learning_rate": 0.21525991500489164,
      "loss": 0.4673,
      "num_input_tokens_seen": 19615984,
      "step": 14270
    },
    {
      "epoch": 0.4569169707445106,
      "grad_norm": 0.003761360188946128,
      "learning_rate": 0.21520687212371362,
      "loss": 0.6633,
      "num_input_tokens_seen": 19622816,
      "step": 14275
    },
    {
      "epoch": 0.45707701171499904,
      "grad_norm": 0.0037608412094414234,
      "learning_rate": 0.21515381918681648,
      "loss": 0.5562,
      "num_input_tokens_seen": 19629424,
      "step": 14280
    },
    {
      "epoch": 0.4572370526854875,
      "grad_norm": 0.003298314521089196,
      "learning_rate": 0.21510075620238167,
      "loss": 0.5971,
      "num_input_tokens_seen": 19636160,
      "step": 14285
    },
    {
      "epoch": 0.4573970936559759,
      "grad_norm": 0.006443415302783251,
      "learning_rate": 0.21504768317859208,
      "loss": 0.6534,
      "num_input_tokens_seen": 19642784,
      "step": 14290
    },
    {
      "epoch": 0.45755713462646436,
      "grad_norm": 0.0020359987393021584,
      "learning_rate": 0.2149946001236323,
      "loss": 0.4476,
      "num_input_tokens_seen": 19649664,
      "step": 14295
    },
    {
      "epoch": 0.4577171755969528,
      "grad_norm": 0.007624128367751837,
      "learning_rate": 0.21494150704568848,
      "loss": 0.6013,
      "num_input_tokens_seen": 19656528,
      "step": 14300
    },
    {
      "epoch": 0.4578772165674413,
      "grad_norm": 0.004642912186682224,
      "learning_rate": 0.21488840395294811,
      "loss": 0.5056,
      "num_input_tokens_seen": 19663584,
      "step": 14305
    },
    {
      "epoch": 0.45803725753792973,
      "grad_norm": 0.00493493489921093,
      "learning_rate": 0.21483529085360042,
      "loss": 0.598,
      "num_input_tokens_seen": 19670640,
      "step": 14310
    },
    {
      "epoch": 0.45819729850841817,
      "grad_norm": 0.004772835876792669,
      "learning_rate": 0.2147821677558361,
      "loss": 0.6039,
      "num_input_tokens_seen": 19677840,
      "step": 14315
    },
    {
      "epoch": 0.4583573394789066,
      "grad_norm": 0.0032236569095402956,
      "learning_rate": 0.2147290346678475,
      "loss": 0.6196,
      "num_input_tokens_seen": 19684880,
      "step": 14320
    },
    {
      "epoch": 0.45851738044939505,
      "grad_norm": 0.004506459925323725,
      "learning_rate": 0.21467589159782827,
      "loss": 0.5542,
      "num_input_tokens_seen": 19692400,
      "step": 14325
    },
    {
      "epoch": 0.4586774214198835,
      "grad_norm": 0.005501519422978163,
      "learning_rate": 0.21462273855397374,
      "loss": 0.5364,
      "num_input_tokens_seen": 19698976,
      "step": 14330
    },
    {
      "epoch": 0.45883746239037193,
      "grad_norm": 0.0034387961495667696,
      "learning_rate": 0.21456957554448083,
      "loss": 0.5698,
      "num_input_tokens_seen": 19705872,
      "step": 14335
    },
    {
      "epoch": 0.45899750336086037,
      "grad_norm": 0.003244242398068309,
      "learning_rate": 0.21451640257754795,
      "loss": 0.6016,
      "num_input_tokens_seen": 19712240,
      "step": 14340
    },
    {
      "epoch": 0.4591575443313488,
      "grad_norm": 0.005115628242492676,
      "learning_rate": 0.21446321966137508,
      "loss": 0.5518,
      "num_input_tokens_seen": 19719152,
      "step": 14345
    },
    {
      "epoch": 0.45931758530183725,
      "grad_norm": 0.007559633813798428,
      "learning_rate": 0.21441002680416354,
      "loss": 0.6564,
      "num_input_tokens_seen": 19725600,
      "step": 14350
    },
    {
      "epoch": 0.4594776262723257,
      "grad_norm": 0.005976091604679823,
      "learning_rate": 0.21435682401411654,
      "loss": 0.5779,
      "num_input_tokens_seen": 19732976,
      "step": 14355
    },
    {
      "epoch": 0.4596376672428142,
      "grad_norm": 0.0038132444024086,
      "learning_rate": 0.2143036112994385,
      "loss": 0.5711,
      "num_input_tokens_seen": 19740128,
      "step": 14360
    },
    {
      "epoch": 0.4597977082133026,
      "grad_norm": 0.005868836771696806,
      "learning_rate": 0.21425038866833548,
      "loss": 0.5919,
      "num_input_tokens_seen": 19746912,
      "step": 14365
    },
    {
      "epoch": 0.45995774918379106,
      "grad_norm": 0.008071482181549072,
      "learning_rate": 0.21419715612901508,
      "loss": 0.9274,
      "num_input_tokens_seen": 19753904,
      "step": 14370
    },
    {
      "epoch": 0.4601177901542795,
      "grad_norm": 0.0031145052053034306,
      "learning_rate": 0.21414391368968652,
      "loss": 0.5336,
      "num_input_tokens_seen": 19760688,
      "step": 14375
    },
    {
      "epoch": 0.46027783112476794,
      "grad_norm": 0.009702183306217194,
      "learning_rate": 0.21409066135856034,
      "loss": 0.7544,
      "num_input_tokens_seen": 19767776,
      "step": 14380
    },
    {
      "epoch": 0.4604378720952564,
      "grad_norm": 0.005970879923552275,
      "learning_rate": 0.21403739914384878,
      "loss": 0.705,
      "num_input_tokens_seen": 19774336,
      "step": 14385
    },
    {
      "epoch": 0.4605979130657448,
      "grad_norm": 0.00370946922339499,
      "learning_rate": 0.21398412705376554,
      "loss": 0.5956,
      "num_input_tokens_seen": 19781568,
      "step": 14390
    },
    {
      "epoch": 0.46075795403623326,
      "grad_norm": 0.003723220666870475,
      "learning_rate": 0.2139308450965258,
      "loss": 0.6643,
      "num_input_tokens_seen": 19788912,
      "step": 14395
    },
    {
      "epoch": 0.4609179950067217,
      "grad_norm": 0.004223753232508898,
      "learning_rate": 0.21387755328034638,
      "loss": 0.5516,
      "num_input_tokens_seen": 19795920,
      "step": 14400
    },
    {
      "epoch": 0.4609179950067217,
      "eval_loss": 0.5729113817214966,
      "eval_runtime": 332.0874,
      "eval_samples_per_second": 41.814,
      "eval_steps_per_second": 20.907,
      "num_input_tokens_seen": 19795920,
      "step": 14400
    },
    {
      "epoch": 0.46107803597721014,
      "grad_norm": 0.00305164628662169,
      "learning_rate": 0.2138242516134455,
      "loss": 0.5676,
      "num_input_tokens_seen": 19802704,
      "step": 14405
    },
    {
      "epoch": 0.4612380769476986,
      "grad_norm": 0.005469270516186953,
      "learning_rate": 0.2137709401040429,
      "loss": 0.5724,
      "num_input_tokens_seen": 19809552,
      "step": 14410
    },
    {
      "epoch": 0.4613981179181871,
      "grad_norm": 0.003257069969549775,
      "learning_rate": 0.21371761876036,
      "loss": 0.4639,
      "num_input_tokens_seen": 19816016,
      "step": 14415
    },
    {
      "epoch": 0.4615581588886755,
      "grad_norm": 0.005052387714385986,
      "learning_rate": 0.21366428759061956,
      "loss": 0.5942,
      "num_input_tokens_seen": 19822928,
      "step": 14420
    },
    {
      "epoch": 0.46171819985916396,
      "grad_norm": 0.006474341731518507,
      "learning_rate": 0.2136109466030459,
      "loss": 0.4962,
      "num_input_tokens_seen": 19829584,
      "step": 14425
    },
    {
      "epoch": 0.4618782408296524,
      "grad_norm": 0.005766233429312706,
      "learning_rate": 0.2135575958058649,
      "loss": 0.7009,
      "num_input_tokens_seen": 19836256,
      "step": 14430
    },
    {
      "epoch": 0.46203828180014084,
      "grad_norm": 0.004003439098596573,
      "learning_rate": 0.2135042352073039,
      "loss": 0.4844,
      "num_input_tokens_seen": 19843040,
      "step": 14435
    },
    {
      "epoch": 0.4621983227706293,
      "grad_norm": 0.005035282112658024,
      "learning_rate": 0.2134508648155918,
      "loss": 0.3422,
      "num_input_tokens_seen": 19849856,
      "step": 14440
    },
    {
      "epoch": 0.4623583637411177,
      "grad_norm": 0.006228466518223286,
      "learning_rate": 0.213397484638959,
      "loss": 0.5527,
      "num_input_tokens_seen": 19856832,
      "step": 14445
    },
    {
      "epoch": 0.46251840471160616,
      "grad_norm": 0.007140811067074537,
      "learning_rate": 0.21334409468563728,
      "loss": 0.6068,
      "num_input_tokens_seen": 19863568,
      "step": 14450
    },
    {
      "epoch": 0.4626784456820946,
      "grad_norm": 0.005498411599546671,
      "learning_rate": 0.2132906949638602,
      "loss": 0.4358,
      "num_input_tokens_seen": 19870448,
      "step": 14455
    },
    {
      "epoch": 0.46283848665258304,
      "grad_norm": 0.007546114735305309,
      "learning_rate": 0.21323728548186255,
      "loss": 0.5263,
      "num_input_tokens_seen": 19877392,
      "step": 14460
    },
    {
      "epoch": 0.46299852762307153,
      "grad_norm": 0.004478450398892164,
      "learning_rate": 0.21318386624788088,
      "loss": 0.5745,
      "num_input_tokens_seen": 19884176,
      "step": 14465
    },
    {
      "epoch": 0.46315856859355997,
      "grad_norm": 0.0047609200701117516,
      "learning_rate": 0.21313043727015288,
      "loss": 0.448,
      "num_input_tokens_seen": 19890896,
      "step": 14470
    },
    {
      "epoch": 0.4633186095640484,
      "grad_norm": 0.0054449718445539474,
      "learning_rate": 0.2130769985569182,
      "loss": 0.7576,
      "num_input_tokens_seen": 19897536,
      "step": 14475
    },
    {
      "epoch": 0.46347865053453685,
      "grad_norm": 0.0036959508433938026,
      "learning_rate": 0.21302355011641766,
      "loss": 0.703,
      "num_input_tokens_seen": 19904144,
      "step": 14480
    },
    {
      "epoch": 0.4636386915050253,
      "grad_norm": 0.00414284085854888,
      "learning_rate": 0.21297009195689365,
      "loss": 0.6456,
      "num_input_tokens_seen": 19911040,
      "step": 14485
    },
    {
      "epoch": 0.46379873247551373,
      "grad_norm": 0.0043324255384504795,
      "learning_rate": 0.21291662408659015,
      "loss": 0.5857,
      "num_input_tokens_seen": 19917936,
      "step": 14490
    },
    {
      "epoch": 0.46395877344600217,
      "grad_norm": 0.0026786490343511105,
      "learning_rate": 0.21286314651375254,
      "loss": 0.5239,
      "num_input_tokens_seen": 19924624,
      "step": 14495
    },
    {
      "epoch": 0.4641188144164906,
      "grad_norm": 0.0033322018571197987,
      "learning_rate": 0.2128096592466278,
      "loss": 0.56,
      "num_input_tokens_seen": 19931216,
      "step": 14500
    },
    {
      "epoch": 0.46427885538697905,
      "grad_norm": 0.002577820559963584,
      "learning_rate": 0.21275616229346428,
      "loss": 0.4234,
      "num_input_tokens_seen": 19938064,
      "step": 14505
    },
    {
      "epoch": 0.4644388963574675,
      "grad_norm": 0.004677499644458294,
      "learning_rate": 0.21270265566251184,
      "loss": 0.5406,
      "num_input_tokens_seen": 19944864,
      "step": 14510
    },
    {
      "epoch": 0.46459893732795593,
      "grad_norm": 0.005318955052644014,
      "learning_rate": 0.21264913936202193,
      "loss": 0.5977,
      "num_input_tokens_seen": 19952160,
      "step": 14515
    },
    {
      "epoch": 0.4647589782984444,
      "grad_norm": 0.00972563587129116,
      "learning_rate": 0.2125956134002475,
      "loss": 0.5149,
      "num_input_tokens_seen": 19958896,
      "step": 14520
    },
    {
      "epoch": 0.46491901926893286,
      "grad_norm": 0.004793182015419006,
      "learning_rate": 0.2125420777854428,
      "loss": 0.4646,
      "num_input_tokens_seen": 19965936,
      "step": 14525
    },
    {
      "epoch": 0.4650790602394213,
      "grad_norm": 0.005402520764619112,
      "learning_rate": 0.21248853252586372,
      "loss": 0.6142,
      "num_input_tokens_seen": 19972896,
      "step": 14530
    },
    {
      "epoch": 0.46523910120990974,
      "grad_norm": 0.008001215755939484,
      "learning_rate": 0.21243497762976774,
      "loss": 0.4914,
      "num_input_tokens_seen": 19980128,
      "step": 14535
    },
    {
      "epoch": 0.4653991421803982,
      "grad_norm": 0.0031805660109966993,
      "learning_rate": 0.21238141310541356,
      "loss": 0.5288,
      "num_input_tokens_seen": 19987024,
      "step": 14540
    },
    {
      "epoch": 0.4655591831508866,
      "grad_norm": 0.006686830893158913,
      "learning_rate": 0.21232783896106153,
      "loss": 0.4949,
      "num_input_tokens_seen": 19993872,
      "step": 14545
    },
    {
      "epoch": 0.46571922412137506,
      "grad_norm": 0.007526444736868143,
      "learning_rate": 0.21227425520497345,
      "loss": 0.4781,
      "num_input_tokens_seen": 20000800,
      "step": 14550
    },
    {
      "epoch": 0.4658792650918635,
      "grad_norm": 0.008107293397188187,
      "learning_rate": 0.2122206618454127,
      "loss": 0.9367,
      "num_input_tokens_seen": 20011056,
      "step": 14555
    },
    {
      "epoch": 0.46603930606235194,
      "grad_norm": 0.005579979158937931,
      "learning_rate": 0.2121670588906439,
      "loss": 0.6746,
      "num_input_tokens_seen": 20018096,
      "step": 14560
    },
    {
      "epoch": 0.4661993470328404,
      "grad_norm": 0.008178089745342731,
      "learning_rate": 0.21211344634893345,
      "loss": 0.4259,
      "num_input_tokens_seen": 20024880,
      "step": 14565
    },
    {
      "epoch": 0.4663593880033289,
      "grad_norm": 0.004046270623803139,
      "learning_rate": 0.21205982422854897,
      "loss": 0.6486,
      "num_input_tokens_seen": 20031696,
      "step": 14570
    },
    {
      "epoch": 0.4665194289738173,
      "grad_norm": 0.004105040803551674,
      "learning_rate": 0.21200619253775974,
      "loss": 0.5154,
      "num_input_tokens_seen": 20038576,
      "step": 14575
    },
    {
      "epoch": 0.46667946994430576,
      "grad_norm": 0.003520325990393758,
      "learning_rate": 0.21195255128483637,
      "loss": 0.5494,
      "num_input_tokens_seen": 20045376,
      "step": 14580
    },
    {
      "epoch": 0.4668395109147942,
      "grad_norm": 0.00483419606462121,
      "learning_rate": 0.21189890047805102,
      "loss": 0.6533,
      "num_input_tokens_seen": 20052544,
      "step": 14585
    },
    {
      "epoch": 0.46699955188528264,
      "grad_norm": 0.006975597236305475,
      "learning_rate": 0.21184524012567735,
      "loss": 0.5514,
      "num_input_tokens_seen": 20059344,
      "step": 14590
    },
    {
      "epoch": 0.4671595928557711,
      "grad_norm": 0.00666079530492425,
      "learning_rate": 0.2117915702359905,
      "loss": 0.5463,
      "num_input_tokens_seen": 20066384,
      "step": 14595
    },
    {
      "epoch": 0.4673196338262595,
      "grad_norm": 0.005959527567028999,
      "learning_rate": 0.211737890817267,
      "loss": 0.5055,
      "num_input_tokens_seen": 20073168,
      "step": 14600
    },
    {
      "epoch": 0.4673196338262595,
      "eval_loss": 0.5889204144477844,
      "eval_runtime": 332.2804,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 20073168,
      "step": 14600
    },
    {
      "epoch": 0.46747967479674796,
      "grad_norm": 0.006988340988755226,
      "learning_rate": 0.21168420187778483,
      "loss": 0.6934,
      "num_input_tokens_seen": 20080384,
      "step": 14605
    },
    {
      "epoch": 0.4676397157672364,
      "grad_norm": 0.020100941881537437,
      "learning_rate": 0.21163050342582362,
      "loss": 0.689,
      "num_input_tokens_seen": 20087568,
      "step": 14610
    },
    {
      "epoch": 0.46779975673772484,
      "grad_norm": 0.0072792102582752705,
      "learning_rate": 0.21157679546966426,
      "loss": 0.6611,
      "num_input_tokens_seen": 20094576,
      "step": 14615
    },
    {
      "epoch": 0.4679597977082133,
      "grad_norm": 0.004811733029782772,
      "learning_rate": 0.2115230780175892,
      "loss": 0.4826,
      "num_input_tokens_seen": 20101408,
      "step": 14620
    },
    {
      "epoch": 0.46811983867870177,
      "grad_norm": 0.004702705889940262,
      "learning_rate": 0.21146935107788237,
      "loss": 0.4992,
      "num_input_tokens_seen": 20108080,
      "step": 14625
    },
    {
      "epoch": 0.4682798796491902,
      "grad_norm": 0.003184698289260268,
      "learning_rate": 0.21141561465882916,
      "loss": 0.4913,
      "num_input_tokens_seen": 20114736,
      "step": 14630
    },
    {
      "epoch": 0.46843992061967865,
      "grad_norm": 0.006130550988018513,
      "learning_rate": 0.21136186876871635,
      "loss": 0.6182,
      "num_input_tokens_seen": 20121760,
      "step": 14635
    },
    {
      "epoch": 0.4685999615901671,
      "grad_norm": 0.013567744754254818,
      "learning_rate": 0.21130811341583225,
      "loss": 0.5946,
      "num_input_tokens_seen": 20128384,
      "step": 14640
    },
    {
      "epoch": 0.46876000256065553,
      "grad_norm": 0.005881201010197401,
      "learning_rate": 0.21125434860846667,
      "loss": 0.4741,
      "num_input_tokens_seen": 20135616,
      "step": 14645
    },
    {
      "epoch": 0.46892004353114397,
      "grad_norm": 0.004227129742503166,
      "learning_rate": 0.2112005743549107,
      "loss": 0.5274,
      "num_input_tokens_seen": 20142496,
      "step": 14650
    },
    {
      "epoch": 0.4690800845016324,
      "grad_norm": 0.004838315770030022,
      "learning_rate": 0.21114679066345707,
      "loss": 0.4265,
      "num_input_tokens_seen": 20149552,
      "step": 14655
    },
    {
      "epoch": 0.46924012547212085,
      "grad_norm": 0.007175454404205084,
      "learning_rate": 0.21109299754239993,
      "loss": 0.6589,
      "num_input_tokens_seen": 20156768,
      "step": 14660
    },
    {
      "epoch": 0.4694001664426093,
      "grad_norm": 0.0057209571823477745,
      "learning_rate": 0.21103919500003482,
      "loss": 0.4415,
      "num_input_tokens_seen": 20163440,
      "step": 14665
    },
    {
      "epoch": 0.46956020741309773,
      "grad_norm": 0.003825499676167965,
      "learning_rate": 0.21098538304465872,
      "loss": 0.5091,
      "num_input_tokens_seen": 20170336,
      "step": 14670
    },
    {
      "epoch": 0.4697202483835862,
      "grad_norm": 0.0047051808796823025,
      "learning_rate": 0.2109315616845702,
      "loss": 0.6001,
      "num_input_tokens_seen": 20177072,
      "step": 14675
    },
    {
      "epoch": 0.46988028935407467,
      "grad_norm": 0.004443014040589333,
      "learning_rate": 0.21087773092806925,
      "loss": 0.4201,
      "num_input_tokens_seen": 20183872,
      "step": 14680
    },
    {
      "epoch": 0.4700403303245631,
      "grad_norm": 0.0075015719048678875,
      "learning_rate": 0.21082389078345704,
      "loss": 0.577,
      "num_input_tokens_seen": 20191104,
      "step": 14685
    },
    {
      "epoch": 0.47020037129505154,
      "grad_norm": 0.0048640016466379166,
      "learning_rate": 0.2107700412590365,
      "loss": 0.5472,
      "num_input_tokens_seen": 20197776,
      "step": 14690
    },
    {
      "epoch": 0.47036041226554,
      "grad_norm": 0.004711032845079899,
      "learning_rate": 0.210716182363112,
      "loss": 0.5001,
      "num_input_tokens_seen": 20204416,
      "step": 14695
    },
    {
      "epoch": 0.4705204532360284,
      "grad_norm": 0.0056556882336735725,
      "learning_rate": 0.2106623141039891,
      "loss": 0.4498,
      "num_input_tokens_seen": 20211520,
      "step": 14700
    },
    {
      "epoch": 0.47068049420651686,
      "grad_norm": 0.005736129358410835,
      "learning_rate": 0.21060843648997507,
      "loss": 0.554,
      "num_input_tokens_seen": 20218864,
      "step": 14705
    },
    {
      "epoch": 0.4708405351770053,
      "grad_norm": 0.004651783034205437,
      "learning_rate": 0.21055454952937844,
      "loss": 0.6775,
      "num_input_tokens_seen": 20225680,
      "step": 14710
    },
    {
      "epoch": 0.47100057614749374,
      "grad_norm": 0.0047839367762207985,
      "learning_rate": 0.21050065323050937,
      "loss": 0.5139,
      "num_input_tokens_seen": 20232144,
      "step": 14715
    },
    {
      "epoch": 0.4711606171179822,
      "grad_norm": 0.0029994207434356213,
      "learning_rate": 0.21044674760167928,
      "loss": 0.5005,
      "num_input_tokens_seen": 20238928,
      "step": 14720
    },
    {
      "epoch": 0.4713206580884706,
      "grad_norm": 0.00748223764821887,
      "learning_rate": 0.210392832651201,
      "loss": 0.4477,
      "num_input_tokens_seen": 20245840,
      "step": 14725
    },
    {
      "epoch": 0.4714806990589591,
      "grad_norm": 0.008709092624485493,
      "learning_rate": 0.210338908387389,
      "loss": 0.5484,
      "num_input_tokens_seen": 20252768,
      "step": 14730
    },
    {
      "epoch": 0.47164074002944756,
      "grad_norm": 0.00593154551461339,
      "learning_rate": 0.21028497481855912,
      "loss": 0.7601,
      "num_input_tokens_seen": 20259984,
      "step": 14735
    },
    {
      "epoch": 0.471800780999936,
      "grad_norm": 0.008359060622751713,
      "learning_rate": 0.21023103195302847,
      "loss": 0.6508,
      "num_input_tokens_seen": 20266816,
      "step": 14740
    },
    {
      "epoch": 0.47196082197042444,
      "grad_norm": 0.00796553399413824,
      "learning_rate": 0.21017707979911582,
      "loss": 0.5884,
      "num_input_tokens_seen": 20273584,
      "step": 14745
    },
    {
      "epoch": 0.4721208629409129,
      "grad_norm": 0.0043874867260456085,
      "learning_rate": 0.21012311836514122,
      "loss": 0.4851,
      "num_input_tokens_seen": 20280464,
      "step": 14750
    },
    {
      "epoch": 0.4722809039114013,
      "grad_norm": 0.00419149873778224,
      "learning_rate": 0.21006914765942622,
      "loss": 0.4271,
      "num_input_tokens_seen": 20287488,
      "step": 14755
    },
    {
      "epoch": 0.47244094488188976,
      "grad_norm": 0.0042138900607824326,
      "learning_rate": 0.2100151676902938,
      "loss": 0.5925,
      "num_input_tokens_seen": 20294624,
      "step": 14760
    },
    {
      "epoch": 0.4726009858523782,
      "grad_norm": 0.00521338265389204,
      "learning_rate": 0.2099611784660683,
      "loss": 0.5467,
      "num_input_tokens_seen": 20301632,
      "step": 14765
    },
    {
      "epoch": 0.47276102682286664,
      "grad_norm": 0.01143896859139204,
      "learning_rate": 0.20990717999507552,
      "loss": 0.5205,
      "num_input_tokens_seen": 20308432,
      "step": 14770
    },
    {
      "epoch": 0.4729210677933551,
      "grad_norm": 0.007045985199511051,
      "learning_rate": 0.20985317228564276,
      "loss": 0.5292,
      "num_input_tokens_seen": 20315248,
      "step": 14775
    },
    {
      "epoch": 0.4730811087638435,
      "grad_norm": 0.008893241174519062,
      "learning_rate": 0.20979915534609872,
      "loss": 0.6739,
      "num_input_tokens_seen": 20322096,
      "step": 14780
    },
    {
      "epoch": 0.473241149734332,
      "grad_norm": 0.0049773394130170345,
      "learning_rate": 0.20974512918477342,
      "loss": 0.4888,
      "num_input_tokens_seen": 20328672,
      "step": 14785
    },
    {
      "epoch": 0.47340119070482045,
      "grad_norm": 0.007249334827065468,
      "learning_rate": 0.2096910938099984,
      "loss": 0.7166,
      "num_input_tokens_seen": 20335296,
      "step": 14790
    },
    {
      "epoch": 0.4735612316753089,
      "grad_norm": 0.004006850533187389,
      "learning_rate": 0.2096370492301066,
      "loss": 0.5164,
      "num_input_tokens_seen": 20341840,
      "step": 14795
    },
    {
      "epoch": 0.47372127264579733,
      "grad_norm": 0.0035423433873802423,
      "learning_rate": 0.2095829954534323,
      "loss": 0.399,
      "num_input_tokens_seen": 20349056,
      "step": 14800
    },
    {
      "epoch": 0.47372127264579733,
      "eval_loss": 0.575423002243042,
      "eval_runtime": 331.4614,
      "eval_samples_per_second": 41.893,
      "eval_steps_per_second": 20.947,
      "num_input_tokens_seen": 20349056,
      "step": 14800
    },
    {
      "epoch": 0.47388131361628577,
      "grad_norm": 0.00426687765866518,
      "learning_rate": 0.2095289324883114,
      "loss": 0.5037,
      "num_input_tokens_seen": 20355936,
      "step": 14805
    },
    {
      "epoch": 0.4740413545867742,
      "grad_norm": 0.0056802877224981785,
      "learning_rate": 0.20947486034308097,
      "loss": 0.4348,
      "num_input_tokens_seen": 20363136,
      "step": 14810
    },
    {
      "epoch": 0.47420139555726265,
      "grad_norm": 0.006532094907015562,
      "learning_rate": 0.2094207790260797,
      "loss": 0.3791,
      "num_input_tokens_seen": 20369856,
      "step": 14815
    },
    {
      "epoch": 0.4743614365277511,
      "grad_norm": 0.010464083403348923,
      "learning_rate": 0.20936668854564758,
      "loss": 0.4645,
      "num_input_tokens_seen": 20376848,
      "step": 14820
    },
    {
      "epoch": 0.47452147749823953,
      "grad_norm": 0.006078263279050589,
      "learning_rate": 0.20931258891012602,
      "loss": 0.5987,
      "num_input_tokens_seen": 20383360,
      "step": 14825
    },
    {
      "epoch": 0.47468151846872797,
      "grad_norm": 0.007304110564291477,
      "learning_rate": 0.20925848012785792,
      "loss": 0.4096,
      "num_input_tokens_seen": 20389920,
      "step": 14830
    },
    {
      "epoch": 0.47484155943921647,
      "grad_norm": 0.0038557674270123243,
      "learning_rate": 0.20920436220718747,
      "loss": 0.5225,
      "num_input_tokens_seen": 20396688,
      "step": 14835
    },
    {
      "epoch": 0.4750016004097049,
      "grad_norm": 0.009120218455791473,
      "learning_rate": 0.20915023515646033,
      "loss": 0.6684,
      "num_input_tokens_seen": 20403520,
      "step": 14840
    },
    {
      "epoch": 0.47516164138019334,
      "grad_norm": 0.006175645161420107,
      "learning_rate": 0.20909609898402368,
      "loss": 0.4121,
      "num_input_tokens_seen": 20410464,
      "step": 14845
    },
    {
      "epoch": 0.4753216823506818,
      "grad_norm": 0.008922236040234566,
      "learning_rate": 0.2090419536982258,
      "loss": 0.5474,
      "num_input_tokens_seen": 20417344,
      "step": 14850
    },
    {
      "epoch": 0.4754817233211702,
      "grad_norm": 0.004771701991558075,
      "learning_rate": 0.2089877993074168,
      "loss": 0.5032,
      "num_input_tokens_seen": 20424368,
      "step": 14855
    },
    {
      "epoch": 0.47564176429165866,
      "grad_norm": 0.004325476475059986,
      "learning_rate": 0.20893363581994784,
      "loss": 0.5676,
      "num_input_tokens_seen": 20431472,
      "step": 14860
    },
    {
      "epoch": 0.4758018052621471,
      "grad_norm": 0.0027397372759878635,
      "learning_rate": 0.2088794632441716,
      "loss": 0.3897,
      "num_input_tokens_seen": 20438176,
      "step": 14865
    },
    {
      "epoch": 0.47596184623263554,
      "grad_norm": 0.0033648035023361444,
      "learning_rate": 0.20882528158844219,
      "loss": 0.3059,
      "num_input_tokens_seen": 20444912,
      "step": 14870
    },
    {
      "epoch": 0.476121887203124,
      "grad_norm": 0.006498902104794979,
      "learning_rate": 0.20877109086111514,
      "loss": 0.5263,
      "num_input_tokens_seen": 20451824,
      "step": 14875
    },
    {
      "epoch": 0.4762819281736124,
      "grad_norm": 0.008187290281057358,
      "learning_rate": 0.2087168910705473,
      "loss": 0.7089,
      "num_input_tokens_seen": 20458736,
      "step": 14880
    },
    {
      "epoch": 0.47644196914410086,
      "grad_norm": 0.010700990445911884,
      "learning_rate": 0.208662682225097,
      "loss": 0.686,
      "num_input_tokens_seen": 20465760,
      "step": 14885
    },
    {
      "epoch": 0.47660201011458936,
      "grad_norm": 0.0043148803524672985,
      "learning_rate": 0.2086084643331239,
      "loss": 0.4634,
      "num_input_tokens_seen": 20472624,
      "step": 14890
    },
    {
      "epoch": 0.4767620510850778,
      "grad_norm": 0.012036063708364964,
      "learning_rate": 0.20855423740298906,
      "loss": 0.52,
      "num_input_tokens_seen": 20479184,
      "step": 14895
    },
    {
      "epoch": 0.47692209205556624,
      "grad_norm": 0.006981853861361742,
      "learning_rate": 0.208500001443055,
      "loss": 0.588,
      "num_input_tokens_seen": 20485680,
      "step": 14900
    },
    {
      "epoch": 0.4770821330260547,
      "grad_norm": 0.003911551553755999,
      "learning_rate": 0.20844575646168553,
      "loss": 0.531,
      "num_input_tokens_seen": 20492432,
      "step": 14905
    },
    {
      "epoch": 0.4772421739965431,
      "grad_norm": 0.005545689258724451,
      "learning_rate": 0.20839150246724594,
      "loss": 0.6714,
      "num_input_tokens_seen": 20499296,
      "step": 14910
    },
    {
      "epoch": 0.47740221496703156,
      "grad_norm": 0.004944461397826672,
      "learning_rate": 0.20833723946810287,
      "loss": 0.5522,
      "num_input_tokens_seen": 20505888,
      "step": 14915
    },
    {
      "epoch": 0.47756225593752,
      "grad_norm": 0.008621599525213242,
      "learning_rate": 0.20828296747262437,
      "loss": 0.6262,
      "num_input_tokens_seen": 20512768,
      "step": 14920
    },
    {
      "epoch": 0.47772229690800844,
      "grad_norm": 0.005445260554552078,
      "learning_rate": 0.20822868648917986,
      "loss": 0.5572,
      "num_input_tokens_seen": 20519808,
      "step": 14925
    },
    {
      "epoch": 0.4778823378784969,
      "grad_norm": 0.0823613628745079,
      "learning_rate": 0.20817439652614017,
      "loss": 0.6231,
      "num_input_tokens_seen": 20526880,
      "step": 14930
    },
    {
      "epoch": 0.4780423788489853,
      "grad_norm": 0.004902184475213289,
      "learning_rate": 0.20812009759187744,
      "loss": 0.5559,
      "num_input_tokens_seen": 20533408,
      "step": 14935
    },
    {
      "epoch": 0.4782024198194738,
      "grad_norm": 0.005068082828074694,
      "learning_rate": 0.2080657896947653,
      "loss": 0.6396,
      "num_input_tokens_seen": 20540528,
      "step": 14940
    },
    {
      "epoch": 0.47836246078996225,
      "grad_norm": 0.005259400233626366,
      "learning_rate": 0.2080114728431787,
      "loss": 0.5669,
      "num_input_tokens_seen": 20547760,
      "step": 14945
    },
    {
      "epoch": 0.4785225017604507,
      "grad_norm": 0.007779786363244057,
      "learning_rate": 0.20795714704549392,
      "loss": 0.3914,
      "num_input_tokens_seen": 20554288,
      "step": 14950
    },
    {
      "epoch": 0.47868254273093913,
      "grad_norm": 0.003416555467993021,
      "learning_rate": 0.20790281231008875,
      "loss": 0.6025,
      "num_input_tokens_seen": 20561392,
      "step": 14955
    },
    {
      "epoch": 0.47884258370142757,
      "grad_norm": 0.006852405145764351,
      "learning_rate": 0.20784846864534226,
      "loss": 0.5861,
      "num_input_tokens_seen": 20568160,
      "step": 14960
    },
    {
      "epoch": 0.479002624671916,
      "grad_norm": 0.008471822366118431,
      "learning_rate": 0.20779411605963496,
      "loss": 0.5938,
      "num_input_tokens_seen": 20574912,
      "step": 14965
    },
    {
      "epoch": 0.47916266564240445,
      "grad_norm": 0.0030823294073343277,
      "learning_rate": 0.2077397545613487,
      "loss": 0.4481,
      "num_input_tokens_seen": 20581680,
      "step": 14970
    },
    {
      "epoch": 0.4793227066128929,
      "grad_norm": 0.003209741087630391,
      "learning_rate": 0.20768538415886661,
      "loss": 0.4816,
      "num_input_tokens_seen": 20588464,
      "step": 14975
    },
    {
      "epoch": 0.47948274758338133,
      "grad_norm": 0.006273317616432905,
      "learning_rate": 0.20763100486057343,
      "loss": 0.5375,
      "num_input_tokens_seen": 20595328,
      "step": 14980
    },
    {
      "epoch": 0.47964278855386977,
      "grad_norm": 0.005054036621004343,
      "learning_rate": 0.20757661667485502,
      "loss": 0.5803,
      "num_input_tokens_seen": 20602128,
      "step": 14985
    },
    {
      "epoch": 0.4798028295243582,
      "grad_norm": 0.007163654081523418,
      "learning_rate": 0.2075222196100988,
      "loss": 0.5736,
      "num_input_tokens_seen": 20609136,
      "step": 14990
    },
    {
      "epoch": 0.4799628704948467,
      "grad_norm": 0.006947923451662064,
      "learning_rate": 0.20746781367469344,
      "loss": 0.4177,
      "num_input_tokens_seen": 20615856,
      "step": 14995
    },
    {
      "epoch": 0.48012291146533514,
      "grad_norm": 0.008447720669209957,
      "learning_rate": 0.207413398877029,
      "loss": 0.5499,
      "num_input_tokens_seen": 20622896,
      "step": 15000
    },
    {
      "epoch": 0.48012291146533514,
      "eval_loss": 0.582549512386322,
      "eval_runtime": 332.2674,
      "eval_samples_per_second": 41.792,
      "eval_steps_per_second": 20.896,
      "num_input_tokens_seen": 20622896,
      "step": 15000
    },
    {
      "epoch": 0.4802829524358236,
      "grad_norm": 0.0022716186940670013,
      "learning_rate": 0.20735897522549698,
      "loss": 0.6528,
      "num_input_tokens_seen": 20629872,
      "step": 15005
    },
    {
      "epoch": 0.480442993406312,
      "grad_norm": 0.0028024024795740843,
      "learning_rate": 0.2073045427284902,
      "loss": 0.3973,
      "num_input_tokens_seen": 20636336,
      "step": 15010
    },
    {
      "epoch": 0.48060303437680046,
      "grad_norm": 0.006032670382410288,
      "learning_rate": 0.2072501013944027,
      "loss": 0.7708,
      "num_input_tokens_seen": 20642640,
      "step": 15015
    },
    {
      "epoch": 0.4807630753472889,
      "grad_norm": 0.006308689247816801,
      "learning_rate": 0.20719565123163017,
      "loss": 0.5396,
      "num_input_tokens_seen": 20649824,
      "step": 15020
    },
    {
      "epoch": 0.48092311631777734,
      "grad_norm": 0.005590109154582024,
      "learning_rate": 0.20714119224856944,
      "loss": 0.7767,
      "num_input_tokens_seen": 20656528,
      "step": 15025
    },
    {
      "epoch": 0.4810831572882658,
      "grad_norm": 0.007603847421705723,
      "learning_rate": 0.2070867244536188,
      "loss": 0.6193,
      "num_input_tokens_seen": 20663280,
      "step": 15030
    },
    {
      "epoch": 0.4812431982587542,
      "grad_norm": 0.0032826305832713842,
      "learning_rate": 0.20703224785517785,
      "loss": 0.5345,
      "num_input_tokens_seen": 20670496,
      "step": 15035
    },
    {
      "epoch": 0.48140323922924266,
      "grad_norm": 0.007361996453255415,
      "learning_rate": 0.20697776246164754,
      "loss": 0.6315,
      "num_input_tokens_seen": 20677120,
      "step": 15040
    },
    {
      "epoch": 0.48156328019973116,
      "grad_norm": 0.004937154706567526,
      "learning_rate": 0.2069232682814303,
      "loss": 0.4527,
      "num_input_tokens_seen": 20683968,
      "step": 15045
    },
    {
      "epoch": 0.4817233211702196,
      "grad_norm": 0.0033493107184767723,
      "learning_rate": 0.20686876532292972,
      "loss": 0.575,
      "num_input_tokens_seen": 20690672,
      "step": 15050
    },
    {
      "epoch": 0.48188336214070804,
      "grad_norm": 0.005989560391753912,
      "learning_rate": 0.20681425359455083,
      "loss": 0.5393,
      "num_input_tokens_seen": 20697088,
      "step": 15055
    },
    {
      "epoch": 0.4820434031111965,
      "grad_norm": 0.010012853890657425,
      "learning_rate": 0.20675973310470008,
      "loss": 0.5424,
      "num_input_tokens_seen": 20704272,
      "step": 15060
    },
    {
      "epoch": 0.4822034440816849,
      "grad_norm": 0.004580677952617407,
      "learning_rate": 0.2067052038617852,
      "loss": 0.7512,
      "num_input_tokens_seen": 20711072,
      "step": 15065
    },
    {
      "epoch": 0.48236348505217336,
      "grad_norm": 0.006009580101817846,
      "learning_rate": 0.2066506658742153,
      "loss": 0.5956,
      "num_input_tokens_seen": 20717872,
      "step": 15070
    },
    {
      "epoch": 0.4825235260226618,
      "grad_norm": 0.007629691623151302,
      "learning_rate": 0.20659611915040077,
      "loss": 0.7494,
      "num_input_tokens_seen": 20724624,
      "step": 15075
    },
    {
      "epoch": 0.48268356699315024,
      "grad_norm": 0.0035890957806259394,
      "learning_rate": 0.20654156369875348,
      "loss": 0.466,
      "num_input_tokens_seen": 20731392,
      "step": 15080
    },
    {
      "epoch": 0.4828436079636387,
      "grad_norm": 0.011104273609817028,
      "learning_rate": 0.20648699952768648,
      "loss": 0.62,
      "num_input_tokens_seen": 20738416,
      "step": 15085
    },
    {
      "epoch": 0.4830036489341271,
      "grad_norm": 0.007934549823403358,
      "learning_rate": 0.20643242664561437,
      "loss": 0.5777,
      "num_input_tokens_seen": 20745216,
      "step": 15090
    },
    {
      "epoch": 0.48316368990461556,
      "grad_norm": 0.00638198759406805,
      "learning_rate": 0.20637784506095277,
      "loss": 0.5467,
      "num_input_tokens_seen": 20752144,
      "step": 15095
    },
    {
      "epoch": 0.48332373087510405,
      "grad_norm": 0.007279231678694487,
      "learning_rate": 0.20632325478211908,
      "loss": 0.538,
      "num_input_tokens_seen": 20758944,
      "step": 15100
    },
    {
      "epoch": 0.4834837718455925,
      "grad_norm": 0.015693821012973785,
      "learning_rate": 0.20626865581753165,
      "loss": 0.5619,
      "num_input_tokens_seen": 20766000,
      "step": 15105
    },
    {
      "epoch": 0.48364381281608093,
      "grad_norm": 0.004016423597931862,
      "learning_rate": 0.2062140481756104,
      "loss": 0.5169,
      "num_input_tokens_seen": 20772688,
      "step": 15110
    },
    {
      "epoch": 0.48380385378656937,
      "grad_norm": 0.004319034982472658,
      "learning_rate": 0.20615943186477648,
      "loss": 0.5415,
      "num_input_tokens_seen": 20779264,
      "step": 15115
    },
    {
      "epoch": 0.4839638947570578,
      "grad_norm": 0.004730228800326586,
      "learning_rate": 0.20610480689345242,
      "loss": 0.5488,
      "num_input_tokens_seen": 20786144,
      "step": 15120
    },
    {
      "epoch": 0.48412393572754625,
      "grad_norm": 0.003693267470225692,
      "learning_rate": 0.2060501732700621,
      "loss": 0.532,
      "num_input_tokens_seen": 20793008,
      "step": 15125
    },
    {
      "epoch": 0.4842839766980347,
      "grad_norm": 0.007782884873449802,
      "learning_rate": 0.20599553100303067,
      "loss": 0.572,
      "num_input_tokens_seen": 20800144,
      "step": 15130
    },
    {
      "epoch": 0.48444401766852313,
      "grad_norm": 0.006590146571397781,
      "learning_rate": 0.20594088010078465,
      "loss": 0.4525,
      "num_input_tokens_seen": 20807056,
      "step": 15135
    },
    {
      "epoch": 0.48460405863901157,
      "grad_norm": 0.006419599521905184,
      "learning_rate": 0.20588622057175196,
      "loss": 0.7716,
      "num_input_tokens_seen": 20814192,
      "step": 15140
    },
    {
      "epoch": 0.4847640996095,
      "grad_norm": 0.0023425656836479902,
      "learning_rate": 0.20583155242436177,
      "loss": 0.6559,
      "num_input_tokens_seen": 20821248,
      "step": 15145
    },
    {
      "epoch": 0.48492414057998845,
      "grad_norm": 0.005508411675691605,
      "learning_rate": 0.20577687566704453,
      "loss": 0.4662,
      "num_input_tokens_seen": 20828496,
      "step": 15150
    },
    {
      "epoch": 0.48508418155047694,
      "grad_norm": 0.0027004098519682884,
      "learning_rate": 0.20572219030823213,
      "loss": 0.7412,
      "num_input_tokens_seen": 20835616,
      "step": 15155
    },
    {
      "epoch": 0.4852442225209654,
      "grad_norm": 0.002839343622326851,
      "learning_rate": 0.20566749635635775,
      "loss": 0.3992,
      "num_input_tokens_seen": 20842560,
      "step": 15160
    },
    {
      "epoch": 0.4854042634914538,
      "grad_norm": 0.004244106821715832,
      "learning_rate": 0.20561279381985587,
      "loss": 0.5362,
      "num_input_tokens_seen": 20849680,
      "step": 15165
    },
    {
      "epoch": 0.48556430446194226,
      "grad_norm": 0.004771473817527294,
      "learning_rate": 0.2055580827071623,
      "loss": 0.4956,
      "num_input_tokens_seen": 20856432,
      "step": 15170
    },
    {
      "epoch": 0.4857243454324307,
      "grad_norm": 0.005421137437224388,
      "learning_rate": 0.20550336302671418,
      "loss": 0.6592,
      "num_input_tokens_seen": 20863296,
      "step": 15175
    },
    {
      "epoch": 0.48588438640291914,
      "grad_norm": 0.0036490699276328087,
      "learning_rate": 0.20544863478695,
      "loss": 0.5491,
      "num_input_tokens_seen": 20870320,
      "step": 15180
    },
    {
      "epoch": 0.4860444273734076,
      "grad_norm": 0.004966201726347208,
      "learning_rate": 0.20539389799630953,
      "loss": 0.7559,
      "num_input_tokens_seen": 20876960,
      "step": 15185
    },
    {
      "epoch": 0.486204468343896,
      "grad_norm": 0.003721121232956648,
      "learning_rate": 0.20533915266323388,
      "loss": 0.5626,
      "num_input_tokens_seen": 20883440,
      "step": 15190
    },
    {
      "epoch": 0.48636450931438446,
      "grad_norm": 0.004189629573374987,
      "learning_rate": 0.20528439879616542,
      "loss": 0.4644,
      "num_input_tokens_seen": 20890096,
      "step": 15195
    },
    {
      "epoch": 0.4865245502848729,
      "grad_norm": 0.003711245721206069,
      "learning_rate": 0.20522963640354794,
      "loss": 0.5207,
      "num_input_tokens_seen": 20896768,
      "step": 15200
    },
    {
      "epoch": 0.4865245502848729,
      "eval_loss": 0.570220410823822,
      "eval_runtime": 331.8698,
      "eval_samples_per_second": 41.842,
      "eval_steps_per_second": 20.921,
      "num_input_tokens_seen": 20896768,
      "step": 15200
    },
    {
      "epoch": 0.4866845912553614,
      "grad_norm": 0.0049219573847949505,
      "learning_rate": 0.20517486549382644,
      "loss": 0.5988,
      "num_input_tokens_seen": 20903488,
      "step": 15205
    },
    {
      "epoch": 0.48684463222584984,
      "grad_norm": 0.005223033018410206,
      "learning_rate": 0.20512008607544735,
      "loss": 0.4859,
      "num_input_tokens_seen": 20910608,
      "step": 15210
    },
    {
      "epoch": 0.4870046731963383,
      "grad_norm": 0.00379561516456306,
      "learning_rate": 0.20506529815685826,
      "loss": 0.5107,
      "num_input_tokens_seen": 20917760,
      "step": 15215
    },
    {
      "epoch": 0.4871647141668267,
      "grad_norm": 0.009498360566794872,
      "learning_rate": 0.2050105017465082,
      "loss": 0.6246,
      "num_input_tokens_seen": 20924448,
      "step": 15220
    },
    {
      "epoch": 0.48732475513731516,
      "grad_norm": 0.006226053927093744,
      "learning_rate": 0.20495569685284754,
      "loss": 0.4552,
      "num_input_tokens_seen": 20931200,
      "step": 15225
    },
    {
      "epoch": 0.4874847961078036,
      "grad_norm": 0.0039151436649262905,
      "learning_rate": 0.20490088348432778,
      "loss": 0.6477,
      "num_input_tokens_seen": 20937696,
      "step": 15230
    },
    {
      "epoch": 0.48764483707829204,
      "grad_norm": 0.003307199105620384,
      "learning_rate": 0.2048460616494018,
      "loss": 0.4363,
      "num_input_tokens_seen": 20944560,
      "step": 15235
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.005709854885935783,
      "learning_rate": 0.2047912313565239,
      "loss": 0.5646,
      "num_input_tokens_seen": 20951312,
      "step": 15240
    },
    {
      "epoch": 0.4879649190192689,
      "grad_norm": 0.002580698812380433,
      "learning_rate": 0.20473639261414958,
      "loss": 0.4219,
      "num_input_tokens_seen": 20958112,
      "step": 15245
    },
    {
      "epoch": 0.48812495998975736,
      "grad_norm": 0.005866439547389746,
      "learning_rate": 0.2046815454307357,
      "loss": 0.6415,
      "num_input_tokens_seen": 20964736,
      "step": 15250
    },
    {
      "epoch": 0.4882850009602458,
      "grad_norm": 0.001894688350148499,
      "learning_rate": 0.20462668981474028,
      "loss": 0.4022,
      "num_input_tokens_seen": 20971456,
      "step": 15255
    },
    {
      "epoch": 0.4884450419307343,
      "grad_norm": 0.004157801158726215,
      "learning_rate": 0.20457182577462288,
      "loss": 0.5654,
      "num_input_tokens_seen": 20978288,
      "step": 15260
    },
    {
      "epoch": 0.48860508290122273,
      "grad_norm": 0.003132233629003167,
      "learning_rate": 0.2045169533188441,
      "loss": 0.4242,
      "num_input_tokens_seen": 20984976,
      "step": 15265
    },
    {
      "epoch": 0.48876512387171117,
      "grad_norm": 0.003249040339142084,
      "learning_rate": 0.20446207245586603,
      "loss": 0.4945,
      "num_input_tokens_seen": 20991984,
      "step": 15270
    },
    {
      "epoch": 0.4889251648421996,
      "grad_norm": 0.008135472424328327,
      "learning_rate": 0.20440718319415196,
      "loss": 0.5581,
      "num_input_tokens_seen": 20998832,
      "step": 15275
    },
    {
      "epoch": 0.48908520581268805,
      "grad_norm": 0.016794249415397644,
      "learning_rate": 0.20435228554216653,
      "loss": 0.7805,
      "num_input_tokens_seen": 21005872,
      "step": 15280
    },
    {
      "epoch": 0.4892452467831765,
      "grad_norm": 0.005354084074497223,
      "learning_rate": 0.20429737950837565,
      "loss": 0.5599,
      "num_input_tokens_seen": 21012672,
      "step": 15285
    },
    {
      "epoch": 0.48940528775366493,
      "grad_norm": 0.0037059616297483444,
      "learning_rate": 0.20424246510124647,
      "loss": 0.5003,
      "num_input_tokens_seen": 21019552,
      "step": 15290
    },
    {
      "epoch": 0.48956532872415337,
      "grad_norm": 0.0063588060438632965,
      "learning_rate": 0.20418754232924755,
      "loss": 0.6101,
      "num_input_tokens_seen": 21026224,
      "step": 15295
    },
    {
      "epoch": 0.4897253696946418,
      "grad_norm": 0.007025472354143858,
      "learning_rate": 0.20413261120084863,
      "loss": 0.7194,
      "num_input_tokens_seen": 21033296,
      "step": 15300
    },
    {
      "epoch": 0.48988541066513025,
      "grad_norm": 0.005895054433494806,
      "learning_rate": 0.2040776717245208,
      "loss": 0.664,
      "num_input_tokens_seen": 21040240,
      "step": 15305
    },
    {
      "epoch": 0.49004545163561875,
      "grad_norm": 0.005959526170045137,
      "learning_rate": 0.2040227239087364,
      "loss": 0.6711,
      "num_input_tokens_seen": 21047936,
      "step": 15310
    },
    {
      "epoch": 0.4902054926061072,
      "grad_norm": 0.003685983130708337,
      "learning_rate": 0.20396776776196904,
      "loss": 0.5429,
      "num_input_tokens_seen": 21054400,
      "step": 15315
    },
    {
      "epoch": 0.4903655335765956,
      "grad_norm": 0.0026217475533485413,
      "learning_rate": 0.20391280329269373,
      "loss": 0.5845,
      "num_input_tokens_seen": 21061088,
      "step": 15320
    },
    {
      "epoch": 0.49052557454708406,
      "grad_norm": 0.010933209210634232,
      "learning_rate": 0.20385783050938663,
      "loss": 0.5838,
      "num_input_tokens_seen": 21067488,
      "step": 15325
    },
    {
      "epoch": 0.4906856155175725,
      "grad_norm": 0.0025822699535638094,
      "learning_rate": 0.20380284942052526,
      "loss": 0.4776,
      "num_input_tokens_seen": 21074304,
      "step": 15330
    },
    {
      "epoch": 0.49084565648806094,
      "grad_norm": 0.005749527830630541,
      "learning_rate": 0.2037478600345884,
      "loss": 0.622,
      "num_input_tokens_seen": 21080928,
      "step": 15335
    },
    {
      "epoch": 0.4910056974585494,
      "grad_norm": 0.0049333879724144936,
      "learning_rate": 0.20369286236005604,
      "loss": 0.6335,
      "num_input_tokens_seen": 21087680,
      "step": 15340
    },
    {
      "epoch": 0.4911657384290378,
      "grad_norm": 0.005234061740338802,
      "learning_rate": 0.20363785640540957,
      "loss": 0.5281,
      "num_input_tokens_seen": 21094960,
      "step": 15345
    },
    {
      "epoch": 0.49132577939952626,
      "grad_norm": 0.00321805477142334,
      "learning_rate": 0.2035828421791316,
      "loss": 0.4805,
      "num_input_tokens_seen": 21102192,
      "step": 15350
    },
    {
      "epoch": 0.4914858203700147,
      "grad_norm": 0.003163263201713562,
      "learning_rate": 0.20352781968970599,
      "loss": 0.4089,
      "num_input_tokens_seen": 21109008,
      "step": 15355
    },
    {
      "epoch": 0.49164586134050314,
      "grad_norm": 0.004701048135757446,
      "learning_rate": 0.2034727889456179,
      "loss": 0.6077,
      "num_input_tokens_seen": 21116240,
      "step": 15360
    },
    {
      "epoch": 0.49180590231099164,
      "grad_norm": 0.008369134739041328,
      "learning_rate": 0.2034177499553538,
      "loss": 0.5773,
      "num_input_tokens_seen": 21123104,
      "step": 15365
    },
    {
      "epoch": 0.4919659432814801,
      "grad_norm": 0.005495659075677395,
      "learning_rate": 0.2033627027274014,
      "loss": 0.5544,
      "num_input_tokens_seen": 21130496,
      "step": 15370
    },
    {
      "epoch": 0.4921259842519685,
      "grad_norm": 0.002720730612054467,
      "learning_rate": 0.20330764727024955,
      "loss": 0.6143,
      "num_input_tokens_seen": 21137248,
      "step": 15375
    },
    {
      "epoch": 0.49228602522245696,
      "grad_norm": 0.0038357051089406013,
      "learning_rate": 0.20325258359238868,
      "loss": 0.5381,
      "num_input_tokens_seen": 21144496,
      "step": 15380
    },
    {
      "epoch": 0.4924460661929454,
      "grad_norm": 0.005887430161237717,
      "learning_rate": 0.20319751170231018,
      "loss": 0.6101,
      "num_input_tokens_seen": 21151632,
      "step": 15385
    },
    {
      "epoch": 0.49260610716343384,
      "grad_norm": 0.004518531262874603,
      "learning_rate": 0.2031424316085068,
      "loss": 0.4794,
      "num_input_tokens_seen": 21158224,
      "step": 15390
    },
    {
      "epoch": 0.4927661481339223,
      "grad_norm": 0.002548045013099909,
      "learning_rate": 0.20308734331947265,
      "loss": 0.5475,
      "num_input_tokens_seen": 21164800,
      "step": 15395
    },
    {
      "epoch": 0.4929261891044107,
      "grad_norm": 0.0023941919207572937,
      "learning_rate": 0.20303224684370305,
      "loss": 0.5521,
      "num_input_tokens_seen": 21171376,
      "step": 15400
    },
    {
      "epoch": 0.4929261891044107,
      "eval_loss": 0.5712263584136963,
      "eval_runtime": 331.9562,
      "eval_samples_per_second": 41.831,
      "eval_steps_per_second": 20.915,
      "num_input_tokens_seen": 21171376,
      "step": 15400
    },
    {
      "epoch": 0.49308623007489916,
      "grad_norm": 0.004637675825506449,
      "learning_rate": 0.20297714218969456,
      "loss": 0.4711,
      "num_input_tokens_seen": 21178352,
      "step": 15405
    },
    {
      "epoch": 0.4932462710453876,
      "grad_norm": 0.003514539683237672,
      "learning_rate": 0.20292202936594497,
      "loss": 0.4698,
      "num_input_tokens_seen": 21184912,
      "step": 15410
    },
    {
      "epoch": 0.4934063120158761,
      "grad_norm": 0.002786531811580062,
      "learning_rate": 0.2028669083809534,
      "loss": 0.4278,
      "num_input_tokens_seen": 21192288,
      "step": 15415
    },
    {
      "epoch": 0.49356635298636453,
      "grad_norm": 0.009131284430623055,
      "learning_rate": 0.20281177924322016,
      "loss": 0.7905,
      "num_input_tokens_seen": 21199024,
      "step": 15420
    },
    {
      "epoch": 0.49372639395685297,
      "grad_norm": 0.0036994963884353638,
      "learning_rate": 0.2027566419612469,
      "loss": 0.7174,
      "num_input_tokens_seen": 21206496,
      "step": 15425
    },
    {
      "epoch": 0.4938864349273414,
      "grad_norm": 0.005079392343759537,
      "learning_rate": 0.20270149654353647,
      "loss": 0.7275,
      "num_input_tokens_seen": 21213424,
      "step": 15430
    },
    {
      "epoch": 0.49404647589782985,
      "grad_norm": 0.009169136174023151,
      "learning_rate": 0.202646342998593,
      "loss": 0.5782,
      "num_input_tokens_seen": 21220480,
      "step": 15435
    },
    {
      "epoch": 0.4942065168683183,
      "grad_norm": 0.0035114865750074387,
      "learning_rate": 0.20259118133492185,
      "loss": 0.4638,
      "num_input_tokens_seen": 21226976,
      "step": 15440
    },
    {
      "epoch": 0.49436655783880673,
      "grad_norm": 0.004709342960268259,
      "learning_rate": 0.20253601156102966,
      "loss": 0.5033,
      "num_input_tokens_seen": 21233728,
      "step": 15445
    },
    {
      "epoch": 0.49452659880929517,
      "grad_norm": 0.008600452914834023,
      "learning_rate": 0.20248083368542422,
      "loss": 0.5146,
      "num_input_tokens_seen": 21240256,
      "step": 15450
    },
    {
      "epoch": 0.4946866397797836,
      "grad_norm": 0.006891382392495871,
      "learning_rate": 0.2024256477166147,
      "loss": 0.5528,
      "num_input_tokens_seen": 21246976,
      "step": 15455
    },
    {
      "epoch": 0.49484668075027205,
      "grad_norm": 0.005086105782538652,
      "learning_rate": 0.2023704536631115,
      "loss": 0.5884,
      "num_input_tokens_seen": 21253888,
      "step": 15460
    },
    {
      "epoch": 0.4950067217207605,
      "grad_norm": 0.008263787254691124,
      "learning_rate": 0.20231525153342625,
      "loss": 0.6405,
      "num_input_tokens_seen": 21260448,
      "step": 15465
    },
    {
      "epoch": 0.495166762691249,
      "grad_norm": 0.003931032493710518,
      "learning_rate": 0.20226004133607173,
      "loss": 0.5357,
      "num_input_tokens_seen": 21267216,
      "step": 15470
    },
    {
      "epoch": 0.4953268036617374,
      "grad_norm": 0.0064331842586398125,
      "learning_rate": 0.20220482307956214,
      "loss": 0.4968,
      "num_input_tokens_seen": 21274144,
      "step": 15475
    },
    {
      "epoch": 0.49548684463222586,
      "grad_norm": 0.004449761006981134,
      "learning_rate": 0.20214959677241276,
      "loss": 0.5883,
      "num_input_tokens_seen": 21281136,
      "step": 15480
    },
    {
      "epoch": 0.4956468856027143,
      "grad_norm": 0.0013905519153922796,
      "learning_rate": 0.20209436242314022,
      "loss": 0.3491,
      "num_input_tokens_seen": 21288112,
      "step": 15485
    },
    {
      "epoch": 0.49580692657320274,
      "grad_norm": 0.004088725429028273,
      "learning_rate": 0.2020391200402623,
      "loss": 0.723,
      "num_input_tokens_seen": 21294928,
      "step": 15490
    },
    {
      "epoch": 0.4959669675436912,
      "grad_norm": 0.00438101077452302,
      "learning_rate": 0.2019838696322981,
      "loss": 0.6245,
      "num_input_tokens_seen": 21302224,
      "step": 15495
    },
    {
      "epoch": 0.4961270085141796,
      "grad_norm": 0.0024198240134865046,
      "learning_rate": 0.20192861120776798,
      "loss": 0.3099,
      "num_input_tokens_seen": 21308736,
      "step": 15500
    },
    {
      "epoch": 0.49628704948466806,
      "grad_norm": 0.004779479466378689,
      "learning_rate": 0.20187334477519345,
      "loss": 0.5031,
      "num_input_tokens_seen": 21315088,
      "step": 15505
    },
    {
      "epoch": 0.4964470904551565,
      "grad_norm": 0.006723963189870119,
      "learning_rate": 0.20181807034309726,
      "loss": 0.5954,
      "num_input_tokens_seen": 21321760,
      "step": 15510
    },
    {
      "epoch": 0.49660713142564494,
      "grad_norm": 0.006477005779743195,
      "learning_rate": 0.2017627879200034,
      "loss": 0.5971,
      "num_input_tokens_seen": 21328624,
      "step": 15515
    },
    {
      "epoch": 0.4967671723961334,
      "grad_norm": 0.004489361308515072,
      "learning_rate": 0.2017074975144372,
      "loss": 0.7512,
      "num_input_tokens_seen": 21335936,
      "step": 15520
    },
    {
      "epoch": 0.4969272133666219,
      "grad_norm": 0.004017229657620192,
      "learning_rate": 0.20165219913492508,
      "loss": 0.3887,
      "num_input_tokens_seen": 21343136,
      "step": 15525
    },
    {
      "epoch": 0.4970872543371103,
      "grad_norm": 0.008399337530136108,
      "learning_rate": 0.20159689278999468,
      "loss": 0.6441,
      "num_input_tokens_seen": 21350448,
      "step": 15530
    },
    {
      "epoch": 0.49724729530759876,
      "grad_norm": 0.0032727746292948723,
      "learning_rate": 0.20154157848817508,
      "loss": 0.4252,
      "num_input_tokens_seen": 21357360,
      "step": 15535
    },
    {
      "epoch": 0.4974073362780872,
      "grad_norm": 0.006922820117324591,
      "learning_rate": 0.20148625623799632,
      "loss": 0.6247,
      "num_input_tokens_seen": 21364480,
      "step": 15540
    },
    {
      "epoch": 0.49756737724857564,
      "grad_norm": 0.008283997885882854,
      "learning_rate": 0.20143092604798984,
      "loss": 0.5487,
      "num_input_tokens_seen": 21371328,
      "step": 15545
    },
    {
      "epoch": 0.4977274182190641,
      "grad_norm": 0.003578796749934554,
      "learning_rate": 0.2013755879266883,
      "loss": 0.6485,
      "num_input_tokens_seen": 21378032,
      "step": 15550
    },
    {
      "epoch": 0.4978874591895525,
      "grad_norm": 0.0032395557500422,
      "learning_rate": 0.20132024188262543,
      "loss": 0.7286,
      "num_input_tokens_seen": 21385248,
      "step": 15555
    },
    {
      "epoch": 0.49804750016004096,
      "grad_norm": 0.00716807134449482,
      "learning_rate": 0.2012648879243363,
      "loss": 0.5148,
      "num_input_tokens_seen": 21392336,
      "step": 15560
    },
    {
      "epoch": 0.4982075411305294,
      "grad_norm": 0.0026158473920077085,
      "learning_rate": 0.20120952606035725,
      "loss": 0.4947,
      "num_input_tokens_seen": 21399136,
      "step": 15565
    },
    {
      "epoch": 0.49836758210101784,
      "grad_norm": 0.016596833243966103,
      "learning_rate": 0.20115415629922576,
      "loss": 0.6808,
      "num_input_tokens_seen": 21406192,
      "step": 15570
    },
    {
      "epoch": 0.49852762307150633,
      "grad_norm": 0.004559489898383617,
      "learning_rate": 0.20109877864948048,
      "loss": 0.6779,
      "num_input_tokens_seen": 21412944,
      "step": 15575
    },
    {
      "epoch": 0.49868766404199477,
      "grad_norm": 0.002988215535879135,
      "learning_rate": 0.20104339311966138,
      "loss": 0.5421,
      "num_input_tokens_seen": 21420112,
      "step": 15580
    },
    {
      "epoch": 0.4988477050124832,
      "grad_norm": 0.006083873566240072,
      "learning_rate": 0.2009879997183097,
      "loss": 0.5964,
      "num_input_tokens_seen": 21427120,
      "step": 15585
    },
    {
      "epoch": 0.49900774598297165,
      "grad_norm": 0.003984491806477308,
      "learning_rate": 0.20093259845396763,
      "loss": 0.4959,
      "num_input_tokens_seen": 21433984,
      "step": 15590
    },
    {
      "epoch": 0.4991677869534601,
      "grad_norm": 0.005441318731755018,
      "learning_rate": 0.20087718933517884,
      "loss": 0.6312,
      "num_input_tokens_seen": 21440928,
      "step": 15595
    },
    {
      "epoch": 0.49932782792394853,
      "grad_norm": 0.004353290423750877,
      "learning_rate": 0.20082177237048807,
      "loss": 0.5371,
      "num_input_tokens_seen": 21447568,
      "step": 15600
    },
    {
      "epoch": 0.49932782792394853,
      "eval_loss": 0.5752123594284058,
      "eval_runtime": 331.3239,
      "eval_samples_per_second": 41.911,
      "eval_steps_per_second": 20.955,
      "num_input_tokens_seen": 21447568,
      "step": 15600
    },
    {
      "epoch": 0.49948786889443697,
      "grad_norm": 0.004492474719882011,
      "learning_rate": 0.20076634756844133,
      "loss": 0.5534,
      "num_input_tokens_seen": 21454848,
      "step": 15605
    },
    {
      "epoch": 0.4996479098649254,
      "grad_norm": 0.007012740708887577,
      "learning_rate": 0.20071091493758586,
      "loss": 0.5257,
      "num_input_tokens_seen": 21461888,
      "step": 15610
    },
    {
      "epoch": 0.49980795083541385,
      "grad_norm": 0.007073978893458843,
      "learning_rate": 0.20065547448647003,
      "loss": 0.5974,
      "num_input_tokens_seen": 21468640,
      "step": 15615
    },
    {
      "epoch": 0.4999679918059023,
      "grad_norm": 0.0037077052984386683,
      "learning_rate": 0.20060002622364348,
      "loss": 0.377,
      "num_input_tokens_seen": 21475280,
      "step": 15620
    },
    {
      "epoch": 0.5001280327763907,
      "grad_norm": 0.004421945195645094,
      "learning_rate": 0.20054457015765695,
      "loss": 0.4948,
      "num_input_tokens_seen": 21481920,
      "step": 15625
    },
    {
      "epoch": 0.5002880737468792,
      "grad_norm": 0.018752846866846085,
      "learning_rate": 0.20048910629706254,
      "loss": 0.7795,
      "num_input_tokens_seen": 21489040,
      "step": 15630
    },
    {
      "epoch": 0.5004481147173676,
      "grad_norm": 0.006970058660954237,
      "learning_rate": 0.20043363465041347,
      "loss": 0.4484,
      "num_input_tokens_seen": 21495504,
      "step": 15635
    },
    {
      "epoch": 0.5006081556878561,
      "grad_norm": 0.00636511342599988,
      "learning_rate": 0.2003781552262641,
      "loss": 0.4482,
      "num_input_tokens_seen": 21502240,
      "step": 15640
    },
    {
      "epoch": 0.5007681966583445,
      "grad_norm": 0.01396650355309248,
      "learning_rate": 0.20032266803317014,
      "loss": 0.5592,
      "num_input_tokens_seen": 21509104,
      "step": 15645
    },
    {
      "epoch": 0.500928237628833,
      "grad_norm": 0.013446572236716747,
      "learning_rate": 0.2002671730796884,
      "loss": 0.6455,
      "num_input_tokens_seen": 21515840,
      "step": 15650
    },
    {
      "epoch": 0.5010882785993215,
      "grad_norm": 0.0037986242678016424,
      "learning_rate": 0.20021167037437684,
      "loss": 0.4114,
      "num_input_tokens_seen": 21522576,
      "step": 15655
    },
    {
      "epoch": 0.5012483195698099,
      "grad_norm": 0.0032802505884319544,
      "learning_rate": 0.20015615992579472,
      "loss": 0.5869,
      "num_input_tokens_seen": 21529568,
      "step": 15660
    },
    {
      "epoch": 0.5014083605402984,
      "grad_norm": 0.002594535006210208,
      "learning_rate": 0.20010064174250244,
      "loss": 0.5647,
      "num_input_tokens_seen": 21536592,
      "step": 15665
    },
    {
      "epoch": 0.5015684015107867,
      "grad_norm": 0.005179567262530327,
      "learning_rate": 0.2000451158330616,
      "loss": 0.6967,
      "num_input_tokens_seen": 21543424,
      "step": 15670
    },
    {
      "epoch": 0.5017284424812752,
      "grad_norm": 0.003928763791918755,
      "learning_rate": 0.199989582206035,
      "loss": 0.603,
      "num_input_tokens_seen": 21550400,
      "step": 15675
    },
    {
      "epoch": 0.5018884834517636,
      "grad_norm": 0.00812840461730957,
      "learning_rate": 0.1999340408699866,
      "loss": 0.6047,
      "num_input_tokens_seen": 21557216,
      "step": 15680
    },
    {
      "epoch": 0.5020485244222521,
      "grad_norm": 0.005990738049149513,
      "learning_rate": 0.19987849183348155,
      "loss": 0.5649,
      "num_input_tokens_seen": 21564624,
      "step": 15685
    },
    {
      "epoch": 0.5022085653927405,
      "grad_norm": 0.004449970554560423,
      "learning_rate": 0.19982293510508628,
      "loss": 0.5503,
      "num_input_tokens_seen": 21571520,
      "step": 15690
    },
    {
      "epoch": 0.502368606363229,
      "grad_norm": 0.007725103292614222,
      "learning_rate": 0.19976737069336833,
      "loss": 0.6863,
      "num_input_tokens_seen": 21578304,
      "step": 15695
    },
    {
      "epoch": 0.5025286473337174,
      "grad_norm": 0.008230676874518394,
      "learning_rate": 0.1997117986068964,
      "loss": 0.5996,
      "num_input_tokens_seen": 21585568,
      "step": 15700
    },
    {
      "epoch": 0.5026886883042059,
      "grad_norm": 0.003492202376946807,
      "learning_rate": 0.19965621885424037,
      "loss": 0.3956,
      "num_input_tokens_seen": 21592432,
      "step": 15705
    },
    {
      "epoch": 0.5028487292746944,
      "grad_norm": 0.002721794880926609,
      "learning_rate": 0.19960063144397142,
      "loss": 0.6287,
      "num_input_tokens_seen": 21599344,
      "step": 15710
    },
    {
      "epoch": 0.5030087702451828,
      "grad_norm": 0.00213025719858706,
      "learning_rate": 0.19954503638466176,
      "loss": 0.5313,
      "num_input_tokens_seen": 21606544,
      "step": 15715
    },
    {
      "epoch": 0.5031688112156713,
      "grad_norm": 0.004502326250076294,
      "learning_rate": 0.1994894336848848,
      "loss": 0.6305,
      "num_input_tokens_seen": 21613344,
      "step": 15720
    },
    {
      "epoch": 0.5033288521861596,
      "grad_norm": 0.002783730160444975,
      "learning_rate": 0.1994338233532153,
      "loss": 0.5949,
      "num_input_tokens_seen": 21620208,
      "step": 15725
    },
    {
      "epoch": 0.5034888931566481,
      "grad_norm": 0.007440091576427221,
      "learning_rate": 0.19937820539822904,
      "loss": 0.5156,
      "num_input_tokens_seen": 21627056,
      "step": 15730
    },
    {
      "epoch": 0.5036489341271365,
      "grad_norm": 0.004033825825899839,
      "learning_rate": 0.199322579828503,
      "loss": 0.6911,
      "num_input_tokens_seen": 21633872,
      "step": 15735
    },
    {
      "epoch": 0.503808975097625,
      "grad_norm": 0.0035245909821242094,
      "learning_rate": 0.19926694665261527,
      "loss": 0.3371,
      "num_input_tokens_seen": 21640432,
      "step": 15740
    },
    {
      "epoch": 0.5039690160681134,
      "grad_norm": 0.007377705071121454,
      "learning_rate": 0.19921130587914526,
      "loss": 0.6113,
      "num_input_tokens_seen": 21647760,
      "step": 15745
    },
    {
      "epoch": 0.5041290570386019,
      "grad_norm": 0.007632804568856955,
      "learning_rate": 0.19915565751667344,
      "loss": 0.6191,
      "num_input_tokens_seen": 21654208,
      "step": 15750
    },
    {
      "epoch": 0.5042890980090903,
      "grad_norm": 0.004981718957424164,
      "learning_rate": 0.19910000157378152,
      "loss": 0.4354,
      "num_input_tokens_seen": 21660896,
      "step": 15755
    },
    {
      "epoch": 0.5044491389795788,
      "grad_norm": 0.004096411634236574,
      "learning_rate": 0.1990443380590523,
      "loss": 0.4954,
      "num_input_tokens_seen": 21667808,
      "step": 15760
    },
    {
      "epoch": 0.5046091799500673,
      "grad_norm": 0.00301181897521019,
      "learning_rate": 0.19898866698106984,
      "loss": 0.3863,
      "num_input_tokens_seen": 21674224,
      "step": 15765
    },
    {
      "epoch": 0.5047692209205557,
      "grad_norm": 0.005637290887534618,
      "learning_rate": 0.19893298834841933,
      "loss": 0.6921,
      "num_input_tokens_seen": 21680832,
      "step": 15770
    },
    {
      "epoch": 0.5049292618910441,
      "grad_norm": 0.008084164001047611,
      "learning_rate": 0.19887730216968705,
      "loss": 0.5088,
      "num_input_tokens_seen": 21687728,
      "step": 15775
    },
    {
      "epoch": 0.5050893028615325,
      "grad_norm": 0.005387009587138891,
      "learning_rate": 0.19882160845346053,
      "loss": 0.5145,
      "num_input_tokens_seen": 21694688,
      "step": 15780
    },
    {
      "epoch": 0.505249343832021,
      "grad_norm": 0.004165842197835445,
      "learning_rate": 0.1987659072083285,
      "loss": 0.606,
      "num_input_tokens_seen": 21701616,
      "step": 15785
    },
    {
      "epoch": 0.5054093848025094,
      "grad_norm": 0.0036687718238681555,
      "learning_rate": 0.1987101984428807,
      "loss": 0.6643,
      "num_input_tokens_seen": 21708288,
      "step": 15790
    },
    {
      "epoch": 0.5055694257729979,
      "grad_norm": 0.004954859148710966,
      "learning_rate": 0.19865448216570822,
      "loss": 0.516,
      "num_input_tokens_seen": 21715056,
      "step": 15795
    },
    {
      "epoch": 0.5057294667434863,
      "grad_norm": 0.0037717714440077543,
      "learning_rate": 0.19859875838540317,
      "loss": 0.6351,
      "num_input_tokens_seen": 21722256,
      "step": 15800
    },
    {
      "epoch": 0.5057294667434863,
      "eval_loss": 0.5704044699668884,
      "eval_runtime": 332.4201,
      "eval_samples_per_second": 41.772,
      "eval_steps_per_second": 20.886,
      "num_input_tokens_seen": 21722256,
      "step": 15800
    },
    {
      "epoch": 0.5058895077139748,
      "grad_norm": 0.0023777366150170565,
      "learning_rate": 0.1985430271105588,
      "loss": 0.5883,
      "num_input_tokens_seen": 21729328,
      "step": 15805
    },
    {
      "epoch": 0.5060495486844632,
      "grad_norm": 0.0065911514684557915,
      "learning_rate": 0.19848728834976961,
      "loss": 0.6434,
      "num_input_tokens_seen": 21736128,
      "step": 15810
    },
    {
      "epoch": 0.5062095896549517,
      "grad_norm": 0.0022249342873692513,
      "learning_rate": 0.19843154211163128,
      "loss": 0.4067,
      "num_input_tokens_seen": 21743328,
      "step": 15815
    },
    {
      "epoch": 0.5063696306254402,
      "grad_norm": 0.009044423699378967,
      "learning_rate": 0.1983757884047405,
      "loss": 0.4859,
      "num_input_tokens_seen": 21749872,
      "step": 15820
    },
    {
      "epoch": 0.5065296715959285,
      "grad_norm": 0.004131062421947718,
      "learning_rate": 0.1983200272376952,
      "loss": 0.6056,
      "num_input_tokens_seen": 21757088,
      "step": 15825
    },
    {
      "epoch": 0.506689712566417,
      "grad_norm": 0.003910768777132034,
      "learning_rate": 0.1982642586190945,
      "loss": 0.5298,
      "num_input_tokens_seen": 21763424,
      "step": 15830
    },
    {
      "epoch": 0.5068497535369054,
      "grad_norm": 0.0038087007123976946,
      "learning_rate": 0.1982084825575386,
      "loss": 0.5802,
      "num_input_tokens_seen": 21770400,
      "step": 15835
    },
    {
      "epoch": 0.5070097945073939,
      "grad_norm": 0.005207827780395746,
      "learning_rate": 0.19815269906162883,
      "loss": 0.5683,
      "num_input_tokens_seen": 21777184,
      "step": 15840
    },
    {
      "epoch": 0.5071698354778823,
      "grad_norm": 0.007101519498974085,
      "learning_rate": 0.19809690813996775,
      "loss": 0.4928,
      "num_input_tokens_seen": 21784512,
      "step": 15845
    },
    {
      "epoch": 0.5073298764483708,
      "grad_norm": 0.005742514971643686,
      "learning_rate": 0.19804110980115905,
      "loss": 0.5779,
      "num_input_tokens_seen": 21791392,
      "step": 15850
    },
    {
      "epoch": 0.5074899174188592,
      "grad_norm": 0.005743726156651974,
      "learning_rate": 0.19798530405380746,
      "loss": 0.6097,
      "num_input_tokens_seen": 21798272,
      "step": 15855
    },
    {
      "epoch": 0.5076499583893477,
      "grad_norm": 0.004111315589398146,
      "learning_rate": 0.19792949090651893,
      "loss": 0.527,
      "num_input_tokens_seen": 21805152,
      "step": 15860
    },
    {
      "epoch": 0.5078099993598362,
      "grad_norm": 0.004171174950897694,
      "learning_rate": 0.19787367036790066,
      "loss": 0.5551,
      "num_input_tokens_seen": 21811632,
      "step": 15865
    },
    {
      "epoch": 0.5079700403303246,
      "grad_norm": 0.0032937591895461082,
      "learning_rate": 0.19781784244656075,
      "loss": 0.5884,
      "num_input_tokens_seen": 21818032,
      "step": 15870
    },
    {
      "epoch": 0.508130081300813,
      "grad_norm": 0.004208238795399666,
      "learning_rate": 0.19776200715110864,
      "loss": 0.5955,
      "num_input_tokens_seen": 21824688,
      "step": 15875
    },
    {
      "epoch": 0.5082901222713014,
      "grad_norm": 0.0036621156614273787,
      "learning_rate": 0.1977061644901548,
      "loss": 0.4411,
      "num_input_tokens_seen": 21831568,
      "step": 15880
    },
    {
      "epoch": 0.5084501632417899,
      "grad_norm": 0.004041642881929874,
      "learning_rate": 0.1976503144723109,
      "loss": 0.6917,
      "num_input_tokens_seen": 21838368,
      "step": 15885
    },
    {
      "epoch": 0.5086102042122783,
      "grad_norm": 0.007018499542027712,
      "learning_rate": 0.19759445710618967,
      "loss": 0.5749,
      "num_input_tokens_seen": 21845136,
      "step": 15890
    },
    {
      "epoch": 0.5087702451827668,
      "grad_norm": 0.004922269843518734,
      "learning_rate": 0.19753859240040508,
      "loss": 0.6583,
      "num_input_tokens_seen": 21852656,
      "step": 15895
    },
    {
      "epoch": 0.5089302861532552,
      "grad_norm": 0.002125919796526432,
      "learning_rate": 0.1974827203635721,
      "loss": 0.4278,
      "num_input_tokens_seen": 21859296,
      "step": 15900
    },
    {
      "epoch": 0.5090903271237437,
      "grad_norm": 0.005188989453017712,
      "learning_rate": 0.19742684100430694,
      "loss": 0.682,
      "num_input_tokens_seen": 21866592,
      "step": 15905
    },
    {
      "epoch": 0.5092503680942321,
      "grad_norm": 0.007504008244723082,
      "learning_rate": 0.19737095433122692,
      "loss": 0.4817,
      "num_input_tokens_seen": 21873440,
      "step": 15910
    },
    {
      "epoch": 0.5094104090647206,
      "grad_norm": 0.004999599885195494,
      "learning_rate": 0.19731506035295046,
      "loss": 0.655,
      "num_input_tokens_seen": 21880080,
      "step": 15915
    },
    {
      "epoch": 0.5095704500352091,
      "grad_norm": 0.0063000088557600975,
      "learning_rate": 0.19725915907809702,
      "loss": 0.6043,
      "num_input_tokens_seen": 21886944,
      "step": 15920
    },
    {
      "epoch": 0.5097304910056975,
      "grad_norm": 0.006972698029130697,
      "learning_rate": 0.1972032505152874,
      "loss": 0.4871,
      "num_input_tokens_seen": 21893584,
      "step": 15925
    },
    {
      "epoch": 0.509890531976186,
      "grad_norm": 0.003566802479326725,
      "learning_rate": 0.19714733467314338,
      "loss": 0.6209,
      "num_input_tokens_seen": 21900752,
      "step": 15930
    },
    {
      "epoch": 0.5100505729466743,
      "grad_norm": 0.006135036703199148,
      "learning_rate": 0.19709141156028784,
      "loss": 0.625,
      "num_input_tokens_seen": 21907904,
      "step": 15935
    },
    {
      "epoch": 0.5102106139171628,
      "grad_norm": 0.00458080368116498,
      "learning_rate": 0.1970354811853448,
      "loss": 0.4222,
      "num_input_tokens_seen": 21914848,
      "step": 15940
    },
    {
      "epoch": 0.5103706548876512,
      "grad_norm": 0.003146293107420206,
      "learning_rate": 0.19697954355693953,
      "loss": 0.5256,
      "num_input_tokens_seen": 21921920,
      "step": 15945
    },
    {
      "epoch": 0.5105306958581397,
      "grad_norm": 0.006453763227909803,
      "learning_rate": 0.19692359868369827,
      "loss": 0.6515,
      "num_input_tokens_seen": 21928784,
      "step": 15950
    },
    {
      "epoch": 0.5106907368286281,
      "grad_norm": 0.003461824031546712,
      "learning_rate": 0.1968676465742484,
      "loss": 0.5107,
      "num_input_tokens_seen": 21936352,
      "step": 15955
    },
    {
      "epoch": 0.5108507777991166,
      "grad_norm": 0.0048082321882247925,
      "learning_rate": 0.19681168723721845,
      "loss": 0.5717,
      "num_input_tokens_seen": 21943472,
      "step": 15960
    },
    {
      "epoch": 0.511010818769605,
      "grad_norm": 0.004893832840025425,
      "learning_rate": 0.19675572068123803,
      "loss": 0.4604,
      "num_input_tokens_seen": 21950144,
      "step": 15965
    },
    {
      "epoch": 0.5111708597400935,
      "grad_norm": 0.0021672791335731745,
      "learning_rate": 0.19669974691493794,
      "loss": 0.5272,
      "num_input_tokens_seen": 21957248,
      "step": 15970
    },
    {
      "epoch": 0.511330900710582,
      "grad_norm": 0.003224777989089489,
      "learning_rate": 0.19664376594695002,
      "loss": 0.3405,
      "num_input_tokens_seen": 21964096,
      "step": 15975
    },
    {
      "epoch": 0.5114909416810703,
      "grad_norm": 0.00305732199922204,
      "learning_rate": 0.19658777778590722,
      "loss": 0.5326,
      "num_input_tokens_seen": 21970992,
      "step": 15980
    },
    {
      "epoch": 0.5116509826515588,
      "grad_norm": 0.005267255008220673,
      "learning_rate": 0.19653178244044364,
      "loss": 0.5795,
      "num_input_tokens_seen": 21977664,
      "step": 15985
    },
    {
      "epoch": 0.5118110236220472,
      "grad_norm": 0.005617485847324133,
      "learning_rate": 0.19647577991919443,
      "loss": 0.8052,
      "num_input_tokens_seen": 21984128,
      "step": 15990
    },
    {
      "epoch": 0.5119710645925357,
      "grad_norm": 0.007865093648433685,
      "learning_rate": 0.1964197702307959,
      "loss": 0.4736,
      "num_input_tokens_seen": 21991184,
      "step": 15995
    },
    {
      "epoch": 0.5121311055630241,
      "grad_norm": 0.004530477803200483,
      "learning_rate": 0.19636375338388545,
      "loss": 0.5437,
      "num_input_tokens_seen": 21998320,
      "step": 16000
    },
    {
      "epoch": 0.5121311055630241,
      "eval_loss": 0.5731611251831055,
      "eval_runtime": 332.9379,
      "eval_samples_per_second": 41.707,
      "eval_steps_per_second": 20.854,
      "num_input_tokens_seen": 21998320,
      "step": 16000
    },
    {
      "epoch": 0.5122911465335126,
      "grad_norm": 0.008644579909741879,
      "learning_rate": 0.1963077293871016,
      "loss": 0.6826,
      "num_input_tokens_seen": 22004960,
      "step": 16005
    },
    {
      "epoch": 0.512451187504001,
      "grad_norm": 0.0072675468400120735,
      "learning_rate": 0.19625169824908395,
      "loss": 0.3903,
      "num_input_tokens_seen": 22011920,
      "step": 16010
    },
    {
      "epoch": 0.5126112284744895,
      "grad_norm": 0.003945550881326199,
      "learning_rate": 0.19619565997847319,
      "loss": 0.5909,
      "num_input_tokens_seen": 22019632,
      "step": 16015
    },
    {
      "epoch": 0.5127712694449779,
      "grad_norm": 0.005393018014729023,
      "learning_rate": 0.19613961458391113,
      "loss": 0.7975,
      "num_input_tokens_seen": 22026576,
      "step": 16020
    },
    {
      "epoch": 0.5129313104154664,
      "grad_norm": 0.00518873892724514,
      "learning_rate": 0.19608356207404065,
      "loss": 0.5885,
      "num_input_tokens_seen": 22033168,
      "step": 16025
    },
    {
      "epoch": 0.5130913513859549,
      "grad_norm": 0.0063039278611540794,
      "learning_rate": 0.1960275024575058,
      "loss": 0.5791,
      "num_input_tokens_seen": 22039648,
      "step": 16030
    },
    {
      "epoch": 0.5132513923564432,
      "grad_norm": 0.004482278600335121,
      "learning_rate": 0.19597143574295164,
      "loss": 0.543,
      "num_input_tokens_seen": 22046176,
      "step": 16035
    },
    {
      "epoch": 0.5134114333269317,
      "grad_norm": 0.006408785004168749,
      "learning_rate": 0.1959153619390244,
      "loss": 0.6574,
      "num_input_tokens_seen": 22053248,
      "step": 16040
    },
    {
      "epoch": 0.5135714742974201,
      "grad_norm": 0.004164599347859621,
      "learning_rate": 0.1958592810543713,
      "loss": 0.4859,
      "num_input_tokens_seen": 22060080,
      "step": 16045
    },
    {
      "epoch": 0.5137315152679086,
      "grad_norm": 0.004019744228571653,
      "learning_rate": 0.19580319309764077,
      "loss": 0.3089,
      "num_input_tokens_seen": 22066608,
      "step": 16050
    },
    {
      "epoch": 0.513891556238397,
      "grad_norm": 0.006393336225301027,
      "learning_rate": 0.1957470980774823,
      "loss": 0.6739,
      "num_input_tokens_seen": 22073376,
      "step": 16055
    },
    {
      "epoch": 0.5140515972088855,
      "grad_norm": 0.003319981275126338,
      "learning_rate": 0.19569099600254639,
      "loss": 0.4525,
      "num_input_tokens_seen": 22080480,
      "step": 16060
    },
    {
      "epoch": 0.5142116381793739,
      "grad_norm": 0.006603216286748648,
      "learning_rate": 0.1956348868814847,
      "loss": 0.4407,
      "num_input_tokens_seen": 22086784,
      "step": 16065
    },
    {
      "epoch": 0.5143716791498624,
      "grad_norm": 0.002906396985054016,
      "learning_rate": 0.19557877072295,
      "loss": 0.5845,
      "num_input_tokens_seen": 22093568,
      "step": 16070
    },
    {
      "epoch": 0.5145317201203508,
      "grad_norm": 0.009130039252340794,
      "learning_rate": 0.19552264753559603,
      "loss": 0.543,
      "num_input_tokens_seen": 22100864,
      "step": 16075
    },
    {
      "epoch": 0.5146917610908393,
      "grad_norm": 0.004611196927726269,
      "learning_rate": 0.19546651732807774,
      "loss": 0.717,
      "num_input_tokens_seen": 22108048,
      "step": 16080
    },
    {
      "epoch": 0.5148518020613277,
      "grad_norm": 0.004125331062823534,
      "learning_rate": 0.19541038010905112,
      "loss": 0.5108,
      "num_input_tokens_seen": 22114544,
      "step": 16085
    },
    {
      "epoch": 0.5150118430318161,
      "grad_norm": 0.00564374728128314,
      "learning_rate": 0.19535423588717324,
      "loss": 0.607,
      "num_input_tokens_seen": 22121280,
      "step": 16090
    },
    {
      "epoch": 0.5151718840023046,
      "grad_norm": 0.013584553264081478,
      "learning_rate": 0.19529808467110224,
      "loss": 0.5341,
      "num_input_tokens_seen": 22128416,
      "step": 16095
    },
    {
      "epoch": 0.515331924972793,
      "grad_norm": 0.004163797479122877,
      "learning_rate": 0.19524192646949734,
      "loss": 0.4627,
      "num_input_tokens_seen": 22134976,
      "step": 16100
    },
    {
      "epoch": 0.5154919659432815,
      "grad_norm": 0.004896792583167553,
      "learning_rate": 0.19518576129101878,
      "loss": 0.6753,
      "num_input_tokens_seen": 22141632,
      "step": 16105
    },
    {
      "epoch": 0.5156520069137699,
      "grad_norm": 0.007610672619193792,
      "learning_rate": 0.19512958914432804,
      "loss": 0.5919,
      "num_input_tokens_seen": 22149056,
      "step": 16110
    },
    {
      "epoch": 0.5158120478842584,
      "grad_norm": 0.00511121517047286,
      "learning_rate": 0.1950734100380875,
      "loss": 0.7165,
      "num_input_tokens_seen": 22156176,
      "step": 16115
    },
    {
      "epoch": 0.5159720888547468,
      "grad_norm": 0.0033864639699459076,
      "learning_rate": 0.19501722398096066,
      "loss": 0.7565,
      "num_input_tokens_seen": 22163136,
      "step": 16120
    },
    {
      "epoch": 0.5161321298252353,
      "grad_norm": 0.005400801543146372,
      "learning_rate": 0.1949610309816122,
      "loss": 0.6342,
      "num_input_tokens_seen": 22169744,
      "step": 16125
    },
    {
      "epoch": 0.5162921707957238,
      "grad_norm": 0.006509678438305855,
      "learning_rate": 0.1949048310487078,
      "loss": 0.4774,
      "num_input_tokens_seen": 22176560,
      "step": 16130
    },
    {
      "epoch": 0.5164522117662121,
      "grad_norm": 0.004955979064106941,
      "learning_rate": 0.19484862419091406,
      "loss": 0.5753,
      "num_input_tokens_seen": 22183552,
      "step": 16135
    },
    {
      "epoch": 0.5166122527367006,
      "grad_norm": 0.004470016807317734,
      "learning_rate": 0.19479241041689893,
      "loss": 0.6353,
      "num_input_tokens_seen": 22190208,
      "step": 16140
    },
    {
      "epoch": 0.516772293707189,
      "grad_norm": 0.004959247075021267,
      "learning_rate": 0.19473618973533116,
      "loss": 0.5968,
      "num_input_tokens_seen": 22197008,
      "step": 16145
    },
    {
      "epoch": 0.5169323346776775,
      "grad_norm": 0.010672681964933872,
      "learning_rate": 0.19467996215488076,
      "loss": 0.7468,
      "num_input_tokens_seen": 22203728,
      "step": 16150
    },
    {
      "epoch": 0.5170923756481659,
      "grad_norm": 0.002769066719338298,
      "learning_rate": 0.1946237276842187,
      "loss": 0.5354,
      "num_input_tokens_seen": 22210480,
      "step": 16155
    },
    {
      "epoch": 0.5172524166186544,
      "grad_norm": 0.005819325335323811,
      "learning_rate": 0.19456748633201712,
      "loss": 0.584,
      "num_input_tokens_seen": 22217744,
      "step": 16160
    },
    {
      "epoch": 0.5174124575891428,
      "grad_norm": 0.004612860735505819,
      "learning_rate": 0.194511238106949,
      "loss": 0.5752,
      "num_input_tokens_seen": 22225440,
      "step": 16165
    },
    {
      "epoch": 0.5175724985596313,
      "grad_norm": 0.00535125657916069,
      "learning_rate": 0.19445498301768863,
      "loss": 0.539,
      "num_input_tokens_seen": 22232720,
      "step": 16170
    },
    {
      "epoch": 0.5177325395301197,
      "grad_norm": 0.006238358095288277,
      "learning_rate": 0.19439872107291126,
      "loss": 0.5601,
      "num_input_tokens_seen": 22239472,
      "step": 16175
    },
    {
      "epoch": 0.5178925805006082,
      "grad_norm": 0.003770931391045451,
      "learning_rate": 0.1943424522812931,
      "loss": 0.4734,
      "num_input_tokens_seen": 22245968,
      "step": 16180
    },
    {
      "epoch": 0.5180526214710967,
      "grad_norm": 0.0048704505898058414,
      "learning_rate": 0.19428617665151157,
      "loss": 0.6113,
      "num_input_tokens_seen": 22253408,
      "step": 16185
    },
    {
      "epoch": 0.518212662441585,
      "grad_norm": 0.0026615301612764597,
      "learning_rate": 0.19422989419224507,
      "loss": 0.6822,
      "num_input_tokens_seen": 22260128,
      "step": 16190
    },
    {
      "epoch": 0.5183727034120735,
      "grad_norm": 0.003604237223044038,
      "learning_rate": 0.19417360491217303,
      "loss": 0.4409,
      "num_input_tokens_seen": 22267104,
      "step": 16195
    },
    {
      "epoch": 0.5185327443825619,
      "grad_norm": 0.00452393526211381,
      "learning_rate": 0.19411730881997605,
      "loss": 0.5628,
      "num_input_tokens_seen": 22273616,
      "step": 16200
    },
    {
      "epoch": 0.5185327443825619,
      "eval_loss": 0.5638567209243774,
      "eval_runtime": 331.1436,
      "eval_samples_per_second": 41.933,
      "eval_steps_per_second": 20.967,
      "num_input_tokens_seen": 22273616,
      "step": 16200
    },
    {
      "epoch": 0.5186927853530504,
      "grad_norm": 0.005197005346417427,
      "learning_rate": 0.1940610059243356,
      "loss": 0.688,
      "num_input_tokens_seen": 22280704,
      "step": 16205
    },
    {
      "epoch": 0.5188528263235388,
      "grad_norm": 0.005468974821269512,
      "learning_rate": 0.19400469623393435,
      "loss": 0.4049,
      "num_input_tokens_seen": 22287344,
      "step": 16210
    },
    {
      "epoch": 0.5190128672940273,
      "grad_norm": 0.0026335623115301132,
      "learning_rate": 0.1939483797574559,
      "loss": 0.6074,
      "num_input_tokens_seen": 22294352,
      "step": 16215
    },
    {
      "epoch": 0.5191729082645157,
      "grad_norm": 0.0035848303232342005,
      "learning_rate": 0.19389205650358504,
      "loss": 0.5806,
      "num_input_tokens_seen": 22301216,
      "step": 16220
    },
    {
      "epoch": 0.5193329492350042,
      "grad_norm": 0.00471110362559557,
      "learning_rate": 0.19383572648100747,
      "loss": 0.5469,
      "num_input_tokens_seen": 22308272,
      "step": 16225
    },
    {
      "epoch": 0.5194929902054926,
      "grad_norm": 0.0024299463257193565,
      "learning_rate": 0.19377938969841,
      "loss": 0.5121,
      "num_input_tokens_seen": 22315024,
      "step": 16230
    },
    {
      "epoch": 0.519653031175981,
      "grad_norm": 0.007666486315429211,
      "learning_rate": 0.1937230461644805,
      "loss": 0.5839,
      "num_input_tokens_seen": 22322384,
      "step": 16235
    },
    {
      "epoch": 0.5198130721464695,
      "grad_norm": 0.006492193788290024,
      "learning_rate": 0.19366669588790777,
      "loss": 0.5965,
      "num_input_tokens_seen": 22329248,
      "step": 16240
    },
    {
      "epoch": 0.5199731131169579,
      "grad_norm": 0.01050378568470478,
      "learning_rate": 0.19361033887738185,
      "loss": 0.4884,
      "num_input_tokens_seen": 22336160,
      "step": 16245
    },
    {
      "epoch": 0.5201331540874464,
      "grad_norm": 0.009970723651349545,
      "learning_rate": 0.19355397514159361,
      "loss": 0.5762,
      "num_input_tokens_seen": 22342912,
      "step": 16250
    },
    {
      "epoch": 0.5202931950579348,
      "grad_norm": 0.006667367182672024,
      "learning_rate": 0.19349760468923508,
      "loss": 0.5461,
      "num_input_tokens_seen": 22349808,
      "step": 16255
    },
    {
      "epoch": 0.5204532360284233,
      "grad_norm": 0.008140746504068375,
      "learning_rate": 0.19344122752899925,
      "loss": 0.6455,
      "num_input_tokens_seen": 22356592,
      "step": 16260
    },
    {
      "epoch": 0.5206132769989117,
      "grad_norm": 0.004436733666807413,
      "learning_rate": 0.1933848436695802,
      "loss": 0.7285,
      "num_input_tokens_seen": 22363008,
      "step": 16265
    },
    {
      "epoch": 0.5207733179694002,
      "grad_norm": 0.008905038237571716,
      "learning_rate": 0.1933284531196731,
      "loss": 0.7215,
      "num_input_tokens_seen": 22370064,
      "step": 16270
    },
    {
      "epoch": 0.5209333589398886,
      "grad_norm": 0.002795109059661627,
      "learning_rate": 0.19327205588797403,
      "loss": 0.595,
      "num_input_tokens_seen": 22377296,
      "step": 16275
    },
    {
      "epoch": 0.5210933999103771,
      "grad_norm": 0.005638009402900934,
      "learning_rate": 0.19321565198318014,
      "loss": 0.6095,
      "num_input_tokens_seen": 22384144,
      "step": 16280
    },
    {
      "epoch": 0.5212534408808654,
      "grad_norm": 0.004452693276107311,
      "learning_rate": 0.1931592414139896,
      "loss": 0.7239,
      "num_input_tokens_seen": 22390912,
      "step": 16285
    },
    {
      "epoch": 0.5214134818513539,
      "grad_norm": 0.003480906132608652,
      "learning_rate": 0.19310282418910169,
      "loss": 0.5164,
      "num_input_tokens_seen": 22397664,
      "step": 16290
    },
    {
      "epoch": 0.5215735228218424,
      "grad_norm": 0.0036114139948040247,
      "learning_rate": 0.1930464003172166,
      "loss": 0.6197,
      "num_input_tokens_seen": 22404816,
      "step": 16295
    },
    {
      "epoch": 0.5217335637923308,
      "grad_norm": 0.0034287255257368088,
      "learning_rate": 0.19298996980703567,
      "loss": 0.5476,
      "num_input_tokens_seen": 22412016,
      "step": 16300
    },
    {
      "epoch": 0.5218936047628193,
      "grad_norm": 0.005010650027543306,
      "learning_rate": 0.19293353266726113,
      "loss": 0.5725,
      "num_input_tokens_seen": 22418784,
      "step": 16305
    },
    {
      "epoch": 0.5220536457333077,
      "grad_norm": 0.004235208034515381,
      "learning_rate": 0.19287708890659633,
      "loss": 0.7504,
      "num_input_tokens_seen": 22425552,
      "step": 16310
    },
    {
      "epoch": 0.5222136867037962,
      "grad_norm": 0.004826229065656662,
      "learning_rate": 0.19282063853374556,
      "loss": 0.4384,
      "num_input_tokens_seen": 22432208,
      "step": 16315
    },
    {
      "epoch": 0.5223737276742846,
      "grad_norm": 0.0033419460523873568,
      "learning_rate": 0.19276418155741423,
      "loss": 0.5419,
      "num_input_tokens_seen": 22439264,
      "step": 16320
    },
    {
      "epoch": 0.5225337686447731,
      "grad_norm": 0.004400718025863171,
      "learning_rate": 0.19270771798630867,
      "loss": 0.5617,
      "num_input_tokens_seen": 22446848,
      "step": 16325
    },
    {
      "epoch": 0.5226938096152615,
      "grad_norm": 0.004150425549596548,
      "learning_rate": 0.1926512478291363,
      "loss": 0.5602,
      "num_input_tokens_seen": 22453408,
      "step": 16330
    },
    {
      "epoch": 0.52285385058575,
      "grad_norm": 0.003683025948703289,
      "learning_rate": 0.19259477109460557,
      "loss": 0.5413,
      "num_input_tokens_seen": 22460320,
      "step": 16335
    },
    {
      "epoch": 0.5230138915562383,
      "grad_norm": 0.0039054597727954388,
      "learning_rate": 0.19253828779142584,
      "loss": 0.4003,
      "num_input_tokens_seen": 22467168,
      "step": 16340
    },
    {
      "epoch": 0.5231739325267268,
      "grad_norm": 0.009833150543272495,
      "learning_rate": 0.19248179792830755,
      "loss": 0.7203,
      "num_input_tokens_seen": 22473504,
      "step": 16345
    },
    {
      "epoch": 0.5233339734972153,
      "grad_norm": 0.005869867280125618,
      "learning_rate": 0.19242530151396217,
      "loss": 0.6164,
      "num_input_tokens_seen": 22480112,
      "step": 16350
    },
    {
      "epoch": 0.5234940144677037,
      "grad_norm": 0.007213095668703318,
      "learning_rate": 0.19236879855710215,
      "loss": 0.5428,
      "num_input_tokens_seen": 22486592,
      "step": 16355
    },
    {
      "epoch": 0.5236540554381922,
      "grad_norm": 0.003861095057800412,
      "learning_rate": 0.19231228906644096,
      "loss": 0.6233,
      "num_input_tokens_seen": 22493584,
      "step": 16360
    },
    {
      "epoch": 0.5238140964086806,
      "grad_norm": 0.004967625718563795,
      "learning_rate": 0.19225577305069302,
      "loss": 0.5372,
      "num_input_tokens_seen": 22500176,
      "step": 16365
    },
    {
      "epoch": 0.5239741373791691,
      "grad_norm": 0.0036937675904482603,
      "learning_rate": 0.1921992505185739,
      "loss": 0.5474,
      "num_input_tokens_seen": 22506848,
      "step": 16370
    },
    {
      "epoch": 0.5241341783496575,
      "grad_norm": 0.00465004239231348,
      "learning_rate": 0.19214272147880004,
      "loss": 0.5797,
      "num_input_tokens_seen": 22514208,
      "step": 16375
    },
    {
      "epoch": 0.524294219320146,
      "grad_norm": 0.00502040283754468,
      "learning_rate": 0.19208618594008892,
      "loss": 0.6211,
      "num_input_tokens_seen": 22521168,
      "step": 16380
    },
    {
      "epoch": 0.5244542602906344,
      "grad_norm": 0.0035737568978220224,
      "learning_rate": 0.19202964391115904,
      "loss": 0.6425,
      "num_input_tokens_seen": 22528512,
      "step": 16385
    },
    {
      "epoch": 0.5246143012611229,
      "grad_norm": 0.004040765576064587,
      "learning_rate": 0.1919730954007299,
      "loss": 0.5037,
      "num_input_tokens_seen": 22535328,
      "step": 16390
    },
    {
      "epoch": 0.5247743422316113,
      "grad_norm": 0.00745677063241601,
      "learning_rate": 0.19191654041752199,
      "loss": 0.677,
      "num_input_tokens_seen": 22542256,
      "step": 16395
    },
    {
      "epoch": 0.5249343832020997,
      "grad_norm": 0.006993425078690052,
      "learning_rate": 0.19185997897025678,
      "loss": 0.626,
      "num_input_tokens_seen": 22549280,
      "step": 16400
    },
    {
      "epoch": 0.5249343832020997,
      "eval_loss": 0.5587302446365356,
      "eval_runtime": 331.9054,
      "eval_samples_per_second": 41.837,
      "eval_steps_per_second": 20.919,
      "num_input_tokens_seen": 22549280,
      "step": 16400
    },
    {
      "epoch": 0.5250944241725882,
      "grad_norm": 0.005451712291687727,
      "learning_rate": 0.19180341106765672,
      "loss": 0.6285,
      "num_input_tokens_seen": 22556208,
      "step": 16405
    },
    {
      "epoch": 0.5252544651430766,
      "grad_norm": 0.006948941387236118,
      "learning_rate": 0.19174683671844536,
      "loss": 0.583,
      "num_input_tokens_seen": 22562608,
      "step": 16410
    },
    {
      "epoch": 0.5254145061135651,
      "grad_norm": 0.003952684812247753,
      "learning_rate": 0.19169025593134717,
      "loss": 0.5835,
      "num_input_tokens_seen": 22569584,
      "step": 16415
    },
    {
      "epoch": 0.5255745470840535,
      "grad_norm": 0.010509693063795567,
      "learning_rate": 0.19163366871508764,
      "loss": 0.6421,
      "num_input_tokens_seen": 22576448,
      "step": 16420
    },
    {
      "epoch": 0.525734588054542,
      "grad_norm": 0.0053841108456254005,
      "learning_rate": 0.19157707507839317,
      "loss": 0.4508,
      "num_input_tokens_seen": 22583552,
      "step": 16425
    },
    {
      "epoch": 0.5258946290250304,
      "grad_norm": 0.006562752183526754,
      "learning_rate": 0.19152047502999123,
      "loss": 0.6223,
      "num_input_tokens_seen": 22590592,
      "step": 16430
    },
    {
      "epoch": 0.5260546699955189,
      "grad_norm": 0.0051033105701208115,
      "learning_rate": 0.19146386857861025,
      "loss": 0.5953,
      "num_input_tokens_seen": 22597328,
      "step": 16435
    },
    {
      "epoch": 0.5262147109660072,
      "grad_norm": 0.0033453002106398344,
      "learning_rate": 0.19140725573297968,
      "loss": 0.355,
      "num_input_tokens_seen": 22604160,
      "step": 16440
    },
    {
      "epoch": 0.5263747519364957,
      "grad_norm": 0.005605675280094147,
      "learning_rate": 0.19135063650182987,
      "loss": 0.608,
      "num_input_tokens_seen": 22610768,
      "step": 16445
    },
    {
      "epoch": 0.5265347929069842,
      "grad_norm": 0.005410805810242891,
      "learning_rate": 0.19129401089389234,
      "loss": 0.4665,
      "num_input_tokens_seen": 22618208,
      "step": 16450
    },
    {
      "epoch": 0.5266948338774726,
      "grad_norm": 0.006197716109454632,
      "learning_rate": 0.19123737891789938,
      "loss": 0.6711,
      "num_input_tokens_seen": 22624896,
      "step": 16455
    },
    {
      "epoch": 0.5268548748479611,
      "grad_norm": 0.006023913621902466,
      "learning_rate": 0.19118074058258439,
      "loss": 0.6295,
      "num_input_tokens_seen": 22631792,
      "step": 16460
    },
    {
      "epoch": 0.5270149158184495,
      "grad_norm": 0.00840598251670599,
      "learning_rate": 0.1911240958966816,
      "loss": 0.6377,
      "num_input_tokens_seen": 22638608,
      "step": 16465
    },
    {
      "epoch": 0.527174956788938,
      "grad_norm": 0.003976970445364714,
      "learning_rate": 0.19106744486892652,
      "loss": 0.5297,
      "num_input_tokens_seen": 22645568,
      "step": 16470
    },
    {
      "epoch": 0.5273349977594264,
      "grad_norm": 0.003700372064486146,
      "learning_rate": 0.1910107875080553,
      "loss": 0.7048,
      "num_input_tokens_seen": 22652320,
      "step": 16475
    },
    {
      "epoch": 0.5274950387299149,
      "grad_norm": 0.003937857691198587,
      "learning_rate": 0.19095412382280533,
      "loss": 0.5866,
      "num_input_tokens_seen": 22659408,
      "step": 16480
    },
    {
      "epoch": 0.5276550797004033,
      "grad_norm": 0.003515251213684678,
      "learning_rate": 0.19089745382191473,
      "loss": 0.5251,
      "num_input_tokens_seen": 22666640,
      "step": 16485
    },
    {
      "epoch": 0.5278151206708918,
      "grad_norm": 0.009492767974734306,
      "learning_rate": 0.19084077751412284,
      "loss": 0.6208,
      "num_input_tokens_seen": 22673232,
      "step": 16490
    },
    {
      "epoch": 0.5279751616413801,
      "grad_norm": 0.00996165070682764,
      "learning_rate": 0.19078409490816986,
      "loss": 0.5596,
      "num_input_tokens_seen": 22679840,
      "step": 16495
    },
    {
      "epoch": 0.5281352026118686,
      "grad_norm": 0.004113904200494289,
      "learning_rate": 0.19072740601279686,
      "loss": 0.6581,
      "num_input_tokens_seen": 22686560,
      "step": 16500
    },
    {
      "epoch": 0.5282952435823571,
      "grad_norm": 0.006772013381123543,
      "learning_rate": 0.19067071083674605,
      "loss": 0.6136,
      "num_input_tokens_seen": 22693584,
      "step": 16505
    },
    {
      "epoch": 0.5284552845528455,
      "grad_norm": 0.0040917666628956795,
      "learning_rate": 0.19061400938876052,
      "loss": 0.5735,
      "num_input_tokens_seen": 22700608,
      "step": 16510
    },
    {
      "epoch": 0.528615325523334,
      "grad_norm": 0.00315575092099607,
      "learning_rate": 0.1905573016775844,
      "loss": 0.4856,
      "num_input_tokens_seen": 22707760,
      "step": 16515
    },
    {
      "epoch": 0.5287753664938224,
      "grad_norm": 0.00394076993688941,
      "learning_rate": 0.19050058771196263,
      "loss": 0.6275,
      "num_input_tokens_seen": 22714800,
      "step": 16520
    },
    {
      "epoch": 0.5289354074643109,
      "grad_norm": 0.005053308792412281,
      "learning_rate": 0.19044386750064132,
      "loss": 0.4948,
      "num_input_tokens_seen": 22721376,
      "step": 16525
    },
    {
      "epoch": 0.5290954484347993,
      "grad_norm": 0.0029116373043507338,
      "learning_rate": 0.19038714105236737,
      "loss": 0.7726,
      "num_input_tokens_seen": 22727616,
      "step": 16530
    },
    {
      "epoch": 0.5292554894052878,
      "grad_norm": 0.0032446738332509995,
      "learning_rate": 0.19033040837588874,
      "loss": 0.5327,
      "num_input_tokens_seen": 22734672,
      "step": 16535
    },
    {
      "epoch": 0.5294155303757762,
      "grad_norm": 0.007246738765388727,
      "learning_rate": 0.1902736694799543,
      "loss": 0.6372,
      "num_input_tokens_seen": 22741856,
      "step": 16540
    },
    {
      "epoch": 0.5295755713462647,
      "grad_norm": 0.007100386545062065,
      "learning_rate": 0.19021692437331392,
      "loss": 0.6197,
      "num_input_tokens_seen": 22748992,
      "step": 16545
    },
    {
      "epoch": 0.529735612316753,
      "grad_norm": 0.005964027717709541,
      "learning_rate": 0.1901601730647184,
      "loss": 0.5878,
      "num_input_tokens_seen": 22755696,
      "step": 16550
    },
    {
      "epoch": 0.5298956532872415,
      "grad_norm": 0.004193414933979511,
      "learning_rate": 0.19010341556291954,
      "loss": 0.4975,
      "num_input_tokens_seen": 22762672,
      "step": 16555
    },
    {
      "epoch": 0.53005569425773,
      "grad_norm": 0.007206378038972616,
      "learning_rate": 0.19004665187667,
      "loss": 0.5608,
      "num_input_tokens_seen": 22769616,
      "step": 16560
    },
    {
      "epoch": 0.5302157352282184,
      "grad_norm": 0.0034456858411431313,
      "learning_rate": 0.1899898820147235,
      "loss": 0.6383,
      "num_input_tokens_seen": 22776864,
      "step": 16565
    },
    {
      "epoch": 0.5303757761987069,
      "grad_norm": 0.011313890106976032,
      "learning_rate": 0.18993310598583465,
      "loss": 0.7688,
      "num_input_tokens_seen": 22783376,
      "step": 16570
    },
    {
      "epoch": 0.5305358171691953,
      "grad_norm": 0.013031781651079655,
      "learning_rate": 0.18987632379875904,
      "loss": 0.5237,
      "num_input_tokens_seen": 22790160,
      "step": 16575
    },
    {
      "epoch": 0.5306958581396838,
      "grad_norm": 0.01117353979498148,
      "learning_rate": 0.18981953546225314,
      "loss": 0.5928,
      "num_input_tokens_seen": 22796544,
      "step": 16580
    },
    {
      "epoch": 0.5308558991101722,
      "grad_norm": 0.011064626276493073,
      "learning_rate": 0.18976274098507445,
      "loss": 0.7169,
      "num_input_tokens_seen": 22803584,
      "step": 16585
    },
    {
      "epoch": 0.5310159400806607,
      "grad_norm": 0.00789149571210146,
      "learning_rate": 0.18970594037598146,
      "loss": 0.4937,
      "num_input_tokens_seen": 22810224,
      "step": 16590
    },
    {
      "epoch": 0.531175981051149,
      "grad_norm": 0.0036124319303780794,
      "learning_rate": 0.1896491336437335,
      "loss": 0.6548,
      "num_input_tokens_seen": 22817152,
      "step": 16595
    },
    {
      "epoch": 0.5313360220216375,
      "grad_norm": 0.003839781740680337,
      "learning_rate": 0.18959232079709085,
      "loss": 0.5518,
      "num_input_tokens_seen": 22823984,
      "step": 16600
    },
    {
      "epoch": 0.5313360220216375,
      "eval_loss": 0.5643008947372437,
      "eval_runtime": 332.0575,
      "eval_samples_per_second": 41.818,
      "eval_steps_per_second": 20.909,
      "num_input_tokens_seen": 22823984,
      "step": 16600
    },
    {
      "epoch": 0.531496062992126,
      "grad_norm": 0.003171872813254595,
      "learning_rate": 0.18953550184481477,
      "loss": 0.5903,
      "num_input_tokens_seen": 22830864,
      "step": 16605
    },
    {
      "epoch": 0.5316561039626144,
      "grad_norm": 0.005606947001069784,
      "learning_rate": 0.18947867679566752,
      "loss": 0.5631,
      "num_input_tokens_seen": 22837648,
      "step": 16610
    },
    {
      "epoch": 0.5318161449331029,
      "grad_norm": 0.007874775677919388,
      "learning_rate": 0.18942184565841216,
      "loss": 0.7049,
      "num_input_tokens_seen": 22845184,
      "step": 16615
    },
    {
      "epoch": 0.5319761859035913,
      "grad_norm": 0.005281705409288406,
      "learning_rate": 0.18936500844181278,
      "loss": 0.539,
      "num_input_tokens_seen": 22851904,
      "step": 16620
    },
    {
      "epoch": 0.5321362268740798,
      "grad_norm": 0.00564982695505023,
      "learning_rate": 0.18930816515463436,
      "loss": 0.6236,
      "num_input_tokens_seen": 22858720,
      "step": 16625
    },
    {
      "epoch": 0.5322962678445682,
      "grad_norm": 0.006486238446086645,
      "learning_rate": 0.18925131580564297,
      "loss": 0.534,
      "num_input_tokens_seen": 22865648,
      "step": 16630
    },
    {
      "epoch": 0.5324563088150567,
      "grad_norm": 0.00616616103798151,
      "learning_rate": 0.1891944604036054,
      "loss": 0.5485,
      "num_input_tokens_seen": 22872512,
      "step": 16635
    },
    {
      "epoch": 0.5326163497855451,
      "grad_norm": 0.0052982838824391365,
      "learning_rate": 0.1891375989572895,
      "loss": 0.5175,
      "num_input_tokens_seen": 22879392,
      "step": 16640
    },
    {
      "epoch": 0.5327763907560336,
      "grad_norm": 0.004827129654586315,
      "learning_rate": 0.18908073147546398,
      "loss": 0.7026,
      "num_input_tokens_seen": 22886384,
      "step": 16645
    },
    {
      "epoch": 0.5329364317265219,
      "grad_norm": 0.004538722336292267,
      "learning_rate": 0.18902385796689858,
      "loss": 0.518,
      "num_input_tokens_seen": 22893184,
      "step": 16650
    },
    {
      "epoch": 0.5330964726970104,
      "grad_norm": 0.003410724690183997,
      "learning_rate": 0.18896697844036384,
      "loss": 0.5625,
      "num_input_tokens_seen": 22900128,
      "step": 16655
    },
    {
      "epoch": 0.5332565136674989,
      "grad_norm": 0.0045714909210801125,
      "learning_rate": 0.18891009290463137,
      "loss": 0.4807,
      "num_input_tokens_seen": 22906832,
      "step": 16660
    },
    {
      "epoch": 0.5334165546379873,
      "grad_norm": 0.005710143130272627,
      "learning_rate": 0.18885320136847353,
      "loss": 0.463,
      "num_input_tokens_seen": 22913456,
      "step": 16665
    },
    {
      "epoch": 0.5335765956084758,
      "grad_norm": 0.005344473756849766,
      "learning_rate": 0.1887963038406639,
      "loss": 0.5557,
      "num_input_tokens_seen": 22919984,
      "step": 16670
    },
    {
      "epoch": 0.5337366365789642,
      "grad_norm": 0.004380507860332727,
      "learning_rate": 0.18873940032997658,
      "loss": 0.3719,
      "num_input_tokens_seen": 22926704,
      "step": 16675
    },
    {
      "epoch": 0.5338966775494527,
      "grad_norm": 0.007938507944345474,
      "learning_rate": 0.18868249084518693,
      "loss": 0.8838,
      "num_input_tokens_seen": 22933616,
      "step": 16680
    },
    {
      "epoch": 0.5340567185199411,
      "grad_norm": 0.0065392241813242435,
      "learning_rate": 0.18862557539507102,
      "loss": 0.6679,
      "num_input_tokens_seen": 22940320,
      "step": 16685
    },
    {
      "epoch": 0.5342167594904296,
      "grad_norm": 0.004834280349314213,
      "learning_rate": 0.18856865398840605,
      "loss": 0.6333,
      "num_input_tokens_seen": 22947344,
      "step": 16690
    },
    {
      "epoch": 0.534376800460918,
      "grad_norm": 0.004794123582541943,
      "learning_rate": 0.18851172663396995,
      "loss": 0.5911,
      "num_input_tokens_seen": 22954560,
      "step": 16695
    },
    {
      "epoch": 0.5345368414314065,
      "grad_norm": 0.0027585255447775126,
      "learning_rate": 0.1884547933405416,
      "loss": 0.5056,
      "num_input_tokens_seen": 22961424,
      "step": 16700
    },
    {
      "epoch": 0.5346968824018948,
      "grad_norm": 0.0020150647033005953,
      "learning_rate": 0.1883978541169009,
      "loss": 0.5964,
      "num_input_tokens_seen": 22968432,
      "step": 16705
    },
    {
      "epoch": 0.5348569233723833,
      "grad_norm": 0.001390174962580204,
      "learning_rate": 0.18834090897182854,
      "loss": 0.4531,
      "num_input_tokens_seen": 22975072,
      "step": 16710
    },
    {
      "epoch": 0.5350169643428718,
      "grad_norm": 0.003536716802045703,
      "learning_rate": 0.1882839579141062,
      "loss": 0.5511,
      "num_input_tokens_seen": 22981728,
      "step": 16715
    },
    {
      "epoch": 0.5351770053133602,
      "grad_norm": 0.002104344544932246,
      "learning_rate": 0.18822700095251646,
      "loss": 0.5878,
      "num_input_tokens_seen": 22988704,
      "step": 16720
    },
    {
      "epoch": 0.5353370462838487,
      "grad_norm": 0.007145702373236418,
      "learning_rate": 0.18817003809584273,
      "loss": 0.5391,
      "num_input_tokens_seen": 22995888,
      "step": 16725
    },
    {
      "epoch": 0.5354970872543371,
      "grad_norm": 0.004683764185756445,
      "learning_rate": 0.1881130693528695,
      "loss": 0.7231,
      "num_input_tokens_seen": 23002544,
      "step": 16730
    },
    {
      "epoch": 0.5356571282248256,
      "grad_norm": 0.005271000321954489,
      "learning_rate": 0.18805609473238197,
      "loss": 0.5804,
      "num_input_tokens_seen": 23009248,
      "step": 16735
    },
    {
      "epoch": 0.535817169195314,
      "grad_norm": 0.005158748012036085,
      "learning_rate": 0.18799911424316643,
      "loss": 0.6032,
      "num_input_tokens_seen": 23016480,
      "step": 16740
    },
    {
      "epoch": 0.5359772101658025,
      "grad_norm": 0.009019174613058567,
      "learning_rate": 0.18794212789400994,
      "loss": 0.4101,
      "num_input_tokens_seen": 23022768,
      "step": 16745
    },
    {
      "epoch": 0.5361372511362908,
      "grad_norm": 0.07212245464324951,
      "learning_rate": 0.18788513569370052,
      "loss": 0.6411,
      "num_input_tokens_seen": 23029296,
      "step": 16750
    },
    {
      "epoch": 0.5362972921067793,
      "grad_norm": 0.003999003209173679,
      "learning_rate": 0.1878281376510271,
      "loss": 0.5127,
      "num_input_tokens_seen": 23035984,
      "step": 16755
    },
    {
      "epoch": 0.5364573330772677,
      "grad_norm": 0.005921054631471634,
      "learning_rate": 0.18777113377477941,
      "loss": 0.5116,
      "num_input_tokens_seen": 23042976,
      "step": 16760
    },
    {
      "epoch": 0.5366173740477562,
      "grad_norm": 0.004130522720515728,
      "learning_rate": 0.1877141240737483,
      "loss": 0.5953,
      "num_input_tokens_seen": 23049856,
      "step": 16765
    },
    {
      "epoch": 0.5367774150182447,
      "grad_norm": 0.006273815408349037,
      "learning_rate": 0.18765710855672527,
      "loss": 0.5899,
      "num_input_tokens_seen": 23056592,
      "step": 16770
    },
    {
      "epoch": 0.5369374559887331,
      "grad_norm": 0.003945595119148493,
      "learning_rate": 0.18760008723250288,
      "loss": 0.5071,
      "num_input_tokens_seen": 23063408,
      "step": 16775
    },
    {
      "epoch": 0.5370974969592216,
      "grad_norm": 0.003922643139958382,
      "learning_rate": 0.18754306010987457,
      "loss": 0.5278,
      "num_input_tokens_seen": 23070688,
      "step": 16780
    },
    {
      "epoch": 0.53725753792971,
      "grad_norm": 0.005512760020792484,
      "learning_rate": 0.18748602719763457,
      "loss": 0.7406,
      "num_input_tokens_seen": 23077936,
      "step": 16785
    },
    {
      "epoch": 0.5374175789001985,
      "grad_norm": 0.003690203418955207,
      "learning_rate": 0.18742898850457804,
      "loss": 0.4531,
      "num_input_tokens_seen": 23084768,
      "step": 16790
    },
    {
      "epoch": 0.5375776198706869,
      "grad_norm": 0.00675995135679841,
      "learning_rate": 0.1873719440395012,
      "loss": 0.4035,
      "num_input_tokens_seen": 23091728,
      "step": 16795
    },
    {
      "epoch": 0.5377376608411754,
      "grad_norm": 0.004455569665879011,
      "learning_rate": 0.1873148938112009,
      "loss": 0.4703,
      "num_input_tokens_seen": 23098384,
      "step": 16800
    },
    {
      "epoch": 0.5377376608411754,
      "eval_loss": 0.5626220107078552,
      "eval_runtime": 332.3422,
      "eval_samples_per_second": 41.782,
      "eval_steps_per_second": 20.891,
      "num_input_tokens_seen": 23098384,
      "step": 16800
    },
    {
      "epoch": 0.5378977018116637,
      "grad_norm": 0.005539002828299999,
      "learning_rate": 0.18725783782847508,
      "loss": 0.5168,
      "num_input_tokens_seen": 23105040,
      "step": 16805
    },
    {
      "epoch": 0.5380577427821522,
      "grad_norm": 0.004273901227861643,
      "learning_rate": 0.1872007761001224,
      "loss": 0.4646,
      "num_input_tokens_seen": 23111536,
      "step": 16810
    },
    {
      "epoch": 0.5382177837526406,
      "grad_norm": 0.004063890781253576,
      "learning_rate": 0.1871437086349426,
      "loss": 0.3615,
      "num_input_tokens_seen": 23118176,
      "step": 16815
    },
    {
      "epoch": 0.5383778247231291,
      "grad_norm": 0.0041692135855555534,
      "learning_rate": 0.18708663544173615,
      "loss": 0.555,
      "num_input_tokens_seen": 23124912,
      "step": 16820
    },
    {
      "epoch": 0.5385378656936176,
      "grad_norm": 0.005595782306045294,
      "learning_rate": 0.18702955652930442,
      "loss": 0.3341,
      "num_input_tokens_seen": 23131552,
      "step": 16825
    },
    {
      "epoch": 0.538697906664106,
      "grad_norm": 0.004227792378515005,
      "learning_rate": 0.18697247190644972,
      "loss": 0.5375,
      "num_input_tokens_seen": 23138592,
      "step": 16830
    },
    {
      "epoch": 0.5388579476345945,
      "grad_norm": 0.006718489807099104,
      "learning_rate": 0.18691538158197527,
      "loss": 0.6938,
      "num_input_tokens_seen": 23145680,
      "step": 16835
    },
    {
      "epoch": 0.5390179886050829,
      "grad_norm": 0.0032633966766297817,
      "learning_rate": 0.1868582855646851,
      "loss": 0.6454,
      "num_input_tokens_seen": 23152400,
      "step": 16840
    },
    {
      "epoch": 0.5391780295755714,
      "grad_norm": 0.00672169029712677,
      "learning_rate": 0.18680118386338404,
      "loss": 0.6182,
      "num_input_tokens_seen": 23159376,
      "step": 16845
    },
    {
      "epoch": 0.5393380705460598,
      "grad_norm": 0.006130763795226812,
      "learning_rate": 0.18674407648687794,
      "loss": 0.5396,
      "num_input_tokens_seen": 23165744,
      "step": 16850
    },
    {
      "epoch": 0.5394981115165483,
      "grad_norm": 0.008866192772984505,
      "learning_rate": 0.1866869634439736,
      "loss": 0.5904,
      "num_input_tokens_seen": 23172656,
      "step": 16855
    },
    {
      "epoch": 0.5396581524870366,
      "grad_norm": 0.004176685586571693,
      "learning_rate": 0.18662984474347838,
      "loss": 0.5852,
      "num_input_tokens_seen": 23179904,
      "step": 16860
    },
    {
      "epoch": 0.5398181934575251,
      "grad_norm": 0.005490063223987818,
      "learning_rate": 0.1865727203942008,
      "loss": 0.6269,
      "num_input_tokens_seen": 23186400,
      "step": 16865
    },
    {
      "epoch": 0.5399782344280136,
      "grad_norm": 0.005609054118394852,
      "learning_rate": 0.1865155904049501,
      "loss": 0.3964,
      "num_input_tokens_seen": 23193296,
      "step": 16870
    },
    {
      "epoch": 0.540138275398502,
      "grad_norm": 0.004662567749619484,
      "learning_rate": 0.1864584547845365,
      "loss": 0.6135,
      "num_input_tokens_seen": 23200816,
      "step": 16875
    },
    {
      "epoch": 0.5402983163689905,
      "grad_norm": 0.0074382382445037365,
      "learning_rate": 0.186401313541771,
      "loss": 0.643,
      "num_input_tokens_seen": 23207904,
      "step": 16880
    },
    {
      "epoch": 0.5404583573394789,
      "grad_norm": 0.003928410820662975,
      "learning_rate": 0.18634416668546552,
      "loss": 0.6473,
      "num_input_tokens_seen": 23214448,
      "step": 16885
    },
    {
      "epoch": 0.5406183983099674,
      "grad_norm": 0.007039809133857489,
      "learning_rate": 0.1862870142244328,
      "loss": 0.6254,
      "num_input_tokens_seen": 23221776,
      "step": 16890
    },
    {
      "epoch": 0.5407784392804558,
      "grad_norm": 0.005686315707862377,
      "learning_rate": 0.1862298561674865,
      "loss": 0.4725,
      "num_input_tokens_seen": 23228720,
      "step": 16895
    },
    {
      "epoch": 0.5409384802509443,
      "grad_norm": 0.007630737032741308,
      "learning_rate": 0.18617269252344104,
      "loss": 0.6978,
      "num_input_tokens_seen": 23235520,
      "step": 16900
    },
    {
      "epoch": 0.5410985212214326,
      "grad_norm": 0.003994850907474756,
      "learning_rate": 0.18611552330111186,
      "loss": 0.4398,
      "num_input_tokens_seen": 23242032,
      "step": 16905
    },
    {
      "epoch": 0.5412585621919211,
      "grad_norm": 0.0067250668071210384,
      "learning_rate": 0.18605834850931507,
      "loss": 0.606,
      "num_input_tokens_seen": 23248720,
      "step": 16910
    },
    {
      "epoch": 0.5414186031624095,
      "grad_norm": 0.0051690638065338135,
      "learning_rate": 0.18600116815686787,
      "loss": 0.695,
      "num_input_tokens_seen": 23255488,
      "step": 16915
    },
    {
      "epoch": 0.541578644132898,
      "grad_norm": 0.004631595686078072,
      "learning_rate": 0.1859439822525881,
      "loss": 0.4961,
      "num_input_tokens_seen": 23262336,
      "step": 16920
    },
    {
      "epoch": 0.5417386851033865,
      "grad_norm": 0.005767499096691608,
      "learning_rate": 0.18588679080529455,
      "loss": 0.467,
      "num_input_tokens_seen": 23268960,
      "step": 16925
    },
    {
      "epoch": 0.5418987260738749,
      "grad_norm": 0.004899156279861927,
      "learning_rate": 0.1858295938238069,
      "loss": 0.3955,
      "num_input_tokens_seen": 23275728,
      "step": 16930
    },
    {
      "epoch": 0.5420587670443634,
      "grad_norm": 0.008098915219306946,
      "learning_rate": 0.18577239131694562,
      "loss": 0.535,
      "num_input_tokens_seen": 23282496,
      "step": 16935
    },
    {
      "epoch": 0.5422188080148518,
      "grad_norm": 0.004445630591362715,
      "learning_rate": 0.18571518329353204,
      "loss": 0.5636,
      "num_input_tokens_seen": 23289520,
      "step": 16940
    },
    {
      "epoch": 0.5423788489853403,
      "grad_norm": 0.0035304450429975986,
      "learning_rate": 0.18565796976238838,
      "loss": 0.4932,
      "num_input_tokens_seen": 23296336,
      "step": 16945
    },
    {
      "epoch": 0.5425388899558287,
      "grad_norm": 0.002871801843866706,
      "learning_rate": 0.18560075073233764,
      "loss": 0.5056,
      "num_input_tokens_seen": 23303264,
      "step": 16950
    },
    {
      "epoch": 0.5426989309263172,
      "grad_norm": 0.0033201503101736307,
      "learning_rate": 0.18554352621220377,
      "loss": 0.445,
      "num_input_tokens_seen": 23310224,
      "step": 16955
    },
    {
      "epoch": 0.5428589718968055,
      "grad_norm": 0.004063247237354517,
      "learning_rate": 0.18548629621081153,
      "loss": 0.5679,
      "num_input_tokens_seen": 23317104,
      "step": 16960
    },
    {
      "epoch": 0.543019012867294,
      "grad_norm": 0.005825176369398832,
      "learning_rate": 0.18542906073698645,
      "loss": 0.6785,
      "num_input_tokens_seen": 23323712,
      "step": 16965
    },
    {
      "epoch": 0.5431790538377824,
      "grad_norm": 0.0038202928844839334,
      "learning_rate": 0.18537181979955494,
      "loss": 0.493,
      "num_input_tokens_seen": 23330464,
      "step": 16970
    },
    {
      "epoch": 0.5433390948082709,
      "grad_norm": 0.0043200780637562275,
      "learning_rate": 0.18531457340734434,
      "loss": 0.5506,
      "num_input_tokens_seen": 23337360,
      "step": 16975
    },
    {
      "epoch": 0.5434991357787594,
      "grad_norm": 0.003217332996428013,
      "learning_rate": 0.1852573215691827,
      "loss": 0.5413,
      "num_input_tokens_seen": 23344416,
      "step": 16980
    },
    {
      "epoch": 0.5436591767492478,
      "grad_norm": 0.005568480119109154,
      "learning_rate": 0.18520006429389904,
      "loss": 0.7814,
      "num_input_tokens_seen": 23351200,
      "step": 16985
    },
    {
      "epoch": 0.5438192177197363,
      "grad_norm": 0.004354110918939114,
      "learning_rate": 0.1851428015903231,
      "loss": 0.6817,
      "num_input_tokens_seen": 23357952,
      "step": 16990
    },
    {
      "epoch": 0.5439792586902247,
      "grad_norm": 0.0032723594922572374,
      "learning_rate": 0.1850855334672855,
      "loss": 0.5678,
      "num_input_tokens_seen": 23364448,
      "step": 16995
    },
    {
      "epoch": 0.5441392996607132,
      "grad_norm": 0.0036621415056288242,
      "learning_rate": 0.1850282599336178,
      "loss": 0.5579,
      "num_input_tokens_seen": 23371136,
      "step": 17000
    },
    {
      "epoch": 0.5441392996607132,
      "eval_loss": 0.5628642439842224,
      "eval_runtime": 331.7357,
      "eval_samples_per_second": 41.859,
      "eval_steps_per_second": 20.929,
      "num_input_tokens_seen": 23371136,
      "step": 17000
    },
    {
      "epoch": 0.5442993406312016,
      "grad_norm": 0.005958563182502985,
      "learning_rate": 0.18497098099815215,
      "loss": 0.6118,
      "num_input_tokens_seen": 23378064,
      "step": 17005
    },
    {
      "epoch": 0.54445938160169,
      "grad_norm": 0.004129853565245867,
      "learning_rate": 0.18491369666972174,
      "loss": 0.6549,
      "num_input_tokens_seen": 23385152,
      "step": 17010
    },
    {
      "epoch": 0.5446194225721784,
      "grad_norm": 0.006397004704922438,
      "learning_rate": 0.1848564069571606,
      "loss": 0.8273,
      "num_input_tokens_seen": 23392304,
      "step": 17015
    },
    {
      "epoch": 0.5447794635426669,
      "grad_norm": 0.00547173572704196,
      "learning_rate": 0.18479911186930348,
      "loss": 0.6102,
      "num_input_tokens_seen": 23399280,
      "step": 17020
    },
    {
      "epoch": 0.5449395045131553,
      "grad_norm": 0.005171711556613445,
      "learning_rate": 0.18474181141498597,
      "loss": 0.5622,
      "num_input_tokens_seen": 23406592,
      "step": 17025
    },
    {
      "epoch": 0.5450995454836438,
      "grad_norm": 0.003895415924489498,
      "learning_rate": 0.18468450560304453,
      "loss": 0.5916,
      "num_input_tokens_seen": 23413712,
      "step": 17030
    },
    {
      "epoch": 0.5452595864541323,
      "grad_norm": 0.002985424594953656,
      "learning_rate": 0.1846271944423165,
      "loss": 0.5844,
      "num_input_tokens_seen": 23420592,
      "step": 17035
    },
    {
      "epoch": 0.5454196274246207,
      "grad_norm": 0.006731249392032623,
      "learning_rate": 0.18456987794163993,
      "loss": 0.6215,
      "num_input_tokens_seen": 23427664,
      "step": 17040
    },
    {
      "epoch": 0.5455796683951092,
      "grad_norm": 0.007219147868454456,
      "learning_rate": 0.18451255610985373,
      "loss": 0.4476,
      "num_input_tokens_seen": 23434864,
      "step": 17045
    },
    {
      "epoch": 0.5457397093655976,
      "grad_norm": 0.00500341085717082,
      "learning_rate": 0.18445522895579766,
      "loss": 0.5265,
      "num_input_tokens_seen": 23441760,
      "step": 17050
    },
    {
      "epoch": 0.5458997503360861,
      "grad_norm": 0.007685074582695961,
      "learning_rate": 0.1843978964883123,
      "loss": 0.5907,
      "num_input_tokens_seen": 23448592,
      "step": 17055
    },
    {
      "epoch": 0.5460597913065744,
      "grad_norm": 0.00456651346758008,
      "learning_rate": 0.18434055871623906,
      "loss": 0.58,
      "num_input_tokens_seen": 23455440,
      "step": 17060
    },
    {
      "epoch": 0.5462198322770629,
      "grad_norm": 0.006563276052474976,
      "learning_rate": 0.18428321564842007,
      "loss": 0.5478,
      "num_input_tokens_seen": 23462400,
      "step": 17065
    },
    {
      "epoch": 0.5463798732475513,
      "grad_norm": 0.004046137910336256,
      "learning_rate": 0.18422586729369841,
      "loss": 0.5036,
      "num_input_tokens_seen": 23469040,
      "step": 17070
    },
    {
      "epoch": 0.5465399142180398,
      "grad_norm": 0.0071657150983810425,
      "learning_rate": 0.1841685136609179,
      "loss": 0.6182,
      "num_input_tokens_seen": 23475568,
      "step": 17075
    },
    {
      "epoch": 0.5466999551885282,
      "grad_norm": 0.008001906797289848,
      "learning_rate": 0.18411115475892326,
      "loss": 0.5095,
      "num_input_tokens_seen": 23482128,
      "step": 17080
    },
    {
      "epoch": 0.5468599961590167,
      "grad_norm": 0.0033480841666460037,
      "learning_rate": 0.18405379059655982,
      "loss": 0.6785,
      "num_input_tokens_seen": 23489216,
      "step": 17085
    },
    {
      "epoch": 0.5470200371295052,
      "grad_norm": 0.006251039449125528,
      "learning_rate": 0.1839964211826739,
      "loss": 0.5793,
      "num_input_tokens_seen": 23495920,
      "step": 17090
    },
    {
      "epoch": 0.5471800780999936,
      "grad_norm": 0.008374329656362534,
      "learning_rate": 0.18393904652611265,
      "loss": 0.6793,
      "num_input_tokens_seen": 23502640,
      "step": 17095
    },
    {
      "epoch": 0.5473401190704821,
      "grad_norm": 0.0033523610327392817,
      "learning_rate": 0.18388166663572392,
      "loss": 0.546,
      "num_input_tokens_seen": 23509296,
      "step": 17100
    },
    {
      "epoch": 0.5475001600409705,
      "grad_norm": 0.0051688761450350285,
      "learning_rate": 0.18382428152035643,
      "loss": 0.5554,
      "num_input_tokens_seen": 23515856,
      "step": 17105
    },
    {
      "epoch": 0.547660201011459,
      "grad_norm": 0.003063978860154748,
      "learning_rate": 0.1837668911888596,
      "loss": 0.3532,
      "num_input_tokens_seen": 23522112,
      "step": 17110
    },
    {
      "epoch": 0.5478202419819473,
      "grad_norm": 0.0041051944717764854,
      "learning_rate": 0.18370949565008388,
      "loss": 0.364,
      "num_input_tokens_seen": 23528736,
      "step": 17115
    },
    {
      "epoch": 0.5479802829524358,
      "grad_norm": 0.004860828630626202,
      "learning_rate": 0.1836520949128803,
      "loss": 0.6164,
      "num_input_tokens_seen": 23535552,
      "step": 17120
    },
    {
      "epoch": 0.5481403239229242,
      "grad_norm": 0.0047182925045490265,
      "learning_rate": 0.18359468898610076,
      "loss": 0.5871,
      "num_input_tokens_seen": 23542448,
      "step": 17125
    },
    {
      "epoch": 0.5483003648934127,
      "grad_norm": 0.004503998439759016,
      "learning_rate": 0.18353727787859797,
      "loss": 0.632,
      "num_input_tokens_seen": 23549632,
      "step": 17130
    },
    {
      "epoch": 0.5484604058639012,
      "grad_norm": 0.00425378605723381,
      "learning_rate": 0.18347986159922552,
      "loss": 0.6426,
      "num_input_tokens_seen": 23556272,
      "step": 17135
    },
    {
      "epoch": 0.5486204468343896,
      "grad_norm": 0.006262375507503748,
      "learning_rate": 0.1834224401568377,
      "loss": 0.5512,
      "num_input_tokens_seen": 23563440,
      "step": 17140
    },
    {
      "epoch": 0.5487804878048781,
      "grad_norm": 0.003703662659972906,
      "learning_rate": 0.1833650135602896,
      "loss": 0.4384,
      "num_input_tokens_seen": 23570528,
      "step": 17145
    },
    {
      "epoch": 0.5489405287753665,
      "grad_norm": 0.004475635942071676,
      "learning_rate": 0.18330758181843707,
      "loss": 0.5234,
      "num_input_tokens_seen": 23577264,
      "step": 17150
    },
    {
      "epoch": 0.549100569745855,
      "grad_norm": 0.003278655232861638,
      "learning_rate": 0.18325014494013686,
      "loss": 0.492,
      "num_input_tokens_seen": 23584384,
      "step": 17155
    },
    {
      "epoch": 0.5492606107163434,
      "grad_norm": 0.0034007092472165823,
      "learning_rate": 0.18319270293424647,
      "loss": 0.5538,
      "num_input_tokens_seen": 23591600,
      "step": 17160
    },
    {
      "epoch": 0.5494206516868319,
      "grad_norm": 0.004250384867191315,
      "learning_rate": 0.18313525580962417,
      "loss": 0.7508,
      "num_input_tokens_seen": 23598944,
      "step": 17165
    },
    {
      "epoch": 0.5495806926573202,
      "grad_norm": 0.005308874417096376,
      "learning_rate": 0.18307780357512896,
      "loss": 0.4648,
      "num_input_tokens_seen": 23606304,
      "step": 17170
    },
    {
      "epoch": 0.5497407336278087,
      "grad_norm": 0.004266240168362856,
      "learning_rate": 0.1830203462396208,
      "loss": 0.5509,
      "num_input_tokens_seen": 23612992,
      "step": 17175
    },
    {
      "epoch": 0.5499007745982971,
      "grad_norm": 0.003997253254055977,
      "learning_rate": 0.18296288381196033,
      "loss": 0.5105,
      "num_input_tokens_seen": 23620176,
      "step": 17180
    },
    {
      "epoch": 0.5500608155687856,
      "grad_norm": 0.005758189596235752,
      "learning_rate": 0.1829054163010089,
      "loss": 0.5713,
      "num_input_tokens_seen": 23627264,
      "step": 17185
    },
    {
      "epoch": 0.5502208565392741,
      "grad_norm": 0.004488000646233559,
      "learning_rate": 0.18284794371562874,
      "loss": 0.627,
      "num_input_tokens_seen": 23634288,
      "step": 17190
    },
    {
      "epoch": 0.5503808975097625,
      "grad_norm": 0.001600253046490252,
      "learning_rate": 0.18279046606468288,
      "loss": 0.6157,
      "num_input_tokens_seen": 23641008,
      "step": 17195
    },
    {
      "epoch": 0.550540938480251,
      "grad_norm": 0.003540879348292947,
      "learning_rate": 0.1827329833570351,
      "loss": 0.4205,
      "num_input_tokens_seen": 23647856,
      "step": 17200
    },
    {
      "epoch": 0.550540938480251,
      "eval_loss": 0.5602400302886963,
      "eval_runtime": 332.0303,
      "eval_samples_per_second": 41.821,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 23647856,
      "step": 17200
    },
    {
      "epoch": 0.5507009794507394,
      "grad_norm": 0.010169092565774918,
      "learning_rate": 0.18267549560154991,
      "loss": 0.5211,
      "num_input_tokens_seen": 23654816,
      "step": 17205
    },
    {
      "epoch": 0.5508610204212279,
      "grad_norm": 0.004575274419039488,
      "learning_rate": 0.18261800280709267,
      "loss": 0.4411,
      "num_input_tokens_seen": 23661936,
      "step": 17210
    },
    {
      "epoch": 0.5510210613917162,
      "grad_norm": 0.004546555690467358,
      "learning_rate": 0.18256050498252957,
      "loss": 0.5452,
      "num_input_tokens_seen": 23668608,
      "step": 17215
    },
    {
      "epoch": 0.5511811023622047,
      "grad_norm": 0.0049612498842179775,
      "learning_rate": 0.18250300213672735,
      "loss": 0.7702,
      "num_input_tokens_seen": 23675600,
      "step": 17220
    },
    {
      "epoch": 0.5513411433326931,
      "grad_norm": 0.007412073668092489,
      "learning_rate": 0.18244549427855378,
      "loss": 0.5773,
      "num_input_tokens_seen": 23682240,
      "step": 17225
    },
    {
      "epoch": 0.5515011843031816,
      "grad_norm": 0.006060408893972635,
      "learning_rate": 0.1823879814168772,
      "loss": 0.5985,
      "num_input_tokens_seen": 23688816,
      "step": 17230
    },
    {
      "epoch": 0.55166122527367,
      "grad_norm": 0.006998094264417887,
      "learning_rate": 0.18233046356056692,
      "loss": 1.0117,
      "num_input_tokens_seen": 23695536,
      "step": 17235
    },
    {
      "epoch": 0.5518212662441585,
      "grad_norm": 0.004837548825889826,
      "learning_rate": 0.18227294071849284,
      "loss": 0.714,
      "num_input_tokens_seen": 23702272,
      "step": 17240
    },
    {
      "epoch": 0.551981307214647,
      "grad_norm": 0.0036296024918556213,
      "learning_rate": 0.18221541289952578,
      "loss": 0.5066,
      "num_input_tokens_seen": 23708832,
      "step": 17245
    },
    {
      "epoch": 0.5521413481851354,
      "grad_norm": 0.004218557383865118,
      "learning_rate": 0.18215788011253717,
      "loss": 0.535,
      "num_input_tokens_seen": 23715616,
      "step": 17250
    },
    {
      "epoch": 0.5523013891556239,
      "grad_norm": 0.002795570995658636,
      "learning_rate": 0.18210034236639935,
      "loss": 0.6012,
      "num_input_tokens_seen": 23722304,
      "step": 17255
    },
    {
      "epoch": 0.5524614301261123,
      "grad_norm": 0.0034887981601059437,
      "learning_rate": 0.1820427996699853,
      "loss": 0.5564,
      "num_input_tokens_seen": 23728752,
      "step": 17260
    },
    {
      "epoch": 0.5526214710966008,
      "grad_norm": 0.005115166772156954,
      "learning_rate": 0.1819852520321689,
      "loss": 0.5296,
      "num_input_tokens_seen": 23735472,
      "step": 17265
    },
    {
      "epoch": 0.5527815120670891,
      "grad_norm": 0.0038836197927594185,
      "learning_rate": 0.18192769946182466,
      "loss": 0.5108,
      "num_input_tokens_seen": 23742464,
      "step": 17270
    },
    {
      "epoch": 0.5529415530375776,
      "grad_norm": 0.005835091695189476,
      "learning_rate": 0.18187014196782794,
      "loss": 0.5335,
      "num_input_tokens_seen": 23749424,
      "step": 17275
    },
    {
      "epoch": 0.553101594008066,
      "grad_norm": 0.004127666354179382,
      "learning_rate": 0.18181257955905486,
      "loss": 0.6075,
      "num_input_tokens_seen": 23756304,
      "step": 17280
    },
    {
      "epoch": 0.5532616349785545,
      "grad_norm": 0.005853566806763411,
      "learning_rate": 0.18175501224438217,
      "loss": 0.5055,
      "num_input_tokens_seen": 23762976,
      "step": 17285
    },
    {
      "epoch": 0.5534216759490429,
      "grad_norm": 0.004537633620202541,
      "learning_rate": 0.18169744003268756,
      "loss": 0.7071,
      "num_input_tokens_seen": 23769808,
      "step": 17290
    },
    {
      "epoch": 0.5535817169195314,
      "grad_norm": 0.004377163015305996,
      "learning_rate": 0.18163986293284937,
      "loss": 0.544,
      "num_input_tokens_seen": 23776688,
      "step": 17295
    },
    {
      "epoch": 0.5537417578900199,
      "grad_norm": 0.00363192823715508,
      "learning_rate": 0.18158228095374673,
      "loss": 0.4609,
      "num_input_tokens_seen": 23783952,
      "step": 17300
    },
    {
      "epoch": 0.5539017988605083,
      "grad_norm": 0.006921117193996906,
      "learning_rate": 0.18152469410425945,
      "loss": 0.4939,
      "num_input_tokens_seen": 23790928,
      "step": 17305
    },
    {
      "epoch": 0.5540618398309968,
      "grad_norm": 0.007376688066869974,
      "learning_rate": 0.18146710239326813,
      "loss": 0.4308,
      "num_input_tokens_seen": 23797888,
      "step": 17310
    },
    {
      "epoch": 0.5542218808014852,
      "grad_norm": 0.003087697783485055,
      "learning_rate": 0.18140950582965423,
      "loss": 0.5426,
      "num_input_tokens_seen": 23804544,
      "step": 17315
    },
    {
      "epoch": 0.5543819217719737,
      "grad_norm": 0.00449490686878562,
      "learning_rate": 0.1813519044222998,
      "loss": 0.397,
      "num_input_tokens_seen": 23811376,
      "step": 17320
    },
    {
      "epoch": 0.554541962742462,
      "grad_norm": 0.0032055883202701807,
      "learning_rate": 0.18129429818008772,
      "loss": 0.4972,
      "num_input_tokens_seen": 23818016,
      "step": 17325
    },
    {
      "epoch": 0.5547020037129505,
      "grad_norm": 0.006438849959522486,
      "learning_rate": 0.18123668711190163,
      "loss": 0.5599,
      "num_input_tokens_seen": 23824752,
      "step": 17330
    },
    {
      "epoch": 0.5548620446834389,
      "grad_norm": 0.0070387269370257854,
      "learning_rate": 0.18117907122662583,
      "loss": 0.7032,
      "num_input_tokens_seen": 23831632,
      "step": 17335
    },
    {
      "epoch": 0.5550220856539274,
      "grad_norm": 0.004021137952804565,
      "learning_rate": 0.1811214505331454,
      "loss": 0.4871,
      "num_input_tokens_seen": 23838240,
      "step": 17340
    },
    {
      "epoch": 0.5551821266244159,
      "grad_norm": 0.003184056142345071,
      "learning_rate": 0.1810638250403462,
      "loss": 0.5512,
      "num_input_tokens_seen": 23845424,
      "step": 17345
    },
    {
      "epoch": 0.5553421675949043,
      "grad_norm": 0.00509802857413888,
      "learning_rate": 0.1810061947571148,
      "loss": 0.5579,
      "num_input_tokens_seen": 23851856,
      "step": 17350
    },
    {
      "epoch": 0.5555022085653928,
      "grad_norm": 0.004948966205120087,
      "learning_rate": 0.1809485596923385,
      "loss": 0.6386,
      "num_input_tokens_seen": 23858624,
      "step": 17355
    },
    {
      "epoch": 0.5556622495358812,
      "grad_norm": 0.004775763489305973,
      "learning_rate": 0.18089091985490546,
      "loss": 0.5064,
      "num_input_tokens_seen": 23865776,
      "step": 17360
    },
    {
      "epoch": 0.5558222905063697,
      "grad_norm": 0.0030047364998608828,
      "learning_rate": 0.18083327525370432,
      "loss": 0.4027,
      "num_input_tokens_seen": 23872160,
      "step": 17365
    },
    {
      "epoch": 0.555982331476858,
      "grad_norm": 0.006932622287422419,
      "learning_rate": 0.18077562589762464,
      "loss": 0.6431,
      "num_input_tokens_seen": 23879152,
      "step": 17370
    },
    {
      "epoch": 0.5561423724473465,
      "grad_norm": 0.0032097860239446163,
      "learning_rate": 0.1807179717955567,
      "loss": 0.6197,
      "num_input_tokens_seen": 23886368,
      "step": 17375
    },
    {
      "epoch": 0.5563024134178349,
      "grad_norm": 0.0041842982172966,
      "learning_rate": 0.1806603129563915,
      "loss": 0.5893,
      "num_input_tokens_seen": 23893088,
      "step": 17380
    },
    {
      "epoch": 0.5564624543883234,
      "grad_norm": 0.0034841636661440134,
      "learning_rate": 0.1806026493890208,
      "loss": 0.5781,
      "num_input_tokens_seen": 23899728,
      "step": 17385
    },
    {
      "epoch": 0.5566224953588118,
      "grad_norm": 0.0054010008461773396,
      "learning_rate": 0.18054498110233688,
      "loss": 0.5905,
      "num_input_tokens_seen": 23906896,
      "step": 17390
    },
    {
      "epoch": 0.5567825363293003,
      "grad_norm": 0.0033525400795042515,
      "learning_rate": 0.1804873081052331,
      "loss": 0.4753,
      "num_input_tokens_seen": 23913776,
      "step": 17395
    },
    {
      "epoch": 0.5569425772997888,
      "grad_norm": 0.00405534915626049,
      "learning_rate": 0.18042963040660326,
      "loss": 0.5438,
      "num_input_tokens_seen": 23921008,
      "step": 17400
    },
    {
      "epoch": 0.5569425772997888,
      "eval_loss": 0.5585953593254089,
      "eval_runtime": 331.6056,
      "eval_samples_per_second": 41.875,
      "eval_steps_per_second": 20.938,
      "num_input_tokens_seen": 23921008,
      "step": 17400
    },
    {
      "epoch": 0.5571026182702772,
      "grad_norm": 0.00484848627820611,
      "learning_rate": 0.180371948015342,
      "loss": 0.6068,
      "num_input_tokens_seen": 23928048,
      "step": 17405
    },
    {
      "epoch": 0.5572626592407657,
      "grad_norm": 0.005739890970289707,
      "learning_rate": 0.18031426094034472,
      "loss": 0.5413,
      "num_input_tokens_seen": 23934816,
      "step": 17410
    },
    {
      "epoch": 0.5574227002112541,
      "grad_norm": 0.007751586381345987,
      "learning_rate": 0.18025656919050737,
      "loss": 0.6792,
      "num_input_tokens_seen": 23941632,
      "step": 17415
    },
    {
      "epoch": 0.5575827411817426,
      "grad_norm": 0.007175160106271505,
      "learning_rate": 0.18019887277472688,
      "loss": 0.696,
      "num_input_tokens_seen": 23948416,
      "step": 17420
    },
    {
      "epoch": 0.5577427821522309,
      "grad_norm": 0.0021795586217194796,
      "learning_rate": 0.18014117170190067,
      "loss": 0.5058,
      "num_input_tokens_seen": 23955104,
      "step": 17425
    },
    {
      "epoch": 0.5579028231227194,
      "grad_norm": 0.013102587312459946,
      "learning_rate": 0.18008346598092703,
      "loss": 0.7466,
      "num_input_tokens_seen": 23961664,
      "step": 17430
    },
    {
      "epoch": 0.5580628640932078,
      "grad_norm": 0.006714201532304287,
      "learning_rate": 0.18002575562070489,
      "loss": 0.6927,
      "num_input_tokens_seen": 23968352,
      "step": 17435
    },
    {
      "epoch": 0.5582229050636963,
      "grad_norm": 0.002694169757887721,
      "learning_rate": 0.1799680406301339,
      "loss": 0.4373,
      "num_input_tokens_seen": 23975344,
      "step": 17440
    },
    {
      "epoch": 0.5583829460341847,
      "grad_norm": 0.005579362623393536,
      "learning_rate": 0.17991032101811447,
      "loss": 0.5757,
      "num_input_tokens_seen": 23981904,
      "step": 17445
    },
    {
      "epoch": 0.5585429870046732,
      "grad_norm": 0.0033328572753816843,
      "learning_rate": 0.1798525967935476,
      "loss": 0.5617,
      "num_input_tokens_seen": 23988912,
      "step": 17450
    },
    {
      "epoch": 0.5587030279751617,
      "grad_norm": 0.006104802247136831,
      "learning_rate": 0.17979486796533517,
      "loss": 0.6145,
      "num_input_tokens_seen": 23995680,
      "step": 17455
    },
    {
      "epoch": 0.5588630689456501,
      "grad_norm": 0.0065618036314845085,
      "learning_rate": 0.1797371345423797,
      "loss": 0.606,
      "num_input_tokens_seen": 24002480,
      "step": 17460
    },
    {
      "epoch": 0.5590231099161386,
      "grad_norm": 0.004141718614846468,
      "learning_rate": 0.17967939653358436,
      "loss": 0.5122,
      "num_input_tokens_seen": 24009360,
      "step": 17465
    },
    {
      "epoch": 0.559183150886627,
      "grad_norm": 0.004218878224492073,
      "learning_rate": 0.17962165394785315,
      "loss": 0.5036,
      "num_input_tokens_seen": 24015856,
      "step": 17470
    },
    {
      "epoch": 0.5593431918571155,
      "grad_norm": 0.005381959956139326,
      "learning_rate": 0.17956390679409057,
      "loss": 0.8028,
      "num_input_tokens_seen": 24022736,
      "step": 17475
    },
    {
      "epoch": 0.5595032328276038,
      "grad_norm": 0.0033223109785467386,
      "learning_rate": 0.1795061550812021,
      "loss": 0.6866,
      "num_input_tokens_seen": 24029728,
      "step": 17480
    },
    {
      "epoch": 0.5596632737980923,
      "grad_norm": 0.00534726632758975,
      "learning_rate": 0.1794483988180937,
      "loss": 0.652,
      "num_input_tokens_seen": 24036608,
      "step": 17485
    },
    {
      "epoch": 0.5598233147685807,
      "grad_norm": 0.004496320150792599,
      "learning_rate": 0.17939063801367214,
      "loss": 0.5536,
      "num_input_tokens_seen": 24043408,
      "step": 17490
    },
    {
      "epoch": 0.5599833557390692,
      "grad_norm": 0.005940320901572704,
      "learning_rate": 0.17933287267684483,
      "loss": 0.5257,
      "num_input_tokens_seen": 24050720,
      "step": 17495
    },
    {
      "epoch": 0.5601433967095576,
      "grad_norm": 0.004043763503432274,
      "learning_rate": 0.17927510281651995,
      "loss": 0.546,
      "num_input_tokens_seen": 24057616,
      "step": 17500
    },
    {
      "epoch": 0.5603034376800461,
      "grad_norm": 0.0036211193073540926,
      "learning_rate": 0.17921732844160634,
      "loss": 0.4953,
      "num_input_tokens_seen": 24064944,
      "step": 17505
    },
    {
      "epoch": 0.5604634786505346,
      "grad_norm": 0.0061036585830152035,
      "learning_rate": 0.17915954956101351,
      "loss": 0.4983,
      "num_input_tokens_seen": 24072128,
      "step": 17510
    },
    {
      "epoch": 0.560623519621023,
      "grad_norm": 0.0040532061830163,
      "learning_rate": 0.17910176618365165,
      "loss": 0.4934,
      "num_input_tokens_seen": 24079136,
      "step": 17515
    },
    {
      "epoch": 0.5607835605915115,
      "grad_norm": 0.004232516046613455,
      "learning_rate": 0.17904397831843177,
      "loss": 0.4699,
      "num_input_tokens_seen": 24085728,
      "step": 17520
    },
    {
      "epoch": 0.5609436015619998,
      "grad_norm": 0.0039073992520570755,
      "learning_rate": 0.17898618597426547,
      "loss": 0.5421,
      "num_input_tokens_seen": 24092768,
      "step": 17525
    },
    {
      "epoch": 0.5611036425324883,
      "grad_norm": 0.004555227234959602,
      "learning_rate": 0.17892838916006495,
      "loss": 0.6218,
      "num_input_tokens_seen": 24100224,
      "step": 17530
    },
    {
      "epoch": 0.5612636835029767,
      "grad_norm": 0.006057188380509615,
      "learning_rate": 0.17887058788474333,
      "loss": 0.5607,
      "num_input_tokens_seen": 24106928,
      "step": 17535
    },
    {
      "epoch": 0.5614237244734652,
      "grad_norm": 0.00509682297706604,
      "learning_rate": 0.17881278215721427,
      "loss": 0.6676,
      "num_input_tokens_seen": 24113744,
      "step": 17540
    },
    {
      "epoch": 0.5615837654439536,
      "grad_norm": 0.006176256574690342,
      "learning_rate": 0.1787549719863921,
      "loss": 0.376,
      "num_input_tokens_seen": 24120832,
      "step": 17545
    },
    {
      "epoch": 0.5617438064144421,
      "grad_norm": 0.003197673475369811,
      "learning_rate": 0.17869715738119188,
      "loss": 0.4032,
      "num_input_tokens_seen": 24127376,
      "step": 17550
    },
    {
      "epoch": 0.5619038473849305,
      "grad_norm": 0.0039826855063438416,
      "learning_rate": 0.17863933835052936,
      "loss": 0.5492,
      "num_input_tokens_seen": 24134032,
      "step": 17555
    },
    {
      "epoch": 0.562063888355419,
      "grad_norm": 0.0049462029710412025,
      "learning_rate": 0.17858151490332097,
      "loss": 0.6135,
      "num_input_tokens_seen": 24140496,
      "step": 17560
    },
    {
      "epoch": 0.5622239293259075,
      "grad_norm": 0.003043057397007942,
      "learning_rate": 0.17852368704848381,
      "loss": 0.4312,
      "num_input_tokens_seen": 24147072,
      "step": 17565
    },
    {
      "epoch": 0.5623839702963959,
      "grad_norm": 0.004288221709430218,
      "learning_rate": 0.17846585479493565,
      "loss": 0.5034,
      "num_input_tokens_seen": 24153328,
      "step": 17570
    },
    {
      "epoch": 0.5625440112668844,
      "grad_norm": 0.0048287734389305115,
      "learning_rate": 0.178408018151595,
      "loss": 0.5923,
      "num_input_tokens_seen": 24160048,
      "step": 17575
    },
    {
      "epoch": 0.5627040522373727,
      "grad_norm": 0.007254378870129585,
      "learning_rate": 0.17835017712738085,
      "loss": 0.6019,
      "num_input_tokens_seen": 24166992,
      "step": 17580
    },
    {
      "epoch": 0.5628640932078612,
      "grad_norm": 0.005147505085915327,
      "learning_rate": 0.17829233173121323,
      "loss": 0.6931,
      "num_input_tokens_seen": 24173840,
      "step": 17585
    },
    {
      "epoch": 0.5630241341783496,
      "grad_norm": 0.00700353505089879,
      "learning_rate": 0.17823448197201244,
      "loss": 0.4296,
      "num_input_tokens_seen": 24180800,
      "step": 17590
    },
    {
      "epoch": 0.5631841751488381,
      "grad_norm": 0.006371034774929285,
      "learning_rate": 0.1781766278586997,
      "loss": 0.3862,
      "num_input_tokens_seen": 24187824,
      "step": 17595
    },
    {
      "epoch": 0.5633442161193265,
      "grad_norm": 0.009305068291723728,
      "learning_rate": 0.1781187694001969,
      "loss": 0.4923,
      "num_input_tokens_seen": 24194480,
      "step": 17600
    },
    {
      "epoch": 0.5633442161193265,
      "eval_loss": 0.5528927445411682,
      "eval_runtime": 332.5779,
      "eval_samples_per_second": 41.753,
      "eval_steps_per_second": 20.876,
      "num_input_tokens_seen": 24194480,
      "step": 17600
    },
    {
      "epoch": 0.563504257089815,
      "grad_norm": 0.002789960475638509,
      "learning_rate": 0.1780609066054265,
      "loss": 0.6442,
      "num_input_tokens_seen": 24201216,
      "step": 17605
    },
    {
      "epoch": 0.5636642980603035,
      "grad_norm": 0.006290818098932505,
      "learning_rate": 0.17800303948331164,
      "loss": 0.6097,
      "num_input_tokens_seen": 24208272,
      "step": 17610
    },
    {
      "epoch": 0.5638243390307919,
      "grad_norm": 0.0064617786556482315,
      "learning_rate": 0.1779451680427762,
      "loss": 0.6157,
      "num_input_tokens_seen": 24215040,
      "step": 17615
    },
    {
      "epoch": 0.5639843800012804,
      "grad_norm": 0.0024732945021241903,
      "learning_rate": 0.17788729229274464,
      "loss": 0.3926,
      "num_input_tokens_seen": 24222224,
      "step": 17620
    },
    {
      "epoch": 0.5641444209717688,
      "grad_norm": 0.004694480914622545,
      "learning_rate": 0.17782941224214222,
      "loss": 0.4798,
      "num_input_tokens_seen": 24228864,
      "step": 17625
    },
    {
      "epoch": 0.5643044619422573,
      "grad_norm": 0.006477726623415947,
      "learning_rate": 0.17777152789989464,
      "loss": 0.8051,
      "num_input_tokens_seen": 24235568,
      "step": 17630
    },
    {
      "epoch": 0.5644645029127456,
      "grad_norm": 0.005611053667962551,
      "learning_rate": 0.17771363927492845,
      "loss": 0.3838,
      "num_input_tokens_seen": 24242208,
      "step": 17635
    },
    {
      "epoch": 0.5646245438832341,
      "grad_norm": 0.004308751784265041,
      "learning_rate": 0.17765574637617085,
      "loss": 0.5041,
      "num_input_tokens_seen": 24249024,
      "step": 17640
    },
    {
      "epoch": 0.5647845848537225,
      "grad_norm": 0.006813515443354845,
      "learning_rate": 0.17759784921254962,
      "loss": 0.5105,
      "num_input_tokens_seen": 24255744,
      "step": 17645
    },
    {
      "epoch": 0.564944625824211,
      "grad_norm": 0.003914745524525642,
      "learning_rate": 0.1775399477929932,
      "loss": 0.4393,
      "num_input_tokens_seen": 24262736,
      "step": 17650
    },
    {
      "epoch": 0.5651046667946994,
      "grad_norm": 0.003856868017464876,
      "learning_rate": 0.17748204212643076,
      "loss": 0.6015,
      "num_input_tokens_seen": 24269776,
      "step": 17655
    },
    {
      "epoch": 0.5652647077651879,
      "grad_norm": 0.00634128600358963,
      "learning_rate": 0.17742413222179204,
      "loss": 0.5755,
      "num_input_tokens_seen": 24276320,
      "step": 17660
    },
    {
      "epoch": 0.5654247487356764,
      "grad_norm": 0.007030562963336706,
      "learning_rate": 0.17736621808800754,
      "loss": 0.7994,
      "num_input_tokens_seen": 24283040,
      "step": 17665
    },
    {
      "epoch": 0.5655847897061648,
      "grad_norm": 0.008785909973084927,
      "learning_rate": 0.17730829973400827,
      "loss": 0.442,
      "num_input_tokens_seen": 24289904,
      "step": 17670
    },
    {
      "epoch": 0.5657448306766533,
      "grad_norm": 0.004406494088470936,
      "learning_rate": 0.17725037716872602,
      "loss": 0.4106,
      "num_input_tokens_seen": 24297280,
      "step": 17675
    },
    {
      "epoch": 0.5659048716471416,
      "grad_norm": 0.002351572038605809,
      "learning_rate": 0.17719245040109313,
      "loss": 0.514,
      "num_input_tokens_seen": 24304160,
      "step": 17680
    },
    {
      "epoch": 0.5660649126176301,
      "grad_norm": 0.004060850478708744,
      "learning_rate": 0.17713451944004271,
      "loss": 0.48,
      "num_input_tokens_seen": 24311136,
      "step": 17685
    },
    {
      "epoch": 0.5662249535881185,
      "grad_norm": 0.005687267053872347,
      "learning_rate": 0.17707658429450843,
      "loss": 0.4985,
      "num_input_tokens_seen": 24318336,
      "step": 17690
    },
    {
      "epoch": 0.566384994558607,
      "grad_norm": 0.0032567286398261786,
      "learning_rate": 0.1770186449734245,
      "loss": 0.5667,
      "num_input_tokens_seen": 24325040,
      "step": 17695
    },
    {
      "epoch": 0.5665450355290954,
      "grad_norm": 0.0027857983950525522,
      "learning_rate": 0.17696070148572599,
      "loss": 0.4933,
      "num_input_tokens_seen": 24331824,
      "step": 17700
    },
    {
      "epoch": 0.5667050764995839,
      "grad_norm": 0.0033297499176114798,
      "learning_rate": 0.17690275384034856,
      "loss": 0.5249,
      "num_input_tokens_seen": 24339024,
      "step": 17705
    },
    {
      "epoch": 0.5668651174700723,
      "grad_norm": 0.006880049593746662,
      "learning_rate": 0.17684480204622835,
      "loss": 0.5677,
      "num_input_tokens_seen": 24345920,
      "step": 17710
    },
    {
      "epoch": 0.5670251584405608,
      "grad_norm": 0.003746644826605916,
      "learning_rate": 0.1767868461123023,
      "loss": 0.5097,
      "num_input_tokens_seen": 24352816,
      "step": 17715
    },
    {
      "epoch": 0.5671851994110493,
      "grad_norm": 0.002719094743952155,
      "learning_rate": 0.176728886047508,
      "loss": 0.6567,
      "num_input_tokens_seen": 24359600,
      "step": 17720
    },
    {
      "epoch": 0.5673452403815377,
      "grad_norm": 0.006838290486484766,
      "learning_rate": 0.17667092186078362,
      "loss": 0.6288,
      "num_input_tokens_seen": 24366672,
      "step": 17725
    },
    {
      "epoch": 0.5675052813520262,
      "grad_norm": 0.00629887031391263,
      "learning_rate": 0.17661295356106785,
      "loss": 0.6386,
      "num_input_tokens_seen": 24373456,
      "step": 17730
    },
    {
      "epoch": 0.5676653223225145,
      "grad_norm": 0.00401453347876668,
      "learning_rate": 0.1765549811573002,
      "loss": 0.5954,
      "num_input_tokens_seen": 24380336,
      "step": 17735
    },
    {
      "epoch": 0.567825363293003,
      "grad_norm": 0.00412351218983531,
      "learning_rate": 0.17649700465842078,
      "loss": 0.4845,
      "num_input_tokens_seen": 24386896,
      "step": 17740
    },
    {
      "epoch": 0.5679854042634914,
      "grad_norm": 0.002464805031195283,
      "learning_rate": 0.17643902407337023,
      "loss": 0.5593,
      "num_input_tokens_seen": 24393808,
      "step": 17745
    },
    {
      "epoch": 0.5681454452339799,
      "grad_norm": 0.004651966970413923,
      "learning_rate": 0.17638103941108993,
      "loss": 0.4987,
      "num_input_tokens_seen": 24400432,
      "step": 17750
    },
    {
      "epoch": 0.5683054862044683,
      "grad_norm": 0.0030024058651179075,
      "learning_rate": 0.1763230506805218,
      "loss": 0.6245,
      "num_input_tokens_seen": 24407424,
      "step": 17755
    },
    {
      "epoch": 0.5684655271749568,
      "grad_norm": 0.004020510241389275,
      "learning_rate": 0.1762650578906085,
      "loss": 0.6156,
      "num_input_tokens_seen": 24413872,
      "step": 17760
    },
    {
      "epoch": 0.5686255681454452,
      "grad_norm": 0.002589642768725753,
      "learning_rate": 0.1762070610502932,
      "loss": 0.5648,
      "num_input_tokens_seen": 24420656,
      "step": 17765
    },
    {
      "epoch": 0.5687856091159337,
      "grad_norm": 0.004624964203685522,
      "learning_rate": 0.17614906016851975,
      "loss": 0.6878,
      "num_input_tokens_seen": 24428000,
      "step": 17770
    },
    {
      "epoch": 0.5689456500864222,
      "grad_norm": 0.005542207043617964,
      "learning_rate": 0.17609105525423258,
      "loss": 0.5177,
      "num_input_tokens_seen": 24434944,
      "step": 17775
    },
    {
      "epoch": 0.5691056910569106,
      "grad_norm": 0.003056481946259737,
      "learning_rate": 0.1760330463163768,
      "loss": 0.6021,
      "num_input_tokens_seen": 24442240,
      "step": 17780
    },
    {
      "epoch": 0.569265732027399,
      "grad_norm": 0.005352030508220196,
      "learning_rate": 0.17597503336389816,
      "loss": 0.5563,
      "num_input_tokens_seen": 24449072,
      "step": 17785
    },
    {
      "epoch": 0.5694257729978874,
      "grad_norm": 0.006303071975708008,
      "learning_rate": 0.17591701640574298,
      "loss": 0.4892,
      "num_input_tokens_seen": 24455568,
      "step": 17790
    },
    {
      "epoch": 0.5695858139683759,
      "grad_norm": 0.0025253156200051308,
      "learning_rate": 0.17585899545085815,
      "loss": 0.4934,
      "num_input_tokens_seen": 24462272,
      "step": 17795
    },
    {
      "epoch": 0.5697458549388643,
      "grad_norm": 0.006142476107925177,
      "learning_rate": 0.17580097050819124,
      "loss": 0.4893,
      "num_input_tokens_seen": 24469312,
      "step": 17800
    },
    {
      "epoch": 0.5697458549388643,
      "eval_loss": 0.5554144978523254,
      "eval_runtime": 332.3192,
      "eval_samples_per_second": 41.785,
      "eval_steps_per_second": 20.893,
      "num_input_tokens_seen": 24469312,
      "step": 17800
    },
    {
      "epoch": 0.5699058959093528,
      "grad_norm": 0.0035626667086035013,
      "learning_rate": 0.17574294158669046,
      "loss": 0.581,
      "num_input_tokens_seen": 24476128,
      "step": 17805
    },
    {
      "epoch": 0.5700659368798412,
      "grad_norm": 0.005558359436690807,
      "learning_rate": 0.17568490869530456,
      "loss": 0.5545,
      "num_input_tokens_seen": 24483152,
      "step": 17810
    },
    {
      "epoch": 0.5702259778503297,
      "grad_norm": 0.003071929095312953,
      "learning_rate": 0.17562687184298295,
      "loss": 0.4429,
      "num_input_tokens_seen": 24489920,
      "step": 17815
    },
    {
      "epoch": 0.5703860188208181,
      "grad_norm": 0.003983143717050552,
      "learning_rate": 0.1755688310386757,
      "loss": 0.5214,
      "num_input_tokens_seen": 24496752,
      "step": 17820
    },
    {
      "epoch": 0.5705460597913066,
      "grad_norm": 0.0027248419355601072,
      "learning_rate": 0.17551078629133335,
      "loss": 0.4299,
      "num_input_tokens_seen": 24502912,
      "step": 17825
    },
    {
      "epoch": 0.5707061007617951,
      "grad_norm": 0.005790607538074255,
      "learning_rate": 0.17545273760990718,
      "loss": 0.3391,
      "num_input_tokens_seen": 24509568,
      "step": 17830
    },
    {
      "epoch": 0.5708661417322834,
      "grad_norm": 0.003917295951396227,
      "learning_rate": 0.17539468500334904,
      "loss": 0.4567,
      "num_input_tokens_seen": 24516784,
      "step": 17835
    },
    {
      "epoch": 0.571026182702772,
      "grad_norm": 0.006508747115731239,
      "learning_rate": 0.17533662848061132,
      "loss": 0.4801,
      "num_input_tokens_seen": 24523600,
      "step": 17840
    },
    {
      "epoch": 0.5711862236732603,
      "grad_norm": 0.006283026188611984,
      "learning_rate": 0.1752785680506471,
      "loss": 0.5231,
      "num_input_tokens_seen": 24530512,
      "step": 17845
    },
    {
      "epoch": 0.5713462646437488,
      "grad_norm": 0.004328989423811436,
      "learning_rate": 0.17522050372241,
      "loss": 0.4132,
      "num_input_tokens_seen": 24537312,
      "step": 17850
    },
    {
      "epoch": 0.5715063056142372,
      "grad_norm": 0.004642547573894262,
      "learning_rate": 0.17516243550485425,
      "loss": 0.4552,
      "num_input_tokens_seen": 24543920,
      "step": 17855
    },
    {
      "epoch": 0.5716663465847257,
      "grad_norm": 0.0030294039752334356,
      "learning_rate": 0.17510436340693478,
      "loss": 0.5928,
      "num_input_tokens_seen": 24550688,
      "step": 17860
    },
    {
      "epoch": 0.5718263875552141,
      "grad_norm": 0.005708435550332069,
      "learning_rate": 0.175046287437607,
      "loss": 0.616,
      "num_input_tokens_seen": 24558048,
      "step": 17865
    },
    {
      "epoch": 0.5719864285257026,
      "grad_norm": 0.004518812522292137,
      "learning_rate": 0.17498820760582695,
      "loss": 0.5238,
      "num_input_tokens_seen": 24564784,
      "step": 17870
    },
    {
      "epoch": 0.5721464694961911,
      "grad_norm": 0.005738791543990374,
      "learning_rate": 0.1749301239205512,
      "loss": 0.6064,
      "num_input_tokens_seen": 24571552,
      "step": 17875
    },
    {
      "epoch": 0.5723065104666795,
      "grad_norm": 0.0036205544602125883,
      "learning_rate": 0.1748720363907371,
      "loss": 0.5261,
      "num_input_tokens_seen": 24578320,
      "step": 17880
    },
    {
      "epoch": 0.572466551437168,
      "grad_norm": 0.006179386749863625,
      "learning_rate": 0.17481394502534242,
      "loss": 0.6109,
      "num_input_tokens_seen": 24585008,
      "step": 17885
    },
    {
      "epoch": 0.5726265924076563,
      "grad_norm": 0.005021568853408098,
      "learning_rate": 0.17475584983332562,
      "loss": 0.4482,
      "num_input_tokens_seen": 24591792,
      "step": 17890
    },
    {
      "epoch": 0.5727866333781448,
      "grad_norm": 0.0038612314965575933,
      "learning_rate": 0.17469775082364558,
      "loss": 0.7431,
      "num_input_tokens_seen": 24598944,
      "step": 17895
    },
    {
      "epoch": 0.5729466743486332,
      "grad_norm": 0.0065106493420898914,
      "learning_rate": 0.17463964800526205,
      "loss": 0.4239,
      "num_input_tokens_seen": 24605520,
      "step": 17900
    },
    {
      "epoch": 0.5731067153191217,
      "grad_norm": 0.004761953838169575,
      "learning_rate": 0.17458154138713522,
      "loss": 0.5897,
      "num_input_tokens_seen": 24612464,
      "step": 17905
    },
    {
      "epoch": 0.5732667562896101,
      "grad_norm": 0.007718234788626432,
      "learning_rate": 0.17452343097822576,
      "loss": 0.5497,
      "num_input_tokens_seen": 24619440,
      "step": 17910
    },
    {
      "epoch": 0.5734267972600986,
      "grad_norm": 0.005175972357392311,
      "learning_rate": 0.17446531678749497,
      "loss": 0.5156,
      "num_input_tokens_seen": 24626384,
      "step": 17915
    },
    {
      "epoch": 0.573586838230587,
      "grad_norm": 0.005446172785013914,
      "learning_rate": 0.17440719882390496,
      "loss": 0.5153,
      "num_input_tokens_seen": 24633392,
      "step": 17920
    },
    {
      "epoch": 0.5737468792010755,
      "grad_norm": 0.00434744730591774,
      "learning_rate": 0.17434907709641814,
      "loss": 0.5083,
      "num_input_tokens_seen": 24640192,
      "step": 17925
    },
    {
      "epoch": 0.573906920171564,
      "grad_norm": 0.0035209960769861937,
      "learning_rate": 0.17429095161399769,
      "loss": 0.6373,
      "num_input_tokens_seen": 24647232,
      "step": 17930
    },
    {
      "epoch": 0.5740669611420524,
      "grad_norm": 0.0035759531892836094,
      "learning_rate": 0.1742328223856072,
      "loss": 0.4692,
      "num_input_tokens_seen": 24654208,
      "step": 17935
    },
    {
      "epoch": 0.5742270021125409,
      "grad_norm": 0.0038467859849333763,
      "learning_rate": 0.174174689420211,
      "loss": 0.5937,
      "num_input_tokens_seen": 24661200,
      "step": 17940
    },
    {
      "epoch": 0.5743870430830292,
      "grad_norm": 0.0038870139978826046,
      "learning_rate": 0.1741165527267739,
      "loss": 0.6586,
      "num_input_tokens_seen": 24667888,
      "step": 17945
    },
    {
      "epoch": 0.5745470840535177,
      "grad_norm": 0.0045510525815188885,
      "learning_rate": 0.17405841231426125,
      "loss": 0.4528,
      "num_input_tokens_seen": 24675472,
      "step": 17950
    },
    {
      "epoch": 0.5747071250240061,
      "grad_norm": 0.005472238175570965,
      "learning_rate": 0.1740002681916391,
      "loss": 0.5809,
      "num_input_tokens_seen": 24682288,
      "step": 17955
    },
    {
      "epoch": 0.5748671659944946,
      "grad_norm": 0.005766722373664379,
      "learning_rate": 0.17394212036787401,
      "loss": 0.6934,
      "num_input_tokens_seen": 24689328,
      "step": 17960
    },
    {
      "epoch": 0.575027206964983,
      "grad_norm": 0.007102601695805788,
      "learning_rate": 0.1738839688519331,
      "loss": 0.6509,
      "num_input_tokens_seen": 24695680,
      "step": 17965
    },
    {
      "epoch": 0.5751872479354715,
      "grad_norm": 0.005593383219093084,
      "learning_rate": 0.17382581365278402,
      "loss": 0.611,
      "num_input_tokens_seen": 24702480,
      "step": 17970
    },
    {
      "epoch": 0.5753472889059599,
      "grad_norm": 0.001802269951440394,
      "learning_rate": 0.17376765477939507,
      "loss": 0.3694,
      "num_input_tokens_seen": 24709312,
      "step": 17975
    },
    {
      "epoch": 0.5755073298764484,
      "grad_norm": 0.002879211213439703,
      "learning_rate": 0.1737094922407351,
      "loss": 0.3892,
      "num_input_tokens_seen": 24715920,
      "step": 17980
    },
    {
      "epoch": 0.5756673708469369,
      "grad_norm": 0.004940110258758068,
      "learning_rate": 0.1736513260457734,
      "loss": 0.384,
      "num_input_tokens_seen": 24722464,
      "step": 17985
    },
    {
      "epoch": 0.5758274118174252,
      "grad_norm": 0.005761794745922089,
      "learning_rate": 0.17359315620348006,
      "loss": 0.6242,
      "num_input_tokens_seen": 24729392,
      "step": 17990
    },
    {
      "epoch": 0.5759874527879137,
      "grad_norm": 0.005084766540676355,
      "learning_rate": 0.17353498272282547,
      "loss": 0.4498,
      "num_input_tokens_seen": 24736496,
      "step": 17995
    },
    {
      "epoch": 0.5761474937584021,
      "grad_norm": 0.00490642711520195,
      "learning_rate": 0.17347680561278087,
      "loss": 0.5432,
      "num_input_tokens_seen": 24743360,
      "step": 18000
    },
    {
      "epoch": 0.5761474937584021,
      "eval_loss": 0.5533514618873596,
      "eval_runtime": 332.757,
      "eval_samples_per_second": 41.73,
      "eval_steps_per_second": 20.865,
      "num_input_tokens_seen": 24743360,
      "step": 18000
    },
    {
      "epoch": 0.5763075347288906,
      "grad_norm": 0.005491395480930805,
      "learning_rate": 0.1734186248823178,
      "loss": 0.3799,
      "num_input_tokens_seen": 24750928,
      "step": 18005
    },
    {
      "epoch": 0.576467575699379,
      "grad_norm": 0.0052959811873734,
      "learning_rate": 0.17336044054040844,
      "loss": 0.471,
      "num_input_tokens_seen": 24758096,
      "step": 18010
    },
    {
      "epoch": 0.5766276166698675,
      "grad_norm": 0.004257375840097666,
      "learning_rate": 0.1733022525960256,
      "loss": 0.4237,
      "num_input_tokens_seen": 24765552,
      "step": 18015
    },
    {
      "epoch": 0.5767876576403559,
      "grad_norm": 0.003552116919308901,
      "learning_rate": 0.1732440610581426,
      "loss": 0.452,
      "num_input_tokens_seen": 24773040,
      "step": 18020
    },
    {
      "epoch": 0.5769476986108444,
      "grad_norm": 0.0062219309620559216,
      "learning_rate": 0.17318586593573326,
      "loss": 0.5638,
      "num_input_tokens_seen": 24779600,
      "step": 18025
    },
    {
      "epoch": 0.5771077395813328,
      "grad_norm": 0.0023091842886060476,
      "learning_rate": 0.17312766723777204,
      "loss": 0.6205,
      "num_input_tokens_seen": 24786368,
      "step": 18030
    },
    {
      "epoch": 0.5772677805518213,
      "grad_norm": 0.0024055761750787497,
      "learning_rate": 0.1730694649732339,
      "loss": 0.6199,
      "num_input_tokens_seen": 24793920,
      "step": 18035
    },
    {
      "epoch": 0.5774278215223098,
      "grad_norm": 0.003739960491657257,
      "learning_rate": 0.17301125915109428,
      "loss": 0.5044,
      "num_input_tokens_seen": 24801232,
      "step": 18040
    },
    {
      "epoch": 0.5775878624927981,
      "grad_norm": 0.009556370787322521,
      "learning_rate": 0.17295304978032938,
      "loss": 0.6812,
      "num_input_tokens_seen": 24808176,
      "step": 18045
    },
    {
      "epoch": 0.5777479034632866,
      "grad_norm": 0.00549929728731513,
      "learning_rate": 0.17289483686991577,
      "loss": 0.6801,
      "num_input_tokens_seen": 24814944,
      "step": 18050
    },
    {
      "epoch": 0.577907944433775,
      "grad_norm": 0.005082262679934502,
      "learning_rate": 0.1728366204288306,
      "loss": 0.5235,
      "num_input_tokens_seen": 24821552,
      "step": 18055
    },
    {
      "epoch": 0.5780679854042635,
      "grad_norm": 0.006477661430835724,
      "learning_rate": 0.17277840046605153,
      "loss": 0.6124,
      "num_input_tokens_seen": 24828432,
      "step": 18060
    },
    {
      "epoch": 0.5782280263747519,
      "grad_norm": 0.004184407647699118,
      "learning_rate": 0.17272017699055686,
      "loss": 0.5394,
      "num_input_tokens_seen": 24835152,
      "step": 18065
    },
    {
      "epoch": 0.5783880673452404,
      "grad_norm": 0.0041054715402424335,
      "learning_rate": 0.17266195001132542,
      "loss": 0.5265,
      "num_input_tokens_seen": 24842224,
      "step": 18070
    },
    {
      "epoch": 0.5785481083157288,
      "grad_norm": 0.0027057684492319822,
      "learning_rate": 0.17260371953733647,
      "loss": 0.5675,
      "num_input_tokens_seen": 24849232,
      "step": 18075
    },
    {
      "epoch": 0.5787081492862173,
      "grad_norm": 0.0033491505309939384,
      "learning_rate": 0.1725454855775699,
      "loss": 0.6419,
      "num_input_tokens_seen": 24855936,
      "step": 18080
    },
    {
      "epoch": 0.5788681902567057,
      "grad_norm": 0.0048465607687830925,
      "learning_rate": 0.17248724814100616,
      "loss": 0.5319,
      "num_input_tokens_seen": 24863024,
      "step": 18085
    },
    {
      "epoch": 0.5790282312271942,
      "grad_norm": 0.004876433406025171,
      "learning_rate": 0.17242900723662619,
      "loss": 0.5588,
      "num_input_tokens_seen": 24869280,
      "step": 18090
    },
    {
      "epoch": 0.5791882721976827,
      "grad_norm": 0.0031509955879300833,
      "learning_rate": 0.1723707628734114,
      "loss": 0.4652,
      "num_input_tokens_seen": 24876096,
      "step": 18095
    },
    {
      "epoch": 0.579348313168171,
      "grad_norm": 0.0052094534039497375,
      "learning_rate": 0.1723125150603438,
      "loss": 0.6199,
      "num_input_tokens_seen": 24882928,
      "step": 18100
    },
    {
      "epoch": 0.5795083541386595,
      "grad_norm": 0.003505605272948742,
      "learning_rate": 0.1722542638064061,
      "loss": 0.5302,
      "num_input_tokens_seen": 24889968,
      "step": 18105
    },
    {
      "epoch": 0.5796683951091479,
      "grad_norm": 0.0025910462718456984,
      "learning_rate": 0.17219600912058117,
      "loss": 0.4968,
      "num_input_tokens_seen": 24896432,
      "step": 18110
    },
    {
      "epoch": 0.5798284360796364,
      "grad_norm": 0.006668312009423971,
      "learning_rate": 0.17213775101185272,
      "loss": 0.6548,
      "num_input_tokens_seen": 24902992,
      "step": 18115
    },
    {
      "epoch": 0.5799884770501248,
      "grad_norm": 0.006597090046852827,
      "learning_rate": 0.17207948948920485,
      "loss": 0.5988,
      "num_input_tokens_seen": 24909856,
      "step": 18120
    },
    {
      "epoch": 0.5801485180206133,
      "grad_norm": 0.011197519488632679,
      "learning_rate": 0.17202122456162228,
      "loss": 0.6592,
      "num_input_tokens_seen": 24916800,
      "step": 18125
    },
    {
      "epoch": 0.5803085589911017,
      "grad_norm": 0.002492629922926426,
      "learning_rate": 0.17196295623809013,
      "loss": 0.3588,
      "num_input_tokens_seen": 24923712,
      "step": 18130
    },
    {
      "epoch": 0.5804685999615902,
      "grad_norm": 0.00431765615940094,
      "learning_rate": 0.1719046845275941,
      "loss": 0.5337,
      "num_input_tokens_seen": 24930256,
      "step": 18135
    },
    {
      "epoch": 0.5806286409320787,
      "grad_norm": 0.002395918359979987,
      "learning_rate": 0.17184640943912044,
      "loss": 0.5415,
      "num_input_tokens_seen": 24936976,
      "step": 18140
    },
    {
      "epoch": 0.580788681902567,
      "grad_norm": 0.004160567652434111,
      "learning_rate": 0.1717881309816559,
      "loss": 0.5414,
      "num_input_tokens_seen": 24943712,
      "step": 18145
    },
    {
      "epoch": 0.5809487228730555,
      "grad_norm": 0.0037960021290928125,
      "learning_rate": 0.1717298491641878,
      "loss": 0.667,
      "num_input_tokens_seen": 24950432,
      "step": 18150
    },
    {
      "epoch": 0.5811087638435439,
      "grad_norm": 0.006136427167803049,
      "learning_rate": 0.17167156399570385,
      "loss": 0.7196,
      "num_input_tokens_seen": 24957360,
      "step": 18155
    },
    {
      "epoch": 0.5812688048140324,
      "grad_norm": 0.005179649218916893,
      "learning_rate": 0.17161327548519242,
      "loss": 0.6109,
      "num_input_tokens_seen": 24964768,
      "step": 18160
    },
    {
      "epoch": 0.5814288457845208,
      "grad_norm": 0.0038737046997994184,
      "learning_rate": 0.1715549836416423,
      "loss": 0.7298,
      "num_input_tokens_seen": 24971952,
      "step": 18165
    },
    {
      "epoch": 0.5815888867550093,
      "grad_norm": 0.004095369018614292,
      "learning_rate": 0.17149668847404279,
      "loss": 0.5536,
      "num_input_tokens_seen": 24978480,
      "step": 18170
    },
    {
      "epoch": 0.5817489277254977,
      "grad_norm": 0.004113973584026098,
      "learning_rate": 0.1714383899913838,
      "loss": 0.5858,
      "num_input_tokens_seen": 24985760,
      "step": 18175
    },
    {
      "epoch": 0.5819089686959862,
      "grad_norm": 0.0034500169567763805,
      "learning_rate": 0.17138008820265563,
      "loss": 0.3967,
      "num_input_tokens_seen": 24993104,
      "step": 18180
    },
    {
      "epoch": 0.5820690096664746,
      "grad_norm": 0.0016887368401512504,
      "learning_rate": 0.17132178311684917,
      "loss": 0.4481,
      "num_input_tokens_seen": 24999632,
      "step": 18185
    },
    {
      "epoch": 0.5822290506369631,
      "grad_norm": 0.004747853614389896,
      "learning_rate": 0.1712634747429559,
      "loss": 0.5263,
      "num_input_tokens_seen": 25006512,
      "step": 18190
    },
    {
      "epoch": 0.5823890916074516,
      "grad_norm": 0.0034021390601992607,
      "learning_rate": 0.17120516308996753,
      "loss": 0.5331,
      "num_input_tokens_seen": 25013360,
      "step": 18195
    },
    {
      "epoch": 0.5825491325779399,
      "grad_norm": 0.003993220627307892,
      "learning_rate": 0.17114684816687653,
      "loss": 0.5548,
      "num_input_tokens_seen": 25020352,
      "step": 18200
    },
    {
      "epoch": 0.5825491325779399,
      "eval_loss": 0.5488320589065552,
      "eval_runtime": 332.3419,
      "eval_samples_per_second": 41.782,
      "eval_steps_per_second": 20.891,
      "num_input_tokens_seen": 25020352,
      "step": 18200
    },
    {
      "epoch": 0.5827091735484284,
      "grad_norm": 0.006845575291663408,
      "learning_rate": 0.17108852998267585,
      "loss": 0.6919,
      "num_input_tokens_seen": 25027344,
      "step": 18205
    },
    {
      "epoch": 0.5828692145189168,
      "grad_norm": 0.0017934410134330392,
      "learning_rate": 0.17103020854635878,
      "loss": 0.4328,
      "num_input_tokens_seen": 25034048,
      "step": 18210
    },
    {
      "epoch": 0.5830292554894053,
      "grad_norm": 0.0029453784227371216,
      "learning_rate": 0.1709718838669193,
      "loss": 0.5133,
      "num_input_tokens_seen": 25040736,
      "step": 18215
    },
    {
      "epoch": 0.5831892964598937,
      "grad_norm": 0.007301813457161188,
      "learning_rate": 0.17091355595335173,
      "loss": 0.4961,
      "num_input_tokens_seen": 25047776,
      "step": 18220
    },
    {
      "epoch": 0.5833493374303822,
      "grad_norm": 0.004443040583282709,
      "learning_rate": 0.17085522481465107,
      "loss": 0.5856,
      "num_input_tokens_seen": 25055280,
      "step": 18225
    },
    {
      "epoch": 0.5835093784008706,
      "grad_norm": 0.005798156838864088,
      "learning_rate": 0.17079689045981264,
      "loss": 0.5828,
      "num_input_tokens_seen": 25062064,
      "step": 18230
    },
    {
      "epoch": 0.5836694193713591,
      "grad_norm": 0.005654516629874706,
      "learning_rate": 0.17073855289783238,
      "loss": 0.4366,
      "num_input_tokens_seen": 25068864,
      "step": 18235
    },
    {
      "epoch": 0.5838294603418475,
      "grad_norm": 0.0055121127516031265,
      "learning_rate": 0.1706802121377066,
      "loss": 0.7207,
      "num_input_tokens_seen": 25075872,
      "step": 18240
    },
    {
      "epoch": 0.583989501312336,
      "grad_norm": 0.0031484756618738174,
      "learning_rate": 0.17062186818843225,
      "loss": 0.3004,
      "num_input_tokens_seen": 25083184,
      "step": 18245
    },
    {
      "epoch": 0.5841495422828245,
      "grad_norm": 0.005197497550398111,
      "learning_rate": 0.17056352105900668,
      "loss": 0.4823,
      "num_input_tokens_seen": 25090000,
      "step": 18250
    },
    {
      "epoch": 0.5843095832533128,
      "grad_norm": 0.007003940176218748,
      "learning_rate": 0.17050517075842772,
      "loss": 0.5454,
      "num_input_tokens_seen": 25096544,
      "step": 18255
    },
    {
      "epoch": 0.5844696242238013,
      "grad_norm": 0.005409471224993467,
      "learning_rate": 0.17044681729569375,
      "loss": 0.5479,
      "num_input_tokens_seen": 25103056,
      "step": 18260
    },
    {
      "epoch": 0.5846296651942897,
      "grad_norm": 0.016058262437582016,
      "learning_rate": 0.17038846067980365,
      "loss": 0.6963,
      "num_input_tokens_seen": 25109920,
      "step": 18265
    },
    {
      "epoch": 0.5847897061647782,
      "grad_norm": 0.006348829250782728,
      "learning_rate": 0.17033010091975664,
      "loss": 0.7051,
      "num_input_tokens_seen": 25116368,
      "step": 18270
    },
    {
      "epoch": 0.5849497471352666,
      "grad_norm": 0.008003294467926025,
      "learning_rate": 0.17027173802455262,
      "loss": 0.7923,
      "num_input_tokens_seen": 25123232,
      "step": 18275
    },
    {
      "epoch": 0.5851097881057551,
      "grad_norm": 0.00357413780875504,
      "learning_rate": 0.1702133720031918,
      "loss": 0.432,
      "num_input_tokens_seen": 25129936,
      "step": 18280
    },
    {
      "epoch": 0.5852698290762435,
      "grad_norm": 0.003962153103202581,
      "learning_rate": 0.17015500286467503,
      "loss": 0.578,
      "num_input_tokens_seen": 25136704,
      "step": 18285
    },
    {
      "epoch": 0.585429870046732,
      "grad_norm": 0.0036469832994043827,
      "learning_rate": 0.17009663061800354,
      "loss": 0.6029,
      "num_input_tokens_seen": 25143136,
      "step": 18290
    },
    {
      "epoch": 0.5855899110172204,
      "grad_norm": 0.0036567384377121925,
      "learning_rate": 0.17003825527217903,
      "loss": 0.5493,
      "num_input_tokens_seen": 25150032,
      "step": 18295
    },
    {
      "epoch": 0.5857499519877089,
      "grad_norm": 0.006947181653231382,
      "learning_rate": 0.16997987683620377,
      "loss": 0.5903,
      "num_input_tokens_seen": 25156864,
      "step": 18300
    },
    {
      "epoch": 0.5859099929581973,
      "grad_norm": 0.006217640824615955,
      "learning_rate": 0.16992149531908043,
      "loss": 0.5342,
      "num_input_tokens_seen": 25163920,
      "step": 18305
    },
    {
      "epoch": 0.5860700339286857,
      "grad_norm": 0.0026402331423014402,
      "learning_rate": 0.16986311072981214,
      "loss": 0.7827,
      "num_input_tokens_seen": 25170816,
      "step": 18310
    },
    {
      "epoch": 0.5862300748991742,
      "grad_norm": 0.0021090598311275244,
      "learning_rate": 0.16980472307740255,
      "loss": 0.4314,
      "num_input_tokens_seen": 25177808,
      "step": 18315
    },
    {
      "epoch": 0.5863901158696626,
      "grad_norm": 0.005635875277221203,
      "learning_rate": 0.1697463323708558,
      "loss": 0.4498,
      "num_input_tokens_seen": 25184848,
      "step": 18320
    },
    {
      "epoch": 0.5865501568401511,
      "grad_norm": 0.004317362792789936,
      "learning_rate": 0.16968793861917641,
      "loss": 0.5774,
      "num_input_tokens_seen": 25191456,
      "step": 18325
    },
    {
      "epoch": 0.5867101978106395,
      "grad_norm": 0.0034212761092931032,
      "learning_rate": 0.16962954183136952,
      "loss": 0.7814,
      "num_input_tokens_seen": 25198544,
      "step": 18330
    },
    {
      "epoch": 0.586870238781128,
      "grad_norm": 0.0027288373094052076,
      "learning_rate": 0.16957114201644058,
      "loss": 0.4681,
      "num_input_tokens_seen": 25205536,
      "step": 18335
    },
    {
      "epoch": 0.5870302797516164,
      "grad_norm": 0.005460051354020834,
      "learning_rate": 0.16951273918339563,
      "loss": 0.6023,
      "num_input_tokens_seen": 25212528,
      "step": 18340
    },
    {
      "epoch": 0.5871903207221049,
      "grad_norm": 0.005464692134410143,
      "learning_rate": 0.16945433334124105,
      "loss": 0.4795,
      "num_input_tokens_seen": 25219648,
      "step": 18345
    },
    {
      "epoch": 0.5873503616925934,
      "grad_norm": 0.006786080542951822,
      "learning_rate": 0.1693959244989838,
      "loss": 0.6238,
      "num_input_tokens_seen": 25226624,
      "step": 18350
    },
    {
      "epoch": 0.5875104026630817,
      "grad_norm": 0.004910113289952278,
      "learning_rate": 0.16933751266563127,
      "loss": 0.5128,
      "num_input_tokens_seen": 25233312,
      "step": 18355
    },
    {
      "epoch": 0.5876704436335702,
      "grad_norm": 0.003971835598349571,
      "learning_rate": 0.16927909785019118,
      "loss": 0.4026,
      "num_input_tokens_seen": 25240176,
      "step": 18360
    },
    {
      "epoch": 0.5878304846040586,
      "grad_norm": 0.001759407576173544,
      "learning_rate": 0.169220680061672,
      "loss": 0.6122,
      "num_input_tokens_seen": 25247552,
      "step": 18365
    },
    {
      "epoch": 0.5879905255745471,
      "grad_norm": 0.011671065352857113,
      "learning_rate": 0.16916225930908244,
      "loss": 0.5737,
      "num_input_tokens_seen": 25253840,
      "step": 18370
    },
    {
      "epoch": 0.5881505665450355,
      "grad_norm": 0.004758489318192005,
      "learning_rate": 0.16910383560143163,
      "loss": 0.7632,
      "num_input_tokens_seen": 25260880,
      "step": 18375
    },
    {
      "epoch": 0.588310607515524,
      "grad_norm": 0.0030543189495801926,
      "learning_rate": 0.16904540894772935,
      "loss": 0.4238,
      "num_input_tokens_seen": 25267520,
      "step": 18380
    },
    {
      "epoch": 0.5884706484860124,
      "grad_norm": 0.011979261413216591,
      "learning_rate": 0.16898697935698562,
      "loss": 0.7022,
      "num_input_tokens_seen": 25274688,
      "step": 18385
    },
    {
      "epoch": 0.5886306894565009,
      "grad_norm": 0.004191938787698746,
      "learning_rate": 0.1689285468382111,
      "loss": 0.6105,
      "num_input_tokens_seen": 25281712,
      "step": 18390
    },
    {
      "epoch": 0.5887907304269893,
      "grad_norm": 0.005834504030644894,
      "learning_rate": 0.16887011140041677,
      "loss": 0.5999,
      "num_input_tokens_seen": 25288832,
      "step": 18395
    },
    {
      "epoch": 0.5889507713974778,
      "grad_norm": 0.0031514144502580166,
      "learning_rate": 0.1688116730526141,
      "loss": 0.5781,
      "num_input_tokens_seen": 25295920,
      "step": 18400
    },
    {
      "epoch": 0.5889507713974778,
      "eval_loss": 0.5530274510383606,
      "eval_runtime": 331.74,
      "eval_samples_per_second": 41.858,
      "eval_steps_per_second": 20.929,
      "num_input_tokens_seen": 25295920,
      "step": 18400
    },
    {
      "epoch": 0.5891108123679663,
      "grad_norm": 0.006009231321513653,
      "learning_rate": 0.1687532318038151,
      "loss": 0.4899,
      "num_input_tokens_seen": 25302784,
      "step": 18405
    },
    {
      "epoch": 0.5892708533384546,
      "grad_norm": 0.0038883707020431757,
      "learning_rate": 0.16869478766303206,
      "loss": 0.3814,
      "num_input_tokens_seen": 25309408,
      "step": 18410
    },
    {
      "epoch": 0.5894308943089431,
      "grad_norm": 0.0042066751047968864,
      "learning_rate": 0.16863634063927788,
      "loss": 0.5177,
      "num_input_tokens_seen": 25316656,
      "step": 18415
    },
    {
      "epoch": 0.5895909352794315,
      "grad_norm": 0.0027121517341583967,
      "learning_rate": 0.16857789074156568,
      "loss": 0.4461,
      "num_input_tokens_seen": 25323376,
      "step": 18420
    },
    {
      "epoch": 0.58975097624992,
      "grad_norm": 0.003263556631281972,
      "learning_rate": 0.16851943797890928,
      "loss": 0.3627,
      "num_input_tokens_seen": 25330288,
      "step": 18425
    },
    {
      "epoch": 0.5899110172204084,
      "grad_norm": 0.0040381839498877525,
      "learning_rate": 0.16846098236032284,
      "loss": 0.7546,
      "num_input_tokens_seen": 25336960,
      "step": 18430
    },
    {
      "epoch": 0.5900710581908969,
      "grad_norm": 0.006303553935140371,
      "learning_rate": 0.16840252389482097,
      "loss": 0.5505,
      "num_input_tokens_seen": 25343808,
      "step": 18435
    },
    {
      "epoch": 0.5902310991613853,
      "grad_norm": 0.004536137450486422,
      "learning_rate": 0.16834406259141857,
      "loss": 0.5894,
      "num_input_tokens_seen": 25350576,
      "step": 18440
    },
    {
      "epoch": 0.5903911401318738,
      "grad_norm": 0.013372865505516529,
      "learning_rate": 0.16828559845913124,
      "loss": 0.5726,
      "num_input_tokens_seen": 25357952,
      "step": 18445
    },
    {
      "epoch": 0.5905511811023622,
      "grad_norm": 0.0038640510756522417,
      "learning_rate": 0.16822713150697488,
      "loss": 0.5539,
      "num_input_tokens_seen": 25364880,
      "step": 18450
    },
    {
      "epoch": 0.5907112220728507,
      "grad_norm": 0.010250034742057323,
      "learning_rate": 0.16816866174396575,
      "loss": 0.4648,
      "num_input_tokens_seen": 25371696,
      "step": 18455
    },
    {
      "epoch": 0.5908712630433391,
      "grad_norm": 0.0036173814442008734,
      "learning_rate": 0.16811018917912057,
      "loss": 0.481,
      "num_input_tokens_seen": 25379056,
      "step": 18460
    },
    {
      "epoch": 0.5910313040138275,
      "grad_norm": 0.004148114938288927,
      "learning_rate": 0.16805171382145673,
      "loss": 0.5771,
      "num_input_tokens_seen": 25385888,
      "step": 18465
    },
    {
      "epoch": 0.591191344984316,
      "grad_norm": 0.005877641960978508,
      "learning_rate": 0.16799323567999175,
      "loss": 0.5143,
      "num_input_tokens_seen": 25392768,
      "step": 18470
    },
    {
      "epoch": 0.5913513859548044,
      "grad_norm": 0.0025393597315996885,
      "learning_rate": 0.16793475476374367,
      "loss": 0.4318,
      "num_input_tokens_seen": 25399744,
      "step": 18475
    },
    {
      "epoch": 0.5915114269252929,
      "grad_norm": 0.0038371323607861996,
      "learning_rate": 0.1678762710817311,
      "loss": 0.57,
      "num_input_tokens_seen": 25406656,
      "step": 18480
    },
    {
      "epoch": 0.5916714678957813,
      "grad_norm": 0.00255717639811337,
      "learning_rate": 0.1678177846429728,
      "loss": 0.6303,
      "num_input_tokens_seen": 25413536,
      "step": 18485
    },
    {
      "epoch": 0.5918315088662698,
      "grad_norm": 0.004836984910070896,
      "learning_rate": 0.16775929545648827,
      "loss": 0.3949,
      "num_input_tokens_seen": 25420032,
      "step": 18490
    },
    {
      "epoch": 0.5919915498367582,
      "grad_norm": 0.004039436113089323,
      "learning_rate": 0.16770080353129715,
      "loss": 0.4627,
      "num_input_tokens_seen": 25426624,
      "step": 18495
    },
    {
      "epoch": 0.5921515908072467,
      "grad_norm": 0.0038851075805723667,
      "learning_rate": 0.16764230887641968,
      "loss": 0.6242,
      "num_input_tokens_seen": 25433728,
      "step": 18500
    },
    {
      "epoch": 0.592311631777735,
      "grad_norm": 0.00466026458889246,
      "learning_rate": 0.1675838115008765,
      "loss": 0.5021,
      "num_input_tokens_seen": 25440528,
      "step": 18505
    },
    {
      "epoch": 0.5924716727482235,
      "grad_norm": 0.0036760868970304728,
      "learning_rate": 0.1675253114136886,
      "loss": 0.4292,
      "num_input_tokens_seen": 25447328,
      "step": 18510
    },
    {
      "epoch": 0.592631713718712,
      "grad_norm": 0.0030887024477124214,
      "learning_rate": 0.16746680862387747,
      "loss": 0.5345,
      "num_input_tokens_seen": 25454688,
      "step": 18515
    },
    {
      "epoch": 0.5927917546892004,
      "grad_norm": 0.008373044431209564,
      "learning_rate": 0.16740830314046493,
      "loss": 0.5968,
      "num_input_tokens_seen": 25461728,
      "step": 18520
    },
    {
      "epoch": 0.5929517956596889,
      "grad_norm": 0.0045136515982449055,
      "learning_rate": 0.1673497949724733,
      "loss": 0.6361,
      "num_input_tokens_seen": 25468624,
      "step": 18525
    },
    {
      "epoch": 0.5931118366301773,
      "grad_norm": 0.0033551896922290325,
      "learning_rate": 0.16729128412892522,
      "loss": 0.4522,
      "num_input_tokens_seen": 25475360,
      "step": 18530
    },
    {
      "epoch": 0.5932718776006658,
      "grad_norm": 0.005954779218882322,
      "learning_rate": 0.16723277061884384,
      "loss": 0.4958,
      "num_input_tokens_seen": 25482224,
      "step": 18535
    },
    {
      "epoch": 0.5934319185711542,
      "grad_norm": 0.007148941978812218,
      "learning_rate": 0.16717425445125267,
      "loss": 0.4298,
      "num_input_tokens_seen": 25488848,
      "step": 18540
    },
    {
      "epoch": 0.5935919595416427,
      "grad_norm": 0.004108482040464878,
      "learning_rate": 0.16711573563517565,
      "loss": 0.6094,
      "num_input_tokens_seen": 25495984,
      "step": 18545
    },
    {
      "epoch": 0.5937520005121311,
      "grad_norm": 0.004915224388241768,
      "learning_rate": 0.1670572141796371,
      "loss": 0.6495,
      "num_input_tokens_seen": 25502432,
      "step": 18550
    },
    {
      "epoch": 0.5939120414826196,
      "grad_norm": 0.006370380520820618,
      "learning_rate": 0.16699869009366175,
      "loss": 0.7119,
      "num_input_tokens_seen": 25509376,
      "step": 18555
    },
    {
      "epoch": 0.5940720824531079,
      "grad_norm": 0.004106251988559961,
      "learning_rate": 0.1669401633862748,
      "loss": 0.7238,
      "num_input_tokens_seen": 25515920,
      "step": 18560
    },
    {
      "epoch": 0.5942321234235964,
      "grad_norm": 0.003329114755615592,
      "learning_rate": 0.16688163406650178,
      "loss": 0.6079,
      "num_input_tokens_seen": 25522720,
      "step": 18565
    },
    {
      "epoch": 0.5943921643940849,
      "grad_norm": 0.005571291781961918,
      "learning_rate": 0.1668231021433686,
      "loss": 0.5248,
      "num_input_tokens_seen": 25529376,
      "step": 18570
    },
    {
      "epoch": 0.5945522053645733,
      "grad_norm": 0.00456958869472146,
      "learning_rate": 0.1667645676259017,
      "loss": 0.4768,
      "num_input_tokens_seen": 25536096,
      "step": 18575
    },
    {
      "epoch": 0.5947122463350618,
      "grad_norm": 0.005377675872296095,
      "learning_rate": 0.1667060305231277,
      "loss": 0.5072,
      "num_input_tokens_seen": 25543216,
      "step": 18580
    },
    {
      "epoch": 0.5948722873055502,
      "grad_norm": 0.007233908399939537,
      "learning_rate": 0.16664749084407396,
      "loss": 0.6971,
      "num_input_tokens_seen": 25550656,
      "step": 18585
    },
    {
      "epoch": 0.5950323282760387,
      "grad_norm": 0.004114589653909206,
      "learning_rate": 0.16658894859776788,
      "loss": 0.6034,
      "num_input_tokens_seen": 25557696,
      "step": 18590
    },
    {
      "epoch": 0.5951923692465271,
      "grad_norm": 0.005769807379692793,
      "learning_rate": 0.16653040379323752,
      "loss": 0.4551,
      "num_input_tokens_seen": 25564592,
      "step": 18595
    },
    {
      "epoch": 0.5953524102170156,
      "grad_norm": 0.0033398610539734364,
      "learning_rate": 0.16647185643951107,
      "loss": 0.5579,
      "num_input_tokens_seen": 25571232,
      "step": 18600
    },
    {
      "epoch": 0.5953524102170156,
      "eval_loss": 0.5462674498558044,
      "eval_runtime": 332.0115,
      "eval_samples_per_second": 41.824,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 25571232,
      "step": 18600
    },
    {
      "epoch": 0.595512451187504,
      "grad_norm": 0.006214276887476444,
      "learning_rate": 0.1664133065456174,
      "loss": 0.6229,
      "num_input_tokens_seen": 25578400,
      "step": 18605
    },
    {
      "epoch": 0.5956724921579925,
      "grad_norm": 0.0044386922381818295,
      "learning_rate": 0.1663547541205856,
      "loss": 0.4808,
      "num_input_tokens_seen": 25584960,
      "step": 18610
    },
    {
      "epoch": 0.595832533128481,
      "grad_norm": 0.0041097248904407024,
      "learning_rate": 0.16629619917344518,
      "loss": 0.4253,
      "num_input_tokens_seen": 25592144,
      "step": 18615
    },
    {
      "epoch": 0.5959925740989693,
      "grad_norm": 0.003034512512385845,
      "learning_rate": 0.16623764171322605,
      "loss": 0.4637,
      "num_input_tokens_seen": 25599040,
      "step": 18620
    },
    {
      "epoch": 0.5961526150694578,
      "grad_norm": 0.003769863396883011,
      "learning_rate": 0.1661790817489585,
      "loss": 0.5567,
      "num_input_tokens_seen": 25606048,
      "step": 18625
    },
    {
      "epoch": 0.5963126560399462,
      "grad_norm": 0.005606465507298708,
      "learning_rate": 0.16612051928967328,
      "loss": 0.5372,
      "num_input_tokens_seen": 25613200,
      "step": 18630
    },
    {
      "epoch": 0.5964726970104347,
      "grad_norm": 0.0032459422945976257,
      "learning_rate": 0.16606195434440138,
      "loss": 0.5529,
      "num_input_tokens_seen": 25619888,
      "step": 18635
    },
    {
      "epoch": 0.5966327379809231,
      "grad_norm": 0.005176421720534563,
      "learning_rate": 0.16600338692217426,
      "loss": 0.4973,
      "num_input_tokens_seen": 25626448,
      "step": 18640
    },
    {
      "epoch": 0.5967927789514116,
      "grad_norm": 0.004052489064633846,
      "learning_rate": 0.16594481703202374,
      "loss": 0.4564,
      "num_input_tokens_seen": 25633216,
      "step": 18645
    },
    {
      "epoch": 0.5969528199219,
      "grad_norm": 0.003853088477626443,
      "learning_rate": 0.1658862446829821,
      "loss": 0.4649,
      "num_input_tokens_seen": 25639840,
      "step": 18650
    },
    {
      "epoch": 0.5971128608923885,
      "grad_norm": 0.003422363195568323,
      "learning_rate": 0.16582766988408187,
      "loss": 0.5529,
      "num_input_tokens_seen": 25646512,
      "step": 18655
    },
    {
      "epoch": 0.5972729018628768,
      "grad_norm": 0.0034893490374088287,
      "learning_rate": 0.16576909264435608,
      "loss": 0.5257,
      "num_input_tokens_seen": 25653152,
      "step": 18660
    },
    {
      "epoch": 0.5974329428333653,
      "grad_norm": 0.00460290489718318,
      "learning_rate": 0.16571051297283798,
      "loss": 0.513,
      "num_input_tokens_seen": 25660016,
      "step": 18665
    },
    {
      "epoch": 0.5975929838038538,
      "grad_norm": 0.005366441793739796,
      "learning_rate": 0.16565193087856137,
      "loss": 0.6947,
      "num_input_tokens_seen": 25666640,
      "step": 18670
    },
    {
      "epoch": 0.5977530247743422,
      "grad_norm": 0.003022613236680627,
      "learning_rate": 0.16559334637056033,
      "loss": 0.5456,
      "num_input_tokens_seen": 25673712,
      "step": 18675
    },
    {
      "epoch": 0.5979130657448307,
      "grad_norm": 0.00392172671854496,
      "learning_rate": 0.16553475945786933,
      "loss": 0.4978,
      "num_input_tokens_seen": 25680704,
      "step": 18680
    },
    {
      "epoch": 0.5980731067153191,
      "grad_norm": 0.0037696862127631903,
      "learning_rate": 0.16547617014952318,
      "loss": 0.5467,
      "num_input_tokens_seen": 25687664,
      "step": 18685
    },
    {
      "epoch": 0.5982331476858076,
      "grad_norm": 0.010882021859288216,
      "learning_rate": 0.1654175784545571,
      "loss": 0.5725,
      "num_input_tokens_seen": 25694448,
      "step": 18690
    },
    {
      "epoch": 0.598393188656296,
      "grad_norm": 0.002636165590956807,
      "learning_rate": 0.1653589843820067,
      "loss": 0.5589,
      "num_input_tokens_seen": 25701456,
      "step": 18695
    },
    {
      "epoch": 0.5985532296267845,
      "grad_norm": 0.0032802193891257048,
      "learning_rate": 0.1653003879409079,
      "loss": 0.5839,
      "num_input_tokens_seen": 25708608,
      "step": 18700
    },
    {
      "epoch": 0.5987132705972729,
      "grad_norm": 0.0034557192120701075,
      "learning_rate": 0.165241789140297,
      "loss": 0.519,
      "num_input_tokens_seen": 25715264,
      "step": 18705
    },
    {
      "epoch": 0.5988733115677614,
      "grad_norm": 0.003666223958134651,
      "learning_rate": 0.16518318798921064,
      "loss": 0.5521,
      "num_input_tokens_seen": 25721936,
      "step": 18710
    },
    {
      "epoch": 0.5990333525382497,
      "grad_norm": 0.004160613752901554,
      "learning_rate": 0.16512458449668593,
      "loss": 0.3886,
      "num_input_tokens_seen": 25729536,
      "step": 18715
    },
    {
      "epoch": 0.5991933935087382,
      "grad_norm": 0.003093911334872246,
      "learning_rate": 0.1650659786717602,
      "loss": 0.6227,
      "num_input_tokens_seen": 25736576,
      "step": 18720
    },
    {
      "epoch": 0.5993534344792267,
      "grad_norm": 0.006761116907000542,
      "learning_rate": 0.1650073705234712,
      "loss": 0.5409,
      "num_input_tokens_seen": 25743632,
      "step": 18725
    },
    {
      "epoch": 0.5995134754497151,
      "grad_norm": 0.003913362510502338,
      "learning_rate": 0.16494876006085712,
      "loss": 0.5436,
      "num_input_tokens_seen": 25750352,
      "step": 18730
    },
    {
      "epoch": 0.5996735164202036,
      "grad_norm": 0.003981831017881632,
      "learning_rate": 0.16489014729295634,
      "loss": 0.5413,
      "num_input_tokens_seen": 25757344,
      "step": 18735
    },
    {
      "epoch": 0.599833557390692,
      "grad_norm": 0.003950800746679306,
      "learning_rate": 0.16483153222880775,
      "loss": 0.6147,
      "num_input_tokens_seen": 25764176,
      "step": 18740
    },
    {
      "epoch": 0.5999935983611805,
      "grad_norm": 0.004355504177510738,
      "learning_rate": 0.16477291487745052,
      "loss": 0.6124,
      "num_input_tokens_seen": 25770768,
      "step": 18745
    },
    {
      "epoch": 0.6001536393316689,
      "grad_norm": 0.003261087229475379,
      "learning_rate": 0.16471429524792416,
      "loss": 0.5228,
      "num_input_tokens_seen": 25777712,
      "step": 18750
    },
    {
      "epoch": 0.6003136803021574,
      "grad_norm": 0.006777916569262743,
      "learning_rate": 0.16465567334926856,
      "loss": 0.6766,
      "num_input_tokens_seen": 25784512,
      "step": 18755
    },
    {
      "epoch": 0.6004737212726458,
      "grad_norm": 0.0035317791625857353,
      "learning_rate": 0.16459704919052395,
      "loss": 0.4432,
      "num_input_tokens_seen": 25791504,
      "step": 18760
    },
    {
      "epoch": 0.6006337622431343,
      "grad_norm": 0.0050454214215278625,
      "learning_rate": 0.16453842278073086,
      "loss": 0.4903,
      "num_input_tokens_seen": 25798512,
      "step": 18765
    },
    {
      "epoch": 0.6007938032136226,
      "grad_norm": 0.003271633293479681,
      "learning_rate": 0.16447979412893038,
      "loss": 0.6224,
      "num_input_tokens_seen": 25805152,
      "step": 18770
    },
    {
      "epoch": 0.6009538441841111,
      "grad_norm": 0.0038395982701331377,
      "learning_rate": 0.16442116324416367,
      "loss": 0.503,
      "num_input_tokens_seen": 25812240,
      "step": 18775
    },
    {
      "epoch": 0.6011138851545996,
      "grad_norm": 0.004005088936537504,
      "learning_rate": 0.1643625301354723,
      "loss": 0.566,
      "num_input_tokens_seen": 25819456,
      "step": 18780
    },
    {
      "epoch": 0.601273926125088,
      "grad_norm": 0.007963480427861214,
      "learning_rate": 0.16430389481189828,
      "loss": 0.3652,
      "num_input_tokens_seen": 25826768,
      "step": 18785
    },
    {
      "epoch": 0.6014339670955765,
      "grad_norm": 0.008632892742753029,
      "learning_rate": 0.164245257282484,
      "loss": 0.6385,
      "num_input_tokens_seen": 25834304,
      "step": 18790
    },
    {
      "epoch": 0.6015940080660649,
      "grad_norm": 0.013229481875896454,
      "learning_rate": 0.16418661755627195,
      "loss": 0.547,
      "num_input_tokens_seen": 25840784,
      "step": 18795
    },
    {
      "epoch": 0.6017540490365534,
      "grad_norm": 0.004784705583006144,
      "learning_rate": 0.16412797564230527,
      "loss": 0.693,
      "num_input_tokens_seen": 25847664,
      "step": 18800
    },
    {
      "epoch": 0.6017540490365534,
      "eval_loss": 0.5611017942428589,
      "eval_runtime": 332.2585,
      "eval_samples_per_second": 41.793,
      "eval_steps_per_second": 20.896,
      "num_input_tokens_seen": 25847664,
      "step": 18800
    },
    {
      "epoch": 0.6019140900070418,
      "grad_norm": 0.007535586133599281,
      "learning_rate": 0.16406933154962713,
      "loss": 0.719,
      "num_input_tokens_seen": 25854608,
      "step": 18805
    },
    {
      "epoch": 0.6020741309775303,
      "grad_norm": 0.006835707928985357,
      "learning_rate": 0.16401068528728133,
      "loss": 0.6031,
      "num_input_tokens_seen": 25861472,
      "step": 18810
    },
    {
      "epoch": 0.6022341719480186,
      "grad_norm": 0.005135117564350367,
      "learning_rate": 0.16395203686431173,
      "loss": 0.464,
      "num_input_tokens_seen": 25868288,
      "step": 18815
    },
    {
      "epoch": 0.6023942129185071,
      "grad_norm": 0.012281331233680248,
      "learning_rate": 0.16389338628976277,
      "loss": 0.6088,
      "num_input_tokens_seen": 25875264,
      "step": 18820
    },
    {
      "epoch": 0.6025542538889955,
      "grad_norm": 0.008121100254356861,
      "learning_rate": 0.163834733572679,
      "loss": 0.66,
      "num_input_tokens_seen": 25882096,
      "step": 18825
    },
    {
      "epoch": 0.602714294859484,
      "grad_norm": 0.00485894363373518,
      "learning_rate": 0.16377607872210545,
      "loss": 0.6738,
      "num_input_tokens_seen": 25888848,
      "step": 18830
    },
    {
      "epoch": 0.6028743358299725,
      "grad_norm": 0.005342515651136637,
      "learning_rate": 0.16371742174708748,
      "loss": 0.6674,
      "num_input_tokens_seen": 25895856,
      "step": 18835
    },
    {
      "epoch": 0.6030343768004609,
      "grad_norm": 0.002592596225440502,
      "learning_rate": 0.16365876265667065,
      "loss": 0.4608,
      "num_input_tokens_seen": 25903008,
      "step": 18840
    },
    {
      "epoch": 0.6031944177709494,
      "grad_norm": 0.006025319918990135,
      "learning_rate": 0.163600101459901,
      "loss": 0.573,
      "num_input_tokens_seen": 25910160,
      "step": 18845
    },
    {
      "epoch": 0.6033544587414378,
      "grad_norm": 0.006717880256474018,
      "learning_rate": 0.16354143816582484,
      "loss": 0.5336,
      "num_input_tokens_seen": 25916768,
      "step": 18850
    },
    {
      "epoch": 0.6035144997119263,
      "grad_norm": 0.006270587909966707,
      "learning_rate": 0.1634827727834887,
      "loss": 0.5805,
      "num_input_tokens_seen": 25923696,
      "step": 18855
    },
    {
      "epoch": 0.6036745406824147,
      "grad_norm": 0.005390053149312735,
      "learning_rate": 0.16342410532193954,
      "loss": 0.6173,
      "num_input_tokens_seen": 25931008,
      "step": 18860
    },
    {
      "epoch": 0.6038345816529032,
      "grad_norm": 0.006442483980208635,
      "learning_rate": 0.16336543579022464,
      "loss": 0.4983,
      "num_input_tokens_seen": 25937920,
      "step": 18865
    },
    {
      "epoch": 0.6039946226233915,
      "grad_norm": 0.0050408244132995605,
      "learning_rate": 0.16330676419739157,
      "loss": 0.6678,
      "num_input_tokens_seen": 25944816,
      "step": 18870
    },
    {
      "epoch": 0.60415466359388,
      "grad_norm": 0.003505922853946686,
      "learning_rate": 0.1632480905524883,
      "loss": 0.5111,
      "num_input_tokens_seen": 25951408,
      "step": 18875
    },
    {
      "epoch": 0.6043147045643685,
      "grad_norm": 0.004814700223505497,
      "learning_rate": 0.16318941486456293,
      "loss": 0.529,
      "num_input_tokens_seen": 25958256,
      "step": 18880
    },
    {
      "epoch": 0.6044747455348569,
      "grad_norm": 0.006421478930860758,
      "learning_rate": 0.16313073714266405,
      "loss": 0.4849,
      "num_input_tokens_seen": 25965088,
      "step": 18885
    },
    {
      "epoch": 0.6046347865053454,
      "grad_norm": 0.004344494082033634,
      "learning_rate": 0.16307205739584052,
      "loss": 0.4642,
      "num_input_tokens_seen": 25972192,
      "step": 18890
    },
    {
      "epoch": 0.6047948274758338,
      "grad_norm": 0.008748299442231655,
      "learning_rate": 0.16301337563314144,
      "loss": 0.5887,
      "num_input_tokens_seen": 25979344,
      "step": 18895
    },
    {
      "epoch": 0.6049548684463223,
      "grad_norm": 0.002644645981490612,
      "learning_rate": 0.1629546918636163,
      "loss": 0.5859,
      "num_input_tokens_seen": 25986720,
      "step": 18900
    },
    {
      "epoch": 0.6051149094168107,
      "grad_norm": 0.004964461550116539,
      "learning_rate": 0.16289600609631485,
      "loss": 0.3709,
      "num_input_tokens_seen": 25993584,
      "step": 18905
    },
    {
      "epoch": 0.6052749503872992,
      "grad_norm": 0.0053706541657447815,
      "learning_rate": 0.16283731834028722,
      "loss": 0.441,
      "num_input_tokens_seen": 26000256,
      "step": 18910
    },
    {
      "epoch": 0.6054349913577876,
      "grad_norm": 0.01134677603840828,
      "learning_rate": 0.16277862860458378,
      "loss": 0.6878,
      "num_input_tokens_seen": 26007088,
      "step": 18915
    },
    {
      "epoch": 0.605595032328276,
      "grad_norm": 0.005576890893280506,
      "learning_rate": 0.16271993689825526,
      "loss": 0.6082,
      "num_input_tokens_seen": 26013856,
      "step": 18920
    },
    {
      "epoch": 0.6057550732987644,
      "grad_norm": 0.006896756123751402,
      "learning_rate": 0.1626612432303526,
      "loss": 0.5994,
      "num_input_tokens_seen": 26020848,
      "step": 18925
    },
    {
      "epoch": 0.6059151142692529,
      "grad_norm": 0.0054159159772098064,
      "learning_rate": 0.1626025476099271,
      "loss": 0.6146,
      "num_input_tokens_seen": 26027696,
      "step": 18930
    },
    {
      "epoch": 0.6060751552397414,
      "grad_norm": 0.003978292923420668,
      "learning_rate": 0.1625438500460304,
      "loss": 0.4921,
      "num_input_tokens_seen": 26034624,
      "step": 18935
    },
    {
      "epoch": 0.6062351962102298,
      "grad_norm": 0.005443072412163019,
      "learning_rate": 0.16248515054771442,
      "loss": 0.6086,
      "num_input_tokens_seen": 26041600,
      "step": 18940
    },
    {
      "epoch": 0.6063952371807183,
      "grad_norm": 0.0039753662422299385,
      "learning_rate": 0.16242644912403123,
      "loss": 0.6206,
      "num_input_tokens_seen": 26048480,
      "step": 18945
    },
    {
      "epoch": 0.6065552781512067,
      "grad_norm": 0.0066193221136927605,
      "learning_rate": 0.1623677457840335,
      "loss": 0.5485,
      "num_input_tokens_seen": 26055328,
      "step": 18950
    },
    {
      "epoch": 0.6067153191216952,
      "grad_norm": 0.005509459879249334,
      "learning_rate": 0.16230904053677397,
      "loss": 0.8288,
      "num_input_tokens_seen": 26062144,
      "step": 18955
    },
    {
      "epoch": 0.6068753600921836,
      "grad_norm": 0.007710503414273262,
      "learning_rate": 0.16225033339130568,
      "loss": 0.6039,
      "num_input_tokens_seen": 26069648,
      "step": 18960
    },
    {
      "epoch": 0.6070354010626721,
      "grad_norm": 0.008254626765847206,
      "learning_rate": 0.16219162435668197,
      "loss": 0.5016,
      "num_input_tokens_seen": 26076464,
      "step": 18965
    },
    {
      "epoch": 0.6071954420331604,
      "grad_norm": 0.005757271312177181,
      "learning_rate": 0.16213291344195666,
      "loss": 0.5886,
      "num_input_tokens_seen": 26083952,
      "step": 18970
    },
    {
      "epoch": 0.6073554830036489,
      "grad_norm": 0.003939670976251364,
      "learning_rate": 0.16207420065618358,
      "loss": 0.5452,
      "num_input_tokens_seen": 26090848,
      "step": 18975
    },
    {
      "epoch": 0.6075155239741373,
      "grad_norm": 0.005309319123625755,
      "learning_rate": 0.16201548600841706,
      "loss": 0.6096,
      "num_input_tokens_seen": 26098048,
      "step": 18980
    },
    {
      "epoch": 0.6076755649446258,
      "grad_norm": 0.005817113909870386,
      "learning_rate": 0.16195676950771154,
      "loss": 0.6836,
      "num_input_tokens_seen": 26104656,
      "step": 18985
    },
    {
      "epoch": 0.6078356059151143,
      "grad_norm": 0.007574467919766903,
      "learning_rate": 0.16189805116312198,
      "loss": 0.8493,
      "num_input_tokens_seen": 26111792,
      "step": 18990
    },
    {
      "epoch": 0.6079956468856027,
      "grad_norm": 0.0044804345816373825,
      "learning_rate": 0.16183933098370337,
      "loss": 0.6185,
      "num_input_tokens_seen": 26118272,
      "step": 18995
    },
    {
      "epoch": 0.6081556878560912,
      "grad_norm": 0.005879140924662352,
      "learning_rate": 0.16178060897851115,
      "loss": 0.6779,
      "num_input_tokens_seen": 26125328,
      "step": 19000
    },
    {
      "epoch": 0.6081556878560912,
      "eval_loss": 0.5611286759376526,
      "eval_runtime": 331.9646,
      "eval_samples_per_second": 41.83,
      "eval_steps_per_second": 20.915,
      "num_input_tokens_seen": 26125328,
      "step": 19000
    },
    {
      "epoch": 0.6083157288265796,
      "grad_norm": 0.002256944542750716,
      "learning_rate": 0.16172188515660096,
      "loss": 0.4919,
      "num_input_tokens_seen": 26131968,
      "step": 19005
    },
    {
      "epoch": 0.6084757697970681,
      "grad_norm": 0.004038271959871054,
      "learning_rate": 0.16166315952702878,
      "loss": 0.4596,
      "num_input_tokens_seen": 26138736,
      "step": 19010
    },
    {
      "epoch": 0.6086358107675565,
      "grad_norm": 0.004811219871044159,
      "learning_rate": 0.16160443209885084,
      "loss": 0.6755,
      "num_input_tokens_seen": 26146352,
      "step": 19015
    },
    {
      "epoch": 0.608795851738045,
      "grad_norm": 0.002691123867407441,
      "learning_rate": 0.16154570288112363,
      "loss": 0.5425,
      "num_input_tokens_seen": 26153760,
      "step": 19020
    },
    {
      "epoch": 0.6089558927085333,
      "grad_norm": 0.0063377870246768,
      "learning_rate": 0.16148697188290395,
      "loss": 0.6671,
      "num_input_tokens_seen": 26160640,
      "step": 19025
    },
    {
      "epoch": 0.6091159336790218,
      "grad_norm": 0.003434934886172414,
      "learning_rate": 0.16142823911324888,
      "loss": 0.7077,
      "num_input_tokens_seen": 26167424,
      "step": 19030
    },
    {
      "epoch": 0.6092759746495102,
      "grad_norm": 0.006945902481675148,
      "learning_rate": 0.16136950458121568,
      "loss": 0.5594,
      "num_input_tokens_seen": 26174304,
      "step": 19035
    },
    {
      "epoch": 0.6094360156199987,
      "grad_norm": 0.0030325280968099833,
      "learning_rate": 0.16131076829586205,
      "loss": 0.7166,
      "num_input_tokens_seen": 26181120,
      "step": 19040
    },
    {
      "epoch": 0.6095960565904872,
      "grad_norm": 0.0044427914544939995,
      "learning_rate": 0.1612520302662457,
      "loss": 0.4782,
      "num_input_tokens_seen": 26187904,
      "step": 19045
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 0.004070748575031757,
      "learning_rate": 0.16119329050142497,
      "loss": 0.7249,
      "num_input_tokens_seen": 26194928,
      "step": 19050
    },
    {
      "epoch": 0.6099161385314641,
      "grad_norm": 0.004799663554877043,
      "learning_rate": 0.16113454901045818,
      "loss": 0.5072,
      "num_input_tokens_seen": 26201792,
      "step": 19055
    },
    {
      "epoch": 0.6100761795019525,
      "grad_norm": 0.015360761433839798,
      "learning_rate": 0.16107580580240397,
      "loss": 0.5177,
      "num_input_tokens_seen": 26208592,
      "step": 19060
    },
    {
      "epoch": 0.610236220472441,
      "grad_norm": 0.0047899214550852776,
      "learning_rate": 0.16101706088632134,
      "loss": 0.4724,
      "num_input_tokens_seen": 26215312,
      "step": 19065
    },
    {
      "epoch": 0.6103962614429294,
      "grad_norm": 0.005695550236850977,
      "learning_rate": 0.16095831427126947,
      "loss": 0.5674,
      "num_input_tokens_seen": 26222208,
      "step": 19070
    },
    {
      "epoch": 0.6105563024134179,
      "grad_norm": 0.0053184456191957,
      "learning_rate": 0.16089956596630783,
      "loss": 0.5519,
      "num_input_tokens_seen": 26228800,
      "step": 19075
    },
    {
      "epoch": 0.6107163433839062,
      "grad_norm": 0.0053987763822078705,
      "learning_rate": 0.16084081598049618,
      "loss": 0.4701,
      "num_input_tokens_seen": 26235632,
      "step": 19080
    },
    {
      "epoch": 0.6108763843543947,
      "grad_norm": 0.003780619241297245,
      "learning_rate": 0.1607820643228944,
      "loss": 0.527,
      "num_input_tokens_seen": 26242368,
      "step": 19085
    },
    {
      "epoch": 0.6110364253248832,
      "grad_norm": 0.006581046152859926,
      "learning_rate": 0.16072331100256285,
      "loss": 0.5492,
      "num_input_tokens_seen": 26249584,
      "step": 19090
    },
    {
      "epoch": 0.6111964662953716,
      "grad_norm": 0.004382728133350611,
      "learning_rate": 0.16066455602856197,
      "loss": 0.4851,
      "num_input_tokens_seen": 26256944,
      "step": 19095
    },
    {
      "epoch": 0.6113565072658601,
      "grad_norm": 0.005034405738115311,
      "learning_rate": 0.16060579940995257,
      "loss": 0.4005,
      "num_input_tokens_seen": 26264336,
      "step": 19100
    },
    {
      "epoch": 0.6115165482363485,
      "grad_norm": 0.005626946222037077,
      "learning_rate": 0.16054704115579557,
      "loss": 0.5862,
      "num_input_tokens_seen": 26271216,
      "step": 19105
    },
    {
      "epoch": 0.611676589206837,
      "grad_norm": 0.004220474511384964,
      "learning_rate": 0.1604882812751523,
      "loss": 0.5152,
      "num_input_tokens_seen": 26278064,
      "step": 19110
    },
    {
      "epoch": 0.6118366301773254,
      "grad_norm": 0.007992624305188656,
      "learning_rate": 0.16042951977708425,
      "loss": 0.6757,
      "num_input_tokens_seen": 26284880,
      "step": 19115
    },
    {
      "epoch": 0.6119966711478139,
      "grad_norm": 0.004625016823410988,
      "learning_rate": 0.16037075667065318,
      "loss": 0.4907,
      "num_input_tokens_seen": 26291984,
      "step": 19120
    },
    {
      "epoch": 0.6121567121183022,
      "grad_norm": 0.004270501434803009,
      "learning_rate": 0.1603119919649211,
      "loss": 0.5634,
      "num_input_tokens_seen": 26299136,
      "step": 19125
    },
    {
      "epoch": 0.6123167530887907,
      "grad_norm": 0.009039634838700294,
      "learning_rate": 0.16025322566895028,
      "loss": 0.5943,
      "num_input_tokens_seen": 26306240,
      "step": 19130
    },
    {
      "epoch": 0.6124767940592791,
      "grad_norm": 0.0038473873864859343,
      "learning_rate": 0.16019445779180322,
      "loss": 0.5666,
      "num_input_tokens_seen": 26313328,
      "step": 19135
    },
    {
      "epoch": 0.6126368350297676,
      "grad_norm": 0.004093735944479704,
      "learning_rate": 0.16013568834254271,
      "loss": 0.4787,
      "num_input_tokens_seen": 26320592,
      "step": 19140
    },
    {
      "epoch": 0.6127968760002561,
      "grad_norm": 0.004492323845624924,
      "learning_rate": 0.1600769173302316,
      "loss": 0.4696,
      "num_input_tokens_seen": 26327648,
      "step": 19145
    },
    {
      "epoch": 0.6129569169707445,
      "grad_norm": 0.0049185482785105705,
      "learning_rate": 0.16001814476393322,
      "loss": 0.6246,
      "num_input_tokens_seen": 26334288,
      "step": 19150
    },
    {
      "epoch": 0.613116957941233,
      "grad_norm": 0.0050048609264194965,
      "learning_rate": 0.15995937065271104,
      "loss": 0.5257,
      "num_input_tokens_seen": 26341264,
      "step": 19155
    },
    {
      "epoch": 0.6132769989117214,
      "grad_norm": 0.003802224528044462,
      "learning_rate": 0.15990059500562873,
      "loss": 0.4871,
      "num_input_tokens_seen": 26348656,
      "step": 19160
    },
    {
      "epoch": 0.6134370398822099,
      "grad_norm": 0.004197648726403713,
      "learning_rate": 0.15984181783175025,
      "loss": 0.5328,
      "num_input_tokens_seen": 26355008,
      "step": 19165
    },
    {
      "epoch": 0.6135970808526983,
      "grad_norm": 0.0030703404918313026,
      "learning_rate": 0.1597830391401398,
      "loss": 0.4483,
      "num_input_tokens_seen": 26361840,
      "step": 19170
    },
    {
      "epoch": 0.6137571218231868,
      "grad_norm": 0.005442138761281967,
      "learning_rate": 0.15972425893986178,
      "loss": 0.6926,
      "num_input_tokens_seen": 26369520,
      "step": 19175
    },
    {
      "epoch": 0.6139171627936751,
      "grad_norm": 0.00419307267293334,
      "learning_rate": 0.15966547723998084,
      "loss": 0.3884,
      "num_input_tokens_seen": 26376304,
      "step": 19180
    },
    {
      "epoch": 0.6140772037641636,
      "grad_norm": 0.003289157524704933,
      "learning_rate": 0.15960669404956176,
      "loss": 0.6394,
      "num_input_tokens_seen": 26383344,
      "step": 19185
    },
    {
      "epoch": 0.614237244734652,
      "grad_norm": 0.006474161986261606,
      "learning_rate": 0.1595479093776698,
      "loss": 0.4607,
      "num_input_tokens_seen": 26390048,
      "step": 19190
    },
    {
      "epoch": 0.6143972857051405,
      "grad_norm": 0.0032726058270782232,
      "learning_rate": 0.15948912323337022,
      "loss": 0.5955,
      "num_input_tokens_seen": 26397040,
      "step": 19195
    },
    {
      "epoch": 0.614557326675629,
      "grad_norm": 0.0070331706665456295,
      "learning_rate": 0.1594303356257286,
      "loss": 0.6462,
      "num_input_tokens_seen": 26404064,
      "step": 19200
    },
    {
      "epoch": 0.614557326675629,
      "eval_loss": 0.5527148842811584,
      "eval_runtime": 331.7611,
      "eval_samples_per_second": 41.855,
      "eval_steps_per_second": 20.928,
      "num_input_tokens_seen": 26404064,
      "step": 19200
    },
    {
      "epoch": 0.6147173676461174,
      "grad_norm": 0.010307886637747288,
      "learning_rate": 0.15937154656381072,
      "loss": 0.564,
      "num_input_tokens_seen": 26411024,
      "step": 19205
    },
    {
      "epoch": 0.6148774086166059,
      "grad_norm": 0.003578662173822522,
      "learning_rate": 0.15931275605668258,
      "loss": 0.5744,
      "num_input_tokens_seen": 26417920,
      "step": 19210
    },
    {
      "epoch": 0.6150374495870943,
      "grad_norm": 0.005616813898086548,
      "learning_rate": 0.1592539641134104,
      "loss": 0.4993,
      "num_input_tokens_seen": 26424960,
      "step": 19215
    },
    {
      "epoch": 0.6151974905575828,
      "grad_norm": 0.004960114136338234,
      "learning_rate": 0.1591951707430607,
      "loss": 0.6781,
      "num_input_tokens_seen": 26431968,
      "step": 19220
    },
    {
      "epoch": 0.6153575315280712,
      "grad_norm": 0.004759480245411396,
      "learning_rate": 0.15913637595470007,
      "loss": 0.7342,
      "num_input_tokens_seen": 26438928,
      "step": 19225
    },
    {
      "epoch": 0.6155175724985597,
      "grad_norm": 0.004506371449679136,
      "learning_rate": 0.15907757975739548,
      "loss": 0.6047,
      "num_input_tokens_seen": 26445392,
      "step": 19230
    },
    {
      "epoch": 0.615677613469048,
      "grad_norm": 0.0046574026346206665,
      "learning_rate": 0.159018782160214,
      "loss": 0.6295,
      "num_input_tokens_seen": 26452016,
      "step": 19235
    },
    {
      "epoch": 0.6158376544395365,
      "grad_norm": 0.004760468378663063,
      "learning_rate": 0.158959983172223,
      "loss": 0.4787,
      "num_input_tokens_seen": 26458624,
      "step": 19240
    },
    {
      "epoch": 0.6159976954100249,
      "grad_norm": 0.004219872877001762,
      "learning_rate": 0.15890118280249,
      "loss": 0.6346,
      "num_input_tokens_seen": 26465440,
      "step": 19245
    },
    {
      "epoch": 0.6161577363805134,
      "grad_norm": 0.0053506274707615376,
      "learning_rate": 0.15884238106008275,
      "loss": 0.6729,
      "num_input_tokens_seen": 26472080,
      "step": 19250
    },
    {
      "epoch": 0.6163177773510019,
      "grad_norm": 0.006223492790013552,
      "learning_rate": 0.15878357795406922,
      "loss": 0.7773,
      "num_input_tokens_seen": 26478976,
      "step": 19255
    },
    {
      "epoch": 0.6164778183214903,
      "grad_norm": 0.004024229012429714,
      "learning_rate": 0.15872477349351757,
      "loss": 0.6175,
      "num_input_tokens_seen": 26485776,
      "step": 19260
    },
    {
      "epoch": 0.6166378592919788,
      "grad_norm": 0.004983413964509964,
      "learning_rate": 0.15866596768749622,
      "loss": 0.681,
      "num_input_tokens_seen": 26492816,
      "step": 19265
    },
    {
      "epoch": 0.6167979002624672,
      "grad_norm": 0.0040796007961034775,
      "learning_rate": 0.15860716054507373,
      "loss": 0.6071,
      "num_input_tokens_seen": 26499488,
      "step": 19270
    },
    {
      "epoch": 0.6169579412329557,
      "grad_norm": 0.004331586416810751,
      "learning_rate": 0.1585483520753189,
      "loss": 0.5923,
      "num_input_tokens_seen": 26506288,
      "step": 19275
    },
    {
      "epoch": 0.617117982203444,
      "grad_norm": 0.0025942909996956587,
      "learning_rate": 0.1584895422873008,
      "loss": 0.7896,
      "num_input_tokens_seen": 26513424,
      "step": 19280
    },
    {
      "epoch": 0.6172780231739325,
      "grad_norm": 0.005859393160790205,
      "learning_rate": 0.1584307311900886,
      "loss": 0.4989,
      "num_input_tokens_seen": 26519808,
      "step": 19285
    },
    {
      "epoch": 0.6174380641444209,
      "grad_norm": 0.0035021118819713593,
      "learning_rate": 0.1583719187927517,
      "loss": 0.3983,
      "num_input_tokens_seen": 26526480,
      "step": 19290
    },
    {
      "epoch": 0.6175981051149094,
      "grad_norm": 0.006616571452468634,
      "learning_rate": 0.15831310510435967,
      "loss": 0.6559,
      "num_input_tokens_seen": 26533008,
      "step": 19295
    },
    {
      "epoch": 0.6177581460853978,
      "grad_norm": 0.00422443263232708,
      "learning_rate": 0.15825429013398243,
      "loss": 0.5419,
      "num_input_tokens_seen": 26539456,
      "step": 19300
    },
    {
      "epoch": 0.6179181870558863,
      "grad_norm": 0.0036737057380378246,
      "learning_rate": 0.15819547389068986,
      "loss": 0.5106,
      "num_input_tokens_seen": 26546704,
      "step": 19305
    },
    {
      "epoch": 0.6180782280263748,
      "grad_norm": 0.006739599164575338,
      "learning_rate": 0.1581366563835522,
      "loss": 0.5443,
      "num_input_tokens_seen": 26554032,
      "step": 19310
    },
    {
      "epoch": 0.6182382689968632,
      "grad_norm": 0.00744269834831357,
      "learning_rate": 0.15807783762163993,
      "loss": 0.692,
      "num_input_tokens_seen": 26561472,
      "step": 19315
    },
    {
      "epoch": 0.6183983099673517,
      "grad_norm": 0.005866185761988163,
      "learning_rate": 0.15801901761402365,
      "loss": 0.5506,
      "num_input_tokens_seen": 26568624,
      "step": 19320
    },
    {
      "epoch": 0.6185583509378401,
      "grad_norm": 0.009055753238499165,
      "learning_rate": 0.157960196369774,
      "loss": 0.7372,
      "num_input_tokens_seen": 26575792,
      "step": 19325
    },
    {
      "epoch": 0.6187183919083286,
      "grad_norm": 0.004838017746806145,
      "learning_rate": 0.157901373897962,
      "loss": 0.5354,
      "num_input_tokens_seen": 26582704,
      "step": 19330
    },
    {
      "epoch": 0.6188784328788169,
      "grad_norm": 0.0058474051766097546,
      "learning_rate": 0.15784255020765892,
      "loss": 0.6549,
      "num_input_tokens_seen": 26589968,
      "step": 19335
    },
    {
      "epoch": 0.6190384738493054,
      "grad_norm": 0.004431513603776693,
      "learning_rate": 0.157783725307936,
      "loss": 0.5559,
      "num_input_tokens_seen": 26596864,
      "step": 19340
    },
    {
      "epoch": 0.6191985148197938,
      "grad_norm": 0.005403112154453993,
      "learning_rate": 0.15772489920786484,
      "loss": 0.6746,
      "num_input_tokens_seen": 26604048,
      "step": 19345
    },
    {
      "epoch": 0.6193585557902823,
      "grad_norm": 0.0053788116201758385,
      "learning_rate": 0.15766607191651713,
      "loss": 0.6886,
      "num_input_tokens_seen": 26611216,
      "step": 19350
    },
    {
      "epoch": 0.6195185967607708,
      "grad_norm": 0.003868329571560025,
      "learning_rate": 0.1576072434429648,
      "loss": 0.5439,
      "num_input_tokens_seen": 26618032,
      "step": 19355
    },
    {
      "epoch": 0.6196786377312592,
      "grad_norm": 0.006218965630978346,
      "learning_rate": 0.15754841379627998,
      "loss": 0.5716,
      "num_input_tokens_seen": 26624704,
      "step": 19360
    },
    {
      "epoch": 0.6198386787017477,
      "grad_norm": 0.005287556443363428,
      "learning_rate": 0.15748958298553484,
      "loss": 0.6328,
      "num_input_tokens_seen": 26631120,
      "step": 19365
    },
    {
      "epoch": 0.6199987196722361,
      "grad_norm": 0.005336710251867771,
      "learning_rate": 0.1574307510198019,
      "loss": 0.6115,
      "num_input_tokens_seen": 26637568,
      "step": 19370
    },
    {
      "epoch": 0.6201587606427246,
      "grad_norm": 0.003957351669669151,
      "learning_rate": 0.15737191790815375,
      "loss": 0.4402,
      "num_input_tokens_seen": 26644240,
      "step": 19375
    },
    {
      "epoch": 0.620318801613213,
      "grad_norm": 0.01328981202095747,
      "learning_rate": 0.15731308365966323,
      "loss": 0.6251,
      "num_input_tokens_seen": 26650800,
      "step": 19380
    },
    {
      "epoch": 0.6204788425837015,
      "grad_norm": 0.004810450132936239,
      "learning_rate": 0.15725424828340331,
      "loss": 0.6226,
      "num_input_tokens_seen": 26657536,
      "step": 19385
    },
    {
      "epoch": 0.6206388835541898,
      "grad_norm": 0.004962344188243151,
      "learning_rate": 0.15719541178844715,
      "loss": 0.5384,
      "num_input_tokens_seen": 26664288,
      "step": 19390
    },
    {
      "epoch": 0.6207989245246783,
      "grad_norm": 0.004025856498628855,
      "learning_rate": 0.15713657418386806,
      "loss": 0.5425,
      "num_input_tokens_seen": 26670816,
      "step": 19395
    },
    {
      "epoch": 0.6209589654951667,
      "grad_norm": 0.005074310582131147,
      "learning_rate": 0.15707773547873957,
      "loss": 0.4534,
      "num_input_tokens_seen": 26677504,
      "step": 19400
    },
    {
      "epoch": 0.6209589654951667,
      "eval_loss": 0.5455572605133057,
      "eval_runtime": 332.0467,
      "eval_samples_per_second": 41.819,
      "eval_steps_per_second": 20.91,
      "num_input_tokens_seen": 26677504,
      "step": 19400
    },
    {
      "epoch": 0.6211190064656552,
      "grad_norm": 0.004790824372321367,
      "learning_rate": 0.1570188956821353,
      "loss": 0.5032,
      "num_input_tokens_seen": 26684512,
      "step": 19405
    },
    {
      "epoch": 0.6212790474361437,
      "grad_norm": 0.004867835436016321,
      "learning_rate": 0.1569600548031291,
      "loss": 0.4892,
      "num_input_tokens_seen": 26691808,
      "step": 19410
    },
    {
      "epoch": 0.6214390884066321,
      "grad_norm": 0.004953769966959953,
      "learning_rate": 0.156901212850795,
      "loss": 0.5319,
      "num_input_tokens_seen": 26698560,
      "step": 19415
    },
    {
      "epoch": 0.6215991293771206,
      "grad_norm": 0.0062316362746059895,
      "learning_rate": 0.15684236983420716,
      "loss": 0.6235,
      "num_input_tokens_seen": 26705376,
      "step": 19420
    },
    {
      "epoch": 0.621759170347609,
      "grad_norm": 0.00871184654533863,
      "learning_rate": 0.1567835257624399,
      "loss": 0.5108,
      "num_input_tokens_seen": 26712304,
      "step": 19425
    },
    {
      "epoch": 0.6219192113180975,
      "grad_norm": 0.003981922287493944,
      "learning_rate": 0.1567246806445677,
      "loss": 0.5161,
      "num_input_tokens_seen": 26719360,
      "step": 19430
    },
    {
      "epoch": 0.6220792522885858,
      "grad_norm": 0.0066596041433513165,
      "learning_rate": 0.15666583448966526,
      "loss": 0.5092,
      "num_input_tokens_seen": 26726096,
      "step": 19435
    },
    {
      "epoch": 0.6222392932590743,
      "grad_norm": 0.004961438477039337,
      "learning_rate": 0.1566069873068074,
      "loss": 0.4345,
      "num_input_tokens_seen": 26732832,
      "step": 19440
    },
    {
      "epoch": 0.6223993342295627,
      "grad_norm": 0.00481338519603014,
      "learning_rate": 0.156548139105069,
      "loss": 0.5038,
      "num_input_tokens_seen": 26739360,
      "step": 19445
    },
    {
      "epoch": 0.6225593752000512,
      "grad_norm": 0.0032011810690164566,
      "learning_rate": 0.15648928989352529,
      "loss": 0.5377,
      "num_input_tokens_seen": 26746432,
      "step": 19450
    },
    {
      "epoch": 0.6227194161705396,
      "grad_norm": 0.0035405338276177645,
      "learning_rate": 0.15643043968125156,
      "loss": 0.7034,
      "num_input_tokens_seen": 26753344,
      "step": 19455
    },
    {
      "epoch": 0.6228794571410281,
      "grad_norm": 0.004374939948320389,
      "learning_rate": 0.15637158847732316,
      "loss": 0.4713,
      "num_input_tokens_seen": 26760000,
      "step": 19460
    },
    {
      "epoch": 0.6230394981115166,
      "grad_norm": 0.003913901746273041,
      "learning_rate": 0.15631273629081582,
      "loss": 0.4573,
      "num_input_tokens_seen": 26767088,
      "step": 19465
    },
    {
      "epoch": 0.623199539082005,
      "grad_norm": 0.006405324209481478,
      "learning_rate": 0.15625388313080518,
      "loss": 0.4643,
      "num_input_tokens_seen": 26774064,
      "step": 19470
    },
    {
      "epoch": 0.6233595800524935,
      "grad_norm": 0.004677442368119955,
      "learning_rate": 0.15619502900636714,
      "loss": 0.5237,
      "num_input_tokens_seen": 26780976,
      "step": 19475
    },
    {
      "epoch": 0.6235196210229819,
      "grad_norm": 0.0041069891303777695,
      "learning_rate": 0.15613617392657783,
      "loss": 0.5405,
      "num_input_tokens_seen": 26787840,
      "step": 19480
    },
    {
      "epoch": 0.6236796619934704,
      "grad_norm": 0.0035718020517379045,
      "learning_rate": 0.15607731790051335,
      "loss": 0.5508,
      "num_input_tokens_seen": 26794592,
      "step": 19485
    },
    {
      "epoch": 0.6238397029639587,
      "grad_norm": 0.004848483484238386,
      "learning_rate": 0.15601846093725008,
      "loss": 0.6445,
      "num_input_tokens_seen": 26801296,
      "step": 19490
    },
    {
      "epoch": 0.6239997439344472,
      "grad_norm": 0.004903629422187805,
      "learning_rate": 0.1559596030458645,
      "loss": 0.4969,
      "num_input_tokens_seen": 26807856,
      "step": 19495
    },
    {
      "epoch": 0.6241597849049356,
      "grad_norm": 0.00240941159427166,
      "learning_rate": 0.1559007442354333,
      "loss": 0.4739,
      "num_input_tokens_seen": 26814672,
      "step": 19500
    },
    {
      "epoch": 0.6243198258754241,
      "grad_norm": 0.006142422556877136,
      "learning_rate": 0.15584188451503314,
      "loss": 0.4864,
      "num_input_tokens_seen": 26821680,
      "step": 19505
    },
    {
      "epoch": 0.6244798668459125,
      "grad_norm": 0.0026405591052025557,
      "learning_rate": 0.15578302389374094,
      "loss": 0.661,
      "num_input_tokens_seen": 26827952,
      "step": 19510
    },
    {
      "epoch": 0.624639907816401,
      "grad_norm": 0.003971456550061703,
      "learning_rate": 0.1557241623806338,
      "loss": 0.3535,
      "num_input_tokens_seen": 26835072,
      "step": 19515
    },
    {
      "epoch": 0.6247999487868895,
      "grad_norm": 0.004246164578944445,
      "learning_rate": 0.15566529998478887,
      "loss": 0.769,
      "num_input_tokens_seen": 26841536,
      "step": 19520
    },
    {
      "epoch": 0.6249599897573779,
      "grad_norm": 0.005581814795732498,
      "learning_rate": 0.15560643671528354,
      "loss": 0.5335,
      "num_input_tokens_seen": 26848832,
      "step": 19525
    },
    {
      "epoch": 0.6251200307278664,
      "grad_norm": 0.004207768943160772,
      "learning_rate": 0.15554757258119514,
      "loss": 0.4538,
      "num_input_tokens_seen": 26855888,
      "step": 19530
    },
    {
      "epoch": 0.6252800716983548,
      "grad_norm": 0.004339316859841347,
      "learning_rate": 0.1554887075916014,
      "loss": 0.5262,
      "num_input_tokens_seen": 26862672,
      "step": 19535
    },
    {
      "epoch": 0.6254401126688433,
      "grad_norm": 0.004156963434070349,
      "learning_rate": 0.15542984175558,
      "loss": 0.5593,
      "num_input_tokens_seen": 26870224,
      "step": 19540
    },
    {
      "epoch": 0.6256001536393316,
      "grad_norm": 0.004578435327857733,
      "learning_rate": 0.1553709750822087,
      "loss": 0.4787,
      "num_input_tokens_seen": 26876736,
      "step": 19545
    },
    {
      "epoch": 0.6257601946098201,
      "grad_norm": 0.004464474506676197,
      "learning_rate": 0.15531210758056554,
      "loss": 0.3731,
      "num_input_tokens_seen": 26883936,
      "step": 19550
    },
    {
      "epoch": 0.6259202355803085,
      "grad_norm": 0.002794765867292881,
      "learning_rate": 0.15525323925972867,
      "loss": 0.4818,
      "num_input_tokens_seen": 26890912,
      "step": 19555
    },
    {
      "epoch": 0.626080276550797,
      "grad_norm": 0.004537642002105713,
      "learning_rate": 0.15519437012877627,
      "loss": 0.5432,
      "num_input_tokens_seen": 26897456,
      "step": 19560
    },
    {
      "epoch": 0.6262403175212854,
      "grad_norm": 0.0036569484509527683,
      "learning_rate": 0.15513550019678676,
      "loss": 0.549,
      "num_input_tokens_seen": 26904368,
      "step": 19565
    },
    {
      "epoch": 0.6264003584917739,
      "grad_norm": 0.007480185013264418,
      "learning_rate": 0.15507662947283854,
      "loss": 0.5726,
      "num_input_tokens_seen": 26911424,
      "step": 19570
    },
    {
      "epoch": 0.6265603994622624,
      "grad_norm": 0.004097169265151024,
      "learning_rate": 0.15501775796601028,
      "loss": 0.4673,
      "num_input_tokens_seen": 26918192,
      "step": 19575
    },
    {
      "epoch": 0.6267204404327508,
      "grad_norm": 0.006154763512313366,
      "learning_rate": 0.15495888568538066,
      "loss": 0.4879,
      "num_input_tokens_seen": 26924704,
      "step": 19580
    },
    {
      "epoch": 0.6268804814032393,
      "grad_norm": 0.00437847850844264,
      "learning_rate": 0.1549000126400286,
      "loss": 0.5239,
      "num_input_tokens_seen": 26931584,
      "step": 19585
    },
    {
      "epoch": 0.6270405223737276,
      "grad_norm": 0.004441389814019203,
      "learning_rate": 0.15484113883903294,
      "loss": 0.7142,
      "num_input_tokens_seen": 26938640,
      "step": 19590
    },
    {
      "epoch": 0.6272005633442161,
      "grad_norm": 0.007987109012901783,
      "learning_rate": 0.15478226429147288,
      "loss": 0.4286,
      "num_input_tokens_seen": 26945568,
      "step": 19595
    },
    {
      "epoch": 0.6273606043147045,
      "grad_norm": 0.0069553363136947155,
      "learning_rate": 0.15472338900642757,
      "loss": 0.5426,
      "num_input_tokens_seen": 26952544,
      "step": 19600
    },
    {
      "epoch": 0.6273606043147045,
      "eval_loss": 0.5538686513900757,
      "eval_runtime": 332.2185,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 26952544,
      "step": 19600
    },
    {
      "epoch": 0.627520645285193,
      "grad_norm": 0.00470711849629879,
      "learning_rate": 0.15466451299297632,
      "loss": 0.4844,
      "num_input_tokens_seen": 26959328,
      "step": 19605
    },
    {
      "epoch": 0.6276806862556814,
      "grad_norm": 0.006952259223908186,
      "learning_rate": 0.15460563626019852,
      "loss": 0.5404,
      "num_input_tokens_seen": 26965984,
      "step": 19610
    },
    {
      "epoch": 0.6278407272261699,
      "grad_norm": 0.004207391291856766,
      "learning_rate": 0.15454675881717375,
      "loss": 0.5101,
      "num_input_tokens_seen": 26972480,
      "step": 19615
    },
    {
      "epoch": 0.6280007681966584,
      "grad_norm": 0.002297166036441922,
      "learning_rate": 0.1544878806729816,
      "loss": 0.3504,
      "num_input_tokens_seen": 26979200,
      "step": 19620
    },
    {
      "epoch": 0.6281608091671468,
      "grad_norm": 0.003993507940322161,
      "learning_rate": 0.1544290018367019,
      "loss": 0.5939,
      "num_input_tokens_seen": 26986000,
      "step": 19625
    },
    {
      "epoch": 0.6283208501376353,
      "grad_norm": 0.01575605943799019,
      "learning_rate": 0.15437012231741445,
      "loss": 0.6277,
      "num_input_tokens_seen": 26993088,
      "step": 19630
    },
    {
      "epoch": 0.6284808911081237,
      "grad_norm": 0.003950702957808971,
      "learning_rate": 0.1543112421241992,
      "loss": 0.443,
      "num_input_tokens_seen": 26999904,
      "step": 19635
    },
    {
      "epoch": 0.6286409320786122,
      "grad_norm": 0.003144593443721533,
      "learning_rate": 0.15425236126613626,
      "loss": 0.6495,
      "num_input_tokens_seen": 27006576,
      "step": 19640
    },
    {
      "epoch": 0.6288009730491005,
      "grad_norm": 0.0038051162846386433,
      "learning_rate": 0.15419347975230577,
      "loss": 0.4178,
      "num_input_tokens_seen": 27013296,
      "step": 19645
    },
    {
      "epoch": 0.628961014019589,
      "grad_norm": 0.008974784053862095,
      "learning_rate": 0.154134597591788,
      "loss": 0.5574,
      "num_input_tokens_seen": 27020096,
      "step": 19650
    },
    {
      "epoch": 0.6291210549900774,
      "grad_norm": 0.0038889404386281967,
      "learning_rate": 0.1540757147936633,
      "loss": 0.5264,
      "num_input_tokens_seen": 27027072,
      "step": 19655
    },
    {
      "epoch": 0.6292810959605659,
      "grad_norm": 0.010173296555876732,
      "learning_rate": 0.1540168313670122,
      "loss": 0.585,
      "num_input_tokens_seen": 27033392,
      "step": 19660
    },
    {
      "epoch": 0.6294411369310543,
      "grad_norm": 0.004672233946621418,
      "learning_rate": 0.1539579473209152,
      "loss": 0.5719,
      "num_input_tokens_seen": 27040176,
      "step": 19665
    },
    {
      "epoch": 0.6296011779015428,
      "grad_norm": 0.0033212287817150354,
      "learning_rate": 0.15389906266445294,
      "loss": 0.4356,
      "num_input_tokens_seen": 27047312,
      "step": 19670
    },
    {
      "epoch": 0.6297612188720313,
      "grad_norm": 0.007506218738853931,
      "learning_rate": 0.15384017740670627,
      "loss": 0.5213,
      "num_input_tokens_seen": 27054144,
      "step": 19675
    },
    {
      "epoch": 0.6299212598425197,
      "grad_norm": 0.0055382270365953445,
      "learning_rate": 0.15378129155675602,
      "loss": 0.4873,
      "num_input_tokens_seen": 27060608,
      "step": 19680
    },
    {
      "epoch": 0.6300813008130082,
      "grad_norm": 0.0037403241731226444,
      "learning_rate": 0.15372240512368307,
      "loss": 0.7565,
      "num_input_tokens_seen": 27067168,
      "step": 19685
    },
    {
      "epoch": 0.6302413417834966,
      "grad_norm": 0.0023648939095437527,
      "learning_rate": 0.1536635181165684,
      "loss": 0.4081,
      "num_input_tokens_seen": 27074112,
      "step": 19690
    },
    {
      "epoch": 0.630401382753985,
      "grad_norm": 0.006531679071485996,
      "learning_rate": 0.15360463054449328,
      "loss": 0.6452,
      "num_input_tokens_seen": 27081040,
      "step": 19695
    },
    {
      "epoch": 0.6305614237244734,
      "grad_norm": 0.00893335696309805,
      "learning_rate": 0.1535457424165388,
      "loss": 0.5664,
      "num_input_tokens_seen": 27087792,
      "step": 19700
    },
    {
      "epoch": 0.6307214646949619,
      "grad_norm": 0.002367465989664197,
      "learning_rate": 0.15348685374178628,
      "loss": 0.3903,
      "num_input_tokens_seen": 27094752,
      "step": 19705
    },
    {
      "epoch": 0.6308815056654503,
      "grad_norm": 0.0018146511865779757,
      "learning_rate": 0.1534279645293171,
      "loss": 0.4741,
      "num_input_tokens_seen": 27101392,
      "step": 19710
    },
    {
      "epoch": 0.6310415466359388,
      "grad_norm": 0.0031041717156767845,
      "learning_rate": 0.1533690747882127,
      "loss": 0.7781,
      "num_input_tokens_seen": 27108624,
      "step": 19715
    },
    {
      "epoch": 0.6312015876064272,
      "grad_norm": 0.003811256028711796,
      "learning_rate": 0.15331018452755465,
      "loss": 0.5174,
      "num_input_tokens_seen": 27115600,
      "step": 19720
    },
    {
      "epoch": 0.6313616285769157,
      "grad_norm": 0.0036068472545593977,
      "learning_rate": 0.15325129375642457,
      "loss": 0.4397,
      "num_input_tokens_seen": 27122432,
      "step": 19725
    },
    {
      "epoch": 0.6315216695474042,
      "grad_norm": 0.004353434778749943,
      "learning_rate": 0.15319240248390406,
      "loss": 0.6742,
      "num_input_tokens_seen": 27129360,
      "step": 19730
    },
    {
      "epoch": 0.6316817105178926,
      "grad_norm": 0.005026424769312143,
      "learning_rate": 0.153133510719075,
      "loss": 0.5661,
      "num_input_tokens_seen": 27136432,
      "step": 19735
    },
    {
      "epoch": 0.6318417514883811,
      "grad_norm": 0.002160028088837862,
      "learning_rate": 0.15307461847101922,
      "loss": 0.3985,
      "num_input_tokens_seen": 27144176,
      "step": 19740
    },
    {
      "epoch": 0.6320017924588694,
      "grad_norm": 0.0033487416803836823,
      "learning_rate": 0.15301572574881864,
      "loss": 0.6347,
      "num_input_tokens_seen": 27150848,
      "step": 19745
    },
    {
      "epoch": 0.6321618334293579,
      "grad_norm": 0.00422600656747818,
      "learning_rate": 0.15295683256155523,
      "loss": 0.482,
      "num_input_tokens_seen": 27157504,
      "step": 19750
    },
    {
      "epoch": 0.6323218743998463,
      "grad_norm": 0.0030133959371596575,
      "learning_rate": 0.15289793891831113,
      "loss": 0.5689,
      "num_input_tokens_seen": 27164816,
      "step": 19755
    },
    {
      "epoch": 0.6324819153703348,
      "grad_norm": 0.0032749068923294544,
      "learning_rate": 0.15283904482816837,
      "loss": 0.5399,
      "num_input_tokens_seen": 27171696,
      "step": 19760
    },
    {
      "epoch": 0.6326419563408232,
      "grad_norm": 0.003016033675521612,
      "learning_rate": 0.15278015030020928,
      "loss": 0.6169,
      "num_input_tokens_seen": 27178288,
      "step": 19765
    },
    {
      "epoch": 0.6328019973113117,
      "grad_norm": 0.00354595179669559,
      "learning_rate": 0.152721255343516,
      "loss": 0.4259,
      "num_input_tokens_seen": 27185040,
      "step": 19770
    },
    {
      "epoch": 0.6329620382818001,
      "grad_norm": 0.007758378516882658,
      "learning_rate": 0.15266235996717098,
      "loss": 0.598,
      "num_input_tokens_seen": 27191824,
      "step": 19775
    },
    {
      "epoch": 0.6331220792522886,
      "grad_norm": 0.0030409134924411774,
      "learning_rate": 0.15260346418025664,
      "loss": 0.5064,
      "num_input_tokens_seen": 27198832,
      "step": 19780
    },
    {
      "epoch": 0.6332821202227771,
      "grad_norm": 0.006025029346346855,
      "learning_rate": 0.15254456799185537,
      "loss": 0.5751,
      "num_input_tokens_seen": 27205888,
      "step": 19785
    },
    {
      "epoch": 0.6334421611932655,
      "grad_norm": 0.008716498501598835,
      "learning_rate": 0.15248567141104974,
      "loss": 0.5757,
      "num_input_tokens_seen": 27213088,
      "step": 19790
    },
    {
      "epoch": 0.633602202163754,
      "grad_norm": 0.004430670291185379,
      "learning_rate": 0.15242677444692232,
      "loss": 0.5303,
      "num_input_tokens_seen": 27220128,
      "step": 19795
    },
    {
      "epoch": 0.6337622431342423,
      "grad_norm": 0.0035375894512981176,
      "learning_rate": 0.15236787710855584,
      "loss": 0.5724,
      "num_input_tokens_seen": 27226896,
      "step": 19800
    },
    {
      "epoch": 0.6337622431342423,
      "eval_loss": 0.542731523513794,
      "eval_runtime": 331.9437,
      "eval_samples_per_second": 41.832,
      "eval_steps_per_second": 20.916,
      "num_input_tokens_seen": 27226896,
      "step": 19800
    },
    {
      "epoch": 0.6339222841047308,
      "grad_norm": 0.005503396037966013,
      "learning_rate": 0.1523089794050329,
      "loss": 0.4917,
      "num_input_tokens_seen": 27233680,
      "step": 19805
    },
    {
      "epoch": 0.6340823250752192,
      "grad_norm": 0.005487470887601376,
      "learning_rate": 0.15225008134543633,
      "loss": 0.4279,
      "num_input_tokens_seen": 27240080,
      "step": 19810
    },
    {
      "epoch": 0.6342423660457077,
      "grad_norm": 0.00485277222469449,
      "learning_rate": 0.15219118293884895,
      "loss": 0.4383,
      "num_input_tokens_seen": 27247072,
      "step": 19815
    },
    {
      "epoch": 0.6344024070161961,
      "grad_norm": 0.004531573969870806,
      "learning_rate": 0.15213228419435362,
      "loss": 0.5112,
      "num_input_tokens_seen": 27253776,
      "step": 19820
    },
    {
      "epoch": 0.6345624479866846,
      "grad_norm": 0.006775734946131706,
      "learning_rate": 0.15207338512103327,
      "loss": 0.5812,
      "num_input_tokens_seen": 27260224,
      "step": 19825
    },
    {
      "epoch": 0.634722488957173,
      "grad_norm": 0.005297208670526743,
      "learning_rate": 0.1520144857279709,
      "loss": 0.4854,
      "num_input_tokens_seen": 27267456,
      "step": 19830
    },
    {
      "epoch": 0.6348825299276615,
      "grad_norm": 0.007001612335443497,
      "learning_rate": 0.1519555860242495,
      "loss": 0.4714,
      "num_input_tokens_seen": 27274096,
      "step": 19835
    },
    {
      "epoch": 0.63504257089815,
      "grad_norm": 0.00466171745210886,
      "learning_rate": 0.15189668601895218,
      "loss": 0.4845,
      "num_input_tokens_seen": 27280560,
      "step": 19840
    },
    {
      "epoch": 0.6352026118686384,
      "grad_norm": 0.0039022325072437525,
      "learning_rate": 0.151837785721162,
      "loss": 0.5826,
      "num_input_tokens_seen": 27287664,
      "step": 19845
    },
    {
      "epoch": 0.6353626528391269,
      "grad_norm": 0.004025063011795282,
      "learning_rate": 0.15177888513996218,
      "loss": 0.475,
      "num_input_tokens_seen": 27294416,
      "step": 19850
    },
    {
      "epoch": 0.6355226938096152,
      "grad_norm": 0.007645810022950172,
      "learning_rate": 0.15171998428443592,
      "loss": 0.5974,
      "num_input_tokens_seen": 27301376,
      "step": 19855
    },
    {
      "epoch": 0.6356827347801037,
      "grad_norm": 0.00502579053863883,
      "learning_rate": 0.1516610831636665,
      "loss": 0.4794,
      "num_input_tokens_seen": 27307968,
      "step": 19860
    },
    {
      "epoch": 0.6358427757505921,
      "grad_norm": 0.003808570560067892,
      "learning_rate": 0.15160218178673715,
      "loss": 0.5885,
      "num_input_tokens_seen": 27315472,
      "step": 19865
    },
    {
      "epoch": 0.6360028167210806,
      "grad_norm": 0.0031436148565262556,
      "learning_rate": 0.15154328016273122,
      "loss": 0.4452,
      "num_input_tokens_seen": 27321936,
      "step": 19870
    },
    {
      "epoch": 0.636162857691569,
      "grad_norm": 0.005259335972368717,
      "learning_rate": 0.1514843783007321,
      "loss": 0.617,
      "num_input_tokens_seen": 27328448,
      "step": 19875
    },
    {
      "epoch": 0.6363228986620575,
      "grad_norm": 0.008941903710365295,
      "learning_rate": 0.15142547620982322,
      "loss": 0.5557,
      "num_input_tokens_seen": 27335376,
      "step": 19880
    },
    {
      "epoch": 0.636482939632546,
      "grad_norm": 0.0044450294226408005,
      "learning_rate": 0.15136657389908797,
      "loss": 0.5251,
      "num_input_tokens_seen": 27342064,
      "step": 19885
    },
    {
      "epoch": 0.6366429806030344,
      "grad_norm": 0.0050749583169817924,
      "learning_rate": 0.15130767137760986,
      "loss": 0.6407,
      "num_input_tokens_seen": 27348800,
      "step": 19890
    },
    {
      "epoch": 0.6368030215735229,
      "grad_norm": 0.004796606954187155,
      "learning_rate": 0.15124876865447243,
      "loss": 0.4259,
      "num_input_tokens_seen": 27355664,
      "step": 19895
    },
    {
      "epoch": 0.6369630625440112,
      "grad_norm": 0.004717948380857706,
      "learning_rate": 0.15118986573875912,
      "loss": 0.64,
      "num_input_tokens_seen": 27362496,
      "step": 19900
    },
    {
      "epoch": 0.6371231035144997,
      "grad_norm": 0.003351303981617093,
      "learning_rate": 0.15113096263955358,
      "loss": 0.5385,
      "num_input_tokens_seen": 27369424,
      "step": 19905
    },
    {
      "epoch": 0.6372831444849881,
      "grad_norm": 0.006423551123589277,
      "learning_rate": 0.1510720593659394,
      "loss": 0.6615,
      "num_input_tokens_seen": 27376304,
      "step": 19910
    },
    {
      "epoch": 0.6374431854554766,
      "grad_norm": 0.005441918037831783,
      "learning_rate": 0.15101315592700015,
      "loss": 0.4279,
      "num_input_tokens_seen": 27383008,
      "step": 19915
    },
    {
      "epoch": 0.637603226425965,
      "grad_norm": 0.003469811985269189,
      "learning_rate": 0.15095425233181956,
      "loss": 0.4623,
      "num_input_tokens_seen": 27390112,
      "step": 19920
    },
    {
      "epoch": 0.6377632673964535,
      "grad_norm": 0.0062620737589895725,
      "learning_rate": 0.15089534858948128,
      "loss": 0.4452,
      "num_input_tokens_seen": 27396768,
      "step": 19925
    },
    {
      "epoch": 0.6379233083669419,
      "grad_norm": 0.003687231568619609,
      "learning_rate": 0.15083644470906898,
      "loss": 0.4867,
      "num_input_tokens_seen": 27403792,
      "step": 19930
    },
    {
      "epoch": 0.6380833493374304,
      "grad_norm": 0.004233143292367458,
      "learning_rate": 0.1507775406996664,
      "loss": 0.6633,
      "num_input_tokens_seen": 27410848,
      "step": 19935
    },
    {
      "epoch": 0.6382433903079189,
      "grad_norm": 0.0036824927665293217,
      "learning_rate": 0.15071863657035725,
      "loss": 0.3921,
      "num_input_tokens_seen": 27417792,
      "step": 19940
    },
    {
      "epoch": 0.6384034312784073,
      "grad_norm": 0.005747613497078419,
      "learning_rate": 0.15065973233022534,
      "loss": 0.3927,
      "num_input_tokens_seen": 27425104,
      "step": 19945
    },
    {
      "epoch": 0.6385634722488958,
      "grad_norm": 0.0031341155990958214,
      "learning_rate": 0.15060082798835442,
      "loss": 0.4132,
      "num_input_tokens_seen": 27431888,
      "step": 19950
    },
    {
      "epoch": 0.6387235132193841,
      "grad_norm": 0.004099000710994005,
      "learning_rate": 0.15054192355382823,
      "loss": 0.6036,
      "num_input_tokens_seen": 27439072,
      "step": 19955
    },
    {
      "epoch": 0.6388835541898726,
      "grad_norm": 0.006591307930648327,
      "learning_rate": 0.15048301903573066,
      "loss": 0.4871,
      "num_input_tokens_seen": 27445712,
      "step": 19960
    },
    {
      "epoch": 0.639043595160361,
      "grad_norm": 0.00631016306579113,
      "learning_rate": 0.15042411444314546,
      "loss": 0.4448,
      "num_input_tokens_seen": 27452576,
      "step": 19965
    },
    {
      "epoch": 0.6392036361308495,
      "grad_norm": 0.005312544759362936,
      "learning_rate": 0.1503652097851565,
      "loss": 0.5328,
      "num_input_tokens_seen": 27459328,
      "step": 19970
    },
    {
      "epoch": 0.6393636771013379,
      "grad_norm": 0.005764713976532221,
      "learning_rate": 0.15030630507084758,
      "loss": 0.7405,
      "num_input_tokens_seen": 27466816,
      "step": 19975
    },
    {
      "epoch": 0.6395237180718264,
      "grad_norm": 0.003921730909496546,
      "learning_rate": 0.1502474003093026,
      "loss": 0.538,
      "num_input_tokens_seen": 27473840,
      "step": 19980
    },
    {
      "epoch": 0.6396837590423148,
      "grad_norm": 0.005625847261399031,
      "learning_rate": 0.15018849550960536,
      "loss": 0.4915,
      "num_input_tokens_seen": 27480688,
      "step": 19985
    },
    {
      "epoch": 0.6398438000128033,
      "grad_norm": 0.007308529689908028,
      "learning_rate": 0.15012959068083975,
      "loss": 0.5104,
      "num_input_tokens_seen": 27487264,
      "step": 19990
    },
    {
      "epoch": 0.6400038409832918,
      "grad_norm": 0.00416825246065855,
      "learning_rate": 0.1500706858320896,
      "loss": 0.4352,
      "num_input_tokens_seen": 27494240,
      "step": 19995
    },
    {
      "epoch": 0.6401638819537802,
      "grad_norm": 0.003011684399098158,
      "learning_rate": 0.15001178097243886,
      "loss": 0.3026,
      "num_input_tokens_seen": 27501216,
      "step": 20000
    },
    {
      "epoch": 0.6401638819537802,
      "eval_loss": 0.5431657433509827,
      "eval_runtime": 331.9986,
      "eval_samples_per_second": 41.825,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 27501216,
      "step": 20000
    },
    {
      "epoch": 0.6403239229242687,
      "grad_norm": 0.002703511156141758,
      "learning_rate": 0.1499528761109713,
      "loss": 0.616,
      "num_input_tokens_seen": 27508000,
      "step": 20005
    },
    {
      "epoch": 0.640483963894757,
      "grad_norm": 0.003143349662423134,
      "learning_rate": 0.14989397125677087,
      "loss": 0.3949,
      "num_input_tokens_seen": 27514640,
      "step": 20010
    },
    {
      "epoch": 0.6406440048652455,
      "grad_norm": 0.004607456736266613,
      "learning_rate": 0.14983506641892141,
      "loss": 0.5045,
      "num_input_tokens_seen": 27521520,
      "step": 20015
    },
    {
      "epoch": 0.6408040458357339,
      "grad_norm": 0.0033372186589986086,
      "learning_rate": 0.14977616160650672,
      "loss": 0.5773,
      "num_input_tokens_seen": 27528176,
      "step": 20020
    },
    {
      "epoch": 0.6409640868062224,
      "grad_norm": 0.0036811085883527994,
      "learning_rate": 0.14971725682861076,
      "loss": 0.5977,
      "num_input_tokens_seen": 27534976,
      "step": 20025
    },
    {
      "epoch": 0.6411241277767108,
      "grad_norm": 0.005001916084438562,
      "learning_rate": 0.14965835209431738,
      "loss": 0.581,
      "num_input_tokens_seen": 27541968,
      "step": 20030
    },
    {
      "epoch": 0.6412841687471993,
      "grad_norm": 0.006111090071499348,
      "learning_rate": 0.14959944741271036,
      "loss": 0.6577,
      "num_input_tokens_seen": 27548640,
      "step": 20035
    },
    {
      "epoch": 0.6414442097176877,
      "grad_norm": 0.006001001223921776,
      "learning_rate": 0.14954054279287363,
      "loss": 0.6119,
      "num_input_tokens_seen": 27555296,
      "step": 20040
    },
    {
      "epoch": 0.6416042506881762,
      "grad_norm": 0.004835387226194143,
      "learning_rate": 0.14948163824389094,
      "loss": 0.3731,
      "num_input_tokens_seen": 27562272,
      "step": 20045
    },
    {
      "epoch": 0.6417642916586647,
      "grad_norm": 0.004345362074673176,
      "learning_rate": 0.14942273377484613,
      "loss": 0.6552,
      "num_input_tokens_seen": 27569072,
      "step": 20050
    },
    {
      "epoch": 0.641924332629153,
      "grad_norm": 0.0031327432952821255,
      "learning_rate": 0.1493638293948231,
      "loss": 0.3352,
      "num_input_tokens_seen": 27575888,
      "step": 20055
    },
    {
      "epoch": 0.6420843735996415,
      "grad_norm": 0.003838020609691739,
      "learning_rate": 0.14930492511290547,
      "loss": 0.5516,
      "num_input_tokens_seen": 27582480,
      "step": 20060
    },
    {
      "epoch": 0.6422444145701299,
      "grad_norm": 0.0056509459391236305,
      "learning_rate": 0.14924602093817715,
      "loss": 0.708,
      "num_input_tokens_seen": 27589696,
      "step": 20065
    },
    {
      "epoch": 0.6424044555406184,
      "grad_norm": 0.004440014250576496,
      "learning_rate": 0.14918711687972194,
      "loss": 0.4817,
      "num_input_tokens_seen": 27596512,
      "step": 20070
    },
    {
      "epoch": 0.6425644965111068,
      "grad_norm": 0.004296732600778341,
      "learning_rate": 0.14912821294662346,
      "loss": 0.5317,
      "num_input_tokens_seen": 27603152,
      "step": 20075
    },
    {
      "epoch": 0.6427245374815953,
      "grad_norm": 0.004800152499228716,
      "learning_rate": 0.14906930914796554,
      "loss": 0.6013,
      "num_input_tokens_seen": 27610256,
      "step": 20080
    },
    {
      "epoch": 0.6428845784520837,
      "grad_norm": 0.0028374367393553257,
      "learning_rate": 0.14901040549283182,
      "loss": 0.6233,
      "num_input_tokens_seen": 27617360,
      "step": 20085
    },
    {
      "epoch": 0.6430446194225722,
      "grad_norm": 0.003649009857326746,
      "learning_rate": 0.148951501990306,
      "loss": 0.405,
      "num_input_tokens_seen": 27624336,
      "step": 20090
    },
    {
      "epoch": 0.6432046603930607,
      "grad_norm": 0.007224451284855604,
      "learning_rate": 0.14889259864947177,
      "loss": 0.6431,
      "num_input_tokens_seen": 27631072,
      "step": 20095
    },
    {
      "epoch": 0.6433647013635491,
      "grad_norm": 0.004572770558297634,
      "learning_rate": 0.14883369547941272,
      "loss": 0.5467,
      "num_input_tokens_seen": 27637936,
      "step": 20100
    },
    {
      "epoch": 0.6435247423340376,
      "grad_norm": 0.0036653780844062567,
      "learning_rate": 0.14877479248921247,
      "loss": 0.5533,
      "num_input_tokens_seen": 27644704,
      "step": 20105
    },
    {
      "epoch": 0.6436847833045259,
      "grad_norm": 0.005683887284249067,
      "learning_rate": 0.14871588968795468,
      "loss": 0.7155,
      "num_input_tokens_seen": 27651184,
      "step": 20110
    },
    {
      "epoch": 0.6438448242750144,
      "grad_norm": 0.006050941068679094,
      "learning_rate": 0.1486569870847228,
      "loss": 0.6736,
      "num_input_tokens_seen": 27658896,
      "step": 20115
    },
    {
      "epoch": 0.6440048652455028,
      "grad_norm": 0.0030512302182614803,
      "learning_rate": 0.1485980846886004,
      "loss": 0.4216,
      "num_input_tokens_seen": 27665712,
      "step": 20120
    },
    {
      "epoch": 0.6441649062159913,
      "grad_norm": 0.0038531292229890823,
      "learning_rate": 0.14853918250867096,
      "loss": 0.5311,
      "num_input_tokens_seen": 27672512,
      "step": 20125
    },
    {
      "epoch": 0.6443249471864797,
      "grad_norm": 0.004472562111914158,
      "learning_rate": 0.1484802805540179,
      "loss": 0.4592,
      "num_input_tokens_seen": 27679968,
      "step": 20130
    },
    {
      "epoch": 0.6444849881569682,
      "grad_norm": 0.004187915939837694,
      "learning_rate": 0.14842137883372472,
      "loss": 0.4763,
      "num_input_tokens_seen": 27686672,
      "step": 20135
    },
    {
      "epoch": 0.6446450291274566,
      "grad_norm": 0.0034177706111222506,
      "learning_rate": 0.14836247735687474,
      "loss": 0.7866,
      "num_input_tokens_seen": 27694128,
      "step": 20140
    },
    {
      "epoch": 0.6448050700979451,
      "grad_norm": 0.0199247058480978,
      "learning_rate": 0.14830357613255132,
      "loss": 0.7333,
      "num_input_tokens_seen": 27701152,
      "step": 20145
    },
    {
      "epoch": 0.6449651110684336,
      "grad_norm": 0.008186361752450466,
      "learning_rate": 0.1482446751698378,
      "loss": 0.5676,
      "num_input_tokens_seen": 27708048,
      "step": 20150
    },
    {
      "epoch": 0.645125152038922,
      "grad_norm": 0.003623457858338952,
      "learning_rate": 0.14818577447781744,
      "loss": 0.5329,
      "num_input_tokens_seen": 27714672,
      "step": 20155
    },
    {
      "epoch": 0.6452851930094105,
      "grad_norm": 0.004610193893313408,
      "learning_rate": 0.14812687406557346,
      "loss": 0.7999,
      "num_input_tokens_seen": 27722048,
      "step": 20160
    },
    {
      "epoch": 0.6454452339798988,
      "grad_norm": 0.005185353569686413,
      "learning_rate": 0.14806797394218899,
      "loss": 0.5647,
      "num_input_tokens_seen": 27728960,
      "step": 20165
    },
    {
      "epoch": 0.6456052749503873,
      "grad_norm": 0.005200582090765238,
      "learning_rate": 0.1480090741167472,
      "loss": 0.5777,
      "num_input_tokens_seen": 27735440,
      "step": 20170
    },
    {
      "epoch": 0.6457653159208757,
      "grad_norm": 0.00766709866002202,
      "learning_rate": 0.1479501745983313,
      "loss": 0.5178,
      "num_input_tokens_seen": 27741952,
      "step": 20175
    },
    {
      "epoch": 0.6459253568913642,
      "grad_norm": 0.0053688096813857555,
      "learning_rate": 0.14789127539602415,
      "loss": 0.5965,
      "num_input_tokens_seen": 27748560,
      "step": 20180
    },
    {
      "epoch": 0.6460853978618526,
      "grad_norm": 0.005461941007524729,
      "learning_rate": 0.14783237651890885,
      "loss": 0.6544,
      "num_input_tokens_seen": 27755136,
      "step": 20185
    },
    {
      "epoch": 0.6462454388323411,
      "grad_norm": 0.003969215322285891,
      "learning_rate": 0.14777347797606838,
      "loss": 0.413,
      "num_input_tokens_seen": 27762064,
      "step": 20190
    },
    {
      "epoch": 0.6464054798028295,
      "grad_norm": 0.0049880570732057095,
      "learning_rate": 0.14771457977658553,
      "loss": 0.6087,
      "num_input_tokens_seen": 27769568,
      "step": 20195
    },
    {
      "epoch": 0.646565520773318,
      "grad_norm": 0.003202598774805665,
      "learning_rate": 0.14765568192954326,
      "loss": 0.5014,
      "num_input_tokens_seen": 27776624,
      "step": 20200
    },
    {
      "epoch": 0.646565520773318,
      "eval_loss": 0.5492203831672668,
      "eval_runtime": 331.7691,
      "eval_samples_per_second": 41.854,
      "eval_steps_per_second": 20.927,
      "num_input_tokens_seen": 27776624,
      "step": 20200
    },
    {
      "epoch": 0.6467255617438065,
      "grad_norm": 0.005833573639392853,
      "learning_rate": 0.14759678444402421,
      "loss": 0.5653,
      "num_input_tokens_seen": 27783584,
      "step": 20205
    },
    {
      "epoch": 0.6468856027142948,
      "grad_norm": 0.00451644929125905,
      "learning_rate": 0.14753788732911122,
      "loss": 0.5479,
      "num_input_tokens_seen": 27790848,
      "step": 20210
    },
    {
      "epoch": 0.6470456436847833,
      "grad_norm": 0.005363274831324816,
      "learning_rate": 0.147478990593887,
      "loss": 0.5217,
      "num_input_tokens_seen": 27797520,
      "step": 20215
    },
    {
      "epoch": 0.6472056846552717,
      "grad_norm": 0.002136073773726821,
      "learning_rate": 0.14742009424743405,
      "loss": 0.423,
      "num_input_tokens_seen": 27804560,
      "step": 20220
    },
    {
      "epoch": 0.6473657256257602,
      "grad_norm": 0.0047526247799396515,
      "learning_rate": 0.14736119829883504,
      "loss": 0.4116,
      "num_input_tokens_seen": 27811344,
      "step": 20225
    },
    {
      "epoch": 0.6475257665962486,
      "grad_norm": 0.002464618533849716,
      "learning_rate": 0.14730230275717243,
      "loss": 0.582,
      "num_input_tokens_seen": 27818320,
      "step": 20230
    },
    {
      "epoch": 0.6476858075667371,
      "grad_norm": 0.003807520493865013,
      "learning_rate": 0.14724340763152854,
      "loss": 0.3968,
      "num_input_tokens_seen": 27825200,
      "step": 20235
    },
    {
      "epoch": 0.6478458485372255,
      "grad_norm": 0.008808376267552376,
      "learning_rate": 0.14718451293098594,
      "loss": 0.6688,
      "num_input_tokens_seen": 27832416,
      "step": 20240
    },
    {
      "epoch": 0.648005889507714,
      "grad_norm": 0.004031692631542683,
      "learning_rate": 0.14712561866462676,
      "loss": 0.6142,
      "num_input_tokens_seen": 27838960,
      "step": 20245
    },
    {
      "epoch": 0.6481659304782024,
      "grad_norm": 0.007176682353019714,
      "learning_rate": 0.1470667248415333,
      "loss": 0.4673,
      "num_input_tokens_seen": 27846112,
      "step": 20250
    },
    {
      "epoch": 0.6483259714486909,
      "grad_norm": 0.005413153674453497,
      "learning_rate": 0.1470078314707878,
      "loss": 0.3901,
      "num_input_tokens_seen": 27853056,
      "step": 20255
    },
    {
      "epoch": 0.6484860124191794,
      "grad_norm": 0.002807577606290579,
      "learning_rate": 0.14694893856147223,
      "loss": 0.4752,
      "num_input_tokens_seen": 27860128,
      "step": 20260
    },
    {
      "epoch": 0.6486460533896677,
      "grad_norm": 0.006361464038491249,
      "learning_rate": 0.14689004612266868,
      "loss": 0.5813,
      "num_input_tokens_seen": 27867024,
      "step": 20265
    },
    {
      "epoch": 0.6488060943601562,
      "grad_norm": 0.006476602517068386,
      "learning_rate": 0.14683115416345913,
      "loss": 0.4777,
      "num_input_tokens_seen": 27874016,
      "step": 20270
    },
    {
      "epoch": 0.6489661353306446,
      "grad_norm": 0.004676918964833021,
      "learning_rate": 0.1467722626929254,
      "loss": 0.7316,
      "num_input_tokens_seen": 27880544,
      "step": 20275
    },
    {
      "epoch": 0.6491261763011331,
      "grad_norm": 0.0037626628763973713,
      "learning_rate": 0.14671337172014937,
      "loss": 0.4839,
      "num_input_tokens_seen": 27887136,
      "step": 20280
    },
    {
      "epoch": 0.6492862172716215,
      "grad_norm": 0.0070534017868340015,
      "learning_rate": 0.14665448125421265,
      "loss": 0.5571,
      "num_input_tokens_seen": 27893856,
      "step": 20285
    },
    {
      "epoch": 0.64944625824211,
      "grad_norm": 0.005505011882632971,
      "learning_rate": 0.146595591304197,
      "loss": 0.6324,
      "num_input_tokens_seen": 27900912,
      "step": 20290
    },
    {
      "epoch": 0.6496062992125984,
      "grad_norm": 0.0044318074360489845,
      "learning_rate": 0.14653670187918397,
      "loss": 0.5924,
      "num_input_tokens_seen": 27907760,
      "step": 20295
    },
    {
      "epoch": 0.6497663401830869,
      "grad_norm": 0.0033206602092832327,
      "learning_rate": 0.14647781298825502,
      "loss": 0.5408,
      "num_input_tokens_seen": 27914560,
      "step": 20300
    },
    {
      "epoch": 0.6499263811535753,
      "grad_norm": 0.002931746654212475,
      "learning_rate": 0.14641892464049153,
      "loss": 0.4841,
      "num_input_tokens_seen": 27921328,
      "step": 20305
    },
    {
      "epoch": 0.6500864221240638,
      "grad_norm": 0.004290154203772545,
      "learning_rate": 0.14636003684497495,
      "loss": 0.4656,
      "num_input_tokens_seen": 27928368,
      "step": 20310
    },
    {
      "epoch": 0.6502464630945523,
      "grad_norm": 0.004051018040627241,
      "learning_rate": 0.14630114961078636,
      "loss": 0.3615,
      "num_input_tokens_seen": 27935552,
      "step": 20315
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 0.0052358065731823444,
      "learning_rate": 0.14624226294700704,
      "loss": 0.6574,
      "num_input_tokens_seen": 27942592,
      "step": 20320
    },
    {
      "epoch": 0.6505665450355291,
      "grad_norm": 0.003594142384827137,
      "learning_rate": 0.14618337686271793,
      "loss": 0.4922,
      "num_input_tokens_seen": 27949216,
      "step": 20325
    },
    {
      "epoch": 0.6507265860060175,
      "grad_norm": 0.0051362169906497,
      "learning_rate": 0.1461244913670001,
      "loss": 0.415,
      "num_input_tokens_seen": 27955952,
      "step": 20330
    },
    {
      "epoch": 0.650886626976506,
      "grad_norm": 0.0036021636333316565,
      "learning_rate": 0.1460656064689344,
      "loss": 0.3866,
      "num_input_tokens_seen": 27962896,
      "step": 20335
    },
    {
      "epoch": 0.6510466679469944,
      "grad_norm": 0.008115858770906925,
      "learning_rate": 0.14600672217760163,
      "loss": 0.4329,
      "num_input_tokens_seen": 27969376,
      "step": 20340
    },
    {
      "epoch": 0.6512067089174829,
      "grad_norm": 0.002803467446938157,
      "learning_rate": 0.14594783850208248,
      "loss": 0.4148,
      "num_input_tokens_seen": 27976048,
      "step": 20345
    },
    {
      "epoch": 0.6513667498879713,
      "grad_norm": 0.004926056135445833,
      "learning_rate": 0.14588895545145758,
      "loss": 0.4495,
      "num_input_tokens_seen": 27982512,
      "step": 20350
    },
    {
      "epoch": 0.6515267908584598,
      "grad_norm": 0.003342719515785575,
      "learning_rate": 0.14583007303480738,
      "loss": 0.5956,
      "num_input_tokens_seen": 27989408,
      "step": 20355
    },
    {
      "epoch": 0.6516868318289483,
      "grad_norm": 0.0056270393542945385,
      "learning_rate": 0.14577119126121235,
      "loss": 0.5305,
      "num_input_tokens_seen": 27996272,
      "step": 20360
    },
    {
      "epoch": 0.6518468727994366,
      "grad_norm": 0.0035199997946619987,
      "learning_rate": 0.14571231013975272,
      "loss": 0.426,
      "num_input_tokens_seen": 28003184,
      "step": 20365
    },
    {
      "epoch": 0.6520069137699251,
      "grad_norm": 0.01117695402354002,
      "learning_rate": 0.1456534296795088,
      "loss": 0.7419,
      "num_input_tokens_seen": 28010256,
      "step": 20370
    },
    {
      "epoch": 0.6521669547404135,
      "grad_norm": 0.005310170352458954,
      "learning_rate": 0.14559454988956066,
      "loss": 0.537,
      "num_input_tokens_seen": 28017200,
      "step": 20375
    },
    {
      "epoch": 0.652326995710902,
      "grad_norm": 0.006064503453671932,
      "learning_rate": 0.1455356707789882,
      "loss": 0.5653,
      "num_input_tokens_seen": 28024096,
      "step": 20380
    },
    {
      "epoch": 0.6524870366813904,
      "grad_norm": 0.003206136403605342,
      "learning_rate": 0.14547679235687147,
      "loss": 0.3836,
      "num_input_tokens_seen": 28030864,
      "step": 20385
    },
    {
      "epoch": 0.6526470776518789,
      "grad_norm": 0.007544087246060371,
      "learning_rate": 0.14541791463229023,
      "loss": 0.7745,
      "num_input_tokens_seen": 28038464,
      "step": 20390
    },
    {
      "epoch": 0.6528071186223673,
      "grad_norm": 0.0023570936173200607,
      "learning_rate": 0.14535903761432406,
      "loss": 0.6786,
      "num_input_tokens_seen": 28045552,
      "step": 20395
    },
    {
      "epoch": 0.6529671595928558,
      "grad_norm": 0.0046915155835449696,
      "learning_rate": 0.1453001613120527,
      "loss": 0.5602,
      "num_input_tokens_seen": 28051872,
      "step": 20400
    },
    {
      "epoch": 0.6529671595928558,
      "eval_loss": 0.5472555160522461,
      "eval_runtime": 332.1509,
      "eval_samples_per_second": 41.806,
      "eval_steps_per_second": 20.903,
      "num_input_tokens_seen": 28051872,
      "step": 20400
    },
    {
      "epoch": 0.6531272005633442,
      "grad_norm": 0.003507008543238044,
      "learning_rate": 0.14524128573455547,
      "loss": 0.6699,
      "num_input_tokens_seen": 28058624,
      "step": 20405
    },
    {
      "epoch": 0.6532872415338327,
      "grad_norm": 0.0049717300571501255,
      "learning_rate": 0.14518241089091177,
      "loss": 0.4645,
      "num_input_tokens_seen": 28065472,
      "step": 20410
    },
    {
      "epoch": 0.6534472825043212,
      "grad_norm": 0.003537873039022088,
      "learning_rate": 0.1451235367902009,
      "loss": 0.4882,
      "num_input_tokens_seen": 28072576,
      "step": 20415
    },
    {
      "epoch": 0.6536073234748095,
      "grad_norm": 0.007430399768054485,
      "learning_rate": 0.1450646634415019,
      "loss": 0.7582,
      "num_input_tokens_seen": 28080464,
      "step": 20420
    },
    {
      "epoch": 0.653767364445298,
      "grad_norm": 0.0036338248755782843,
      "learning_rate": 0.1450057908538938,
      "loss": 0.5876,
      "num_input_tokens_seen": 28087888,
      "step": 20425
    },
    {
      "epoch": 0.6539274054157864,
      "grad_norm": 0.007758373394608498,
      "learning_rate": 0.14494691903645557,
      "loss": 0.5379,
      "num_input_tokens_seen": 28094736,
      "step": 20430
    },
    {
      "epoch": 0.6540874463862749,
      "grad_norm": 0.003017149632796645,
      "learning_rate": 0.14488804799826588,
      "loss": 0.5578,
      "num_input_tokens_seen": 28101936,
      "step": 20435
    },
    {
      "epoch": 0.6542474873567633,
      "grad_norm": 0.003348940284922719,
      "learning_rate": 0.14482917774840348,
      "loss": 0.6628,
      "num_input_tokens_seen": 28108576,
      "step": 20440
    },
    {
      "epoch": 0.6544075283272518,
      "grad_norm": 0.0041002389043569565,
      "learning_rate": 0.14477030829594684,
      "loss": 0.6272,
      "num_input_tokens_seen": 28115680,
      "step": 20445
    },
    {
      "epoch": 0.6545675692977402,
      "grad_norm": 0.0047668698243796825,
      "learning_rate": 0.14471143964997432,
      "loss": 0.6891,
      "num_input_tokens_seen": 28122416,
      "step": 20450
    },
    {
      "epoch": 0.6547276102682287,
      "grad_norm": 0.0072285765781998634,
      "learning_rate": 0.14465257181956434,
      "loss": 0.6577,
      "num_input_tokens_seen": 28129136,
      "step": 20455
    },
    {
      "epoch": 0.6548876512387171,
      "grad_norm": 0.005993380211293697,
      "learning_rate": 0.1445937048137949,
      "loss": 0.6955,
      "num_input_tokens_seen": 28136016,
      "step": 20460
    },
    {
      "epoch": 0.6550476922092056,
      "grad_norm": 0.006365607958287001,
      "learning_rate": 0.14453483864174416,
      "loss": 0.4571,
      "num_input_tokens_seen": 28142640,
      "step": 20465
    },
    {
      "epoch": 0.655207733179694,
      "grad_norm": 0.003138079773634672,
      "learning_rate": 0.14447597331249,
      "loss": 0.6075,
      "num_input_tokens_seen": 28149680,
      "step": 20470
    },
    {
      "epoch": 0.6553677741501824,
      "grad_norm": 0.004626384470611811,
      "learning_rate": 0.1444171088351102,
      "loss": 0.4659,
      "num_input_tokens_seen": 28156256,
      "step": 20475
    },
    {
      "epoch": 0.6555278151206709,
      "grad_norm": 0.003031849628314376,
      "learning_rate": 0.14435824521868235,
      "loss": 0.3823,
      "num_input_tokens_seen": 28163024,
      "step": 20480
    },
    {
      "epoch": 0.6556878560911593,
      "grad_norm": 0.004499088507145643,
      "learning_rate": 0.14429938247228397,
      "loss": 0.4837,
      "num_input_tokens_seen": 28170112,
      "step": 20485
    },
    {
      "epoch": 0.6558478970616478,
      "grad_norm": 0.006202454213052988,
      "learning_rate": 0.14424052060499243,
      "loss": 0.4953,
      "num_input_tokens_seen": 28176736,
      "step": 20490
    },
    {
      "epoch": 0.6560079380321362,
      "grad_norm": 0.0032803660724312067,
      "learning_rate": 0.14418165962588506,
      "loss": 0.4659,
      "num_input_tokens_seen": 28183328,
      "step": 20495
    },
    {
      "epoch": 0.6561679790026247,
      "grad_norm": 0.002532265381887555,
      "learning_rate": 0.1441227995440388,
      "loss": 0.5929,
      "num_input_tokens_seen": 28190112,
      "step": 20500
    },
    {
      "epoch": 0.6563280199731131,
      "grad_norm": 0.003148794872686267,
      "learning_rate": 0.14406394036853082,
      "loss": 0.3348,
      "num_input_tokens_seen": 28196640,
      "step": 20505
    },
    {
      "epoch": 0.6564880609436016,
      "grad_norm": 0.007862764410674572,
      "learning_rate": 0.14400508210843774,
      "loss": 0.5118,
      "num_input_tokens_seen": 28203296,
      "step": 20510
    },
    {
      "epoch": 0.65664810191409,
      "grad_norm": 0.006400418467819691,
      "learning_rate": 0.1439462247728364,
      "loss": 0.601,
      "num_input_tokens_seen": 28210016,
      "step": 20515
    },
    {
      "epoch": 0.6568081428845784,
      "grad_norm": 0.0038117333315312862,
      "learning_rate": 0.14388736837080326,
      "loss": 0.5944,
      "num_input_tokens_seen": 28216512,
      "step": 20520
    },
    {
      "epoch": 0.656968183855067,
      "grad_norm": 0.006184394471347332,
      "learning_rate": 0.14382851291141469,
      "loss": 0.5132,
      "num_input_tokens_seen": 28223248,
      "step": 20525
    },
    {
      "epoch": 0.6571282248255553,
      "grad_norm": 0.0049441843293607235,
      "learning_rate": 0.14376965840374697,
      "loss": 0.5025,
      "num_input_tokens_seen": 28229968,
      "step": 20530
    },
    {
      "epoch": 0.6572882657960438,
      "grad_norm": 0.004952819552272558,
      "learning_rate": 0.14371080485687632,
      "loss": 0.4158,
      "num_input_tokens_seen": 28236768,
      "step": 20535
    },
    {
      "epoch": 0.6574483067665322,
      "grad_norm": 0.003702745772898197,
      "learning_rate": 0.1436519522798785,
      "loss": 0.5079,
      "num_input_tokens_seen": 28243424,
      "step": 20540
    },
    {
      "epoch": 0.6576083477370207,
      "grad_norm": 0.0051313117146492004,
      "learning_rate": 0.14359310068182948,
      "loss": 0.652,
      "num_input_tokens_seen": 28250464,
      "step": 20545
    },
    {
      "epoch": 0.6577683887075091,
      "grad_norm": 0.0046921041794121265,
      "learning_rate": 0.14353425007180484,
      "loss": 0.5983,
      "num_input_tokens_seen": 28257552,
      "step": 20550
    },
    {
      "epoch": 0.6579284296779976,
      "grad_norm": 0.003551506670191884,
      "learning_rate": 0.14347540045888005,
      "loss": 0.608,
      "num_input_tokens_seen": 28264416,
      "step": 20555
    },
    {
      "epoch": 0.658088470648486,
      "grad_norm": 0.004052640870213509,
      "learning_rate": 0.14341655185213056,
      "loss": 0.6002,
      "num_input_tokens_seen": 28271168,
      "step": 20560
    },
    {
      "epoch": 0.6582485116189745,
      "grad_norm": 0.002800090704113245,
      "learning_rate": 0.14335770426063144,
      "loss": 0.3799,
      "num_input_tokens_seen": 28277712,
      "step": 20565
    },
    {
      "epoch": 0.6584085525894628,
      "grad_norm": 0.0039388928562402725,
      "learning_rate": 0.1432988576934578,
      "loss": 0.4382,
      "num_input_tokens_seen": 28284416,
      "step": 20570
    },
    {
      "epoch": 0.6585685935599513,
      "grad_norm": 0.0032344236969947815,
      "learning_rate": 0.14324001215968457,
      "loss": 0.3365,
      "num_input_tokens_seen": 28291184,
      "step": 20575
    },
    {
      "epoch": 0.6587286345304398,
      "grad_norm": 0.005505464505404234,
      "learning_rate": 0.14318116766838637,
      "loss": 0.3993,
      "num_input_tokens_seen": 28298048,
      "step": 20580
    },
    {
      "epoch": 0.6588886755009282,
      "grad_norm": 0.009119297377765179,
      "learning_rate": 0.14312232422863788,
      "loss": 0.5611,
      "num_input_tokens_seen": 28304704,
      "step": 20585
    },
    {
      "epoch": 0.6590487164714167,
      "grad_norm": 0.008060352876782417,
      "learning_rate": 0.14306348184951334,
      "loss": 0.7472,
      "num_input_tokens_seen": 28311872,
      "step": 20590
    },
    {
      "epoch": 0.6592087574419051,
      "grad_norm": 0.00706221628934145,
      "learning_rate": 0.1430046405400871,
      "loss": 0.504,
      "num_input_tokens_seen": 28318576,
      "step": 20595
    },
    {
      "epoch": 0.6593687984123936,
      "grad_norm": 0.004316028673201799,
      "learning_rate": 0.14294580030943324,
      "loss": 0.6304,
      "num_input_tokens_seen": 28325632,
      "step": 20600
    },
    {
      "epoch": 0.6593687984123936,
      "eval_loss": 0.5416109561920166,
      "eval_runtime": 332.2772,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 28325632,
      "step": 20600
    },
    {
      "epoch": 0.659528839382882,
      "grad_norm": 0.008557579480111599,
      "learning_rate": 0.14288696116662553,
      "loss": 0.6734,
      "num_input_tokens_seen": 28332336,
      "step": 20605
    },
    {
      "epoch": 0.6596888803533705,
      "grad_norm": 0.004908262751996517,
      "learning_rate": 0.1428281231207378,
      "loss": 0.4024,
      "num_input_tokens_seen": 28339344,
      "step": 20610
    },
    {
      "epoch": 0.6598489213238589,
      "grad_norm": 0.005127147771418095,
      "learning_rate": 0.1427692861808437,
      "loss": 0.5025,
      "num_input_tokens_seen": 28346208,
      "step": 20615
    },
    {
      "epoch": 0.6600089622943474,
      "grad_norm": 0.00405283086001873,
      "learning_rate": 0.1427104503560165,
      "loss": 0.38,
      "num_input_tokens_seen": 28353312,
      "step": 20620
    },
    {
      "epoch": 0.6601690032648359,
      "grad_norm": 0.004445621278136969,
      "learning_rate": 0.14265161565532947,
      "loss": 0.4697,
      "num_input_tokens_seen": 28360304,
      "step": 20625
    },
    {
      "epoch": 0.6603290442353242,
      "grad_norm": 0.005539124831557274,
      "learning_rate": 0.14259278208785564,
      "loss": 0.6163,
      "num_input_tokens_seen": 28367184,
      "step": 20630
    },
    {
      "epoch": 0.6604890852058127,
      "grad_norm": 0.005993111990392208,
      "learning_rate": 0.14253394966266789,
      "loss": 0.5433,
      "num_input_tokens_seen": 28373968,
      "step": 20635
    },
    {
      "epoch": 0.6606491261763011,
      "grad_norm": 0.0065610348246991634,
      "learning_rate": 0.14247511838883894,
      "loss": 0.7026,
      "num_input_tokens_seen": 28380672,
      "step": 20640
    },
    {
      "epoch": 0.6608091671467896,
      "grad_norm": 0.00638887844979763,
      "learning_rate": 0.14241628827544126,
      "loss": 0.5717,
      "num_input_tokens_seen": 28387264,
      "step": 20645
    },
    {
      "epoch": 0.660969208117278,
      "grad_norm": 0.005395316053181887,
      "learning_rate": 0.14235745933154723,
      "loss": 0.5035,
      "num_input_tokens_seen": 28393584,
      "step": 20650
    },
    {
      "epoch": 0.6611292490877665,
      "grad_norm": 0.005130899604409933,
      "learning_rate": 0.14229863156622907,
      "loss": 0.5021,
      "num_input_tokens_seen": 28400560,
      "step": 20655
    },
    {
      "epoch": 0.6612892900582549,
      "grad_norm": 0.0031878864392638206,
      "learning_rate": 0.14223980498855868,
      "loss": 0.4711,
      "num_input_tokens_seen": 28407232,
      "step": 20660
    },
    {
      "epoch": 0.6614493310287434,
      "grad_norm": 0.0051816427148878574,
      "learning_rate": 0.14218097960760792,
      "loss": 0.4562,
      "num_input_tokens_seen": 28413936,
      "step": 20665
    },
    {
      "epoch": 0.6616093719992318,
      "grad_norm": 0.00612757308408618,
      "learning_rate": 0.1421221554324483,
      "loss": 0.5174,
      "num_input_tokens_seen": 28420880,
      "step": 20670
    },
    {
      "epoch": 0.6617694129697202,
      "grad_norm": 0.0033944975584745407,
      "learning_rate": 0.1420633324721513,
      "loss": 0.5712,
      "num_input_tokens_seen": 28427568,
      "step": 20675
    },
    {
      "epoch": 0.6619294539402087,
      "grad_norm": 0.0038730984088033438,
      "learning_rate": 0.14200451073578824,
      "loss": 0.4305,
      "num_input_tokens_seen": 28434400,
      "step": 20680
    },
    {
      "epoch": 0.6620894949106971,
      "grad_norm": 0.007936517708003521,
      "learning_rate": 0.14194569023243003,
      "loss": 0.8052,
      "num_input_tokens_seen": 28441360,
      "step": 20685
    },
    {
      "epoch": 0.6622495358811856,
      "grad_norm": 0.005371889099478722,
      "learning_rate": 0.14188687097114766,
      "loss": 0.5689,
      "num_input_tokens_seen": 28448160,
      "step": 20690
    },
    {
      "epoch": 0.662409576851674,
      "grad_norm": 0.0036231481935828924,
      "learning_rate": 0.14182805296101172,
      "loss": 0.4946,
      "num_input_tokens_seen": 28455152,
      "step": 20695
    },
    {
      "epoch": 0.6625696178221625,
      "grad_norm": 0.004775784909725189,
      "learning_rate": 0.14176923621109272,
      "loss": 0.487,
      "num_input_tokens_seen": 28461760,
      "step": 20700
    },
    {
      "epoch": 0.6627296587926509,
      "grad_norm": 0.007578725926578045,
      "learning_rate": 0.14171042073046097,
      "loss": 0.4921,
      "num_input_tokens_seen": 28468416,
      "step": 20705
    },
    {
      "epoch": 0.6628896997631394,
      "grad_norm": 0.0036662451457232237,
      "learning_rate": 0.14165160652818642,
      "loss": 0.6567,
      "num_input_tokens_seen": 28475536,
      "step": 20710
    },
    {
      "epoch": 0.6630497407336278,
      "grad_norm": 0.008098173886537552,
      "learning_rate": 0.14159279361333907,
      "loss": 0.7953,
      "num_input_tokens_seen": 28482496,
      "step": 20715
    },
    {
      "epoch": 0.6632097817041163,
      "grad_norm": 0.004194494336843491,
      "learning_rate": 0.14153398199498868,
      "loss": 0.4279,
      "num_input_tokens_seen": 28489568,
      "step": 20720
    },
    {
      "epoch": 0.6633698226746046,
      "grad_norm": 0.003691134275868535,
      "learning_rate": 0.14147517168220458,
      "loss": 0.5234,
      "num_input_tokens_seen": 28496624,
      "step": 20725
    },
    {
      "epoch": 0.6635298636450931,
      "grad_norm": 0.004997500218451023,
      "learning_rate": 0.14141636268405616,
      "loss": 0.5499,
      "num_input_tokens_seen": 28503696,
      "step": 20730
    },
    {
      "epoch": 0.6636899046155816,
      "grad_norm": 0.005068746395409107,
      "learning_rate": 0.14135755500961253,
      "loss": 0.461,
      "num_input_tokens_seen": 28510480,
      "step": 20735
    },
    {
      "epoch": 0.66384994558607,
      "grad_norm": 0.00542415538802743,
      "learning_rate": 0.14129874866794245,
      "loss": 0.4662,
      "num_input_tokens_seen": 28517136,
      "step": 20740
    },
    {
      "epoch": 0.6640099865565585,
      "grad_norm": 0.002607134636491537,
      "learning_rate": 0.14123994366811476,
      "loss": 0.3912,
      "num_input_tokens_seen": 28524144,
      "step": 20745
    },
    {
      "epoch": 0.6641700275270469,
      "grad_norm": 0.0032468887511640787,
      "learning_rate": 0.14118114001919774,
      "loss": 0.4899,
      "num_input_tokens_seen": 28530976,
      "step": 20750
    },
    {
      "epoch": 0.6643300684975354,
      "grad_norm": 0.005784339737147093,
      "learning_rate": 0.14112233773025978,
      "loss": 0.4792,
      "num_input_tokens_seen": 28537952,
      "step": 20755
    },
    {
      "epoch": 0.6644901094680238,
      "grad_norm": 0.0063229999504983425,
      "learning_rate": 0.14106353681036896,
      "loss": 0.6913,
      "num_input_tokens_seen": 28544816,
      "step": 20760
    },
    {
      "epoch": 0.6646501504385123,
      "grad_norm": 0.003179202089086175,
      "learning_rate": 0.14100473726859303,
      "loss": 0.4272,
      "num_input_tokens_seen": 28551136,
      "step": 20765
    },
    {
      "epoch": 0.6648101914090007,
      "grad_norm": 0.0038595860823988914,
      "learning_rate": 0.14094593911399964,
      "loss": 0.4536,
      "num_input_tokens_seen": 28558304,
      "step": 20770
    },
    {
      "epoch": 0.6649702323794892,
      "grad_norm": 0.0035756065044552088,
      "learning_rate": 0.14088714235565625,
      "loss": 0.4455,
      "num_input_tokens_seen": 28564848,
      "step": 20775
    },
    {
      "epoch": 0.6651302733499775,
      "grad_norm": 0.0035401317290961742,
      "learning_rate": 0.14082834700263,
      "loss": 0.4282,
      "num_input_tokens_seen": 28572032,
      "step": 20780
    },
    {
      "epoch": 0.665290314320466,
      "grad_norm": 0.003673882456496358,
      "learning_rate": 0.14076955306398795,
      "loss": 0.45,
      "num_input_tokens_seen": 28578736,
      "step": 20785
    },
    {
      "epoch": 0.6654503552909545,
      "grad_norm": 0.0042344992980360985,
      "learning_rate": 0.14071076054879675,
      "loss": 0.3903,
      "num_input_tokens_seen": 28585456,
      "step": 20790
    },
    {
      "epoch": 0.6656103962614429,
      "grad_norm": 0.0028979319613426924,
      "learning_rate": 0.14065196946612302,
      "loss": 0.5142,
      "num_input_tokens_seen": 28592432,
      "step": 20795
    },
    {
      "epoch": 0.6657704372319314,
      "grad_norm": 0.003646980505436659,
      "learning_rate": 0.1405931798250331,
      "loss": 0.5525,
      "num_input_tokens_seen": 28598784,
      "step": 20800
    },
    {
      "epoch": 0.6657704372319314,
      "eval_loss": 0.550563395023346,
      "eval_runtime": 331.9577,
      "eval_samples_per_second": 41.831,
      "eval_steps_per_second": 20.915,
      "num_input_tokens_seen": 28598784,
      "step": 20800
    },
    {
      "epoch": 0.6659304782024198,
      "grad_norm": 0.006464644335210323,
      "learning_rate": 0.14053439163459308,
      "loss": 0.6036,
      "num_input_tokens_seen": 28605568,
      "step": 20805
    },
    {
      "epoch": 0.6660905191729083,
      "grad_norm": 0.004812302999198437,
      "learning_rate": 0.14047560490386876,
      "loss": 0.7748,
      "num_input_tokens_seen": 28613184,
      "step": 20810
    },
    {
      "epoch": 0.6662505601433967,
      "grad_norm": 0.0037433214019984007,
      "learning_rate": 0.14041681964192593,
      "loss": 0.4629,
      "num_input_tokens_seen": 28619872,
      "step": 20815
    },
    {
      "epoch": 0.6664106011138852,
      "grad_norm": 0.0039202980697155,
      "learning_rate": 0.14035803585782988,
      "loss": 0.3806,
      "num_input_tokens_seen": 28626960,
      "step": 20820
    },
    {
      "epoch": 0.6665706420843736,
      "grad_norm": 0.008079049177467823,
      "learning_rate": 0.14029925356064593,
      "loss": 0.7637,
      "num_input_tokens_seen": 28633776,
      "step": 20825
    },
    {
      "epoch": 0.666730683054862,
      "grad_norm": 0.004740612115710974,
      "learning_rate": 0.1402404727594389,
      "loss": 0.487,
      "num_input_tokens_seen": 28640576,
      "step": 20830
    },
    {
      "epoch": 0.6668907240253504,
      "grad_norm": 0.00398792140185833,
      "learning_rate": 0.1401816934632737,
      "loss": 0.3894,
      "num_input_tokens_seen": 28647552,
      "step": 20835
    },
    {
      "epoch": 0.6670507649958389,
      "grad_norm": 0.005302645266056061,
      "learning_rate": 0.1401229156812147,
      "loss": 0.4645,
      "num_input_tokens_seen": 28654608,
      "step": 20840
    },
    {
      "epoch": 0.6672108059663274,
      "grad_norm": 0.003033103421330452,
      "learning_rate": 0.14006413942232626,
      "loss": 0.4858,
      "num_input_tokens_seen": 28661520,
      "step": 20845
    },
    {
      "epoch": 0.6673708469368158,
      "grad_norm": 0.0033966689370572567,
      "learning_rate": 0.14000536469567235,
      "loss": 0.5278,
      "num_input_tokens_seen": 28668304,
      "step": 20850
    },
    {
      "epoch": 0.6675308879073043,
      "grad_norm": 0.00812900997698307,
      "learning_rate": 0.13994659151031685,
      "loss": 0.6251,
      "num_input_tokens_seen": 28675328,
      "step": 20855
    },
    {
      "epoch": 0.6676909288777927,
      "grad_norm": 0.005666601005941629,
      "learning_rate": 0.13988781987532323,
      "loss": 0.5166,
      "num_input_tokens_seen": 28682432,
      "step": 20860
    },
    {
      "epoch": 0.6678509698482812,
      "grad_norm": 0.005306627135723829,
      "learning_rate": 0.1398290497997549,
      "loss": 0.4533,
      "num_input_tokens_seen": 28689008,
      "step": 20865
    },
    {
      "epoch": 0.6680110108187696,
      "grad_norm": 0.004493619780987501,
      "learning_rate": 0.13977028129267488,
      "loss": 0.4823,
      "num_input_tokens_seen": 28695568,
      "step": 20870
    },
    {
      "epoch": 0.6681710517892581,
      "grad_norm": 0.003879901487380266,
      "learning_rate": 0.13971151436314605,
      "loss": 0.7518,
      "num_input_tokens_seen": 28702512,
      "step": 20875
    },
    {
      "epoch": 0.6683310927597464,
      "grad_norm": 0.004232262726873159,
      "learning_rate": 0.13965274902023103,
      "loss": 0.6152,
      "num_input_tokens_seen": 28709168,
      "step": 20880
    },
    {
      "epoch": 0.6684911337302349,
      "grad_norm": 0.003277064301073551,
      "learning_rate": 0.13959398527299208,
      "loss": 0.4349,
      "num_input_tokens_seen": 28715888,
      "step": 20885
    },
    {
      "epoch": 0.6686511747007234,
      "grad_norm": 0.005668079946190119,
      "learning_rate": 0.13953522313049138,
      "loss": 0.4688,
      "num_input_tokens_seen": 28722672,
      "step": 20890
    },
    {
      "epoch": 0.6688112156712118,
      "grad_norm": 0.003868643194437027,
      "learning_rate": 0.13947646260179083,
      "loss": 0.6229,
      "num_input_tokens_seen": 28729456,
      "step": 20895
    },
    {
      "epoch": 0.6689712566417003,
      "grad_norm": 0.0035560321994125843,
      "learning_rate": 0.13941770369595194,
      "loss": 0.4591,
      "num_input_tokens_seen": 28735904,
      "step": 20900
    },
    {
      "epoch": 0.6691312976121887,
      "grad_norm": 0.0044164578430354595,
      "learning_rate": 0.1393589464220362,
      "loss": 0.3711,
      "num_input_tokens_seen": 28743024,
      "step": 20905
    },
    {
      "epoch": 0.6692913385826772,
      "grad_norm": 0.002475676592439413,
      "learning_rate": 0.13930019078910455,
      "loss": 0.4138,
      "num_input_tokens_seen": 28750112,
      "step": 20910
    },
    {
      "epoch": 0.6694513795531656,
      "grad_norm": 0.006959996651858091,
      "learning_rate": 0.139241436806218,
      "loss": 0.6429,
      "num_input_tokens_seen": 28756880,
      "step": 20915
    },
    {
      "epoch": 0.6696114205236541,
      "grad_norm": 0.006394444033503532,
      "learning_rate": 0.13918268448243712,
      "loss": 0.5254,
      "num_input_tokens_seen": 28763584,
      "step": 20920
    },
    {
      "epoch": 0.6697714614941425,
      "grad_norm": 0.010175780393183231,
      "learning_rate": 0.13912393382682217,
      "loss": 0.7705,
      "num_input_tokens_seen": 28770720,
      "step": 20925
    },
    {
      "epoch": 0.669931502464631,
      "grad_norm": 0.005030809435993433,
      "learning_rate": 0.1390651848484333,
      "loss": 0.5512,
      "num_input_tokens_seen": 28777648,
      "step": 20930
    },
    {
      "epoch": 0.6700915434351193,
      "grad_norm": 0.003682781709358096,
      "learning_rate": 0.1390064375563304,
      "loss": 0.6504,
      "num_input_tokens_seen": 28784576,
      "step": 20935
    },
    {
      "epoch": 0.6702515844056078,
      "grad_norm": 0.005938766524195671,
      "learning_rate": 0.13894769195957293,
      "loss": 0.5527,
      "num_input_tokens_seen": 28791376,
      "step": 20940
    },
    {
      "epoch": 0.6704116253760963,
      "grad_norm": 0.0021515153348445892,
      "learning_rate": 0.13888894806722032,
      "loss": 0.3502,
      "num_input_tokens_seen": 28798144,
      "step": 20945
    },
    {
      "epoch": 0.6705716663465847,
      "grad_norm": 0.0037630225997418165,
      "learning_rate": 0.1388302058883315,
      "loss": 0.6265,
      "num_input_tokens_seen": 28805072,
      "step": 20950
    },
    {
      "epoch": 0.6707317073170732,
      "grad_norm": 0.006391898263245821,
      "learning_rate": 0.13877146543196528,
      "loss": 0.6325,
      "num_input_tokens_seen": 28812064,
      "step": 20955
    },
    {
      "epoch": 0.6708917482875616,
      "grad_norm": 0.008020552806556225,
      "learning_rate": 0.13871272670718027,
      "loss": 0.6544,
      "num_input_tokens_seen": 28819040,
      "step": 20960
    },
    {
      "epoch": 0.6710517892580501,
      "grad_norm": 0.00308881513774395,
      "learning_rate": 0.13865398972303455,
      "loss": 0.5167,
      "num_input_tokens_seen": 28826208,
      "step": 20965
    },
    {
      "epoch": 0.6712118302285385,
      "grad_norm": 0.004417970776557922,
      "learning_rate": 0.13859525448858623,
      "loss": 0.5777,
      "num_input_tokens_seen": 28833216,
      "step": 20970
    },
    {
      "epoch": 0.671371871199027,
      "grad_norm": 0.002318910788744688,
      "learning_rate": 0.13853652101289304,
      "loss": 0.5386,
      "num_input_tokens_seen": 28840112,
      "step": 20975
    },
    {
      "epoch": 0.6715319121695154,
      "grad_norm": 0.00399438664317131,
      "learning_rate": 0.13847778930501234,
      "loss": 0.6854,
      "num_input_tokens_seen": 28846656,
      "step": 20980
    },
    {
      "epoch": 0.6716919531400039,
      "grad_norm": 0.0029809316620230675,
      "learning_rate": 0.1384190593740013,
      "loss": 0.5915,
      "num_input_tokens_seen": 28853872,
      "step": 20985
    },
    {
      "epoch": 0.6718519941104922,
      "grad_norm": 0.003402155125513673,
      "learning_rate": 0.13836033122891686,
      "loss": 0.4136,
      "num_input_tokens_seen": 28860816,
      "step": 20990
    },
    {
      "epoch": 0.6720120350809807,
      "grad_norm": 0.004738356918096542,
      "learning_rate": 0.1383016048788156,
      "loss": 0.4009,
      "num_input_tokens_seen": 28868000,
      "step": 20995
    },
    {
      "epoch": 0.6721720760514692,
      "grad_norm": 0.007414720486849546,
      "learning_rate": 0.13824288033275392,
      "loss": 0.5865,
      "num_input_tokens_seen": 28874800,
      "step": 21000
    },
    {
      "epoch": 0.6721720760514692,
      "eval_loss": 0.5343495011329651,
      "eval_runtime": 331.9112,
      "eval_samples_per_second": 41.836,
      "eval_steps_per_second": 20.918,
      "num_input_tokens_seen": 28874800,
      "step": 21000
    },
    {
      "epoch": 0.6723321170219576,
      "grad_norm": 0.003788368310779333,
      "learning_rate": 0.1381841575997878,
      "loss": 0.4948,
      "num_input_tokens_seen": 28881904,
      "step": 21005
    },
    {
      "epoch": 0.6724921579924461,
      "grad_norm": 0.0031163243111222982,
      "learning_rate": 0.13812543668897306,
      "loss": 0.5901,
      "num_input_tokens_seen": 28888784,
      "step": 21010
    },
    {
      "epoch": 0.6726521989629345,
      "grad_norm": 0.0032624497544020414,
      "learning_rate": 0.13806671760936526,
      "loss": 0.3924,
      "num_input_tokens_seen": 28896304,
      "step": 21015
    },
    {
      "epoch": 0.672812239933423,
      "grad_norm": 0.005429091397672892,
      "learning_rate": 0.13800800037001956,
      "loss": 0.46,
      "num_input_tokens_seen": 28903376,
      "step": 21020
    },
    {
      "epoch": 0.6729722809039114,
      "grad_norm": 0.005877424031496048,
      "learning_rate": 0.13794928497999087,
      "loss": 0.5059,
      "num_input_tokens_seen": 28909856,
      "step": 21025
    },
    {
      "epoch": 0.6731323218743999,
      "grad_norm": 0.005731526296585798,
      "learning_rate": 0.1378905714483339,
      "loss": 0.8394,
      "num_input_tokens_seen": 28916528,
      "step": 21030
    },
    {
      "epoch": 0.6732923628448882,
      "grad_norm": 0.003394638653844595,
      "learning_rate": 0.13783185978410295,
      "loss": 0.582,
      "num_input_tokens_seen": 28923536,
      "step": 21035
    },
    {
      "epoch": 0.6734524038153767,
      "grad_norm": 0.0037726929876953363,
      "learning_rate": 0.13777314999635218,
      "loss": 0.5109,
      "num_input_tokens_seen": 28930448,
      "step": 21040
    },
    {
      "epoch": 0.6736124447858651,
      "grad_norm": 0.00492625217884779,
      "learning_rate": 0.1377144420941353,
      "loss": 0.5474,
      "num_input_tokens_seen": 28937296,
      "step": 21045
    },
    {
      "epoch": 0.6737724857563536,
      "grad_norm": 0.005100585520267487,
      "learning_rate": 0.13765573608650586,
      "loss": 0.5341,
      "num_input_tokens_seen": 28944336,
      "step": 21050
    },
    {
      "epoch": 0.6739325267268421,
      "grad_norm": 0.002916100202128291,
      "learning_rate": 0.13759703198251702,
      "loss": 0.5023,
      "num_input_tokens_seen": 28950960,
      "step": 21055
    },
    {
      "epoch": 0.6740925676973305,
      "grad_norm": 0.003875237889587879,
      "learning_rate": 0.13753832979122174,
      "loss": 0.484,
      "num_input_tokens_seen": 28958384,
      "step": 21060
    },
    {
      "epoch": 0.674252608667819,
      "grad_norm": 0.004485597368329763,
      "learning_rate": 0.13747962952167264,
      "loss": 0.6271,
      "num_input_tokens_seen": 28964896,
      "step": 21065
    },
    {
      "epoch": 0.6744126496383074,
      "grad_norm": 0.004814330022782087,
      "learning_rate": 0.13742093118292192,
      "loss": 0.5644,
      "num_input_tokens_seen": 28971952,
      "step": 21070
    },
    {
      "epoch": 0.6745726906087959,
      "grad_norm": 0.0047497996129095554,
      "learning_rate": 0.13736223478402174,
      "loss": 0.7031,
      "num_input_tokens_seen": 28978912,
      "step": 21075
    },
    {
      "epoch": 0.6747327315792843,
      "grad_norm": 0.004250241443514824,
      "learning_rate": 0.1373035403340238,
      "loss": 0.5273,
      "num_input_tokens_seen": 28985920,
      "step": 21080
    },
    {
      "epoch": 0.6748927725497728,
      "grad_norm": 0.0042314291931688786,
      "learning_rate": 0.13724484784197943,
      "loss": 0.4664,
      "num_input_tokens_seen": 28993200,
      "step": 21085
    },
    {
      "epoch": 0.6750528135202611,
      "grad_norm": 0.004856852348893881,
      "learning_rate": 0.13718615731693987,
      "loss": 0.4708,
      "num_input_tokens_seen": 29000096,
      "step": 21090
    },
    {
      "epoch": 0.6752128544907496,
      "grad_norm": 0.0030121784657239914,
      "learning_rate": 0.13712746876795587,
      "loss": 0.4117,
      "num_input_tokens_seen": 29007232,
      "step": 21095
    },
    {
      "epoch": 0.6753728954612381,
      "grad_norm": 0.012201286852359772,
      "learning_rate": 0.13706878220407792,
      "loss": 0.4773,
      "num_input_tokens_seen": 29013968,
      "step": 21100
    },
    {
      "epoch": 0.6755329364317265,
      "grad_norm": 0.005975651554763317,
      "learning_rate": 0.13701009763435631,
      "loss": 0.4609,
      "num_input_tokens_seen": 29021232,
      "step": 21105
    },
    {
      "epoch": 0.675692977402215,
      "grad_norm": 0.006085206288844347,
      "learning_rate": 0.13695141506784084,
      "loss": 0.49,
      "num_input_tokens_seen": 29027744,
      "step": 21110
    },
    {
      "epoch": 0.6758530183727034,
      "grad_norm": 0.005113503430038691,
      "learning_rate": 0.13689273451358114,
      "loss": 0.5579,
      "num_input_tokens_seen": 29034880,
      "step": 21115
    },
    {
      "epoch": 0.6760130593431919,
      "grad_norm": 0.006393501069396734,
      "learning_rate": 0.13683405598062653,
      "loss": 0.5291,
      "num_input_tokens_seen": 29041936,
      "step": 21120
    },
    {
      "epoch": 0.6761731003136803,
      "grad_norm": 0.00851181335747242,
      "learning_rate": 0.1367753794780259,
      "loss": 0.7741,
      "num_input_tokens_seen": 29048912,
      "step": 21125
    },
    {
      "epoch": 0.6763331412841688,
      "grad_norm": 0.004138168878853321,
      "learning_rate": 0.13671670501482802,
      "loss": 0.3775,
      "num_input_tokens_seen": 29056112,
      "step": 21130
    },
    {
      "epoch": 0.6764931822546572,
      "grad_norm": 0.008214644156396389,
      "learning_rate": 0.1366580326000811,
      "loss": 0.4679,
      "num_input_tokens_seen": 29062816,
      "step": 21135
    },
    {
      "epoch": 0.6766532232251457,
      "grad_norm": 0.007638027425855398,
      "learning_rate": 0.1365993622428332,
      "loss": 0.6614,
      "num_input_tokens_seen": 29069664,
      "step": 21140
    },
    {
      "epoch": 0.676813264195634,
      "grad_norm": 0.010877637192606926,
      "learning_rate": 0.13654069395213211,
      "loss": 0.5931,
      "num_input_tokens_seen": 29076944,
      "step": 21145
    },
    {
      "epoch": 0.6769733051661225,
      "grad_norm": 0.004776325076818466,
      "learning_rate": 0.13648202773702509,
      "loss": 0.4689,
      "num_input_tokens_seen": 29083456,
      "step": 21150
    },
    {
      "epoch": 0.677133346136611,
      "grad_norm": 0.005830613896250725,
      "learning_rate": 0.13642336360655927,
      "loss": 0.7063,
      "num_input_tokens_seen": 29089952,
      "step": 21155
    },
    {
      "epoch": 0.6772933871070994,
      "grad_norm": 0.005298457108438015,
      "learning_rate": 0.13636470156978145,
      "loss": 0.713,
      "num_input_tokens_seen": 29096672,
      "step": 21160
    },
    {
      "epoch": 0.6774534280775879,
      "grad_norm": 0.006184582598507404,
      "learning_rate": 0.13630604163573798,
      "loss": 0.6028,
      "num_input_tokens_seen": 29103184,
      "step": 21165
    },
    {
      "epoch": 0.6776134690480763,
      "grad_norm": 0.011109293438494205,
      "learning_rate": 0.13624738381347495,
      "loss": 0.4561,
      "num_input_tokens_seen": 29109872,
      "step": 21170
    },
    {
      "epoch": 0.6777735100185648,
      "grad_norm": 0.0064604138024151325,
      "learning_rate": 0.1361887281120382,
      "loss": 0.58,
      "num_input_tokens_seen": 29116416,
      "step": 21175
    },
    {
      "epoch": 0.6779335509890532,
      "grad_norm": 0.008693324401974678,
      "learning_rate": 0.13613007454047307,
      "loss": 0.61,
      "num_input_tokens_seen": 29122928,
      "step": 21180
    },
    {
      "epoch": 0.6780935919595417,
      "grad_norm": 0.023853126913309097,
      "learning_rate": 0.13607142310782486,
      "loss": 0.5195,
      "num_input_tokens_seen": 29130160,
      "step": 21185
    },
    {
      "epoch": 0.67825363293003,
      "grad_norm": 0.004213184118270874,
      "learning_rate": 0.13601277382313814,
      "loss": 0.4969,
      "num_input_tokens_seen": 29137504,
      "step": 21190
    },
    {
      "epoch": 0.6784136739005185,
      "grad_norm": 0.0037257159128785133,
      "learning_rate": 0.1359541266954575,
      "loss": 0.5443,
      "num_input_tokens_seen": 29144368,
      "step": 21195
    },
    {
      "epoch": 0.6785737148710069,
      "grad_norm": 0.004103382118046284,
      "learning_rate": 0.13589548173382707,
      "loss": 0.5382,
      "num_input_tokens_seen": 29151312,
      "step": 21200
    },
    {
      "epoch": 0.6785737148710069,
      "eval_loss": 0.5438513159751892,
      "eval_runtime": 332.1541,
      "eval_samples_per_second": 41.806,
      "eval_steps_per_second": 20.903,
      "num_input_tokens_seen": 29151312,
      "step": 21200
    },
    {
      "epoch": 0.6787337558414954,
      "grad_norm": 0.003782594110816717,
      "learning_rate": 0.1358368389472906,
      "loss": 0.4953,
      "num_input_tokens_seen": 29158000,
      "step": 21205
    },
    {
      "epoch": 0.6788937968119839,
      "grad_norm": 0.004109243396669626,
      "learning_rate": 0.13577819834489155,
      "loss": 0.6801,
      "num_input_tokens_seen": 29164800,
      "step": 21210
    },
    {
      "epoch": 0.6790538377824723,
      "grad_norm": 0.0057392544113099575,
      "learning_rate": 0.135719559935673,
      "loss": 0.5571,
      "num_input_tokens_seen": 29172352,
      "step": 21215
    },
    {
      "epoch": 0.6792138787529608,
      "grad_norm": 0.006399322766810656,
      "learning_rate": 0.13566092372867775,
      "loss": 0.5888,
      "num_input_tokens_seen": 29179712,
      "step": 21220
    },
    {
      "epoch": 0.6793739197234492,
      "grad_norm": 0.005346665158867836,
      "learning_rate": 0.13560228973294833,
      "loss": 0.7865,
      "num_input_tokens_seen": 29186672,
      "step": 21225
    },
    {
      "epoch": 0.6795339606939377,
      "grad_norm": 0.006705811712890863,
      "learning_rate": 0.13554365795752668,
      "loss": 0.6105,
      "num_input_tokens_seen": 29193632,
      "step": 21230
    },
    {
      "epoch": 0.6796940016644261,
      "grad_norm": 0.00541487243026495,
      "learning_rate": 0.1354850284114547,
      "loss": 0.6307,
      "num_input_tokens_seen": 29200512,
      "step": 21235
    },
    {
      "epoch": 0.6798540426349146,
      "grad_norm": 0.00440460117533803,
      "learning_rate": 0.13542640110377374,
      "loss": 0.4398,
      "num_input_tokens_seen": 29207424,
      "step": 21240
    },
    {
      "epoch": 0.6800140836054029,
      "grad_norm": 0.0060231126844882965,
      "learning_rate": 0.13536777604352487,
      "loss": 0.6014,
      "num_input_tokens_seen": 29214368,
      "step": 21245
    },
    {
      "epoch": 0.6801741245758914,
      "grad_norm": 0.0052259014919400215,
      "learning_rate": 0.13530915323974887,
      "loss": 0.3237,
      "num_input_tokens_seen": 29221152,
      "step": 21250
    },
    {
      "epoch": 0.6803341655463798,
      "grad_norm": 0.005237159319221973,
      "learning_rate": 0.13525053270148596,
      "loss": 0.6587,
      "num_input_tokens_seen": 29228000,
      "step": 21255
    },
    {
      "epoch": 0.6804942065168683,
      "grad_norm": 0.0040972610004246235,
      "learning_rate": 0.13519191443777628,
      "loss": 0.5432,
      "num_input_tokens_seen": 29234880,
      "step": 21260
    },
    {
      "epoch": 0.6806542474873568,
      "grad_norm": 0.0036788962315768003,
      "learning_rate": 0.13513329845765953,
      "loss": 0.4897,
      "num_input_tokens_seen": 29241952,
      "step": 21265
    },
    {
      "epoch": 0.6808142884578452,
      "grad_norm": 0.0032910644076764584,
      "learning_rate": 0.13507468477017495,
      "loss": 0.5543,
      "num_input_tokens_seen": 29248912,
      "step": 21270
    },
    {
      "epoch": 0.6809743294283337,
      "grad_norm": 0.0046062348410487175,
      "learning_rate": 0.13501607338436153,
      "loss": 0.4508,
      "num_input_tokens_seen": 29255776,
      "step": 21275
    },
    {
      "epoch": 0.6811343703988221,
      "grad_norm": 0.005074721295386553,
      "learning_rate": 0.13495746430925798,
      "loss": 0.7768,
      "num_input_tokens_seen": 29262672,
      "step": 21280
    },
    {
      "epoch": 0.6812944113693106,
      "grad_norm": 0.005571441724896431,
      "learning_rate": 0.13489885755390238,
      "loss": 0.5203,
      "num_input_tokens_seen": 29269424,
      "step": 21285
    },
    {
      "epoch": 0.681454452339799,
      "grad_norm": 0.0038787368685007095,
      "learning_rate": 0.13484025312733275,
      "loss": 0.4759,
      "num_input_tokens_seen": 29275808,
      "step": 21290
    },
    {
      "epoch": 0.6816144933102875,
      "grad_norm": 0.0067487298510968685,
      "learning_rate": 0.13478165103858658,
      "loss": 0.4682,
      "num_input_tokens_seen": 29282720,
      "step": 21295
    },
    {
      "epoch": 0.6817745342807758,
      "grad_norm": 0.00413237139582634,
      "learning_rate": 0.13472305129670106,
      "loss": 0.5705,
      "num_input_tokens_seen": 29289360,
      "step": 21300
    },
    {
      "epoch": 0.6819345752512643,
      "grad_norm": 0.00497261481359601,
      "learning_rate": 0.13466445391071305,
      "loss": 0.5771,
      "num_input_tokens_seen": 29295824,
      "step": 21305
    },
    {
      "epoch": 0.6820946162217527,
      "grad_norm": 0.005493517499417067,
      "learning_rate": 0.13460585888965895,
      "loss": 0.5734,
      "num_input_tokens_seen": 29302592,
      "step": 21310
    },
    {
      "epoch": 0.6822546571922412,
      "grad_norm": 0.004627089481800795,
      "learning_rate": 0.13454726624257482,
      "loss": 0.5414,
      "num_input_tokens_seen": 29309472,
      "step": 21315
    },
    {
      "epoch": 0.6824146981627297,
      "grad_norm": 0.004965395200997591,
      "learning_rate": 0.1344886759784965,
      "loss": 0.3613,
      "num_input_tokens_seen": 29316240,
      "step": 21320
    },
    {
      "epoch": 0.6825747391332181,
      "grad_norm": 0.003481067717075348,
      "learning_rate": 0.13443008810645923,
      "loss": 0.6955,
      "num_input_tokens_seen": 29323360,
      "step": 21325
    },
    {
      "epoch": 0.6827347801037066,
      "grad_norm": 0.005916815251111984,
      "learning_rate": 0.13437150263549807,
      "loss": 0.753,
      "num_input_tokens_seen": 29330096,
      "step": 21330
    },
    {
      "epoch": 0.682894821074195,
      "grad_norm": 0.004314270801842213,
      "learning_rate": 0.13431291957464755,
      "loss": 0.4114,
      "num_input_tokens_seen": 29336928,
      "step": 21335
    },
    {
      "epoch": 0.6830548620446835,
      "grad_norm": 0.005389268044382334,
      "learning_rate": 0.13425433893294197,
      "loss": 0.6275,
      "num_input_tokens_seen": 29343584,
      "step": 21340
    },
    {
      "epoch": 0.6832149030151718,
      "grad_norm": 0.0029426799155771732,
      "learning_rate": 0.13419576071941525,
      "loss": 0.4824,
      "num_input_tokens_seen": 29350176,
      "step": 21345
    },
    {
      "epoch": 0.6833749439856603,
      "grad_norm": 0.004960406571626663,
      "learning_rate": 0.1341371849431008,
      "loss": 0.4862,
      "num_input_tokens_seen": 29356912,
      "step": 21350
    },
    {
      "epoch": 0.6835349849561487,
      "grad_norm": 0.003411137731745839,
      "learning_rate": 0.13407861161303178,
      "loss": 0.3241,
      "num_input_tokens_seen": 29363712,
      "step": 21355
    },
    {
      "epoch": 0.6836950259266372,
      "grad_norm": 0.003926464822143316,
      "learning_rate": 0.13402004073824098,
      "loss": 0.4696,
      "num_input_tokens_seen": 29370816,
      "step": 21360
    },
    {
      "epoch": 0.6838550668971257,
      "grad_norm": 0.005945575423538685,
      "learning_rate": 0.13396147232776062,
      "loss": 0.4713,
      "num_input_tokens_seen": 29377664,
      "step": 21365
    },
    {
      "epoch": 0.6840151078676141,
      "grad_norm": 0.002629320602864027,
      "learning_rate": 0.13390290639062288,
      "loss": 0.4367,
      "num_input_tokens_seen": 29384560,
      "step": 21370
    },
    {
      "epoch": 0.6841751488381026,
      "grad_norm": 0.007124952506273985,
      "learning_rate": 0.13384434293585917,
      "loss": 0.4672,
      "num_input_tokens_seen": 29391408,
      "step": 21375
    },
    {
      "epoch": 0.684335189808591,
      "grad_norm": 0.0030754522886127234,
      "learning_rate": 0.13378578197250088,
      "loss": 0.645,
      "num_input_tokens_seen": 29398544,
      "step": 21380
    },
    {
      "epoch": 0.6844952307790795,
      "grad_norm": 0.004814988002181053,
      "learning_rate": 0.13372722350957872,
      "loss": 0.4278,
      "num_input_tokens_seen": 29405264,
      "step": 21385
    },
    {
      "epoch": 0.6846552717495679,
      "grad_norm": 0.002894802950322628,
      "learning_rate": 0.13366866755612322,
      "loss": 0.4817,
      "num_input_tokens_seen": 29412240,
      "step": 21390
    },
    {
      "epoch": 0.6848153127200564,
      "grad_norm": 0.005509062670171261,
      "learning_rate": 0.13361011412116436,
      "loss": 0.5878,
      "num_input_tokens_seen": 29419072,
      "step": 21395
    },
    {
      "epoch": 0.6849753536905447,
      "grad_norm": 0.005488331895321608,
      "learning_rate": 0.13355156321373196,
      "loss": 0.6467,
      "num_input_tokens_seen": 29425936,
      "step": 21400
    },
    {
      "epoch": 0.6849753536905447,
      "eval_loss": 0.541100800037384,
      "eval_runtime": 331.7645,
      "eval_samples_per_second": 41.855,
      "eval_steps_per_second": 20.927,
      "num_input_tokens_seen": 29425936,
      "step": 21400
    },
    {
      "epoch": 0.6851353946610332,
      "grad_norm": 0.004085480235517025,
      "learning_rate": 0.13349301484285514,
      "loss": 0.3875,
      "num_input_tokens_seen": 29433248,
      "step": 21405
    },
    {
      "epoch": 0.6852954356315216,
      "grad_norm": 0.00467375572770834,
      "learning_rate": 0.13343446901756295,
      "loss": 0.6553,
      "num_input_tokens_seen": 29440336,
      "step": 21410
    },
    {
      "epoch": 0.6854554766020101,
      "grad_norm": 0.005061016883701086,
      "learning_rate": 0.13337592574688376,
      "loss": 0.5563,
      "num_input_tokens_seen": 29447968,
      "step": 21415
    },
    {
      "epoch": 0.6856155175724986,
      "grad_norm": 0.00856689177453518,
      "learning_rate": 0.13331738503984572,
      "loss": 0.4256,
      "num_input_tokens_seen": 29454800,
      "step": 21420
    },
    {
      "epoch": 0.685775558542987,
      "grad_norm": 0.0015793605707585812,
      "learning_rate": 0.1332588469054766,
      "loss": 0.5289,
      "num_input_tokens_seen": 29462144,
      "step": 21425
    },
    {
      "epoch": 0.6859355995134755,
      "grad_norm": 0.0023276559077203274,
      "learning_rate": 0.1332003113528036,
      "loss": 0.5247,
      "num_input_tokens_seen": 29468816,
      "step": 21430
    },
    {
      "epoch": 0.6860956404839639,
      "grad_norm": 0.014720883220434189,
      "learning_rate": 0.13314177839085373,
      "loss": 0.4251,
      "num_input_tokens_seen": 29475792,
      "step": 21435
    },
    {
      "epoch": 0.6862556814544524,
      "grad_norm": 0.00814051739871502,
      "learning_rate": 0.13308324802865354,
      "loss": 0.4243,
      "num_input_tokens_seen": 29482512,
      "step": 21440
    },
    {
      "epoch": 0.6864157224249408,
      "grad_norm": 0.004097375553101301,
      "learning_rate": 0.13302472027522905,
      "loss": 0.5895,
      "num_input_tokens_seen": 29489232,
      "step": 21445
    },
    {
      "epoch": 0.6865757633954293,
      "grad_norm": 0.0073575121350586414,
      "learning_rate": 0.13296619513960606,
      "loss": 0.6068,
      "num_input_tokens_seen": 29496288,
      "step": 21450
    },
    {
      "epoch": 0.6867358043659176,
      "grad_norm": 0.004717991687357426,
      "learning_rate": 0.1329076726308098,
      "loss": 0.6629,
      "num_input_tokens_seen": 29502992,
      "step": 21455
    },
    {
      "epoch": 0.6868958453364061,
      "grad_norm": 0.0030479864217340946,
      "learning_rate": 0.13284915275786519,
      "loss": 0.3737,
      "num_input_tokens_seen": 29510144,
      "step": 21460
    },
    {
      "epoch": 0.6870558863068945,
      "grad_norm": 0.0048057385720312595,
      "learning_rate": 0.1327906355297968,
      "loss": 0.5701,
      "num_input_tokens_seen": 29516976,
      "step": 21465
    },
    {
      "epoch": 0.687215927277383,
      "grad_norm": 0.0030141754541546106,
      "learning_rate": 0.13273212095562867,
      "loss": 0.4425,
      "num_input_tokens_seen": 29524032,
      "step": 21470
    },
    {
      "epoch": 0.6873759682478715,
      "grad_norm": 0.005347480997443199,
      "learning_rate": 0.13267360904438444,
      "loss": 0.4054,
      "num_input_tokens_seen": 29530912,
      "step": 21475
    },
    {
      "epoch": 0.6875360092183599,
      "grad_norm": 0.004930577706545591,
      "learning_rate": 0.1326150998050875,
      "loss": 0.5004,
      "num_input_tokens_seen": 29537872,
      "step": 21480
    },
    {
      "epoch": 0.6876960501888484,
      "grad_norm": 0.002387054730206728,
      "learning_rate": 0.1325565932467606,
      "loss": 0.4977,
      "num_input_tokens_seen": 29544848,
      "step": 21485
    },
    {
      "epoch": 0.6878560911593368,
      "grad_norm": 0.005533142480999231,
      "learning_rate": 0.13249808937842628,
      "loss": 0.8801,
      "num_input_tokens_seen": 29551904,
      "step": 21490
    },
    {
      "epoch": 0.6880161321298253,
      "grad_norm": 0.004399573430418968,
      "learning_rate": 0.1324395882091065,
      "loss": 0.5579,
      "num_input_tokens_seen": 29558512,
      "step": 21495
    },
    {
      "epoch": 0.6881761731003136,
      "grad_norm": 0.0026388675905764103,
      "learning_rate": 0.13238108974782284,
      "loss": 0.5387,
      "num_input_tokens_seen": 29565232,
      "step": 21500
    },
    {
      "epoch": 0.6883362140708021,
      "grad_norm": 0.005817824508994818,
      "learning_rate": 0.13232259400359664,
      "loss": 0.7952,
      "num_input_tokens_seen": 29572368,
      "step": 21505
    },
    {
      "epoch": 0.6884962550412905,
      "grad_norm": 0.004432172514498234,
      "learning_rate": 0.13226410098544852,
      "loss": 0.5651,
      "num_input_tokens_seen": 29579152,
      "step": 21510
    },
    {
      "epoch": 0.688656296011779,
      "grad_norm": 0.006406520493328571,
      "learning_rate": 0.13220561070239892,
      "loss": 0.6934,
      "num_input_tokens_seen": 29586032,
      "step": 21515
    },
    {
      "epoch": 0.6888163369822674,
      "grad_norm": 0.004118368960916996,
      "learning_rate": 0.13214712316346783,
      "loss": 0.51,
      "num_input_tokens_seen": 29592976,
      "step": 21520
    },
    {
      "epoch": 0.6889763779527559,
      "grad_norm": 0.008332411758601665,
      "learning_rate": 0.13208863837767465,
      "loss": 0.6236,
      "num_input_tokens_seen": 29599664,
      "step": 21525
    },
    {
      "epoch": 0.6891364189232444,
      "grad_norm": 0.004436344373971224,
      "learning_rate": 0.13203015635403856,
      "loss": 0.494,
      "num_input_tokens_seen": 29606512,
      "step": 21530
    },
    {
      "epoch": 0.6892964598937328,
      "grad_norm": 0.004103250335901976,
      "learning_rate": 0.13197167710157817,
      "loss": 0.4469,
      "num_input_tokens_seen": 29613136,
      "step": 21535
    },
    {
      "epoch": 0.6894565008642213,
      "grad_norm": 0.0054748947732150555,
      "learning_rate": 0.13191320062931167,
      "loss": 0.5025,
      "num_input_tokens_seen": 29620192,
      "step": 21540
    },
    {
      "epoch": 0.6896165418347097,
      "grad_norm": 0.0034381765872240067,
      "learning_rate": 0.13185472694625702,
      "loss": 0.4318,
      "num_input_tokens_seen": 29626960,
      "step": 21545
    },
    {
      "epoch": 0.6897765828051982,
      "grad_norm": 0.005933019798249006,
      "learning_rate": 0.13179625606143142,
      "loss": 0.4671,
      "num_input_tokens_seen": 29633440,
      "step": 21550
    },
    {
      "epoch": 0.6899366237756865,
      "grad_norm": 0.005212028976529837,
      "learning_rate": 0.13173778798385188,
      "loss": 0.645,
      "num_input_tokens_seen": 29640288,
      "step": 21555
    },
    {
      "epoch": 0.690096664746175,
      "grad_norm": 0.004061408340930939,
      "learning_rate": 0.13167932272253505,
      "loss": 0.5496,
      "num_input_tokens_seen": 29646976,
      "step": 21560
    },
    {
      "epoch": 0.6902567057166634,
      "grad_norm": 0.004231381695717573,
      "learning_rate": 0.1316208602864968,
      "loss": 0.4059,
      "num_input_tokens_seen": 29653904,
      "step": 21565
    },
    {
      "epoch": 0.6904167466871519,
      "grad_norm": 0.006222291383892298,
      "learning_rate": 0.13156240068475292,
      "loss": 0.5592,
      "num_input_tokens_seen": 29661136,
      "step": 21570
    },
    {
      "epoch": 0.6905767876576403,
      "grad_norm": 0.004501586779952049,
      "learning_rate": 0.1315039439263185,
      "loss": 0.5231,
      "num_input_tokens_seen": 29668032,
      "step": 21575
    },
    {
      "epoch": 0.6907368286281288,
      "grad_norm": 0.004233778920024633,
      "learning_rate": 0.13144549002020833,
      "loss": 0.5048,
      "num_input_tokens_seen": 29675232,
      "step": 21580
    },
    {
      "epoch": 0.6908968695986173,
      "grad_norm": 0.0038856372702866793,
      "learning_rate": 0.13138703897543688,
      "loss": 0.4227,
      "num_input_tokens_seen": 29682160,
      "step": 21585
    },
    {
      "epoch": 0.6910569105691057,
      "grad_norm": 0.003067540004849434,
      "learning_rate": 0.1313285908010178,
      "loss": 0.4506,
      "num_input_tokens_seen": 29688976,
      "step": 21590
    },
    {
      "epoch": 0.6912169515395942,
      "grad_norm": 0.006778191775083542,
      "learning_rate": 0.13127014550596475,
      "loss": 0.4817,
      "num_input_tokens_seen": 29695808,
      "step": 21595
    },
    {
      "epoch": 0.6913769925100826,
      "grad_norm": 0.005162088666111231,
      "learning_rate": 0.1312117030992906,
      "loss": 0.565,
      "num_input_tokens_seen": 29702784,
      "step": 21600
    },
    {
      "epoch": 0.6913769925100826,
      "eval_loss": 0.5406631231307983,
      "eval_runtime": 331.8067,
      "eval_samples_per_second": 41.85,
      "eval_steps_per_second": 20.925,
      "num_input_tokens_seen": 29702784,
      "step": 21600
    },
    {
      "epoch": 0.691537033480571,
      "grad_norm": 0.008251921273767948,
      "learning_rate": 0.13115326359000795,
      "loss": 0.5512,
      "num_input_tokens_seen": 29709568,
      "step": 21605
    },
    {
      "epoch": 0.6916970744510594,
      "grad_norm": 0.0030683635268360376,
      "learning_rate": 0.13109482698712896,
      "loss": 0.4457,
      "num_input_tokens_seen": 29716304,
      "step": 21610
    },
    {
      "epoch": 0.6918571154215479,
      "grad_norm": 0.0028900050092488527,
      "learning_rate": 0.1310363932996651,
      "loss": 0.5544,
      "num_input_tokens_seen": 29723104,
      "step": 21615
    },
    {
      "epoch": 0.6920171563920363,
      "grad_norm": 0.00255665834993124,
      "learning_rate": 0.13097796253662775,
      "loss": 0.7114,
      "num_input_tokens_seen": 29730032,
      "step": 21620
    },
    {
      "epoch": 0.6921771973625248,
      "grad_norm": 0.0061723231337964535,
      "learning_rate": 0.1309195347070277,
      "loss": 0.6199,
      "num_input_tokens_seen": 29737040,
      "step": 21625
    },
    {
      "epoch": 0.6923372383330133,
      "grad_norm": 0.002397523494437337,
      "learning_rate": 0.13086110981987506,
      "loss": 0.419,
      "num_input_tokens_seen": 29744080,
      "step": 21630
    },
    {
      "epoch": 0.6924972793035017,
      "grad_norm": 0.00299266679212451,
      "learning_rate": 0.13080268788417987,
      "loss": 0.4474,
      "num_input_tokens_seen": 29750768,
      "step": 21635
    },
    {
      "epoch": 0.6926573202739902,
      "grad_norm": 0.006095734890550375,
      "learning_rate": 0.1307442689089515,
      "loss": 0.5256,
      "num_input_tokens_seen": 29757744,
      "step": 21640
    },
    {
      "epoch": 0.6928173612444786,
      "grad_norm": 0.0029973164200782776,
      "learning_rate": 0.13068585290319873,
      "loss": 0.6375,
      "num_input_tokens_seen": 29764688,
      "step": 21645
    },
    {
      "epoch": 0.6929774022149671,
      "grad_norm": 0.0038049076683819294,
      "learning_rate": 0.13062743987593026,
      "loss": 0.4729,
      "num_input_tokens_seen": 29771536,
      "step": 21650
    },
    {
      "epoch": 0.6931374431854554,
      "grad_norm": 0.005806330591440201,
      "learning_rate": 0.13056902983615395,
      "loss": 0.4758,
      "num_input_tokens_seen": 29778512,
      "step": 21655
    },
    {
      "epoch": 0.6932974841559439,
      "grad_norm": 0.002145415171980858,
      "learning_rate": 0.13051062279287742,
      "loss": 0.3502,
      "num_input_tokens_seen": 29785552,
      "step": 21660
    },
    {
      "epoch": 0.6934575251264323,
      "grad_norm": 0.008165122009813786,
      "learning_rate": 0.13045221875510782,
      "loss": 0.557,
      "num_input_tokens_seen": 29792288,
      "step": 21665
    },
    {
      "epoch": 0.6936175660969208,
      "grad_norm": 0.0072020855732262135,
      "learning_rate": 0.13039381773185174,
      "loss": 0.3991,
      "num_input_tokens_seen": 29799040,
      "step": 21670
    },
    {
      "epoch": 0.6937776070674092,
      "grad_norm": 0.0024924236349761486,
      "learning_rate": 0.1303354197321153,
      "loss": 0.4281,
      "num_input_tokens_seen": 29806176,
      "step": 21675
    },
    {
      "epoch": 0.6939376480378977,
      "grad_norm": 0.0035502715036273003,
      "learning_rate": 0.13027702476490433,
      "loss": 0.511,
      "num_input_tokens_seen": 29813040,
      "step": 21680
    },
    {
      "epoch": 0.6940976890083862,
      "grad_norm": 0.002615329110994935,
      "learning_rate": 0.1302186328392239,
      "loss": 0.4518,
      "num_input_tokens_seen": 29819728,
      "step": 21685
    },
    {
      "epoch": 0.6942577299788746,
      "grad_norm": 0.00613014679402113,
      "learning_rate": 0.130160243964079,
      "loss": 0.7394,
      "num_input_tokens_seen": 29826752,
      "step": 21690
    },
    {
      "epoch": 0.6944177709493631,
      "grad_norm": 0.004098621662706137,
      "learning_rate": 0.13010185814847372,
      "loss": 0.4689,
      "num_input_tokens_seen": 29833392,
      "step": 21695
    },
    {
      "epoch": 0.6945778119198515,
      "grad_norm": 0.0034830074291676283,
      "learning_rate": 0.13004347540141192,
      "loss": 0.4575,
      "num_input_tokens_seen": 29840112,
      "step": 21700
    },
    {
      "epoch": 0.69473785289034,
      "grad_norm": 0.005345372948795557,
      "learning_rate": 0.12998509573189712,
      "loss": 0.5238,
      "num_input_tokens_seen": 29846848,
      "step": 21705
    },
    {
      "epoch": 0.6948978938608283,
      "grad_norm": 0.003802721854299307,
      "learning_rate": 0.12992671914893203,
      "loss": 0.6564,
      "num_input_tokens_seen": 29853520,
      "step": 21710
    },
    {
      "epoch": 0.6950579348313168,
      "grad_norm": 0.0029410729184746742,
      "learning_rate": 0.12986834566151909,
      "loss": 0.5133,
      "num_input_tokens_seen": 29860176,
      "step": 21715
    },
    {
      "epoch": 0.6952179758018052,
      "grad_norm": 0.003004864789545536,
      "learning_rate": 0.12980997527866028,
      "loss": 0.5041,
      "num_input_tokens_seen": 29866816,
      "step": 21720
    },
    {
      "epoch": 0.6953780167722937,
      "grad_norm": 0.003932887688279152,
      "learning_rate": 0.12975160800935692,
      "loss": 0.5972,
      "num_input_tokens_seen": 29873472,
      "step": 21725
    },
    {
      "epoch": 0.6955380577427821,
      "grad_norm": 0.002574885729700327,
      "learning_rate": 0.12969324386261016,
      "loss": 0.4639,
      "num_input_tokens_seen": 29881344,
      "step": 21730
    },
    {
      "epoch": 0.6956980987132706,
      "grad_norm": 0.006952138151973486,
      "learning_rate": 0.12963488284742034,
      "loss": 0.6172,
      "num_input_tokens_seen": 29888592,
      "step": 21735
    },
    {
      "epoch": 0.6958581396837591,
      "grad_norm": 0.005687180906534195,
      "learning_rate": 0.12957652497278752,
      "loss": 0.5855,
      "num_input_tokens_seen": 29896000,
      "step": 21740
    },
    {
      "epoch": 0.6960181806542475,
      "grad_norm": 0.004737763199955225,
      "learning_rate": 0.12951817024771117,
      "loss": 0.3747,
      "num_input_tokens_seen": 29903248,
      "step": 21745
    },
    {
      "epoch": 0.696178221624736,
      "grad_norm": 0.010247262194752693,
      "learning_rate": 0.12945981868119041,
      "loss": 0.6205,
      "num_input_tokens_seen": 29910048,
      "step": 21750
    },
    {
      "epoch": 0.6963382625952244,
      "grad_norm": 0.002815922489389777,
      "learning_rate": 0.12940147028222376,
      "loss": 0.4258,
      "num_input_tokens_seen": 29917280,
      "step": 21755
    },
    {
      "epoch": 0.6964983035657129,
      "grad_norm": 0.007760562002658844,
      "learning_rate": 0.12934312505980916,
      "loss": 0.6523,
      "num_input_tokens_seen": 29924192,
      "step": 21760
    },
    {
      "epoch": 0.6966583445362012,
      "grad_norm": 0.004894857294857502,
      "learning_rate": 0.1292847830229443,
      "loss": 0.5998,
      "num_input_tokens_seen": 29930816,
      "step": 21765
    },
    {
      "epoch": 0.6968183855066897,
      "grad_norm": 0.004715834744274616,
      "learning_rate": 0.12922644418062626,
      "loss": 0.487,
      "num_input_tokens_seen": 29937824,
      "step": 21770
    },
    {
      "epoch": 0.6969784264771781,
      "grad_norm": 0.0033514651004225016,
      "learning_rate": 0.1291681085418515,
      "loss": 0.6846,
      "num_input_tokens_seen": 29945120,
      "step": 21775
    },
    {
      "epoch": 0.6971384674476666,
      "grad_norm": 0.004549511708319187,
      "learning_rate": 0.12910977611561628,
      "loss": 0.6381,
      "num_input_tokens_seen": 29951984,
      "step": 21780
    },
    {
      "epoch": 0.697298508418155,
      "grad_norm": 0.004082785453647375,
      "learning_rate": 0.1290514469109161,
      "loss": 0.5375,
      "num_input_tokens_seen": 29958720,
      "step": 21785
    },
    {
      "epoch": 0.6974585493886435,
      "grad_norm": 0.005378464236855507,
      "learning_rate": 0.128993120936746,
      "loss": 0.5543,
      "num_input_tokens_seen": 29965472,
      "step": 21790
    },
    {
      "epoch": 0.697618590359132,
      "grad_norm": 0.002906029811128974,
      "learning_rate": 0.12893479820210071,
      "loss": 0.3299,
      "num_input_tokens_seen": 29972464,
      "step": 21795
    },
    {
      "epoch": 0.6977786313296204,
      "grad_norm": 0.008076222613453865,
      "learning_rate": 0.1288764787159742,
      "loss": 0.6301,
      "num_input_tokens_seen": 29979824,
      "step": 21800
    },
    {
      "epoch": 0.6977786313296204,
      "eval_loss": 0.5329286456108093,
      "eval_runtime": 331.4587,
      "eval_samples_per_second": 41.894,
      "eval_steps_per_second": 20.947,
      "num_input_tokens_seen": 29979824,
      "step": 21800
    },
    {
      "epoch": 0.6979386723001089,
      "grad_norm": 0.0034561657812446356,
      "learning_rate": 0.1288181624873601,
      "loss": 0.5325,
      "num_input_tokens_seen": 29986688,
      "step": 21805
    },
    {
      "epoch": 0.6980987132705972,
      "grad_norm": 0.004296391271054745,
      "learning_rate": 0.12875984952525163,
      "loss": 0.5045,
      "num_input_tokens_seen": 29993408,
      "step": 21810
    },
    {
      "epoch": 0.6982587542410857,
      "grad_norm": 0.0015722739044576883,
      "learning_rate": 0.12870153983864122,
      "loss": 0.5134,
      "num_input_tokens_seen": 30000272,
      "step": 21815
    },
    {
      "epoch": 0.6984187952115741,
      "grad_norm": 0.006317676976323128,
      "learning_rate": 0.12864323343652104,
      "loss": 0.7127,
      "num_input_tokens_seen": 30006704,
      "step": 21820
    },
    {
      "epoch": 0.6985788361820626,
      "grad_norm": 0.0038353034760802984,
      "learning_rate": 0.12858493032788268,
      "loss": 0.3604,
      "num_input_tokens_seen": 30013680,
      "step": 21825
    },
    {
      "epoch": 0.698738877152551,
      "grad_norm": 0.0065573519095778465,
      "learning_rate": 0.12852663052171714,
      "loss": 0.5265,
      "num_input_tokens_seen": 30021008,
      "step": 21830
    },
    {
      "epoch": 0.6988989181230395,
      "grad_norm": 0.004441387485712767,
      "learning_rate": 0.12846833402701507,
      "loss": 0.5938,
      "num_input_tokens_seen": 30027936,
      "step": 21835
    },
    {
      "epoch": 0.699058959093528,
      "grad_norm": 0.004370640963315964,
      "learning_rate": 0.12841004085276642,
      "loss": 0.6756,
      "num_input_tokens_seen": 30034704,
      "step": 21840
    },
    {
      "epoch": 0.6992190000640164,
      "grad_norm": 0.006813535466790199,
      "learning_rate": 0.12835175100796076,
      "loss": 0.5788,
      "num_input_tokens_seen": 30041728,
      "step": 21845
    },
    {
      "epoch": 0.6993790410345049,
      "grad_norm": 0.006504692602902651,
      "learning_rate": 0.12829346450158724,
      "loss": 0.4388,
      "num_input_tokens_seen": 30048160,
      "step": 21850
    },
    {
      "epoch": 0.6995390820049933,
      "grad_norm": 0.004411282017827034,
      "learning_rate": 0.12823518134263423,
      "loss": 0.4185,
      "num_input_tokens_seen": 30054560,
      "step": 21855
    },
    {
      "epoch": 0.6996991229754818,
      "grad_norm": 0.004053784999996424,
      "learning_rate": 0.12817690154008973,
      "loss": 0.4362,
      "num_input_tokens_seen": 30062160,
      "step": 21860
    },
    {
      "epoch": 0.6998591639459701,
      "grad_norm": 0.004768786020576954,
      "learning_rate": 0.12811862510294134,
      "loss": 0.4926,
      "num_input_tokens_seen": 30069008,
      "step": 21865
    },
    {
      "epoch": 0.7000192049164586,
      "grad_norm": 0.0046904622577130795,
      "learning_rate": 0.12806035204017585,
      "loss": 0.536,
      "num_input_tokens_seen": 30075856,
      "step": 21870
    },
    {
      "epoch": 0.700179245886947,
      "grad_norm": 0.004996160510927439,
      "learning_rate": 0.12800208236077987,
      "loss": 0.6701,
      "num_input_tokens_seen": 30082976,
      "step": 21875
    },
    {
      "epoch": 0.7003392868574355,
      "grad_norm": 0.004600906744599342,
      "learning_rate": 0.12794381607373917,
      "loss": 0.5081,
      "num_input_tokens_seen": 30090016,
      "step": 21880
    },
    {
      "epoch": 0.7004993278279239,
      "grad_norm": 0.003957144916057587,
      "learning_rate": 0.12788555318803924,
      "loss": 0.5397,
      "num_input_tokens_seen": 30096576,
      "step": 21885
    },
    {
      "epoch": 0.7006593687984124,
      "grad_norm": 0.007063776720315218,
      "learning_rate": 0.1278272937126649,
      "loss": 0.5005,
      "num_input_tokens_seen": 30103456,
      "step": 21890
    },
    {
      "epoch": 0.7008194097689009,
      "grad_norm": 0.004385913256555796,
      "learning_rate": 0.1277690376566005,
      "loss": 0.4188,
      "num_input_tokens_seen": 30110384,
      "step": 21895
    },
    {
      "epoch": 0.7009794507393893,
      "grad_norm": 0.004095693118870258,
      "learning_rate": 0.12771078502882985,
      "loss": 0.5524,
      "num_input_tokens_seen": 30117776,
      "step": 21900
    },
    {
      "epoch": 0.7011394917098778,
      "grad_norm": 0.003885513637214899,
      "learning_rate": 0.12765253583833633,
      "loss": 0.5843,
      "num_input_tokens_seen": 30124624,
      "step": 21905
    },
    {
      "epoch": 0.7012995326803662,
      "grad_norm": 0.0034614403266459703,
      "learning_rate": 0.12759429009410256,
      "loss": 0.6066,
      "num_input_tokens_seen": 30131872,
      "step": 21910
    },
    {
      "epoch": 0.7014595736508547,
      "grad_norm": 0.007573915645480156,
      "learning_rate": 0.12753604780511085,
      "loss": 0.559,
      "num_input_tokens_seen": 30138736,
      "step": 21915
    },
    {
      "epoch": 0.701619614621343,
      "grad_norm": 0.004344714339822531,
      "learning_rate": 0.12747780898034283,
      "loss": 0.5318,
      "num_input_tokens_seen": 30145840,
      "step": 21920
    },
    {
      "epoch": 0.7017796555918315,
      "grad_norm": 0.003302167169749737,
      "learning_rate": 0.12741957362877973,
      "loss": 0.6366,
      "num_input_tokens_seen": 30152672,
      "step": 21925
    },
    {
      "epoch": 0.7019396965623199,
      "grad_norm": 0.004850419238209724,
      "learning_rate": 0.12736134175940214,
      "loss": 0.6498,
      "num_input_tokens_seen": 30159136,
      "step": 21930
    },
    {
      "epoch": 0.7020997375328084,
      "grad_norm": 0.0058421543799340725,
      "learning_rate": 0.12730311338119016,
      "loss": 0.5979,
      "num_input_tokens_seen": 30165344,
      "step": 21935
    },
    {
      "epoch": 0.7022597785032968,
      "grad_norm": 0.003811925183981657,
      "learning_rate": 0.12724488850312327,
      "loss": 0.4661,
      "num_input_tokens_seen": 30172272,
      "step": 21940
    },
    {
      "epoch": 0.7024198194737853,
      "grad_norm": 0.0031847283244132996,
      "learning_rate": 0.1271866671341806,
      "loss": 0.5092,
      "num_input_tokens_seen": 30179168,
      "step": 21945
    },
    {
      "epoch": 0.7025798604442738,
      "grad_norm": 0.0045278025791049,
      "learning_rate": 0.12712844928334047,
      "loss": 0.9212,
      "num_input_tokens_seen": 30185936,
      "step": 21950
    },
    {
      "epoch": 0.7027399014147622,
      "grad_norm": 0.004665226209908724,
      "learning_rate": 0.12707023495958095,
      "loss": 0.3749,
      "num_input_tokens_seen": 30192784,
      "step": 21955
    },
    {
      "epoch": 0.7028999423852507,
      "grad_norm": 0.0031324257142841816,
      "learning_rate": 0.12701202417187932,
      "loss": 0.4167,
      "num_input_tokens_seen": 30200000,
      "step": 21960
    },
    {
      "epoch": 0.703059983355739,
      "grad_norm": 0.00159674440510571,
      "learning_rate": 0.12695381692921243,
      "loss": 0.3841,
      "num_input_tokens_seen": 30206544,
      "step": 21965
    },
    {
      "epoch": 0.7032200243262275,
      "grad_norm": 0.007518807891756296,
      "learning_rate": 0.12689561324055665,
      "loss": 0.6246,
      "num_input_tokens_seen": 30213328,
      "step": 21970
    },
    {
      "epoch": 0.7033800652967159,
      "grad_norm": 0.0048501440323889256,
      "learning_rate": 0.12683741311488758,
      "loss": 0.5451,
      "num_input_tokens_seen": 30220528,
      "step": 21975
    },
    {
      "epoch": 0.7035401062672044,
      "grad_norm": 0.010939151979982853,
      "learning_rate": 0.1267792165611805,
      "loss": 0.6074,
      "num_input_tokens_seen": 30228080,
      "step": 21980
    },
    {
      "epoch": 0.7037001472376928,
      "grad_norm": 0.009424583055078983,
      "learning_rate": 0.1267210235884101,
      "loss": 0.4496,
      "num_input_tokens_seen": 30234848,
      "step": 21985
    },
    {
      "epoch": 0.7038601882081813,
      "grad_norm": 0.0040415083058178425,
      "learning_rate": 0.12666283420555033,
      "loss": 0.6097,
      "num_input_tokens_seen": 30241840,
      "step": 21990
    },
    {
      "epoch": 0.7040202291786697,
      "grad_norm": 0.005968274548649788,
      "learning_rate": 0.12660464842157487,
      "loss": 0.6815,
      "num_input_tokens_seen": 30248752,
      "step": 21995
    },
    {
      "epoch": 0.7041802701491582,
      "grad_norm": 0.01002202183008194,
      "learning_rate": 0.1265464662454566,
      "loss": 0.4458,
      "num_input_tokens_seen": 30256128,
      "step": 22000
    },
    {
      "epoch": 0.7041802701491582,
      "eval_loss": 0.5339456796646118,
      "eval_runtime": 332.2683,
      "eval_samples_per_second": 41.792,
      "eval_steps_per_second": 20.896,
      "num_input_tokens_seen": 30256128,
      "step": 22000
    },
    {
      "epoch": 0.7043403111196467,
      "grad_norm": 0.003453341545537114,
      "learning_rate": 0.12648828768616793,
      "loss": 0.4824,
      "num_input_tokens_seen": 30262960,
      "step": 22005
    },
    {
      "epoch": 0.7045003520901351,
      "grad_norm": 0.0045168581418693066,
      "learning_rate": 0.12643011275268085,
      "loss": 0.5954,
      "num_input_tokens_seen": 30269584,
      "step": 22010
    },
    {
      "epoch": 0.7046603930606236,
      "grad_norm": 0.004318471997976303,
      "learning_rate": 0.1263719414539665,
      "loss": 0.4655,
      "num_input_tokens_seen": 30276624,
      "step": 22015
    },
    {
      "epoch": 0.7048204340311119,
      "grad_norm": 0.005507009569555521,
      "learning_rate": 0.1263137737989957,
      "loss": 0.5726,
      "num_input_tokens_seen": 30283936,
      "step": 22020
    },
    {
      "epoch": 0.7049804750016004,
      "grad_norm": 0.0044115642085671425,
      "learning_rate": 0.1262556097967387,
      "loss": 0.5642,
      "num_input_tokens_seen": 30291440,
      "step": 22025
    },
    {
      "epoch": 0.7051405159720888,
      "grad_norm": 0.004592353943735361,
      "learning_rate": 0.126197449456165,
      "loss": 0.4348,
      "num_input_tokens_seen": 30298224,
      "step": 22030
    },
    {
      "epoch": 0.7053005569425773,
      "grad_norm": 0.007557590026408434,
      "learning_rate": 0.12613929278624378,
      "loss": 0.6003,
      "num_input_tokens_seen": 30304848,
      "step": 22035
    },
    {
      "epoch": 0.7054605979130657,
      "grad_norm": 0.006506884004920721,
      "learning_rate": 0.12608113979594343,
      "loss": 0.4632,
      "num_input_tokens_seen": 30311040,
      "step": 22040
    },
    {
      "epoch": 0.7056206388835542,
      "grad_norm": 0.0036733057349920273,
      "learning_rate": 0.1260229904942319,
      "loss": 0.5472,
      "num_input_tokens_seen": 30317984,
      "step": 22045
    },
    {
      "epoch": 0.7057806798540426,
      "grad_norm": 0.005290931090712547,
      "learning_rate": 0.12596484489007662,
      "loss": 0.6159,
      "num_input_tokens_seen": 30324512,
      "step": 22050
    },
    {
      "epoch": 0.7059407208245311,
      "grad_norm": 0.0029390440322458744,
      "learning_rate": 0.1259067029924442,
      "loss": 0.4118,
      "num_input_tokens_seen": 30331648,
      "step": 22055
    },
    {
      "epoch": 0.7061007617950196,
      "grad_norm": 0.003953814040869474,
      "learning_rate": 0.12584856481030096,
      "loss": 0.6069,
      "num_input_tokens_seen": 30338688,
      "step": 22060
    },
    {
      "epoch": 0.706260802765508,
      "grad_norm": 0.004458283074200153,
      "learning_rate": 0.12579043035261261,
      "loss": 0.4214,
      "num_input_tokens_seen": 30345216,
      "step": 22065
    },
    {
      "epoch": 0.7064208437359965,
      "grad_norm": 0.003292601788416505,
      "learning_rate": 0.1257322996283441,
      "loss": 0.6078,
      "num_input_tokens_seen": 30352480,
      "step": 22070
    },
    {
      "epoch": 0.7065808847064848,
      "grad_norm": 0.004820235539227724,
      "learning_rate": 0.12567417264645994,
      "loss": 0.5325,
      "num_input_tokens_seen": 30359152,
      "step": 22075
    },
    {
      "epoch": 0.7067409256769733,
      "grad_norm": 0.0044559454545378685,
      "learning_rate": 0.12561604941592408,
      "loss": 0.6108,
      "num_input_tokens_seen": 30365984,
      "step": 22080
    },
    {
      "epoch": 0.7069009666474617,
      "grad_norm": 0.008081074804067612,
      "learning_rate": 0.12555792994569978,
      "loss": 0.4454,
      "num_input_tokens_seen": 30372912,
      "step": 22085
    },
    {
      "epoch": 0.7070610076179502,
      "grad_norm": 0.003502162639051676,
      "learning_rate": 0.1254998142447499,
      "loss": 0.6268,
      "num_input_tokens_seen": 30380528,
      "step": 22090
    },
    {
      "epoch": 0.7072210485884386,
      "grad_norm": 0.005108211655169725,
      "learning_rate": 0.1254417023220365,
      "loss": 0.4573,
      "num_input_tokens_seen": 30387376,
      "step": 22095
    },
    {
      "epoch": 0.7073810895589271,
      "grad_norm": 0.004964611027389765,
      "learning_rate": 0.12538359418652126,
      "loss": 0.3721,
      "num_input_tokens_seen": 30394240,
      "step": 22100
    },
    {
      "epoch": 0.7075411305294156,
      "grad_norm": 0.003762179985642433,
      "learning_rate": 0.12532548984716513,
      "loss": 0.3654,
      "num_input_tokens_seen": 30400784,
      "step": 22105
    },
    {
      "epoch": 0.707701171499904,
      "grad_norm": 0.0015598498284816742,
      "learning_rate": 0.12526738931292855,
      "loss": 0.3423,
      "num_input_tokens_seen": 30407584,
      "step": 22110
    },
    {
      "epoch": 0.7078612124703925,
      "grad_norm": 0.0042604426853358746,
      "learning_rate": 0.1252092925927714,
      "loss": 0.4164,
      "num_input_tokens_seen": 30414176,
      "step": 22115
    },
    {
      "epoch": 0.7080212534408808,
      "grad_norm": 0.004380104597657919,
      "learning_rate": 0.12515119969565278,
      "loss": 0.469,
      "num_input_tokens_seen": 30420880,
      "step": 22120
    },
    {
      "epoch": 0.7081812944113693,
      "grad_norm": 0.005479603074491024,
      "learning_rate": 0.12509311063053144,
      "loss": 0.5475,
      "num_input_tokens_seen": 30427616,
      "step": 22125
    },
    {
      "epoch": 0.7083413353818577,
      "grad_norm": 0.006856982130557299,
      "learning_rate": 0.1250350254063655,
      "loss": 0.4895,
      "num_input_tokens_seen": 30434576,
      "step": 22130
    },
    {
      "epoch": 0.7085013763523462,
      "grad_norm": 0.005362532567232847,
      "learning_rate": 0.1249769440321123,
      "loss": 0.6007,
      "num_input_tokens_seen": 30441312,
      "step": 22135
    },
    {
      "epoch": 0.7086614173228346,
      "grad_norm": 0.0030856034718453884,
      "learning_rate": 0.12491886651672884,
      "loss": 0.3321,
      "num_input_tokens_seen": 30448000,
      "step": 22140
    },
    {
      "epoch": 0.7088214582933231,
      "grad_norm": 0.005557321477681398,
      "learning_rate": 0.12486079286917139,
      "loss": 0.5222,
      "num_input_tokens_seen": 30454672,
      "step": 22145
    },
    {
      "epoch": 0.7089814992638115,
      "grad_norm": 0.006577275227755308,
      "learning_rate": 0.12480272309839553,
      "loss": 0.6301,
      "num_input_tokens_seen": 30461296,
      "step": 22150
    },
    {
      "epoch": 0.7091415402343,
      "grad_norm": 0.003771289484575391,
      "learning_rate": 0.12474465721335648,
      "loss": 0.6239,
      "num_input_tokens_seen": 30468144,
      "step": 22155
    },
    {
      "epoch": 0.7093015812047885,
      "grad_norm": 0.0038809787947684526,
      "learning_rate": 0.12468659522300861,
      "loss": 0.4249,
      "num_input_tokens_seen": 30474784,
      "step": 22160
    },
    {
      "epoch": 0.7094616221752769,
      "grad_norm": 0.005098714958876371,
      "learning_rate": 0.12462853713630584,
      "loss": 0.4014,
      "num_input_tokens_seen": 30481920,
      "step": 22165
    },
    {
      "epoch": 0.7096216631457654,
      "grad_norm": 0.003642935771495104,
      "learning_rate": 0.12457048296220156,
      "loss": 0.3805,
      "num_input_tokens_seen": 30488416,
      "step": 22170
    },
    {
      "epoch": 0.7097817041162537,
      "grad_norm": 0.005179346539080143,
      "learning_rate": 0.12451243270964832,
      "loss": 0.424,
      "num_input_tokens_seen": 30495072,
      "step": 22175
    },
    {
      "epoch": 0.7099417450867422,
      "grad_norm": 0.002779822563752532,
      "learning_rate": 0.12445438638759827,
      "loss": 0.4099,
      "num_input_tokens_seen": 30502064,
      "step": 22180
    },
    {
      "epoch": 0.7101017860572306,
      "grad_norm": 0.006813079118728638,
      "learning_rate": 0.1243963440050029,
      "loss": 0.5307,
      "num_input_tokens_seen": 30508672,
      "step": 22185
    },
    {
      "epoch": 0.7102618270277191,
      "grad_norm": 0.012959908694028854,
      "learning_rate": 0.12433830557081298,
      "loss": 0.5771,
      "num_input_tokens_seen": 30515136,
      "step": 22190
    },
    {
      "epoch": 0.7104218679982075,
      "grad_norm": 0.0038125677965581417,
      "learning_rate": 0.12428027109397889,
      "loss": 0.5231,
      "num_input_tokens_seen": 30521744,
      "step": 22195
    },
    {
      "epoch": 0.710581908968696,
      "grad_norm": 0.0037017955910414457,
      "learning_rate": 0.12422224058345015,
      "loss": 0.452,
      "num_input_tokens_seen": 30528032,
      "step": 22200
    },
    {
      "epoch": 0.710581908968696,
      "eval_loss": 0.5391675233840942,
      "eval_runtime": 331.5891,
      "eval_samples_per_second": 41.877,
      "eval_steps_per_second": 20.939,
      "num_input_tokens_seen": 30528032,
      "step": 22200
    },
    {
      "epoch": 0.7107419499391844,
      "grad_norm": 0.007560868747532368,
      "learning_rate": 0.12416421404817583,
      "loss": 0.5513,
      "num_input_tokens_seen": 30535056,
      "step": 22205
    },
    {
      "epoch": 0.7109019909096729,
      "grad_norm": 0.009526748210191727,
      "learning_rate": 0.12410619149710447,
      "loss": 0.4232,
      "num_input_tokens_seen": 30541520,
      "step": 22210
    },
    {
      "epoch": 0.7110620318801614,
      "grad_norm": 0.005275679286569357,
      "learning_rate": 0.12404817293918374,
      "loss": 0.4231,
      "num_input_tokens_seen": 30548336,
      "step": 22215
    },
    {
      "epoch": 0.7112220728506498,
      "grad_norm": 0.0034485734067857265,
      "learning_rate": 0.12399015838336086,
      "loss": 0.4113,
      "num_input_tokens_seen": 30555152,
      "step": 22220
    },
    {
      "epoch": 0.7113821138211383,
      "grad_norm": 0.011947451159358025,
      "learning_rate": 0.12393214783858246,
      "loss": 0.8073,
      "num_input_tokens_seen": 30562256,
      "step": 22225
    },
    {
      "epoch": 0.7115421547916266,
      "grad_norm": 0.00566583126783371,
      "learning_rate": 0.1238741413137944,
      "loss": 0.5011,
      "num_input_tokens_seen": 30569232,
      "step": 22230
    },
    {
      "epoch": 0.7117021957621151,
      "grad_norm": 0.004841635935008526,
      "learning_rate": 0.12381613881794212,
      "loss": 0.6875,
      "num_input_tokens_seen": 30575904,
      "step": 22235
    },
    {
      "epoch": 0.7118622367326035,
      "grad_norm": 0.0030621180776506662,
      "learning_rate": 0.12375814035997022,
      "loss": 0.3995,
      "num_input_tokens_seen": 30583312,
      "step": 22240
    },
    {
      "epoch": 0.712022277703092,
      "grad_norm": 0.00479919696226716,
      "learning_rate": 0.12370014594882285,
      "loss": 0.5722,
      "num_input_tokens_seen": 30589904,
      "step": 22245
    },
    {
      "epoch": 0.7121823186735804,
      "grad_norm": 0.005496045108884573,
      "learning_rate": 0.12364215559344356,
      "loss": 0.5463,
      "num_input_tokens_seen": 30596288,
      "step": 22250
    },
    {
      "epoch": 0.7123423596440689,
      "grad_norm": 0.005014161579310894,
      "learning_rate": 0.12358416930277506,
      "loss": 0.5739,
      "num_input_tokens_seen": 30603248,
      "step": 22255
    },
    {
      "epoch": 0.7125024006145573,
      "grad_norm": 0.003464675508439541,
      "learning_rate": 0.1235261870857596,
      "loss": 0.419,
      "num_input_tokens_seen": 30609904,
      "step": 22260
    },
    {
      "epoch": 0.7126624415850458,
      "grad_norm": 0.0037145500537008047,
      "learning_rate": 0.12346820895133884,
      "loss": 0.3281,
      "num_input_tokens_seen": 30616784,
      "step": 22265
    },
    {
      "epoch": 0.7128224825555343,
      "grad_norm": 0.0035880310460925102,
      "learning_rate": 0.12341023490845361,
      "loss": 0.5641,
      "num_input_tokens_seen": 30624176,
      "step": 22270
    },
    {
      "epoch": 0.7129825235260226,
      "grad_norm": 0.005808493122458458,
      "learning_rate": 0.12335226496604437,
      "loss": 0.6439,
      "num_input_tokens_seen": 30630864,
      "step": 22275
    },
    {
      "epoch": 0.7131425644965111,
      "grad_norm": 0.0037859869189560413,
      "learning_rate": 0.12329429913305069,
      "loss": 0.5078,
      "num_input_tokens_seen": 30637616,
      "step": 22280
    },
    {
      "epoch": 0.7133026054669995,
      "grad_norm": 0.0022164774127304554,
      "learning_rate": 0.12323633741841171,
      "loss": 0.5106,
      "num_input_tokens_seen": 30644256,
      "step": 22285
    },
    {
      "epoch": 0.713462646437488,
      "grad_norm": 0.004806910175830126,
      "learning_rate": 0.12317837983106583,
      "loss": 0.6436,
      "num_input_tokens_seen": 30650752,
      "step": 22290
    },
    {
      "epoch": 0.7136226874079764,
      "grad_norm": 0.004868002608418465,
      "learning_rate": 0.12312042637995087,
      "loss": 0.5358,
      "num_input_tokens_seen": 30657424,
      "step": 22295
    },
    {
      "epoch": 0.7137827283784649,
      "grad_norm": 0.006269209086894989,
      "learning_rate": 0.12306247707400389,
      "loss": 0.4603,
      "num_input_tokens_seen": 30664192,
      "step": 22300
    },
    {
      "epoch": 0.7139427693489533,
      "grad_norm": 0.00679905666038394,
      "learning_rate": 0.12300453192216154,
      "loss": 0.6024,
      "num_input_tokens_seen": 30670832,
      "step": 22305
    },
    {
      "epoch": 0.7141028103194418,
      "grad_norm": 0.0024078001733869314,
      "learning_rate": 0.12294659093335956,
      "loss": 0.4821,
      "num_input_tokens_seen": 30678208,
      "step": 22310
    },
    {
      "epoch": 0.7142628512899302,
      "grad_norm": 0.006084134336560965,
      "learning_rate": 0.12288865411653327,
      "loss": 0.6125,
      "num_input_tokens_seen": 30684976,
      "step": 22315
    },
    {
      "epoch": 0.7144228922604187,
      "grad_norm": 0.008133797906339169,
      "learning_rate": 0.12283072148061717,
      "loss": 0.5563,
      "num_input_tokens_seen": 30691952,
      "step": 22320
    },
    {
      "epoch": 0.7145829332309072,
      "grad_norm": 0.006765895988792181,
      "learning_rate": 0.12277279303454529,
      "loss": 0.4728,
      "num_input_tokens_seen": 30698912,
      "step": 22325
    },
    {
      "epoch": 0.7147429742013955,
      "grad_norm": 0.002641635946929455,
      "learning_rate": 0.12271486878725091,
      "loss": 0.4576,
      "num_input_tokens_seen": 30705616,
      "step": 22330
    },
    {
      "epoch": 0.714903015171884,
      "grad_norm": 0.004530943930149078,
      "learning_rate": 0.12265694874766658,
      "loss": 0.4373,
      "num_input_tokens_seen": 30712272,
      "step": 22335
    },
    {
      "epoch": 0.7150630561423724,
      "grad_norm": 0.005456211976706982,
      "learning_rate": 0.12259903292472435,
      "loss": 0.5143,
      "num_input_tokens_seen": 30719632,
      "step": 22340
    },
    {
      "epoch": 0.7152230971128609,
      "grad_norm": 0.0057774498127400875,
      "learning_rate": 0.12254112132735567,
      "loss": 0.5067,
      "num_input_tokens_seen": 30726368,
      "step": 22345
    },
    {
      "epoch": 0.7153831380833493,
      "grad_norm": 0.004010878037661314,
      "learning_rate": 0.12248321396449108,
      "loss": 0.5294,
      "num_input_tokens_seen": 30733392,
      "step": 22350
    },
    {
      "epoch": 0.7155431790538378,
      "grad_norm": 0.0038782116025686264,
      "learning_rate": 0.12242531084506075,
      "loss": 0.6798,
      "num_input_tokens_seen": 30740656,
      "step": 22355
    },
    {
      "epoch": 0.7157032200243262,
      "grad_norm": 0.009325983002781868,
      "learning_rate": 0.122367411977994,
      "loss": 0.6002,
      "num_input_tokens_seen": 30747232,
      "step": 22360
    },
    {
      "epoch": 0.7158632609948147,
      "grad_norm": 0.004302721470594406,
      "learning_rate": 0.12230951737221954,
      "loss": 0.578,
      "num_input_tokens_seen": 30754416,
      "step": 22365
    },
    {
      "epoch": 0.7160233019653032,
      "grad_norm": 0.007641791366040707,
      "learning_rate": 0.12225162703666555,
      "loss": 0.6238,
      "num_input_tokens_seen": 30761872,
      "step": 22370
    },
    {
      "epoch": 0.7161833429357916,
      "grad_norm": 0.0031430849339812994,
      "learning_rate": 0.1221937409802593,
      "loss": 0.5713,
      "num_input_tokens_seen": 30768496,
      "step": 22375
    },
    {
      "epoch": 0.71634338390628,
      "grad_norm": 0.008628533221781254,
      "learning_rate": 0.12213585921192768,
      "loss": 0.5313,
      "num_input_tokens_seen": 30775616,
      "step": 22380
    },
    {
      "epoch": 0.7165034248767684,
      "grad_norm": 0.0025936963502317667,
      "learning_rate": 0.1220779817405967,
      "loss": 0.336,
      "num_input_tokens_seen": 30782464,
      "step": 22385
    },
    {
      "epoch": 0.7166634658472569,
      "grad_norm": 0.002901102416217327,
      "learning_rate": 0.12202010857519181,
      "loss": 0.6364,
      "num_input_tokens_seen": 30789872,
      "step": 22390
    },
    {
      "epoch": 0.7168235068177453,
      "grad_norm": 0.00302097643725574,
      "learning_rate": 0.12196223972463785,
      "loss": 0.5383,
      "num_input_tokens_seen": 30797072,
      "step": 22395
    },
    {
      "epoch": 0.7169835477882338,
      "grad_norm": 0.006227423902601004,
      "learning_rate": 0.12190437519785885,
      "loss": 0.608,
      "num_input_tokens_seen": 30803904,
      "step": 22400
    },
    {
      "epoch": 0.7169835477882338,
      "eval_loss": 0.5369259715080261,
      "eval_runtime": 331.6191,
      "eval_samples_per_second": 41.873,
      "eval_steps_per_second": 20.937,
      "num_input_tokens_seen": 30803904,
      "step": 22400
    },
    {
      "epoch": 0.7171435887587222,
      "grad_norm": 0.004423102363944054,
      "learning_rate": 0.12184651500377823,
      "loss": 0.6647,
      "num_input_tokens_seen": 30810976,
      "step": 22405
    },
    {
      "epoch": 0.7173036297292107,
      "grad_norm": 0.004121426958590746,
      "learning_rate": 0.12178865915131885,
      "loss": 0.5607,
      "num_input_tokens_seen": 30817920,
      "step": 22410
    },
    {
      "epoch": 0.7174636706996991,
      "grad_norm": 0.00632835878059268,
      "learning_rate": 0.1217308076494027,
      "loss": 0.5962,
      "num_input_tokens_seen": 30825328,
      "step": 22415
    },
    {
      "epoch": 0.7176237116701876,
      "grad_norm": 0.004894710611552,
      "learning_rate": 0.12167296050695134,
      "loss": 0.4162,
      "num_input_tokens_seen": 30832576,
      "step": 22420
    },
    {
      "epoch": 0.7177837526406761,
      "grad_norm": 0.004557478241622448,
      "learning_rate": 0.12161511773288536,
      "loss": 0.6606,
      "num_input_tokens_seen": 30839648,
      "step": 22425
    },
    {
      "epoch": 0.7179437936111644,
      "grad_norm": 0.0025935498997569084,
      "learning_rate": 0.121557279336125,
      "loss": 0.3909,
      "num_input_tokens_seen": 30846336,
      "step": 22430
    },
    {
      "epoch": 0.7181038345816529,
      "grad_norm": 0.005692094564437866,
      "learning_rate": 0.12149944532558957,
      "loss": 0.4936,
      "num_input_tokens_seen": 30853312,
      "step": 22435
    },
    {
      "epoch": 0.7182638755521413,
      "grad_norm": 0.005541136488318443,
      "learning_rate": 0.12144161571019785,
      "loss": 0.6186,
      "num_input_tokens_seen": 30860128,
      "step": 22440
    },
    {
      "epoch": 0.7184239165226298,
      "grad_norm": 0.005297125782817602,
      "learning_rate": 0.12138379049886781,
      "loss": 0.6221,
      "num_input_tokens_seen": 30867232,
      "step": 22445
    },
    {
      "epoch": 0.7185839574931182,
      "grad_norm": 0.007940540090203285,
      "learning_rate": 0.12132596970051697,
      "loss": 0.4549,
      "num_input_tokens_seen": 30874000,
      "step": 22450
    },
    {
      "epoch": 0.7187439984636067,
      "grad_norm": 0.004030456766486168,
      "learning_rate": 0.12126815332406189,
      "loss": 0.6338,
      "num_input_tokens_seen": 30881008,
      "step": 22455
    },
    {
      "epoch": 0.7189040394340951,
      "grad_norm": 0.003977833781391382,
      "learning_rate": 0.12121034137841868,
      "loss": 0.6478,
      "num_input_tokens_seen": 30887648,
      "step": 22460
    },
    {
      "epoch": 0.7190640804045836,
      "grad_norm": 0.003612960921600461,
      "learning_rate": 0.12115253387250258,
      "loss": 0.6235,
      "num_input_tokens_seen": 30894112,
      "step": 22465
    },
    {
      "epoch": 0.719224121375072,
      "grad_norm": 0.005281766876578331,
      "learning_rate": 0.12109473081522831,
      "loss": 0.3912,
      "num_input_tokens_seen": 30900816,
      "step": 22470
    },
    {
      "epoch": 0.7193841623455605,
      "grad_norm": 0.0036345242988318205,
      "learning_rate": 0.12103693221550982,
      "loss": 0.4937,
      "num_input_tokens_seen": 30907712,
      "step": 22475
    },
    {
      "epoch": 0.719544203316049,
      "grad_norm": 0.004104630555957556,
      "learning_rate": 0.12097913808226027,
      "loss": 0.7652,
      "num_input_tokens_seen": 30914432,
      "step": 22480
    },
    {
      "epoch": 0.7197042442865373,
      "grad_norm": 0.0032885223627090454,
      "learning_rate": 0.12092134842439234,
      "loss": 0.4016,
      "num_input_tokens_seen": 30921504,
      "step": 22485
    },
    {
      "epoch": 0.7198642852570258,
      "grad_norm": 0.003124308306723833,
      "learning_rate": 0.12086356325081798,
      "loss": 0.3481,
      "num_input_tokens_seen": 30928032,
      "step": 22490
    },
    {
      "epoch": 0.7200243262275142,
      "grad_norm": 0.0034800905268639326,
      "learning_rate": 0.12080578257044824,
      "loss": 0.433,
      "num_input_tokens_seen": 30934944,
      "step": 22495
    },
    {
      "epoch": 0.7201843671980027,
      "grad_norm": 0.004787159152328968,
      "learning_rate": 0.12074800639219378,
      "loss": 0.4496,
      "num_input_tokens_seen": 30942000,
      "step": 22500
    },
    {
      "epoch": 0.7203444081684911,
      "grad_norm": 0.0047556497156620026,
      "learning_rate": 0.12069023472496428,
      "loss": 0.7112,
      "num_input_tokens_seen": 30948656,
      "step": 22505
    },
    {
      "epoch": 0.7205044491389796,
      "grad_norm": 0.004154772963374853,
      "learning_rate": 0.12063246757766893,
      "loss": 0.5651,
      "num_input_tokens_seen": 30955504,
      "step": 22510
    },
    {
      "epoch": 0.720664490109468,
      "grad_norm": 0.005149948876351118,
      "learning_rate": 0.12057470495921618,
      "loss": 0.3811,
      "num_input_tokens_seen": 30962000,
      "step": 22515
    },
    {
      "epoch": 0.7208245310799565,
      "grad_norm": 0.005500344093888998,
      "learning_rate": 0.12051694687851364,
      "loss": 0.6119,
      "num_input_tokens_seen": 30968848,
      "step": 22520
    },
    {
      "epoch": 0.7209845720504449,
      "grad_norm": 0.0025735129602253437,
      "learning_rate": 0.12045919334446839,
      "loss": 0.4972,
      "num_input_tokens_seen": 30975584,
      "step": 22525
    },
    {
      "epoch": 0.7211446130209334,
      "grad_norm": 0.0052067288197577,
      "learning_rate": 0.12040144436598683,
      "loss": 0.474,
      "num_input_tokens_seen": 30982464,
      "step": 22530
    },
    {
      "epoch": 0.7213046539914219,
      "grad_norm": 0.006131328642368317,
      "learning_rate": 0.12034369995197444,
      "loss": 0.5197,
      "num_input_tokens_seen": 30989120,
      "step": 22535
    },
    {
      "epoch": 0.7214646949619102,
      "grad_norm": 0.006424741353839636,
      "learning_rate": 0.12028596011133627,
      "loss": 0.4752,
      "num_input_tokens_seen": 30995632,
      "step": 22540
    },
    {
      "epoch": 0.7216247359323987,
      "grad_norm": 0.006900355219841003,
      "learning_rate": 0.12022822485297643,
      "loss": 0.707,
      "num_input_tokens_seen": 31002448,
      "step": 22545
    },
    {
      "epoch": 0.7217847769028871,
      "grad_norm": 0.004413241520524025,
      "learning_rate": 0.12017049418579843,
      "loss": 0.3537,
      "num_input_tokens_seen": 31008880,
      "step": 22550
    },
    {
      "epoch": 0.7219448178733756,
      "grad_norm": 0.007651384919881821,
      "learning_rate": 0.12011276811870514,
      "loss": 0.7037,
      "num_input_tokens_seen": 31015232,
      "step": 22555
    },
    {
      "epoch": 0.722104858843864,
      "grad_norm": 0.003559782402589917,
      "learning_rate": 0.12005504666059852,
      "loss": 0.5566,
      "num_input_tokens_seen": 31022128,
      "step": 22560
    },
    {
      "epoch": 0.7222648998143525,
      "grad_norm": 0.004283577669411898,
      "learning_rate": 0.11999732982038003,
      "loss": 0.5183,
      "num_input_tokens_seen": 31029120,
      "step": 22565
    },
    {
      "epoch": 0.7224249407848409,
      "grad_norm": 0.0056988997384905815,
      "learning_rate": 0.11993961760695038,
      "loss": 0.4593,
      "num_input_tokens_seen": 31035984,
      "step": 22570
    },
    {
      "epoch": 0.7225849817553294,
      "grad_norm": 0.004399907775223255,
      "learning_rate": 0.11988191002920942,
      "loss": 0.4291,
      "num_input_tokens_seen": 31043344,
      "step": 22575
    },
    {
      "epoch": 0.7227450227258178,
      "grad_norm": 0.0036552725359797478,
      "learning_rate": 0.11982420709605641,
      "loss": 0.5184,
      "num_input_tokens_seen": 31050096,
      "step": 22580
    },
    {
      "epoch": 0.7229050636963062,
      "grad_norm": 0.00616455776616931,
      "learning_rate": 0.11976650881638991,
      "loss": 0.4156,
      "num_input_tokens_seen": 31057040,
      "step": 22585
    },
    {
      "epoch": 0.7230651046667947,
      "grad_norm": 0.005482041276991367,
      "learning_rate": 0.11970881519910764,
      "loss": 0.5609,
      "num_input_tokens_seen": 31063744,
      "step": 22590
    },
    {
      "epoch": 0.7232251456372831,
      "grad_norm": 0.0034746367018669844,
      "learning_rate": 0.1196511262531068,
      "loss": 0.4483,
      "num_input_tokens_seen": 31070688,
      "step": 22595
    },
    {
      "epoch": 0.7233851866077716,
      "grad_norm": 0.00426492327824235,
      "learning_rate": 0.11959344198728361,
      "loss": 0.7056,
      "num_input_tokens_seen": 31077632,
      "step": 22600
    },
    {
      "epoch": 0.7233851866077716,
      "eval_loss": 0.5328890681266785,
      "eval_runtime": 331.942,
      "eval_samples_per_second": 41.833,
      "eval_steps_per_second": 20.916,
      "num_input_tokens_seen": 31077632,
      "step": 22600
    },
    {
      "epoch": 0.72354522757826,
      "grad_norm": 0.004953298717737198,
      "learning_rate": 0.11953576241053378,
      "loss": 0.6019,
      "num_input_tokens_seen": 31084560,
      "step": 22605
    },
    {
      "epoch": 0.7237052685487485,
      "grad_norm": 0.0038547536823898554,
      "learning_rate": 0.11947808753175228,
      "loss": 0.5111,
      "num_input_tokens_seen": 31091792,
      "step": 22610
    },
    {
      "epoch": 0.7238653095192369,
      "grad_norm": 0.005733686499297619,
      "learning_rate": 0.1194204173598332,
      "loss": 0.4511,
      "num_input_tokens_seen": 31098464,
      "step": 22615
    },
    {
      "epoch": 0.7240253504897254,
      "grad_norm": 0.0051308502443134785,
      "learning_rate": 0.11936275190367007,
      "loss": 0.6431,
      "num_input_tokens_seen": 31105584,
      "step": 22620
    },
    {
      "epoch": 0.7241853914602138,
      "grad_norm": 0.0067002479918301105,
      "learning_rate": 0.11930509117215563,
      "loss": 0.6686,
      "num_input_tokens_seen": 31112320,
      "step": 22625
    },
    {
      "epoch": 0.7243454324307023,
      "grad_norm": 0.006892415229231119,
      "learning_rate": 0.11924743517418179,
      "loss": 0.6008,
      "num_input_tokens_seen": 31119184,
      "step": 22630
    },
    {
      "epoch": 0.7245054734011908,
      "grad_norm": 0.00544470502063632,
      "learning_rate": 0.11918978391864,
      "loss": 0.4818,
      "num_input_tokens_seen": 31126192,
      "step": 22635
    },
    {
      "epoch": 0.7246655143716791,
      "grad_norm": 0.005240310914814472,
      "learning_rate": 0.11913213741442065,
      "loss": 0.5747,
      "num_input_tokens_seen": 31133552,
      "step": 22640
    },
    {
      "epoch": 0.7248255553421676,
      "grad_norm": 0.0024629791732877493,
      "learning_rate": 0.11907449567041364,
      "loss": 0.4935,
      "num_input_tokens_seen": 31140112,
      "step": 22645
    },
    {
      "epoch": 0.724985596312656,
      "grad_norm": 0.0030684955418109894,
      "learning_rate": 0.11901685869550803,
      "loss": 0.5231,
      "num_input_tokens_seen": 31147136,
      "step": 22650
    },
    {
      "epoch": 0.7251456372831445,
      "grad_norm": 0.003395436331629753,
      "learning_rate": 0.1189592264985922,
      "loss": 0.4567,
      "num_input_tokens_seen": 31153984,
      "step": 22655
    },
    {
      "epoch": 0.7253056782536329,
      "grad_norm": 0.004497291054576635,
      "learning_rate": 0.11890159908855373,
      "loss": 0.4939,
      "num_input_tokens_seen": 31160624,
      "step": 22660
    },
    {
      "epoch": 0.7254657192241214,
      "grad_norm": 0.004297100473195314,
      "learning_rate": 0.11884397647427941,
      "loss": 0.4325,
      "num_input_tokens_seen": 31167424,
      "step": 22665
    },
    {
      "epoch": 0.7256257601946098,
      "grad_norm": 0.006790527608245611,
      "learning_rate": 0.11878635866465546,
      "loss": 0.6008,
      "num_input_tokens_seen": 31174208,
      "step": 22670
    },
    {
      "epoch": 0.7257858011650983,
      "grad_norm": 0.006429414730519056,
      "learning_rate": 0.11872874566856734,
      "loss": 0.3938,
      "num_input_tokens_seen": 31181040,
      "step": 22675
    },
    {
      "epoch": 0.7259458421355867,
      "grad_norm": 0.001937663066200912,
      "learning_rate": 0.11867113749489955,
      "loss": 0.5556,
      "num_input_tokens_seen": 31187776,
      "step": 22680
    },
    {
      "epoch": 0.7261058831060752,
      "grad_norm": 0.003090126207098365,
      "learning_rate": 0.11861353415253607,
      "loss": 0.4932,
      "num_input_tokens_seen": 31194304,
      "step": 22685
    },
    {
      "epoch": 0.7262659240765637,
      "grad_norm": 0.0027974764816462994,
      "learning_rate": 0.11855593565036011,
      "loss": 0.6049,
      "num_input_tokens_seen": 31201280,
      "step": 22690
    },
    {
      "epoch": 0.726425965047052,
      "grad_norm": 0.0018798564560711384,
      "learning_rate": 0.11849834199725394,
      "loss": 0.4533,
      "num_input_tokens_seen": 31208160,
      "step": 22695
    },
    {
      "epoch": 0.7265860060175405,
      "grad_norm": 0.0042821308597922325,
      "learning_rate": 0.1184407532020994,
      "loss": 0.6575,
      "num_input_tokens_seen": 31215056,
      "step": 22700
    },
    {
      "epoch": 0.7267460469880289,
      "grad_norm": 0.003309437073767185,
      "learning_rate": 0.11838316927377723,
      "loss": 0.3837,
      "num_input_tokens_seen": 31221856,
      "step": 22705
    },
    {
      "epoch": 0.7269060879585174,
      "grad_norm": 0.0058858972042799,
      "learning_rate": 0.11832559022116766,
      "loss": 0.5669,
      "num_input_tokens_seen": 31228688,
      "step": 22710
    },
    {
      "epoch": 0.7270661289290058,
      "grad_norm": 0.005636887159198523,
      "learning_rate": 0.11826801605315022,
      "loss": 0.5417,
      "num_input_tokens_seen": 31235728,
      "step": 22715
    },
    {
      "epoch": 0.7272261698994943,
      "grad_norm": 0.0076821851544082165,
      "learning_rate": 0.1182104467786034,
      "loss": 0.725,
      "num_input_tokens_seen": 31242512,
      "step": 22720
    },
    {
      "epoch": 0.7273862108699827,
      "grad_norm": 0.003417073981836438,
      "learning_rate": 0.1181528824064052,
      "loss": 0.442,
      "num_input_tokens_seen": 31249312,
      "step": 22725
    },
    {
      "epoch": 0.7275462518404712,
      "grad_norm": 0.005036122165620327,
      "learning_rate": 0.11809532294543279,
      "loss": 0.707,
      "num_input_tokens_seen": 31256784,
      "step": 22730
    },
    {
      "epoch": 0.7277062928109596,
      "grad_norm": 0.0035227418411523104,
      "learning_rate": 0.11803776840456245,
      "loss": 0.5277,
      "num_input_tokens_seen": 31264048,
      "step": 22735
    },
    {
      "epoch": 0.727866333781448,
      "grad_norm": 0.004334829747676849,
      "learning_rate": 0.11798021879266997,
      "loss": 0.4349,
      "num_input_tokens_seen": 31270832,
      "step": 22740
    },
    {
      "epoch": 0.7280263747519365,
      "grad_norm": 0.003284081583842635,
      "learning_rate": 0.11792267411863006,
      "loss": 0.3593,
      "num_input_tokens_seen": 31277728,
      "step": 22745
    },
    {
      "epoch": 0.7281864157224249,
      "grad_norm": 0.007319828495383263,
      "learning_rate": 0.1178651343913169,
      "loss": 0.575,
      "num_input_tokens_seen": 31284672,
      "step": 22750
    },
    {
      "epoch": 0.7283464566929134,
      "grad_norm": 0.005338010378181934,
      "learning_rate": 0.11780759961960392,
      "loss": 0.502,
      "num_input_tokens_seen": 31291248,
      "step": 22755
    },
    {
      "epoch": 0.7285064976634018,
      "grad_norm": 0.005309192463755608,
      "learning_rate": 0.1177500698123636,
      "loss": 0.3845,
      "num_input_tokens_seen": 31298272,
      "step": 22760
    },
    {
      "epoch": 0.7286665386338903,
      "grad_norm": 0.0026438923086971045,
      "learning_rate": 0.11769254497846778,
      "loss": 0.4114,
      "num_input_tokens_seen": 31304976,
      "step": 22765
    },
    {
      "epoch": 0.7288265796043787,
      "grad_norm": 0.004866008181124926,
      "learning_rate": 0.11763502512678758,
      "loss": 0.4712,
      "num_input_tokens_seen": 31312096,
      "step": 22770
    },
    {
      "epoch": 0.7289866205748672,
      "grad_norm": 0.003774001495912671,
      "learning_rate": 0.11757751026619315,
      "loss": 0.4947,
      "num_input_tokens_seen": 31318976,
      "step": 22775
    },
    {
      "epoch": 0.7291466615453556,
      "grad_norm": 0.0032038716599345207,
      "learning_rate": 0.11752000040555416,
      "loss": 0.6146,
      "num_input_tokens_seen": 31326464,
      "step": 22780
    },
    {
      "epoch": 0.7293067025158441,
      "grad_norm": 0.003312087384983897,
      "learning_rate": 0.11746249555373921,
      "loss": 0.5242,
      "num_input_tokens_seen": 31333232,
      "step": 22785
    },
    {
      "epoch": 0.7294667434863324,
      "grad_norm": 0.005090189632028341,
      "learning_rate": 0.11740499571961638,
      "loss": 0.5263,
      "num_input_tokens_seen": 31340656,
      "step": 22790
    },
    {
      "epoch": 0.7296267844568209,
      "grad_norm": 0.005160207860171795,
      "learning_rate": 0.11734750091205279,
      "loss": 0.4114,
      "num_input_tokens_seen": 31347664,
      "step": 22795
    },
    {
      "epoch": 0.7297868254273094,
      "grad_norm": 0.006524462252855301,
      "learning_rate": 0.11729001113991493,
      "loss": 0.5536,
      "num_input_tokens_seen": 31354544,
      "step": 22800
    },
    {
      "epoch": 0.7297868254273094,
      "eval_loss": 0.530288815498352,
      "eval_runtime": 331.9039,
      "eval_samples_per_second": 41.837,
      "eval_steps_per_second": 20.919,
      "num_input_tokens_seen": 31354544,
      "step": 22800
    },
    {
      "epoch": 0.7299468663977978,
      "grad_norm": 0.004365478176623583,
      "learning_rate": 0.11723252641206837,
      "loss": 0.418,
      "num_input_tokens_seen": 31361424,
      "step": 22805
    },
    {
      "epoch": 0.7301069073682863,
      "grad_norm": 0.00637094909325242,
      "learning_rate": 0.11717504673737808,
      "loss": 0.404,
      "num_input_tokens_seen": 31368384,
      "step": 22810
    },
    {
      "epoch": 0.7302669483387747,
      "grad_norm": 0.0036403241101652384,
      "learning_rate": 0.11711757212470802,
      "loss": 0.5525,
      "num_input_tokens_seen": 31374640,
      "step": 22815
    },
    {
      "epoch": 0.7304269893092632,
      "grad_norm": 0.0037817100528627634,
      "learning_rate": 0.11706010258292165,
      "loss": 0.5278,
      "num_input_tokens_seen": 31381104,
      "step": 22820
    },
    {
      "epoch": 0.7305870302797516,
      "grad_norm": 0.0024608667008578777,
      "learning_rate": 0.11700263812088131,
      "loss": 0.4738,
      "num_input_tokens_seen": 31387536,
      "step": 22825
    },
    {
      "epoch": 0.7307470712502401,
      "grad_norm": 0.005112083628773689,
      "learning_rate": 0.11694517874744892,
      "loss": 0.5805,
      "num_input_tokens_seen": 31394128,
      "step": 22830
    },
    {
      "epoch": 0.7309071122207285,
      "grad_norm": 0.00428067846223712,
      "learning_rate": 0.11688772447148532,
      "loss": 0.5391,
      "num_input_tokens_seen": 31400976,
      "step": 22835
    },
    {
      "epoch": 0.731067153191217,
      "grad_norm": 0.003136588493362069,
      "learning_rate": 0.11683027530185074,
      "loss": 0.568,
      "num_input_tokens_seen": 31408000,
      "step": 22840
    },
    {
      "epoch": 0.7312271941617055,
      "grad_norm": 0.004582615103572607,
      "learning_rate": 0.11677283124740451,
      "loss": 0.4497,
      "num_input_tokens_seen": 31415072,
      "step": 22845
    },
    {
      "epoch": 0.7313872351321938,
      "grad_norm": 0.002676158444955945,
      "learning_rate": 0.11671539231700531,
      "loss": 0.541,
      "num_input_tokens_seen": 31421840,
      "step": 22850
    },
    {
      "epoch": 0.7315472761026823,
      "grad_norm": 0.003951875492930412,
      "learning_rate": 0.11665795851951084,
      "loss": 0.4431,
      "num_input_tokens_seen": 31428464,
      "step": 22855
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.00421429518610239,
      "learning_rate": 0.11660052986377825,
      "loss": 0.4269,
      "num_input_tokens_seen": 31435504,
      "step": 22860
    },
    {
      "epoch": 0.7318673580436592,
      "grad_norm": 0.0025041664484888315,
      "learning_rate": 0.1165431063586636,
      "loss": 0.3696,
      "num_input_tokens_seen": 31442208,
      "step": 22865
    },
    {
      "epoch": 0.7320273990141476,
      "grad_norm": 0.004379799589514732,
      "learning_rate": 0.11648568801302245,
      "loss": 0.5156,
      "num_input_tokens_seen": 31448944,
      "step": 22870
    },
    {
      "epoch": 0.7321874399846361,
      "grad_norm": 0.004530338104814291,
      "learning_rate": 0.11642827483570937,
      "loss": 0.3932,
      "num_input_tokens_seen": 31455888,
      "step": 22875
    },
    {
      "epoch": 0.7323474809551245,
      "grad_norm": 0.006200017407536507,
      "learning_rate": 0.11637086683557815,
      "loss": 0.486,
      "num_input_tokens_seen": 31462864,
      "step": 22880
    },
    {
      "epoch": 0.732507521925613,
      "grad_norm": 0.004079591948539019,
      "learning_rate": 0.11631346402148188,
      "loss": 0.4975,
      "num_input_tokens_seen": 31469504,
      "step": 22885
    },
    {
      "epoch": 0.7326675628961014,
      "grad_norm": 0.008735904470086098,
      "learning_rate": 0.11625606640227285,
      "loss": 0.6127,
      "num_input_tokens_seen": 31476336,
      "step": 22890
    },
    {
      "epoch": 0.7328276038665898,
      "grad_norm": 0.005515082739293575,
      "learning_rate": 0.11619867398680238,
      "loss": 0.6757,
      "num_input_tokens_seen": 31483168,
      "step": 22895
    },
    {
      "epoch": 0.7329876448370783,
      "grad_norm": 0.007880435325205326,
      "learning_rate": 0.11614128678392119,
      "loss": 0.7081,
      "num_input_tokens_seen": 31490272,
      "step": 22900
    },
    {
      "epoch": 0.7331476858075667,
      "grad_norm": 0.007279742043465376,
      "learning_rate": 0.11608390480247906,
      "loss": 0.6121,
      "num_input_tokens_seen": 31496832,
      "step": 22905
    },
    {
      "epoch": 0.7333077267780552,
      "grad_norm": 0.004283972084522247,
      "learning_rate": 0.11602652805132499,
      "loss": 0.5437,
      "num_input_tokens_seen": 31503536,
      "step": 22910
    },
    {
      "epoch": 0.7334677677485436,
      "grad_norm": 0.002778289606794715,
      "learning_rate": 0.11596915653930731,
      "loss": 0.424,
      "num_input_tokens_seen": 31510256,
      "step": 22915
    },
    {
      "epoch": 0.7336278087190321,
      "grad_norm": 0.0032583074644207954,
      "learning_rate": 0.11591179027527328,
      "loss": 0.542,
      "num_input_tokens_seen": 31517296,
      "step": 22920
    },
    {
      "epoch": 0.7337878496895205,
      "grad_norm": 0.006179861258715391,
      "learning_rate": 0.11585442926806956,
      "loss": 0.6301,
      "num_input_tokens_seen": 31524000,
      "step": 22925
    },
    {
      "epoch": 0.733947890660009,
      "grad_norm": 0.004916607867926359,
      "learning_rate": 0.11579707352654202,
      "loss": 0.5614,
      "num_input_tokens_seen": 31530912,
      "step": 22930
    },
    {
      "epoch": 0.7341079316304974,
      "grad_norm": 0.006952032912522554,
      "learning_rate": 0.11573972305953548,
      "loss": 0.4497,
      "num_input_tokens_seen": 31537808,
      "step": 22935
    },
    {
      "epoch": 0.7342679726009859,
      "grad_norm": 0.005049984902143478,
      "learning_rate": 0.11568237787589426,
      "loss": 0.5881,
      "num_input_tokens_seen": 31544816,
      "step": 22940
    },
    {
      "epoch": 0.7344280135714742,
      "grad_norm": 0.00515425018966198,
      "learning_rate": 0.11562503798446161,
      "loss": 0.4141,
      "num_input_tokens_seen": 31551584,
      "step": 22945
    },
    {
      "epoch": 0.7345880545419627,
      "grad_norm": 0.005193233024328947,
      "learning_rate": 0.11556770339408005,
      "loss": 0.4552,
      "num_input_tokens_seen": 31558544,
      "step": 22950
    },
    {
      "epoch": 0.7347480955124512,
      "grad_norm": 0.005577665753662586,
      "learning_rate": 0.1155103741135914,
      "loss": 0.5705,
      "num_input_tokens_seen": 31565360,
      "step": 22955
    },
    {
      "epoch": 0.7349081364829396,
      "grad_norm": 0.008727485314011574,
      "learning_rate": 0.1154530501518364,
      "loss": 0.6777,
      "num_input_tokens_seen": 31571888,
      "step": 22960
    },
    {
      "epoch": 0.7350681774534281,
      "grad_norm": 0.005474078003317118,
      "learning_rate": 0.11539573151765523,
      "loss": 0.5911,
      "num_input_tokens_seen": 31578576,
      "step": 22965
    },
    {
      "epoch": 0.7352282184239165,
      "grad_norm": 0.004391213413327932,
      "learning_rate": 0.11533841821988719,
      "loss": 0.4971,
      "num_input_tokens_seen": 31585264,
      "step": 22970
    },
    {
      "epoch": 0.735388259394405,
      "grad_norm": 0.007711675018072128,
      "learning_rate": 0.11528111026737059,
      "loss": 0.4979,
      "num_input_tokens_seen": 31592096,
      "step": 22975
    },
    {
      "epoch": 0.7355483003648934,
      "grad_norm": 0.003634247463196516,
      "learning_rate": 0.11522380766894312,
      "loss": 0.5893,
      "num_input_tokens_seen": 31598992,
      "step": 22980
    },
    {
      "epoch": 0.7357083413353819,
      "grad_norm": 0.004434159956872463,
      "learning_rate": 0.11516651043344152,
      "loss": 0.5025,
      "num_input_tokens_seen": 31605824,
      "step": 22985
    },
    {
      "epoch": 0.7358683823058703,
      "grad_norm": 0.003085073083639145,
      "learning_rate": 0.11510921856970172,
      "loss": 0.4664,
      "num_input_tokens_seen": 31612976,
      "step": 22990
    },
    {
      "epoch": 0.7360284232763588,
      "grad_norm": 0.0048600537702441216,
      "learning_rate": 0.11505193208655895,
      "loss": 0.5228,
      "num_input_tokens_seen": 31620224,
      "step": 22995
    },
    {
      "epoch": 0.7361884642468471,
      "grad_norm": 0.008527729660272598,
      "learning_rate": 0.11499465099284738,
      "loss": 0.594,
      "num_input_tokens_seen": 31626736,
      "step": 23000
    },
    {
      "epoch": 0.7361884642468471,
      "eval_loss": 0.527244508266449,
      "eval_runtime": 332.0115,
      "eval_samples_per_second": 41.824,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 31626736,
      "step": 23000
    },
    {
      "epoch": 0.7363485052173356,
      "grad_norm": 0.0032084574922919273,
      "learning_rate": 0.1149373752974006,
      "loss": 0.5745,
      "num_input_tokens_seen": 31633360,
      "step": 23005
    },
    {
      "epoch": 0.7365085461878241,
      "grad_norm": 0.0029659250285476446,
      "learning_rate": 0.11488010500905109,
      "loss": 0.5989,
      "num_input_tokens_seen": 31640144,
      "step": 23010
    },
    {
      "epoch": 0.7366685871583125,
      "grad_norm": 0.002684955485165119,
      "learning_rate": 0.11482284013663077,
      "loss": 0.6016,
      "num_input_tokens_seen": 31647216,
      "step": 23015
    },
    {
      "epoch": 0.736828628128801,
      "grad_norm": 0.0043228138238191605,
      "learning_rate": 0.11476558068897061,
      "loss": 0.7091,
      "num_input_tokens_seen": 31654416,
      "step": 23020
    },
    {
      "epoch": 0.7369886690992894,
      "grad_norm": 0.003980796784162521,
      "learning_rate": 0.11470832667490061,
      "loss": 0.639,
      "num_input_tokens_seen": 31661600,
      "step": 23025
    },
    {
      "epoch": 0.7371487100697779,
      "grad_norm": 0.002240376314148307,
      "learning_rate": 0.11465107810325013,
      "loss": 0.4448,
      "num_input_tokens_seen": 31668176,
      "step": 23030
    },
    {
      "epoch": 0.7373087510402663,
      "grad_norm": 0.0035769061651080847,
      "learning_rate": 0.11459383498284771,
      "loss": 0.5001,
      "num_input_tokens_seen": 31676336,
      "step": 23035
    },
    {
      "epoch": 0.7374687920107548,
      "grad_norm": 0.0036376004572957754,
      "learning_rate": 0.11453659732252082,
      "loss": 0.4974,
      "num_input_tokens_seen": 31683072,
      "step": 23040
    },
    {
      "epoch": 0.7376288329812432,
      "grad_norm": 0.006851377431303263,
      "learning_rate": 0.11447936513109633,
      "loss": 0.5025,
      "num_input_tokens_seen": 31690352,
      "step": 23045
    },
    {
      "epoch": 0.7377888739517316,
      "grad_norm": 0.0036730917636305094,
      "learning_rate": 0.11442213841740011,
      "loss": 0.4226,
      "num_input_tokens_seen": 31697088,
      "step": 23050
    },
    {
      "epoch": 0.73794891492222,
      "grad_norm": 0.005778862629085779,
      "learning_rate": 0.1143649171902572,
      "loss": 0.4614,
      "num_input_tokens_seen": 31703648,
      "step": 23055
    },
    {
      "epoch": 0.7381089558927085,
      "grad_norm": 0.0049844891764223576,
      "learning_rate": 0.11430770145849194,
      "loss": 0.6172,
      "num_input_tokens_seen": 31710304,
      "step": 23060
    },
    {
      "epoch": 0.738268996863197,
      "grad_norm": 0.003437704173848033,
      "learning_rate": 0.11425049123092756,
      "loss": 0.4188,
      "num_input_tokens_seen": 31717248,
      "step": 23065
    },
    {
      "epoch": 0.7384290378336854,
      "grad_norm": 0.0023116045631468296,
      "learning_rate": 0.11419328651638674,
      "loss": 0.3146,
      "num_input_tokens_seen": 31723856,
      "step": 23070
    },
    {
      "epoch": 0.7385890788041739,
      "grad_norm": 0.002759914379566908,
      "learning_rate": 0.11413608732369115,
      "loss": 0.4316,
      "num_input_tokens_seen": 31730800,
      "step": 23075
    },
    {
      "epoch": 0.7387491197746623,
      "grad_norm": 0.0029720363672822714,
      "learning_rate": 0.11407889366166153,
      "loss": 0.5189,
      "num_input_tokens_seen": 31737504,
      "step": 23080
    },
    {
      "epoch": 0.7389091607451508,
      "grad_norm": 0.0053105964325368404,
      "learning_rate": 0.11402170553911797,
      "loss": 0.5195,
      "num_input_tokens_seen": 31744288,
      "step": 23085
    },
    {
      "epoch": 0.7390692017156392,
      "grad_norm": 0.0032305107451975346,
      "learning_rate": 0.11396452296487955,
      "loss": 0.5367,
      "num_input_tokens_seen": 31751088,
      "step": 23090
    },
    {
      "epoch": 0.7392292426861277,
      "grad_norm": 0.004336309619247913,
      "learning_rate": 0.11390734594776449,
      "loss": 0.5356,
      "num_input_tokens_seen": 31758288,
      "step": 23095
    },
    {
      "epoch": 0.739389283656616,
      "grad_norm": 0.004305638372898102,
      "learning_rate": 0.11385017449659031,
      "loss": 0.5533,
      "num_input_tokens_seen": 31764528,
      "step": 23100
    },
    {
      "epoch": 0.7395493246271045,
      "grad_norm": 0.004798075184226036,
      "learning_rate": 0.11379300862017344,
      "loss": 0.4729,
      "num_input_tokens_seen": 31771200,
      "step": 23105
    },
    {
      "epoch": 0.739709365597593,
      "grad_norm": 0.004602794535458088,
      "learning_rate": 0.11373584832732966,
      "loss": 0.4463,
      "num_input_tokens_seen": 31777792,
      "step": 23110
    },
    {
      "epoch": 0.7398694065680814,
      "grad_norm": 0.0065804957412183285,
      "learning_rate": 0.11367869362687386,
      "loss": 0.7586,
      "num_input_tokens_seen": 31784576,
      "step": 23115
    },
    {
      "epoch": 0.7400294475385699,
      "grad_norm": 0.007684122771024704,
      "learning_rate": 0.11362154452761988,
      "loss": 0.4782,
      "num_input_tokens_seen": 31791136,
      "step": 23120
    },
    {
      "epoch": 0.7401894885090583,
      "grad_norm": 0.003306918079033494,
      "learning_rate": 0.11356440103838095,
      "loss": 0.4078,
      "num_input_tokens_seen": 31798048,
      "step": 23125
    },
    {
      "epoch": 0.7403495294795468,
      "grad_norm": 0.0024990644305944443,
      "learning_rate": 0.11350726316796922,
      "loss": 0.34,
      "num_input_tokens_seen": 31804752,
      "step": 23130
    },
    {
      "epoch": 0.7405095704500352,
      "grad_norm": 0.0061267768032848835,
      "learning_rate": 0.11345013092519607,
      "loss": 0.5774,
      "num_input_tokens_seen": 31811696,
      "step": 23135
    },
    {
      "epoch": 0.7406696114205237,
      "grad_norm": 0.004779368173331022,
      "learning_rate": 0.11339300431887213,
      "loss": 0.6116,
      "num_input_tokens_seen": 31818704,
      "step": 23140
    },
    {
      "epoch": 0.7408296523910121,
      "grad_norm": 0.008281279355287552,
      "learning_rate": 0.11333588335780687,
      "loss": 0.5594,
      "num_input_tokens_seen": 31825552,
      "step": 23145
    },
    {
      "epoch": 0.7409896933615006,
      "grad_norm": 0.004317398648709059,
      "learning_rate": 0.11327876805080916,
      "loss": 0.4143,
      "num_input_tokens_seen": 31832592,
      "step": 23150
    },
    {
      "epoch": 0.7411497343319889,
      "grad_norm": 0.004594513680785894,
      "learning_rate": 0.11322165840668696,
      "loss": 0.5767,
      "num_input_tokens_seen": 31839296,
      "step": 23155
    },
    {
      "epoch": 0.7413097753024774,
      "grad_norm": 0.006462369579821825,
      "learning_rate": 0.11316455443424717,
      "loss": 0.4681,
      "num_input_tokens_seen": 31845760,
      "step": 23160
    },
    {
      "epoch": 0.7414698162729659,
      "grad_norm": 0.0037924994248896837,
      "learning_rate": 0.11310745614229603,
      "loss": 0.6553,
      "num_input_tokens_seen": 31852576,
      "step": 23165
    },
    {
      "epoch": 0.7416298572434543,
      "grad_norm": 0.004647198133170605,
      "learning_rate": 0.1130503635396387,
      "loss": 0.588,
      "num_input_tokens_seen": 31859760,
      "step": 23170
    },
    {
      "epoch": 0.7417898982139428,
      "grad_norm": 0.004700062330812216,
      "learning_rate": 0.11299327663507966,
      "loss": 0.4283,
      "num_input_tokens_seen": 31866496,
      "step": 23175
    },
    {
      "epoch": 0.7419499391844312,
      "grad_norm": 0.01070236787199974,
      "learning_rate": 0.11293619543742246,
      "loss": 0.5125,
      "num_input_tokens_seen": 31873616,
      "step": 23180
    },
    {
      "epoch": 0.7421099801549197,
      "grad_norm": 0.0035676101688295603,
      "learning_rate": 0.11287911995546965,
      "loss": 0.5867,
      "num_input_tokens_seen": 31880608,
      "step": 23185
    },
    {
      "epoch": 0.7422700211254081,
      "grad_norm": 0.002541943686082959,
      "learning_rate": 0.11282205019802308,
      "loss": 0.5326,
      "num_input_tokens_seen": 31887440,
      "step": 23190
    },
    {
      "epoch": 0.7424300620958966,
      "grad_norm": 0.00526055321097374,
      "learning_rate": 0.11276498617388354,
      "loss": 0.5612,
      "num_input_tokens_seen": 31894560,
      "step": 23195
    },
    {
      "epoch": 0.742590103066385,
      "grad_norm": 0.0034504635259509087,
      "learning_rate": 0.11270792789185109,
      "loss": 0.4026,
      "num_input_tokens_seen": 31901472,
      "step": 23200
    },
    {
      "epoch": 0.742590103066385,
      "eval_loss": 0.5333110094070435,
      "eval_runtime": 332.1114,
      "eval_samples_per_second": 41.811,
      "eval_steps_per_second": 20.906,
      "num_input_tokens_seen": 31901472,
      "step": 23200
    },
    {
      "epoch": 0.7427501440368734,
      "grad_norm": 0.0031829779036343098,
      "learning_rate": 0.11265087536072482,
      "loss": 0.4915,
      "num_input_tokens_seen": 31908592,
      "step": 23205
    },
    {
      "epoch": 0.7429101850073618,
      "grad_norm": 0.005592056550085545,
      "learning_rate": 0.11259382858930288,
      "loss": 0.476,
      "num_input_tokens_seen": 31915488,
      "step": 23210
    },
    {
      "epoch": 0.7430702259778503,
      "grad_norm": 0.0049525233916938305,
      "learning_rate": 0.11253678758638262,
      "loss": 0.5709,
      "num_input_tokens_seen": 31922336,
      "step": 23215
    },
    {
      "epoch": 0.7432302669483388,
      "grad_norm": 0.0037091311533004045,
      "learning_rate": 0.11247975236076059,
      "loss": 0.505,
      "num_input_tokens_seen": 31929504,
      "step": 23220
    },
    {
      "epoch": 0.7433903079188272,
      "grad_norm": 0.007807297166436911,
      "learning_rate": 0.11242272292123218,
      "loss": 0.8008,
      "num_input_tokens_seen": 31936608,
      "step": 23225
    },
    {
      "epoch": 0.7435503488893157,
      "grad_norm": 0.00475909560918808,
      "learning_rate": 0.11236569927659217,
      "loss": 0.5088,
      "num_input_tokens_seen": 31943216,
      "step": 23230
    },
    {
      "epoch": 0.7437103898598041,
      "grad_norm": 0.004044638480991125,
      "learning_rate": 0.11230868143563429,
      "loss": 0.6525,
      "num_input_tokens_seen": 31950240,
      "step": 23235
    },
    {
      "epoch": 0.7438704308302926,
      "grad_norm": 0.005018346942961216,
      "learning_rate": 0.11225166940715131,
      "loss": 0.3714,
      "num_input_tokens_seen": 31957040,
      "step": 23240
    },
    {
      "epoch": 0.744030471800781,
      "grad_norm": 0.007090817205607891,
      "learning_rate": 0.11219466319993537,
      "loss": 0.594,
      "num_input_tokens_seen": 31964608,
      "step": 23245
    },
    {
      "epoch": 0.7441905127712695,
      "grad_norm": 0.0034560374915599823,
      "learning_rate": 0.11213766282277739,
      "loss": 0.7438,
      "num_input_tokens_seen": 31971584,
      "step": 23250
    },
    {
      "epoch": 0.7443505537417578,
      "grad_norm": 0.0027998811565339565,
      "learning_rate": 0.11208066828446761,
      "loss": 0.3648,
      "num_input_tokens_seen": 31978352,
      "step": 23255
    },
    {
      "epoch": 0.7445105947122463,
      "grad_norm": 0.0037456071004271507,
      "learning_rate": 0.11202367959379537,
      "loss": 0.5268,
      "num_input_tokens_seen": 31984672,
      "step": 23260
    },
    {
      "epoch": 0.7446706356827347,
      "grad_norm": 0.004736834671348333,
      "learning_rate": 0.11196669675954894,
      "loss": 0.6092,
      "num_input_tokens_seen": 31991472,
      "step": 23265
    },
    {
      "epoch": 0.7448306766532232,
      "grad_norm": 0.005580897442996502,
      "learning_rate": 0.1119097197905158,
      "loss": 0.4757,
      "num_input_tokens_seen": 31998256,
      "step": 23270
    },
    {
      "epoch": 0.7449907176237117,
      "grad_norm": 0.005391440819948912,
      "learning_rate": 0.11185274869548259,
      "loss": 0.3487,
      "num_input_tokens_seen": 32005424,
      "step": 23275
    },
    {
      "epoch": 0.7451507585942001,
      "grad_norm": 0.0031395405530929565,
      "learning_rate": 0.11179578348323486,
      "loss": 0.4677,
      "num_input_tokens_seen": 32012480,
      "step": 23280
    },
    {
      "epoch": 0.7453107995646886,
      "grad_norm": 0.002977053401991725,
      "learning_rate": 0.1117388241625575,
      "loss": 0.5627,
      "num_input_tokens_seen": 32019808,
      "step": 23285
    },
    {
      "epoch": 0.745470840535177,
      "grad_norm": 0.004499363712966442,
      "learning_rate": 0.11168187074223421,
      "loss": 0.5382,
      "num_input_tokens_seen": 32026496,
      "step": 23290
    },
    {
      "epoch": 0.7456308815056655,
      "grad_norm": 0.005079113878309727,
      "learning_rate": 0.11162492323104796,
      "loss": 0.7088,
      "num_input_tokens_seen": 32033360,
      "step": 23295
    },
    {
      "epoch": 0.7457909224761539,
      "grad_norm": 0.00593687454238534,
      "learning_rate": 0.11156798163778091,
      "loss": 0.5299,
      "num_input_tokens_seen": 32040304,
      "step": 23300
    },
    {
      "epoch": 0.7459509634466424,
      "grad_norm": 0.006819567643105984,
      "learning_rate": 0.11151104597121399,
      "loss": 0.6451,
      "num_input_tokens_seen": 32047456,
      "step": 23305
    },
    {
      "epoch": 0.7461110044171307,
      "grad_norm": 0.004792673047631979,
      "learning_rate": 0.11145411624012742,
      "loss": 0.5601,
      "num_input_tokens_seen": 32054688,
      "step": 23310
    },
    {
      "epoch": 0.7462710453876192,
      "grad_norm": 0.00374649278819561,
      "learning_rate": 0.11139719245330063,
      "loss": 0.3969,
      "num_input_tokens_seen": 32061584,
      "step": 23315
    },
    {
      "epoch": 0.7464310863581076,
      "grad_norm": 0.005386375822126865,
      "learning_rate": 0.11134027461951179,
      "loss": 0.2568,
      "num_input_tokens_seen": 32068464,
      "step": 23320
    },
    {
      "epoch": 0.7465911273285961,
      "grad_norm": 0.003025546669960022,
      "learning_rate": 0.11128336274753849,
      "loss": 0.593,
      "num_input_tokens_seen": 32075424,
      "step": 23325
    },
    {
      "epoch": 0.7467511682990846,
      "grad_norm": 0.005513148847967386,
      "learning_rate": 0.11122645684615715,
      "loss": 0.6478,
      "num_input_tokens_seen": 32082176,
      "step": 23330
    },
    {
      "epoch": 0.746911209269573,
      "grad_norm": 0.0030398936942219734,
      "learning_rate": 0.11116955692414345,
      "loss": 0.459,
      "num_input_tokens_seen": 32089616,
      "step": 23335
    },
    {
      "epoch": 0.7470712502400615,
      "grad_norm": 0.005496103782206774,
      "learning_rate": 0.11111266299027203,
      "loss": 0.5652,
      "num_input_tokens_seen": 32096800,
      "step": 23340
    },
    {
      "epoch": 0.7472312912105499,
      "grad_norm": 0.005498453509062529,
      "learning_rate": 0.11105577505331668,
      "loss": 0.6842,
      "num_input_tokens_seen": 32103200,
      "step": 23345
    },
    {
      "epoch": 0.7473913321810384,
      "grad_norm": 0.004028089344501495,
      "learning_rate": 0.11099889312205018,
      "loss": 0.5134,
      "num_input_tokens_seen": 32110384,
      "step": 23350
    },
    {
      "epoch": 0.7475513731515268,
      "grad_norm": 0.003442407352849841,
      "learning_rate": 0.11094201720524455,
      "loss": 0.485,
      "num_input_tokens_seen": 32116960,
      "step": 23355
    },
    {
      "epoch": 0.7477114141220152,
      "grad_norm": 0.002899385057389736,
      "learning_rate": 0.11088514731167064,
      "loss": 0.6344,
      "num_input_tokens_seen": 32123776,
      "step": 23360
    },
    {
      "epoch": 0.7478714550925036,
      "grad_norm": 0.0025472580455243587,
      "learning_rate": 0.11082828345009862,
      "loss": 0.6582,
      "num_input_tokens_seen": 32130384,
      "step": 23365
    },
    {
      "epoch": 0.7480314960629921,
      "grad_norm": 0.00515768863260746,
      "learning_rate": 0.11077142562929748,
      "loss": 0.52,
      "num_input_tokens_seen": 32137664,
      "step": 23370
    },
    {
      "epoch": 0.7481915370334806,
      "grad_norm": 0.008598633110523224,
      "learning_rate": 0.11071457385803554,
      "loss": 0.6846,
      "num_input_tokens_seen": 32144592,
      "step": 23375
    },
    {
      "epoch": 0.748351578003969,
      "grad_norm": 0.005082926247268915,
      "learning_rate": 0.11065772814508001,
      "loss": 0.4741,
      "num_input_tokens_seen": 32151680,
      "step": 23380
    },
    {
      "epoch": 0.7485116189744575,
      "grad_norm": 0.002992135938256979,
      "learning_rate": 0.11060088849919715,
      "loss": 0.4873,
      "num_input_tokens_seen": 32158640,
      "step": 23385
    },
    {
      "epoch": 0.7486716599449459,
      "grad_norm": 0.0034754350781440735,
      "learning_rate": 0.11054405492915244,
      "loss": 0.4169,
      "num_input_tokens_seen": 32165632,
      "step": 23390
    },
    {
      "epoch": 0.7488317009154344,
      "grad_norm": 0.0022615455090999603,
      "learning_rate": 0.11048722744371031,
      "loss": 0.5269,
      "num_input_tokens_seen": 32172880,
      "step": 23395
    },
    {
      "epoch": 0.7489917418859228,
      "grad_norm": 0.00611089775338769,
      "learning_rate": 0.1104304060516342,
      "loss": 0.4775,
      "num_input_tokens_seen": 32179968,
      "step": 23400
    },
    {
      "epoch": 0.7489917418859228,
      "eval_loss": 0.5250540971755981,
      "eval_runtime": 332.0249,
      "eval_samples_per_second": 41.822,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 32179968,
      "step": 23400
    },
    {
      "epoch": 0.7491517828564113,
      "grad_norm": 0.008232650347054005,
      "learning_rate": 0.11037359076168682,
      "loss": 0.6569,
      "num_input_tokens_seen": 32186928,
      "step": 23405
    },
    {
      "epoch": 0.7493118238268996,
      "grad_norm": 0.005531429313123226,
      "learning_rate": 0.11031678158262966,
      "loss": 0.4818,
      "num_input_tokens_seen": 32193776,
      "step": 23410
    },
    {
      "epoch": 0.7494718647973881,
      "grad_norm": 0.007085063960403204,
      "learning_rate": 0.11025997852322349,
      "loss": 0.5997,
      "num_input_tokens_seen": 32200352,
      "step": 23415
    },
    {
      "epoch": 0.7496319057678765,
      "grad_norm": 0.005481385625898838,
      "learning_rate": 0.11020318159222807,
      "loss": 0.5624,
      "num_input_tokens_seen": 32207568,
      "step": 23420
    },
    {
      "epoch": 0.749791946738365,
      "grad_norm": 0.003189771668985486,
      "learning_rate": 0.1101463907984021,
      "loss": 0.4303,
      "num_input_tokens_seen": 32214480,
      "step": 23425
    },
    {
      "epoch": 0.7499519877088535,
      "grad_norm": 0.004777433350682259,
      "learning_rate": 0.11008960615050352,
      "loss": 0.5623,
      "num_input_tokens_seen": 32221024,
      "step": 23430
    },
    {
      "epoch": 0.7501120286793419,
      "grad_norm": 0.005945923272520304,
      "learning_rate": 0.11003282765728925,
      "loss": 0.4952,
      "num_input_tokens_seen": 32227840,
      "step": 23435
    },
    {
      "epoch": 0.7502720696498304,
      "grad_norm": 0.006948897149413824,
      "learning_rate": 0.10997605532751518,
      "loss": 0.4913,
      "num_input_tokens_seen": 32234752,
      "step": 23440
    },
    {
      "epoch": 0.7504321106203188,
      "grad_norm": 0.0029883345123380423,
      "learning_rate": 0.1099192891699364,
      "loss": 0.5804,
      "num_input_tokens_seen": 32241552,
      "step": 23445
    },
    {
      "epoch": 0.7505921515908073,
      "grad_norm": 0.0031739911064505577,
      "learning_rate": 0.10986252919330687,
      "loss": 0.5286,
      "num_input_tokens_seen": 32248368,
      "step": 23450
    },
    {
      "epoch": 0.7507521925612957,
      "grad_norm": 0.004022277891635895,
      "learning_rate": 0.10980577540637973,
      "loss": 0.5122,
      "num_input_tokens_seen": 32255664,
      "step": 23455
    },
    {
      "epoch": 0.7509122335317842,
      "grad_norm": 0.006023390684276819,
      "learning_rate": 0.10974902781790719,
      "loss": 0.5464,
      "num_input_tokens_seen": 32262048,
      "step": 23460
    },
    {
      "epoch": 0.7510722745022725,
      "grad_norm": 0.004751662258058786,
      "learning_rate": 0.10969228643664032,
      "loss": 0.5288,
      "num_input_tokens_seen": 32269216,
      "step": 23465
    },
    {
      "epoch": 0.751232315472761,
      "grad_norm": 0.0032300930470228195,
      "learning_rate": 0.10963555127132942,
      "loss": 0.5409,
      "num_input_tokens_seen": 32276272,
      "step": 23470
    },
    {
      "epoch": 0.7513923564432494,
      "grad_norm": 0.003896591253578663,
      "learning_rate": 0.10957882233072382,
      "loss": 0.4249,
      "num_input_tokens_seen": 32283360,
      "step": 23475
    },
    {
      "epoch": 0.7515523974137379,
      "grad_norm": 0.004784835036844015,
      "learning_rate": 0.10952209962357176,
      "loss": 0.5608,
      "num_input_tokens_seen": 32290400,
      "step": 23480
    },
    {
      "epoch": 0.7517124383842264,
      "grad_norm": 0.004705514758825302,
      "learning_rate": 0.10946538315862062,
      "loss": 0.608,
      "num_input_tokens_seen": 32297200,
      "step": 23485
    },
    {
      "epoch": 0.7518724793547148,
      "grad_norm": 0.005002094432711601,
      "learning_rate": 0.10940867294461679,
      "loss": 0.5022,
      "num_input_tokens_seen": 32304160,
      "step": 23490
    },
    {
      "epoch": 0.7520325203252033,
      "grad_norm": 0.0022350000217556953,
      "learning_rate": 0.10935196899030565,
      "loss": 0.4058,
      "num_input_tokens_seen": 32311440,
      "step": 23495
    },
    {
      "epoch": 0.7521925612956917,
      "grad_norm": 0.003966890275478363,
      "learning_rate": 0.10929527130443177,
      "loss": 0.4906,
      "num_input_tokens_seen": 32319072,
      "step": 23500
    },
    {
      "epoch": 0.7523526022661802,
      "grad_norm": 0.004805440083146095,
      "learning_rate": 0.1092385798957385,
      "loss": 0.5563,
      "num_input_tokens_seen": 32325680,
      "step": 23505
    },
    {
      "epoch": 0.7525126432366686,
      "grad_norm": 0.0035940425004810095,
      "learning_rate": 0.10918189477296848,
      "loss": 0.487,
      "num_input_tokens_seen": 32332320,
      "step": 23510
    },
    {
      "epoch": 0.752672684207157,
      "grad_norm": 0.0039032031781971455,
      "learning_rate": 0.1091252159448633,
      "loss": 0.5177,
      "num_input_tokens_seen": 32339216,
      "step": 23515
    },
    {
      "epoch": 0.7528327251776454,
      "grad_norm": 0.006932584568858147,
      "learning_rate": 0.10906854342016345,
      "loss": 0.5967,
      "num_input_tokens_seen": 32345680,
      "step": 23520
    },
    {
      "epoch": 0.7529927661481339,
      "grad_norm": 0.003930500242859125,
      "learning_rate": 0.10901187720760858,
      "loss": 0.4527,
      "num_input_tokens_seen": 32352688,
      "step": 23525
    },
    {
      "epoch": 0.7531528071186223,
      "grad_norm": 0.003063844982534647,
      "learning_rate": 0.10895521731593734,
      "loss": 0.3652,
      "num_input_tokens_seen": 32359728,
      "step": 23530
    },
    {
      "epoch": 0.7533128480891108,
      "grad_norm": 0.005065892357379198,
      "learning_rate": 0.10889856375388733,
      "loss": 0.5705,
      "num_input_tokens_seen": 32367360,
      "step": 23535
    },
    {
      "epoch": 0.7534728890595993,
      "grad_norm": 0.0044392747804522514,
      "learning_rate": 0.1088419165301954,
      "loss": 0.4674,
      "num_input_tokens_seen": 32374512,
      "step": 23540
    },
    {
      "epoch": 0.7536329300300877,
      "grad_norm": 0.005588581785559654,
      "learning_rate": 0.1087852756535971,
      "loss": 0.4764,
      "num_input_tokens_seen": 32381408,
      "step": 23545
    },
    {
      "epoch": 0.7537929710005762,
      "grad_norm": 0.003900444833561778,
      "learning_rate": 0.10872864113282725,
      "loss": 0.478,
      "num_input_tokens_seen": 32388256,
      "step": 23550
    },
    {
      "epoch": 0.7539530119710646,
      "grad_norm": 0.006336525548249483,
      "learning_rate": 0.10867201297661958,
      "loss": 0.6065,
      "num_input_tokens_seen": 32395536,
      "step": 23555
    },
    {
      "epoch": 0.7541130529415531,
      "grad_norm": 0.0037786338943988085,
      "learning_rate": 0.10861539119370689,
      "loss": 0.4378,
      "num_input_tokens_seen": 32402464,
      "step": 23560
    },
    {
      "epoch": 0.7542730939120414,
      "grad_norm": 0.007851816713809967,
      "learning_rate": 0.10855877579282096,
      "loss": 0.5668,
      "num_input_tokens_seen": 32409664,
      "step": 23565
    },
    {
      "epoch": 0.7544331348825299,
      "grad_norm": 0.005238791927695274,
      "learning_rate": 0.10850216678269252,
      "loss": 0.443,
      "num_input_tokens_seen": 32416704,
      "step": 23570
    },
    {
      "epoch": 0.7545931758530183,
      "grad_norm": 0.004814666695892811,
      "learning_rate": 0.10844556417205146,
      "loss": 0.5884,
      "num_input_tokens_seen": 32423568,
      "step": 23575
    },
    {
      "epoch": 0.7547532168235068,
      "grad_norm": 0.009123525582253933,
      "learning_rate": 0.10838896796962669,
      "loss": 0.8251,
      "num_input_tokens_seen": 32430368,
      "step": 23580
    },
    {
      "epoch": 0.7549132577939953,
      "grad_norm": 0.007239544298499823,
      "learning_rate": 0.1083323781841459,
      "loss": 0.5974,
      "num_input_tokens_seen": 32437088,
      "step": 23585
    },
    {
      "epoch": 0.7550732987644837,
      "grad_norm": 0.0038199545815587044,
      "learning_rate": 0.10827579482433607,
      "loss": 0.5015,
      "num_input_tokens_seen": 32444176,
      "step": 23590
    },
    {
      "epoch": 0.7552333397349722,
      "grad_norm": 0.0049401866272091866,
      "learning_rate": 0.10821921789892304,
      "loss": 0.4708,
      "num_input_tokens_seen": 32450704,
      "step": 23595
    },
    {
      "epoch": 0.7553933807054606,
      "grad_norm": 0.007896163500845432,
      "learning_rate": 0.10816264741663158,
      "loss": 0.8036,
      "num_input_tokens_seen": 32457728,
      "step": 23600
    },
    {
      "epoch": 0.7553933807054606,
      "eval_loss": 0.5234057903289795,
      "eval_runtime": 331.9815,
      "eval_samples_per_second": 41.828,
      "eval_steps_per_second": 20.914,
      "num_input_tokens_seen": 32457728,
      "step": 23600
    },
    {
      "epoch": 0.7555534216759491,
      "grad_norm": 0.00178397202398628,
      "learning_rate": 0.10810608338618573,
      "loss": 0.4629,
      "num_input_tokens_seen": 32464224,
      "step": 23605
    },
    {
      "epoch": 0.7557134626464375,
      "grad_norm": 0.004712054505944252,
      "learning_rate": 0.10804952581630821,
      "loss": 0.6089,
      "num_input_tokens_seen": 32471312,
      "step": 23610
    },
    {
      "epoch": 0.755873503616926,
      "grad_norm": 0.004716184921562672,
      "learning_rate": 0.10799297471572102,
      "loss": 0.5988,
      "num_input_tokens_seen": 32478336,
      "step": 23615
    },
    {
      "epoch": 0.7560335445874143,
      "grad_norm": 0.004171342123299837,
      "learning_rate": 0.10793643009314507,
      "loss": 0.4739,
      "num_input_tokens_seen": 32485168,
      "step": 23620
    },
    {
      "epoch": 0.7561935855579028,
      "grad_norm": 0.004666878841817379,
      "learning_rate": 0.10787989195730015,
      "loss": 0.3798,
      "num_input_tokens_seen": 32491424,
      "step": 23625
    },
    {
      "epoch": 0.7563536265283912,
      "grad_norm": 0.006228539161384106,
      "learning_rate": 0.10782336031690525,
      "loss": 0.6186,
      "num_input_tokens_seen": 32498832,
      "step": 23630
    },
    {
      "epoch": 0.7565136674988797,
      "grad_norm": 0.0032120186369866133,
      "learning_rate": 0.10776683518067821,
      "loss": 0.4539,
      "num_input_tokens_seen": 32505520,
      "step": 23635
    },
    {
      "epoch": 0.7566737084693682,
      "grad_norm": 0.0036488750483840704,
      "learning_rate": 0.10771031655733587,
      "loss": 0.4814,
      "num_input_tokens_seen": 32512384,
      "step": 23640
    },
    {
      "epoch": 0.7568337494398566,
      "grad_norm": 0.005905428901314735,
      "learning_rate": 0.10765380445559422,
      "loss": 0.4906,
      "num_input_tokens_seen": 32519056,
      "step": 23645
    },
    {
      "epoch": 0.7569937904103451,
      "grad_norm": 0.003158945357427001,
      "learning_rate": 0.10759729888416801,
      "loss": 0.4409,
      "num_input_tokens_seen": 32525952,
      "step": 23650
    },
    {
      "epoch": 0.7571538313808335,
      "grad_norm": 0.0068109286949038506,
      "learning_rate": 0.10754079985177119,
      "loss": 0.6154,
      "num_input_tokens_seen": 32532720,
      "step": 23655
    },
    {
      "epoch": 0.757313872351322,
      "grad_norm": 0.0032902704551815987,
      "learning_rate": 0.10748430736711667,
      "loss": 0.4507,
      "num_input_tokens_seen": 32539648,
      "step": 23660
    },
    {
      "epoch": 0.7574739133218104,
      "grad_norm": 0.005561427213251591,
      "learning_rate": 0.10742782143891623,
      "loss": 0.5832,
      "num_input_tokens_seen": 32546320,
      "step": 23665
    },
    {
      "epoch": 0.7576339542922989,
      "grad_norm": 0.005281665828078985,
      "learning_rate": 0.10737134207588069,
      "loss": 0.2498,
      "num_input_tokens_seen": 32553232,
      "step": 23670
    },
    {
      "epoch": 0.7577939952627872,
      "grad_norm": 0.0038296226412057877,
      "learning_rate": 0.10731486928671992,
      "loss": 0.5559,
      "num_input_tokens_seen": 32559952,
      "step": 23675
    },
    {
      "epoch": 0.7579540362332757,
      "grad_norm": 0.003173388307914138,
      "learning_rate": 0.10725840308014269,
      "loss": 0.5939,
      "num_input_tokens_seen": 32566944,
      "step": 23680
    },
    {
      "epoch": 0.7581140772037641,
      "grad_norm": 0.004605639260262251,
      "learning_rate": 0.10720194346485688,
      "loss": 0.4197,
      "num_input_tokens_seen": 32573632,
      "step": 23685
    },
    {
      "epoch": 0.7582741181742526,
      "grad_norm": 0.006954639218747616,
      "learning_rate": 0.10714549044956918,
      "loss": 0.6347,
      "num_input_tokens_seen": 32580352,
      "step": 23690
    },
    {
      "epoch": 0.7584341591447411,
      "grad_norm": 0.004796850960701704,
      "learning_rate": 0.10708904404298542,
      "loss": 0.4653,
      "num_input_tokens_seen": 32587280,
      "step": 23695
    },
    {
      "epoch": 0.7585942001152295,
      "grad_norm": 0.004295811522752047,
      "learning_rate": 0.1070326042538103,
      "loss": 0.631,
      "num_input_tokens_seen": 32594000,
      "step": 23700
    },
    {
      "epoch": 0.758754241085718,
      "grad_norm": 0.00437525799497962,
      "learning_rate": 0.10697617109074758,
      "loss": 0.5785,
      "num_input_tokens_seen": 32600624,
      "step": 23705
    },
    {
      "epoch": 0.7589142820562064,
      "grad_norm": 0.003962505608797073,
      "learning_rate": 0.10691974456249999,
      "loss": 0.4074,
      "num_input_tokens_seen": 32607472,
      "step": 23710
    },
    {
      "epoch": 0.7590743230266949,
      "grad_norm": 0.005573108792304993,
      "learning_rate": 0.10686332467776909,
      "loss": 0.752,
      "num_input_tokens_seen": 32614368,
      "step": 23715
    },
    {
      "epoch": 0.7592343639971832,
      "grad_norm": 0.005728152114897966,
      "learning_rate": 0.10680691144525563,
      "loss": 0.5748,
      "num_input_tokens_seen": 32620976,
      "step": 23720
    },
    {
      "epoch": 0.7593944049676717,
      "grad_norm": 0.007831727154552937,
      "learning_rate": 0.10675050487365928,
      "loss": 0.6277,
      "num_input_tokens_seen": 32627872,
      "step": 23725
    },
    {
      "epoch": 0.7595544459381601,
      "grad_norm": 0.004864500369876623,
      "learning_rate": 0.10669410497167851,
      "loss": 0.5241,
      "num_input_tokens_seen": 32635120,
      "step": 23730
    },
    {
      "epoch": 0.7597144869086486,
      "grad_norm": 0.0036656437441706657,
      "learning_rate": 0.10663771174801102,
      "loss": 0.4155,
      "num_input_tokens_seen": 32641568,
      "step": 23735
    },
    {
      "epoch": 0.759874527879137,
      "grad_norm": 0.005116717424243689,
      "learning_rate": 0.10658132521135329,
      "loss": 0.5895,
      "num_input_tokens_seen": 32648528,
      "step": 23740
    },
    {
      "epoch": 0.7600345688496255,
      "grad_norm": 0.0064493282698094845,
      "learning_rate": 0.10652494537040084,
      "loss": 0.3759,
      "num_input_tokens_seen": 32655552,
      "step": 23745
    },
    {
      "epoch": 0.760194609820114,
      "grad_norm": 0.00617881678044796,
      "learning_rate": 0.1064685722338482,
      "loss": 0.5812,
      "num_input_tokens_seen": 32662704,
      "step": 23750
    },
    {
      "epoch": 0.7603546507906024,
      "grad_norm": 0.008798317052423954,
      "learning_rate": 0.10641220581038871,
      "loss": 0.6744,
      "num_input_tokens_seen": 32669536,
      "step": 23755
    },
    {
      "epoch": 0.7605146917610909,
      "grad_norm": 0.004148604813963175,
      "learning_rate": 0.10635584610871483,
      "loss": 0.4569,
      "num_input_tokens_seen": 32676304,
      "step": 23760
    },
    {
      "epoch": 0.7606747327315793,
      "grad_norm": 0.005446512717753649,
      "learning_rate": 0.10629949313751803,
      "loss": 0.6701,
      "num_input_tokens_seen": 32683296,
      "step": 23765
    },
    {
      "epoch": 0.7608347737020678,
      "grad_norm": 0.0033094179816544056,
      "learning_rate": 0.10624314690548849,
      "loss": 0.5364,
      "num_input_tokens_seen": 32690352,
      "step": 23770
    },
    {
      "epoch": 0.7609948146725561,
      "grad_norm": 0.0034667900763452053,
      "learning_rate": 0.1061868074213156,
      "loss": 0.4927,
      "num_input_tokens_seen": 32697504,
      "step": 23775
    },
    {
      "epoch": 0.7611548556430446,
      "grad_norm": 0.0026620279531925917,
      "learning_rate": 0.10613047469368765,
      "loss": 0.5507,
      "num_input_tokens_seen": 32704272,
      "step": 23780
    },
    {
      "epoch": 0.761314896613533,
      "grad_norm": 0.0035000108182430267,
      "learning_rate": 0.10607414873129171,
      "loss": 0.6534,
      "num_input_tokens_seen": 32711808,
      "step": 23785
    },
    {
      "epoch": 0.7614749375840215,
      "grad_norm": 0.006265026982873678,
      "learning_rate": 0.10601782954281413,
      "loss": 0.5943,
      "num_input_tokens_seen": 32718512,
      "step": 23790
    },
    {
      "epoch": 0.7616349785545099,
      "grad_norm": 0.0073075913824141026,
      "learning_rate": 0.1059615171369399,
      "loss": 0.6089,
      "num_input_tokens_seen": 32725360,
      "step": 23795
    },
    {
      "epoch": 0.7617950195249984,
      "grad_norm": 0.008841676637530327,
      "learning_rate": 0.10590521152235312,
      "loss": 0.5097,
      "num_input_tokens_seen": 32732288,
      "step": 23800
    },
    {
      "epoch": 0.7617950195249984,
      "eval_loss": 0.5252019166946411,
      "eval_runtime": 331.9615,
      "eval_samples_per_second": 41.83,
      "eval_steps_per_second": 20.915,
      "num_input_tokens_seen": 32732288,
      "step": 23800
    },
    {
      "epoch": 0.7619550604954869,
      "grad_norm": 0.006259195040911436,
      "learning_rate": 0.1058489127077369,
      "loss": 0.5928,
      "num_input_tokens_seen": 32739216,
      "step": 23805
    },
    {
      "epoch": 0.7621151014659753,
      "grad_norm": 0.006046810187399387,
      "learning_rate": 0.1057926207017732,
      "loss": 0.7894,
      "num_input_tokens_seen": 32746448,
      "step": 23810
    },
    {
      "epoch": 0.7622751424364638,
      "grad_norm": 0.004403335973620415,
      "learning_rate": 0.10573633551314285,
      "loss": 0.5619,
      "num_input_tokens_seen": 32753344,
      "step": 23815
    },
    {
      "epoch": 0.7624351834069522,
      "grad_norm": 0.010415323078632355,
      "learning_rate": 0.1056800571505259,
      "loss": 0.6461,
      "num_input_tokens_seen": 32760096,
      "step": 23820
    },
    {
      "epoch": 0.7625952243774407,
      "grad_norm": 0.0032516885548830032,
      "learning_rate": 0.10562378562260105,
      "loss": 0.5133,
      "num_input_tokens_seen": 32766912,
      "step": 23825
    },
    {
      "epoch": 0.762755265347929,
      "grad_norm": 0.006276856642216444,
      "learning_rate": 0.10556752093804615,
      "loss": 0.4801,
      "num_input_tokens_seen": 32773376,
      "step": 23830
    },
    {
      "epoch": 0.7629153063184175,
      "grad_norm": 0.004583241418004036,
      "learning_rate": 0.10551126310553786,
      "loss": 0.4766,
      "num_input_tokens_seen": 32780560,
      "step": 23835
    },
    {
      "epoch": 0.7630753472889059,
      "grad_norm": 0.002494577784091234,
      "learning_rate": 0.10545501213375187,
      "loss": 0.4105,
      "num_input_tokens_seen": 32787904,
      "step": 23840
    },
    {
      "epoch": 0.7632353882593944,
      "grad_norm": 0.005292195826768875,
      "learning_rate": 0.10539876803136287,
      "loss": 0.5736,
      "num_input_tokens_seen": 32794928,
      "step": 23845
    },
    {
      "epoch": 0.7633954292298829,
      "grad_norm": 0.001077477470971644,
      "learning_rate": 0.10534253080704428,
      "loss": 0.5166,
      "num_input_tokens_seen": 32802208,
      "step": 23850
    },
    {
      "epoch": 0.7635554702003713,
      "grad_norm": 0.0040268367156386375,
      "learning_rate": 0.10528630046946862,
      "loss": 0.4899,
      "num_input_tokens_seen": 32809040,
      "step": 23855
    },
    {
      "epoch": 0.7637155111708598,
      "grad_norm": 0.008519050665199757,
      "learning_rate": 0.1052300770273074,
      "loss": 0.6744,
      "num_input_tokens_seen": 32815904,
      "step": 23860
    },
    {
      "epoch": 0.7638755521413482,
      "grad_norm": 0.006110725458711386,
      "learning_rate": 0.10517386048923086,
      "loss": 0.374,
      "num_input_tokens_seen": 32822800,
      "step": 23865
    },
    {
      "epoch": 0.7640355931118367,
      "grad_norm": 0.007589409593492746,
      "learning_rate": 0.10511765086390841,
      "loss": 0.5051,
      "num_input_tokens_seen": 32829488,
      "step": 23870
    },
    {
      "epoch": 0.764195634082325,
      "grad_norm": 0.00362498895265162,
      "learning_rate": 0.10506144816000816,
      "loss": 0.5436,
      "num_input_tokens_seen": 32836112,
      "step": 23875
    },
    {
      "epoch": 0.7643556750528135,
      "grad_norm": 0.005793282762169838,
      "learning_rate": 0.10500525238619736,
      "loss": 0.3685,
      "num_input_tokens_seen": 32843008,
      "step": 23880
    },
    {
      "epoch": 0.7645157160233019,
      "grad_norm": 0.004484988283365965,
      "learning_rate": 0.10494906355114209,
      "loss": 0.4056,
      "num_input_tokens_seen": 32850128,
      "step": 23885
    },
    {
      "epoch": 0.7646757569937904,
      "grad_norm": 0.006737179588526487,
      "learning_rate": 0.10489288166350737,
      "loss": 0.5849,
      "num_input_tokens_seen": 32857488,
      "step": 23890
    },
    {
      "epoch": 0.7648357979642788,
      "grad_norm": 0.005789239425212145,
      "learning_rate": 0.10483670673195711,
      "loss": 0.503,
      "num_input_tokens_seen": 32864288,
      "step": 23895
    },
    {
      "epoch": 0.7649958389347673,
      "grad_norm": 0.004834441933780909,
      "learning_rate": 0.10478053876515431,
      "loss": 0.4102,
      "num_input_tokens_seen": 32871344,
      "step": 23900
    },
    {
      "epoch": 0.7651558799052558,
      "grad_norm": 0.00537822674959898,
      "learning_rate": 0.10472437777176061,
      "loss": 0.4739,
      "num_input_tokens_seen": 32878256,
      "step": 23905
    },
    {
      "epoch": 0.7653159208757442,
      "grad_norm": 0.0042223697528243065,
      "learning_rate": 0.1046682237604369,
      "loss": 0.7044,
      "num_input_tokens_seen": 32884896,
      "step": 23910
    },
    {
      "epoch": 0.7654759618462327,
      "grad_norm": 0.007123555522412062,
      "learning_rate": 0.1046120767398427,
      "loss": 0.702,
      "num_input_tokens_seen": 32891632,
      "step": 23915
    },
    {
      "epoch": 0.7656360028167211,
      "grad_norm": 0.005737745203077793,
      "learning_rate": 0.10455593671863667,
      "loss": 0.5273,
      "num_input_tokens_seen": 32898496,
      "step": 23920
    },
    {
      "epoch": 0.7657960437872096,
      "grad_norm": 0.0037105761002749205,
      "learning_rate": 0.1044998037054763,
      "loss": 0.462,
      "num_input_tokens_seen": 32905024,
      "step": 23925
    },
    {
      "epoch": 0.7659560847576979,
      "grad_norm": 0.005894902627915144,
      "learning_rate": 0.10444367770901794,
      "loss": 0.6643,
      "num_input_tokens_seen": 32911712,
      "step": 23930
    },
    {
      "epoch": 0.7661161257281864,
      "grad_norm": 0.006314700935035944,
      "learning_rate": 0.10438755873791698,
      "loss": 0.5764,
      "num_input_tokens_seen": 32918368,
      "step": 23935
    },
    {
      "epoch": 0.7662761666986748,
      "grad_norm": 0.004035291727632284,
      "learning_rate": 0.10433144680082775,
      "loss": 0.5104,
      "num_input_tokens_seen": 32925248,
      "step": 23940
    },
    {
      "epoch": 0.7664362076691633,
      "grad_norm": 0.002194770611822605,
      "learning_rate": 0.10427534190640322,
      "loss": 0.5762,
      "num_input_tokens_seen": 32932048,
      "step": 23945
    },
    {
      "epoch": 0.7665962486396517,
      "grad_norm": 0.008787997998297215,
      "learning_rate": 0.10421924406329568,
      "loss": 0.6254,
      "num_input_tokens_seen": 32939040,
      "step": 23950
    },
    {
      "epoch": 0.7667562896101402,
      "grad_norm": 0.005567027721554041,
      "learning_rate": 0.10416315328015598,
      "loss": 0.3577,
      "num_input_tokens_seen": 32945744,
      "step": 23955
    },
    {
      "epoch": 0.7669163305806287,
      "grad_norm": 0.0032542524859309196,
      "learning_rate": 0.10410706956563402,
      "loss": 0.5086,
      "num_input_tokens_seen": 32952560,
      "step": 23960
    },
    {
      "epoch": 0.7670763715511171,
      "grad_norm": 0.00572668993845582,
      "learning_rate": 0.10405099292837874,
      "loss": 0.6317,
      "num_input_tokens_seen": 32959472,
      "step": 23965
    },
    {
      "epoch": 0.7672364125216056,
      "grad_norm": 0.006705798674374819,
      "learning_rate": 0.10399492337703771,
      "loss": 0.5135,
      "num_input_tokens_seen": 32966672,
      "step": 23970
    },
    {
      "epoch": 0.767396453492094,
      "grad_norm": 0.0038609213661402464,
      "learning_rate": 0.10393886092025764,
      "loss": 0.5543,
      "num_input_tokens_seen": 32973536,
      "step": 23975
    },
    {
      "epoch": 0.7675564944625825,
      "grad_norm": 0.0042970068752765656,
      "learning_rate": 0.10388280556668412,
      "loss": 0.5307,
      "num_input_tokens_seen": 32980144,
      "step": 23980
    },
    {
      "epoch": 0.7677165354330708,
      "grad_norm": 0.003610296407714486,
      "learning_rate": 0.10382675732496145,
      "loss": 0.5211,
      "num_input_tokens_seen": 32987264,
      "step": 23985
    },
    {
      "epoch": 0.7678765764035593,
      "grad_norm": 0.0046787685714662075,
      "learning_rate": 0.10377071620373311,
      "loss": 0.594,
      "num_input_tokens_seen": 32994032,
      "step": 23990
    },
    {
      "epoch": 0.7680366173740477,
      "grad_norm": 0.00713600218296051,
      "learning_rate": 0.10371468221164128,
      "loss": 0.6004,
      "num_input_tokens_seen": 33000768,
      "step": 23995
    },
    {
      "epoch": 0.7681966583445362,
      "grad_norm": 0.003444960806518793,
      "learning_rate": 0.10365865535732706,
      "loss": 0.3575,
      "num_input_tokens_seen": 33007504,
      "step": 24000
    },
    {
      "epoch": 0.7681966583445362,
      "eval_loss": 0.5279513001441956,
      "eval_runtime": 332.2712,
      "eval_samples_per_second": 41.791,
      "eval_steps_per_second": 20.896,
      "num_input_tokens_seen": 33007504,
      "step": 24000
    },
    {
      "epoch": 0.7683566993150246,
      "grad_norm": 0.010585414245724678,
      "learning_rate": 0.10360263564943062,
      "loss": 0.4966,
      "num_input_tokens_seen": 33014096,
      "step": 24005
    },
    {
      "epoch": 0.7685167402855131,
      "grad_norm": 0.009349755942821503,
      "learning_rate": 0.10354662309659075,
      "loss": 0.4747,
      "num_input_tokens_seen": 33020880,
      "step": 24010
    },
    {
      "epoch": 0.7686767812560016,
      "grad_norm": 0.00476035475730896,
      "learning_rate": 0.10349061770744537,
      "loss": 0.6529,
      "num_input_tokens_seen": 33027696,
      "step": 24015
    },
    {
      "epoch": 0.76883682222649,
      "grad_norm": 0.003869037376716733,
      "learning_rate": 0.10343461949063128,
      "loss": 0.503,
      "num_input_tokens_seen": 33034768,
      "step": 24020
    },
    {
      "epoch": 0.7689968631969785,
      "grad_norm": 0.006682123988866806,
      "learning_rate": 0.103378628454784,
      "loss": 0.5633,
      "num_input_tokens_seen": 33041808,
      "step": 24025
    },
    {
      "epoch": 0.7691569041674668,
      "grad_norm": 0.0044747935608029366,
      "learning_rate": 0.10332264460853811,
      "loss": 0.6705,
      "num_input_tokens_seen": 33048448,
      "step": 24030
    },
    {
      "epoch": 0.7693169451379553,
      "grad_norm": 0.005842548329383135,
      "learning_rate": 0.10326666796052701,
      "loss": 0.4872,
      "num_input_tokens_seen": 33055536,
      "step": 24035
    },
    {
      "epoch": 0.7694769861084437,
      "grad_norm": 0.003993530757725239,
      "learning_rate": 0.10321069851938296,
      "loss": 0.576,
      "num_input_tokens_seen": 33062032,
      "step": 24040
    },
    {
      "epoch": 0.7696370270789322,
      "grad_norm": 0.0027145103085786104,
      "learning_rate": 0.10315473629373724,
      "loss": 0.4392,
      "num_input_tokens_seen": 33068784,
      "step": 24045
    },
    {
      "epoch": 0.7697970680494206,
      "grad_norm": 0.008786173537373543,
      "learning_rate": 0.10309878129221982,
      "loss": 0.6,
      "num_input_tokens_seen": 33075392,
      "step": 24050
    },
    {
      "epoch": 0.7699571090199091,
      "grad_norm": 0.005088538397103548,
      "learning_rate": 0.10304283352345973,
      "loss": 0.4992,
      "num_input_tokens_seen": 33082112,
      "step": 24055
    },
    {
      "epoch": 0.7701171499903975,
      "grad_norm": 0.007883574813604355,
      "learning_rate": 0.10298689299608486,
      "loss": 0.6706,
      "num_input_tokens_seen": 33089120,
      "step": 24060
    },
    {
      "epoch": 0.770277190960886,
      "grad_norm": 0.006204643752425909,
      "learning_rate": 0.10293095971872188,
      "loss": 0.6229,
      "num_input_tokens_seen": 33095920,
      "step": 24065
    },
    {
      "epoch": 0.7704372319313745,
      "grad_norm": 0.004702119622379541,
      "learning_rate": 0.10287503369999645,
      "loss": 0.5398,
      "num_input_tokens_seen": 33102672,
      "step": 24070
    },
    {
      "epoch": 0.7705972729018629,
      "grad_norm": 0.00753770349547267,
      "learning_rate": 0.10281911494853295,
      "loss": 0.5822,
      "num_input_tokens_seen": 33109328,
      "step": 24075
    },
    {
      "epoch": 0.7707573138723514,
      "grad_norm": 0.0038640133570879698,
      "learning_rate": 0.10276320347295485,
      "loss": 0.5157,
      "num_input_tokens_seen": 33116048,
      "step": 24080
    },
    {
      "epoch": 0.7709173548428397,
      "grad_norm": 0.004109593108296394,
      "learning_rate": 0.10270729928188446,
      "loss": 0.5448,
      "num_input_tokens_seen": 33122736,
      "step": 24085
    },
    {
      "epoch": 0.7710773958133282,
      "grad_norm": 0.006061702035367489,
      "learning_rate": 0.10265140238394276,
      "loss": 0.496,
      "num_input_tokens_seen": 33129776,
      "step": 24090
    },
    {
      "epoch": 0.7712374367838166,
      "grad_norm": 0.004225756041705608,
      "learning_rate": 0.10259551278774988,
      "loss": 0.5359,
      "num_input_tokens_seen": 33136816,
      "step": 24095
    },
    {
      "epoch": 0.7713974777543051,
      "grad_norm": 0.014245700091123581,
      "learning_rate": 0.10253963050192462,
      "loss": 0.5445,
      "num_input_tokens_seen": 33143376,
      "step": 24100
    },
    {
      "epoch": 0.7715575187247935,
      "grad_norm": 0.006356760859489441,
      "learning_rate": 0.10248375553508478,
      "loss": 0.7686,
      "num_input_tokens_seen": 33150256,
      "step": 24105
    },
    {
      "epoch": 0.771717559695282,
      "grad_norm": 0.003746755886822939,
      "learning_rate": 0.102427887895847,
      "loss": 0.4511,
      "num_input_tokens_seen": 33156816,
      "step": 24110
    },
    {
      "epoch": 0.7718776006657705,
      "grad_norm": 0.0032016057521104813,
      "learning_rate": 0.10237202759282668,
      "loss": 0.4769,
      "num_input_tokens_seen": 33163920,
      "step": 24115
    },
    {
      "epoch": 0.7720376416362589,
      "grad_norm": 0.0033589093945920467,
      "learning_rate": 0.10231617463463821,
      "loss": 0.5703,
      "num_input_tokens_seen": 33170928,
      "step": 24120
    },
    {
      "epoch": 0.7721976826067474,
      "grad_norm": 0.005272758658975363,
      "learning_rate": 0.10226032902989492,
      "loss": 0.4515,
      "num_input_tokens_seen": 33177712,
      "step": 24125
    },
    {
      "epoch": 0.7723577235772358,
      "grad_norm": 0.006506805773824453,
      "learning_rate": 0.10220449078720877,
      "loss": 0.4832,
      "num_input_tokens_seen": 33185344,
      "step": 24130
    },
    {
      "epoch": 0.7725177645477243,
      "grad_norm": 0.0042031751945614815,
      "learning_rate": 0.1021486599151908,
      "loss": 0.4919,
      "num_input_tokens_seen": 33192400,
      "step": 24135
    },
    {
      "epoch": 0.7726778055182126,
      "grad_norm": 0.0025429166853427887,
      "learning_rate": 0.10209283642245084,
      "loss": 0.6377,
      "num_input_tokens_seen": 33199392,
      "step": 24140
    },
    {
      "epoch": 0.7728378464887011,
      "grad_norm": 0.007425665855407715,
      "learning_rate": 0.10203702031759748,
      "loss": 0.7117,
      "num_input_tokens_seen": 33206512,
      "step": 24145
    },
    {
      "epoch": 0.7729978874591895,
      "grad_norm": 0.005119276233017445,
      "learning_rate": 0.1019812116092384,
      "loss": 0.4926,
      "num_input_tokens_seen": 33213104,
      "step": 24150
    },
    {
      "epoch": 0.773157928429678,
      "grad_norm": 0.004803550895303488,
      "learning_rate": 0.10192541030597986,
      "loss": 0.53,
      "num_input_tokens_seen": 33220176,
      "step": 24155
    },
    {
      "epoch": 0.7733179694001664,
      "grad_norm": 0.004028937313705683,
      "learning_rate": 0.1018696164164272,
      "loss": 0.5855,
      "num_input_tokens_seen": 33226688,
      "step": 24160
    },
    {
      "epoch": 0.7734780103706549,
      "grad_norm": 0.0047318944707512856,
      "learning_rate": 0.10181382994918459,
      "loss": 0.4445,
      "num_input_tokens_seen": 33233616,
      "step": 24165
    },
    {
      "epoch": 0.7736380513411434,
      "grad_norm": 0.004116765223443508,
      "learning_rate": 0.10175805091285492,
      "loss": 0.565,
      "num_input_tokens_seen": 33240480,
      "step": 24170
    },
    {
      "epoch": 0.7737980923116318,
      "grad_norm": 0.0040125250816345215,
      "learning_rate": 0.10170227931603999,
      "loss": 0.4651,
      "num_input_tokens_seen": 33246992,
      "step": 24175
    },
    {
      "epoch": 0.7739581332821203,
      "grad_norm": 0.00458461744710803,
      "learning_rate": 0.10164651516734062,
      "loss": 0.4583,
      "num_input_tokens_seen": 33253936,
      "step": 24180
    },
    {
      "epoch": 0.7741181742526086,
      "grad_norm": 0.004098680801689625,
      "learning_rate": 0.1015907584753562,
      "loss": 0.487,
      "num_input_tokens_seen": 33261408,
      "step": 24185
    },
    {
      "epoch": 0.7742782152230971,
      "grad_norm": 0.0036958742421120405,
      "learning_rate": 0.10153500924868523,
      "loss": 0.4062,
      "num_input_tokens_seen": 33267856,
      "step": 24190
    },
    {
      "epoch": 0.7744382561935855,
      "grad_norm": 0.0065236869268119335,
      "learning_rate": 0.10147926749592483,
      "loss": 0.4619,
      "num_input_tokens_seen": 33274912,
      "step": 24195
    },
    {
      "epoch": 0.774598297164074,
      "grad_norm": 0.004231298808008432,
      "learning_rate": 0.10142353322567112,
      "loss": 0.6269,
      "num_input_tokens_seen": 33281968,
      "step": 24200
    },
    {
      "epoch": 0.774598297164074,
      "eval_loss": 0.5303714275360107,
      "eval_runtime": 332.3005,
      "eval_samples_per_second": 41.787,
      "eval_steps_per_second": 20.894,
      "num_input_tokens_seen": 33281968,
      "step": 24200
    },
    {
      "epoch": 0.7747583381345624,
      "grad_norm": 0.005357945337891579,
      "learning_rate": 0.1013678064465191,
      "loss": 0.6367,
      "num_input_tokens_seen": 33288800,
      "step": 24205
    },
    {
      "epoch": 0.7749183791050509,
      "grad_norm": 0.0030523224268108606,
      "learning_rate": 0.10131208716706244,
      "loss": 0.6471,
      "num_input_tokens_seen": 33295984,
      "step": 24210
    },
    {
      "epoch": 0.7750784200755393,
      "grad_norm": 0.0029699956066906452,
      "learning_rate": 0.10125637539589379,
      "loss": 0.4104,
      "num_input_tokens_seen": 33302560,
      "step": 24215
    },
    {
      "epoch": 0.7752384610460278,
      "grad_norm": 0.0037353732623159885,
      "learning_rate": 0.10120067114160464,
      "loss": 0.4903,
      "num_input_tokens_seen": 33309312,
      "step": 24220
    },
    {
      "epoch": 0.7753985020165163,
      "grad_norm": 0.004627125337719917,
      "learning_rate": 0.10114497441278517,
      "loss": 0.3936,
      "num_input_tokens_seen": 33316032,
      "step": 24225
    },
    {
      "epoch": 0.7755585429870047,
      "grad_norm": 0.0051997750997543335,
      "learning_rate": 0.10108928521802468,
      "loss": 0.5575,
      "num_input_tokens_seen": 33322752,
      "step": 24230
    },
    {
      "epoch": 0.7757185839574932,
      "grad_norm": 0.006095576100051403,
      "learning_rate": 0.101033603565911,
      "loss": 0.4702,
      "num_input_tokens_seen": 33329808,
      "step": 24235
    },
    {
      "epoch": 0.7758786249279815,
      "grad_norm": 0.0023416357580572367,
      "learning_rate": 0.10097792946503102,
      "loss": 0.7039,
      "num_input_tokens_seen": 33336880,
      "step": 24240
    },
    {
      "epoch": 0.77603866589847,
      "grad_norm": 0.005975750740617514,
      "learning_rate": 0.10092226292397039,
      "loss": 0.3844,
      "num_input_tokens_seen": 33343728,
      "step": 24245
    },
    {
      "epoch": 0.7761987068689584,
      "grad_norm": 0.0023791370913386345,
      "learning_rate": 0.10086660395131354,
      "loss": 0.5413,
      "num_input_tokens_seen": 33350384,
      "step": 24250
    },
    {
      "epoch": 0.7763587478394469,
      "grad_norm": 0.0031477762386202812,
      "learning_rate": 0.10081095255564385,
      "loss": 0.401,
      "num_input_tokens_seen": 33357024,
      "step": 24255
    },
    {
      "epoch": 0.7765187888099353,
      "grad_norm": 0.004100964870303869,
      "learning_rate": 0.10075530874554335,
      "loss": 0.4756,
      "num_input_tokens_seen": 33364496,
      "step": 24260
    },
    {
      "epoch": 0.7766788297804238,
      "grad_norm": 0.007746510673314333,
      "learning_rate": 0.10069967252959311,
      "loss": 0.4236,
      "num_input_tokens_seen": 33371680,
      "step": 24265
    },
    {
      "epoch": 0.7768388707509122,
      "grad_norm": 0.0021456493996083736,
      "learning_rate": 0.10064404391637297,
      "loss": 0.4858,
      "num_input_tokens_seen": 33378848,
      "step": 24270
    },
    {
      "epoch": 0.7769989117214007,
      "grad_norm": 0.004576570354402065,
      "learning_rate": 0.10058842291446145,
      "loss": 0.5165,
      "num_input_tokens_seen": 33385952,
      "step": 24275
    },
    {
      "epoch": 0.7771589526918892,
      "grad_norm": 0.004441666882485151,
      "learning_rate": 0.10053280953243608,
      "loss": 0.5121,
      "num_input_tokens_seen": 33392528,
      "step": 24280
    },
    {
      "epoch": 0.7773189936623776,
      "grad_norm": 0.005023427773267031,
      "learning_rate": 0.10047720377887315,
      "loss": 0.6321,
      "num_input_tokens_seen": 33399488,
      "step": 24285
    },
    {
      "epoch": 0.777479034632866,
      "grad_norm": 0.00423751724883914,
      "learning_rate": 0.10042160566234767,
      "loss": 0.5803,
      "num_input_tokens_seen": 33406208,
      "step": 24290
    },
    {
      "epoch": 0.7776390756033544,
      "grad_norm": 0.002929102862253785,
      "learning_rate": 0.10036601519143372,
      "loss": 0.3644,
      "num_input_tokens_seen": 33413136,
      "step": 24295
    },
    {
      "epoch": 0.7777991165738429,
      "grad_norm": 0.004751512315124273,
      "learning_rate": 0.1003104323747039,
      "loss": 0.6813,
      "num_input_tokens_seen": 33420048,
      "step": 24300
    },
    {
      "epoch": 0.7779591575443313,
      "grad_norm": 0.005446856375783682,
      "learning_rate": 0.10025485722072984,
      "loss": 0.502,
      "num_input_tokens_seen": 33426784,
      "step": 24305
    },
    {
      "epoch": 0.7781191985148198,
      "grad_norm": 0.004194737412035465,
      "learning_rate": 0.10019928973808201,
      "loss": 0.4323,
      "num_input_tokens_seen": 33433680,
      "step": 24310
    },
    {
      "epoch": 0.7782792394853082,
      "grad_norm": 0.008740151301026344,
      "learning_rate": 0.10014372993532945,
      "loss": 0.6044,
      "num_input_tokens_seen": 33441088,
      "step": 24315
    },
    {
      "epoch": 0.7784392804557967,
      "grad_norm": 0.006073057185858488,
      "learning_rate": 0.1000881778210403,
      "loss": 0.4711,
      "num_input_tokens_seen": 33447840,
      "step": 24320
    },
    {
      "epoch": 0.7785993214262851,
      "grad_norm": 0.007394514512270689,
      "learning_rate": 0.10003263340378142,
      "loss": 0.4745,
      "num_input_tokens_seen": 33454992,
      "step": 24325
    },
    {
      "epoch": 0.7787593623967736,
      "grad_norm": 0.004423193167895079,
      "learning_rate": 0.09997709669211834,
      "loss": 0.4236,
      "num_input_tokens_seen": 33461840,
      "step": 24330
    },
    {
      "epoch": 0.7789194033672621,
      "grad_norm": 0.004096449818462133,
      "learning_rate": 0.0999215676946156,
      "loss": 0.3678,
      "num_input_tokens_seen": 33468416,
      "step": 24335
    },
    {
      "epoch": 0.7790794443377504,
      "grad_norm": 0.004294292069971561,
      "learning_rate": 0.0998660464198364,
      "loss": 0.5227,
      "num_input_tokens_seen": 33475856,
      "step": 24340
    },
    {
      "epoch": 0.7792394853082389,
      "grad_norm": 0.004010593984276056,
      "learning_rate": 0.09981053287634288,
      "loss": 0.4257,
      "num_input_tokens_seen": 33482624,
      "step": 24345
    },
    {
      "epoch": 0.7793995262787273,
      "grad_norm": 0.0036739942152053118,
      "learning_rate": 0.09975502707269596,
      "loss": 0.6398,
      "num_input_tokens_seen": 33489376,
      "step": 24350
    },
    {
      "epoch": 0.7795595672492158,
      "grad_norm": 0.0036040223203599453,
      "learning_rate": 0.09969952901745524,
      "loss": 0.5377,
      "num_input_tokens_seen": 33496112,
      "step": 24355
    },
    {
      "epoch": 0.7797196082197042,
      "grad_norm": 0.004156568553298712,
      "learning_rate": 0.09964403871917925,
      "loss": 0.5935,
      "num_input_tokens_seen": 33502544,
      "step": 24360
    },
    {
      "epoch": 0.7798796491901927,
      "grad_norm": 0.005025472957640886,
      "learning_rate": 0.09958855618642536,
      "loss": 0.5688,
      "num_input_tokens_seen": 33510032,
      "step": 24365
    },
    {
      "epoch": 0.7800396901606811,
      "grad_norm": 0.0048709348775446415,
      "learning_rate": 0.09953308142774955,
      "loss": 0.4241,
      "num_input_tokens_seen": 33516768,
      "step": 24370
    },
    {
      "epoch": 0.7801997311311696,
      "grad_norm": 0.006323124747723341,
      "learning_rate": 0.09947761445170686,
      "loss": 0.6976,
      "num_input_tokens_seen": 33523520,
      "step": 24375
    },
    {
      "epoch": 0.7803597721016581,
      "grad_norm": 0.007068118546158075,
      "learning_rate": 0.09942215526685086,
      "loss": 0.5428,
      "num_input_tokens_seen": 33530896,
      "step": 24380
    },
    {
      "epoch": 0.7805198130721465,
      "grad_norm": 0.003533395705744624,
      "learning_rate": 0.09936670388173414,
      "loss": 0.528,
      "num_input_tokens_seen": 33538032,
      "step": 24385
    },
    {
      "epoch": 0.780679854042635,
      "grad_norm": 0.005301983095705509,
      "learning_rate": 0.09931126030490799,
      "loss": 0.4539,
      "num_input_tokens_seen": 33544768,
      "step": 24390
    },
    {
      "epoch": 0.7808398950131233,
      "grad_norm": 0.0034074024297297,
      "learning_rate": 0.0992558245449225,
      "loss": 0.5394,
      "num_input_tokens_seen": 33551760,
      "step": 24395
    },
    {
      "epoch": 0.7809999359836118,
      "grad_norm": 0.0036688712425529957,
      "learning_rate": 0.09920039661032651,
      "loss": 0.6274,
      "num_input_tokens_seen": 33558736,
      "step": 24400
    },
    {
      "epoch": 0.7809999359836118,
      "eval_loss": 0.5319734811782837,
      "eval_runtime": 332.0532,
      "eval_samples_per_second": 41.819,
      "eval_steps_per_second": 20.909,
      "num_input_tokens_seen": 33558736,
      "step": 24400
    },
    {
      "epoch": 0.7811599769541002,
      "grad_norm": 0.0016162741230800748,
      "learning_rate": 0.09914497650966782,
      "loss": 0.4724,
      "num_input_tokens_seen": 33564992,
      "step": 24405
    },
    {
      "epoch": 0.7813200179245887,
      "grad_norm": 0.003442785469815135,
      "learning_rate": 0.09908956425149276,
      "loss": 0.4513,
      "num_input_tokens_seen": 33571600,
      "step": 24410
    },
    {
      "epoch": 0.7814800588950771,
      "grad_norm": 0.0053422097116708755,
      "learning_rate": 0.09903415984434677,
      "loss": 0.3722,
      "num_input_tokens_seen": 33578384,
      "step": 24415
    },
    {
      "epoch": 0.7816400998655656,
      "grad_norm": 0.00488536199554801,
      "learning_rate": 0.09897876329677373,
      "loss": 0.5924,
      "num_input_tokens_seen": 33585152,
      "step": 24420
    },
    {
      "epoch": 0.781800140836054,
      "grad_norm": 0.0059975930489599705,
      "learning_rate": 0.09892337461731658,
      "loss": 0.5982,
      "num_input_tokens_seen": 33592144,
      "step": 24425
    },
    {
      "epoch": 0.7819601818065425,
      "grad_norm": 0.0026947767473757267,
      "learning_rate": 0.09886799381451693,
      "loss": 0.4731,
      "num_input_tokens_seen": 33599056,
      "step": 24430
    },
    {
      "epoch": 0.782120222777031,
      "grad_norm": 0.002410922199487686,
      "learning_rate": 0.09881262089691521,
      "loss": 0.3819,
      "num_input_tokens_seen": 33606256,
      "step": 24435
    },
    {
      "epoch": 0.7822802637475194,
      "grad_norm": 0.002915076445788145,
      "learning_rate": 0.09875725587305059,
      "loss": 0.4801,
      "num_input_tokens_seen": 33613040,
      "step": 24440
    },
    {
      "epoch": 0.7824403047180079,
      "grad_norm": 0.0034339018166065216,
      "learning_rate": 0.09870189875146111,
      "loss": 0.4982,
      "num_input_tokens_seen": 33619552,
      "step": 24445
    },
    {
      "epoch": 0.7826003456884962,
      "grad_norm": 0.004740595351904631,
      "learning_rate": 0.09864654954068346,
      "loss": 0.4699,
      "num_input_tokens_seen": 33626352,
      "step": 24450
    },
    {
      "epoch": 0.7827603866589847,
      "grad_norm": 0.0031161492224782705,
      "learning_rate": 0.09859120824925326,
      "loss": 0.4787,
      "num_input_tokens_seen": 33633344,
      "step": 24455
    },
    {
      "epoch": 0.7829204276294731,
      "grad_norm": 0.004556220956146717,
      "learning_rate": 0.09853587488570474,
      "loss": 0.598,
      "num_input_tokens_seen": 33640288,
      "step": 24460
    },
    {
      "epoch": 0.7830804685999616,
      "grad_norm": 0.005688574630767107,
      "learning_rate": 0.09848054945857107,
      "loss": 0.4307,
      "num_input_tokens_seen": 33647056,
      "step": 24465
    },
    {
      "epoch": 0.78324050957045,
      "grad_norm": 0.002730408450588584,
      "learning_rate": 0.09842523197638416,
      "loss": 0.4585,
      "num_input_tokens_seen": 33653920,
      "step": 24470
    },
    {
      "epoch": 0.7834005505409385,
      "grad_norm": 0.005120495799928904,
      "learning_rate": 0.09836992244767452,
      "loss": 0.454,
      "num_input_tokens_seen": 33660832,
      "step": 24475
    },
    {
      "epoch": 0.7835605915114269,
      "grad_norm": 0.003962913062423468,
      "learning_rate": 0.09831462088097168,
      "loss": 0.3124,
      "num_input_tokens_seen": 33667776,
      "step": 24480
    },
    {
      "epoch": 0.7837206324819154,
      "grad_norm": 0.004201010335236788,
      "learning_rate": 0.09825932728480385,
      "loss": 0.6533,
      "num_input_tokens_seen": 33675280,
      "step": 24485
    },
    {
      "epoch": 0.7838806734524039,
      "grad_norm": 0.005655007436871529,
      "learning_rate": 0.09820404166769794,
      "loss": 0.5527,
      "num_input_tokens_seen": 33682416,
      "step": 24490
    },
    {
      "epoch": 0.7840407144228922,
      "grad_norm": 0.0037577205803245306,
      "learning_rate": 0.09814876403817978,
      "loss": 0.4277,
      "num_input_tokens_seen": 33689392,
      "step": 24495
    },
    {
      "epoch": 0.7842007553933807,
      "grad_norm": 0.0044019403867423534,
      "learning_rate": 0.09809349440477376,
      "loss": 0.4415,
      "num_input_tokens_seen": 33696144,
      "step": 24500
    },
    {
      "epoch": 0.7843607963638691,
      "grad_norm": 0.004468062426894903,
      "learning_rate": 0.09803823277600317,
      "loss": 0.4576,
      "num_input_tokens_seen": 33702624,
      "step": 24505
    },
    {
      "epoch": 0.7845208373343576,
      "grad_norm": 0.00328546529635787,
      "learning_rate": 0.09798297916039014,
      "loss": 0.4884,
      "num_input_tokens_seen": 33709280,
      "step": 24510
    },
    {
      "epoch": 0.784680878304846,
      "grad_norm": 0.00536393141373992,
      "learning_rate": 0.09792773356645534,
      "loss": 0.3887,
      "num_input_tokens_seen": 33716144,
      "step": 24515
    },
    {
      "epoch": 0.7848409192753345,
      "grad_norm": 0.005482586566358805,
      "learning_rate": 0.09787249600271843,
      "loss": 0.4034,
      "num_input_tokens_seen": 33722944,
      "step": 24520
    },
    {
      "epoch": 0.7850009602458229,
      "grad_norm": 0.0033993322867900133,
      "learning_rate": 0.09781726647769776,
      "loss": 0.4645,
      "num_input_tokens_seen": 33729760,
      "step": 24525
    },
    {
      "epoch": 0.7851610012163114,
      "grad_norm": 0.001965167000889778,
      "learning_rate": 0.0977620449999103,
      "loss": 0.4615,
      "num_input_tokens_seen": 33737008,
      "step": 24530
    },
    {
      "epoch": 0.7853210421867998,
      "grad_norm": 0.005090078804641962,
      "learning_rate": 0.09770683157787204,
      "loss": 0.507,
      "num_input_tokens_seen": 33743968,
      "step": 24535
    },
    {
      "epoch": 0.7854810831572883,
      "grad_norm": 0.0027149326633661985,
      "learning_rate": 0.09765162622009745,
      "loss": 0.5733,
      "num_input_tokens_seen": 33750864,
      "step": 24540
    },
    {
      "epoch": 0.7856411241277768,
      "grad_norm": 0.004767951089888811,
      "learning_rate": 0.09759642893509995,
      "loss": 0.7629,
      "num_input_tokens_seen": 33757264,
      "step": 24545
    },
    {
      "epoch": 0.7858011650982651,
      "grad_norm": 0.0033452578354626894,
      "learning_rate": 0.09754123973139169,
      "loss": 0.5178,
      "num_input_tokens_seen": 33763904,
      "step": 24550
    },
    {
      "epoch": 0.7859612060687536,
      "grad_norm": 0.0038583555724471807,
      "learning_rate": 0.09748605861748345,
      "loss": 0.4577,
      "num_input_tokens_seen": 33770464,
      "step": 24555
    },
    {
      "epoch": 0.786121247039242,
      "grad_norm": 0.004098902456462383,
      "learning_rate": 0.0974308856018849,
      "loss": 0.6166,
      "num_input_tokens_seen": 33777168,
      "step": 24560
    },
    {
      "epoch": 0.7862812880097305,
      "grad_norm": 0.004064918961375952,
      "learning_rate": 0.09737572069310449,
      "loss": 0.5286,
      "num_input_tokens_seen": 33783808,
      "step": 24565
    },
    {
      "epoch": 0.7864413289802189,
      "grad_norm": 0.004532911814749241,
      "learning_rate": 0.09732056389964922,
      "loss": 0.5237,
      "num_input_tokens_seen": 33790272,
      "step": 24570
    },
    {
      "epoch": 0.7866013699507074,
      "grad_norm": 0.004518440458923578,
      "learning_rate": 0.097265415230025,
      "loss": 0.5906,
      "num_input_tokens_seen": 33797632,
      "step": 24575
    },
    {
      "epoch": 0.7867614109211958,
      "grad_norm": 0.0021410908084362745,
      "learning_rate": 0.09721027469273648,
      "loss": 0.4637,
      "num_input_tokens_seen": 33804176,
      "step": 24580
    },
    {
      "epoch": 0.7869214518916843,
      "grad_norm": 0.005884411744773388,
      "learning_rate": 0.09715514229628695,
      "loss": 0.482,
      "num_input_tokens_seen": 33810576,
      "step": 24585
    },
    {
      "epoch": 0.7870814928621728,
      "grad_norm": 0.006846964359283447,
      "learning_rate": 0.09710001804917864,
      "loss": 0.6776,
      "num_input_tokens_seen": 33817280,
      "step": 24590
    },
    {
      "epoch": 0.7872415338326612,
      "grad_norm": 0.005897311959415674,
      "learning_rate": 0.09704490195991226,
      "loss": 0.5838,
      "num_input_tokens_seen": 33824144,
      "step": 24595
    },
    {
      "epoch": 0.7874015748031497,
      "grad_norm": 0.0028760165441781282,
      "learning_rate": 0.09698979403698753,
      "loss": 0.5539,
      "num_input_tokens_seen": 33830832,
      "step": 24600
    },
    {
      "epoch": 0.7874015748031497,
      "eval_loss": 0.5232862234115601,
      "eval_runtime": 332.0849,
      "eval_samples_per_second": 41.815,
      "eval_steps_per_second": 20.907,
      "num_input_tokens_seen": 33830832,
      "step": 24600
    },
    {
      "epoch": 0.787561615773638,
      "grad_norm": 0.007173052988946438,
      "learning_rate": 0.0969346942889027,
      "loss": 0.3394,
      "num_input_tokens_seen": 33837600,
      "step": 24605
    },
    {
      "epoch": 0.7877216567441265,
      "grad_norm": 0.008842142298817635,
      "learning_rate": 0.09687960272415487,
      "loss": 0.5023,
      "num_input_tokens_seen": 33844992,
      "step": 24610
    },
    {
      "epoch": 0.7878816977146149,
      "grad_norm": 0.00508269015699625,
      "learning_rate": 0.0968245193512399,
      "loss": 0.3836,
      "num_input_tokens_seen": 33852160,
      "step": 24615
    },
    {
      "epoch": 0.7880417386851034,
      "grad_norm": 0.002970246598124504,
      "learning_rate": 0.09676944417865221,
      "loss": 0.4967,
      "num_input_tokens_seen": 33858688,
      "step": 24620
    },
    {
      "epoch": 0.7882017796555918,
      "grad_norm": 0.004523683339357376,
      "learning_rate": 0.09671437721488517,
      "loss": 0.6373,
      "num_input_tokens_seen": 33865792,
      "step": 24625
    },
    {
      "epoch": 0.7883618206260803,
      "grad_norm": 0.006456195842474699,
      "learning_rate": 0.09665931846843086,
      "loss": 0.3692,
      "num_input_tokens_seen": 33872960,
      "step": 24630
    },
    {
      "epoch": 0.7885218615965687,
      "grad_norm": 0.003729834221303463,
      "learning_rate": 0.0966042679477799,
      "loss": 0.5615,
      "num_input_tokens_seen": 33879808,
      "step": 24635
    },
    {
      "epoch": 0.7886819025670572,
      "grad_norm": 0.0009214961319230497,
      "learning_rate": 0.09654922566142186,
      "loss": 0.6093,
      "num_input_tokens_seen": 33886560,
      "step": 24640
    },
    {
      "epoch": 0.7888419435375457,
      "grad_norm": 0.004854454193264246,
      "learning_rate": 0.09649419161784498,
      "loss": 0.5726,
      "num_input_tokens_seen": 33893232,
      "step": 24645
    },
    {
      "epoch": 0.789001984508034,
      "grad_norm": 0.005271600093692541,
      "learning_rate": 0.09643916582553606,
      "loss": 0.6198,
      "num_input_tokens_seen": 33900304,
      "step": 24650
    },
    {
      "epoch": 0.7891620254785225,
      "grad_norm": 0.004691849462687969,
      "learning_rate": 0.09638414829298093,
      "loss": 0.4448,
      "num_input_tokens_seen": 33907296,
      "step": 24655
    },
    {
      "epoch": 0.7893220664490109,
      "grad_norm": 0.004449313972145319,
      "learning_rate": 0.09632913902866386,
      "loss": 0.5688,
      "num_input_tokens_seen": 33914032,
      "step": 24660
    },
    {
      "epoch": 0.7894821074194994,
      "grad_norm": 0.0055535477586090565,
      "learning_rate": 0.096274138041068,
      "loss": 0.3401,
      "num_input_tokens_seen": 33920608,
      "step": 24665
    },
    {
      "epoch": 0.7896421483899878,
      "grad_norm": 0.008040804415941238,
      "learning_rate": 0.09621914533867527,
      "loss": 0.6457,
      "num_input_tokens_seen": 33927248,
      "step": 24670
    },
    {
      "epoch": 0.7898021893604763,
      "grad_norm": 0.004228982143104076,
      "learning_rate": 0.09616416092996616,
      "loss": 0.4913,
      "num_input_tokens_seen": 33934016,
      "step": 24675
    },
    {
      "epoch": 0.7899622303309647,
      "grad_norm": 0.0029847787227481604,
      "learning_rate": 0.09610918482342,
      "loss": 0.4277,
      "num_input_tokens_seen": 33940720,
      "step": 24680
    },
    {
      "epoch": 0.7901222713014532,
      "grad_norm": 0.005576231516897678,
      "learning_rate": 0.09605421702751478,
      "loss": 0.5263,
      "num_input_tokens_seen": 33947424,
      "step": 24685
    },
    {
      "epoch": 0.7902823122719416,
      "grad_norm": 0.003972083330154419,
      "learning_rate": 0.09599925755072718,
      "loss": 0.4437,
      "num_input_tokens_seen": 33954128,
      "step": 24690
    },
    {
      "epoch": 0.7904423532424301,
      "grad_norm": 0.004992392845451832,
      "learning_rate": 0.09594430640153273,
      "loss": 0.627,
      "num_input_tokens_seen": 33961376,
      "step": 24695
    },
    {
      "epoch": 0.7906023942129186,
      "grad_norm": 0.009724493138492107,
      "learning_rate": 0.09588936358840547,
      "loss": 0.4618,
      "num_input_tokens_seen": 33968576,
      "step": 24700
    },
    {
      "epoch": 0.7907624351834069,
      "grad_norm": 0.0029119839891791344,
      "learning_rate": 0.09583442911981836,
      "loss": 0.5532,
      "num_input_tokens_seen": 33975392,
      "step": 24705
    },
    {
      "epoch": 0.7909224761538954,
      "grad_norm": 0.007365334779024124,
      "learning_rate": 0.09577950300424302,
      "loss": 0.5822,
      "num_input_tokens_seen": 33982112,
      "step": 24710
    },
    {
      "epoch": 0.7910825171243838,
      "grad_norm": 0.0048444801941514015,
      "learning_rate": 0.09572458525014967,
      "loss": 0.467,
      "num_input_tokens_seen": 33989040,
      "step": 24715
    },
    {
      "epoch": 0.7912425580948723,
      "grad_norm": 0.002800499554723501,
      "learning_rate": 0.0956696758660073,
      "loss": 0.4893,
      "num_input_tokens_seen": 33995728,
      "step": 24720
    },
    {
      "epoch": 0.7914025990653607,
      "grad_norm": 0.0046144709922373295,
      "learning_rate": 0.09561477486028373,
      "loss": 0.4134,
      "num_input_tokens_seen": 34002736,
      "step": 24725
    },
    {
      "epoch": 0.7915626400358492,
      "grad_norm": 0.0028720907866954803,
      "learning_rate": 0.09555988224144528,
      "loss": 0.4785,
      "num_input_tokens_seen": 34009104,
      "step": 24730
    },
    {
      "epoch": 0.7917226810063376,
      "grad_norm": 0.0037569517735391855,
      "learning_rate": 0.09550499801795717,
      "loss": 0.5704,
      "num_input_tokens_seen": 34016192,
      "step": 24735
    },
    {
      "epoch": 0.7918827219768261,
      "grad_norm": 0.00388193316757679,
      "learning_rate": 0.09545012219828314,
      "loss": 0.4089,
      "num_input_tokens_seen": 34023152,
      "step": 24740
    },
    {
      "epoch": 0.7920427629473145,
      "grad_norm": 0.0027277113404124975,
      "learning_rate": 0.09539525479088577,
      "loss": 0.4824,
      "num_input_tokens_seen": 34030016,
      "step": 24745
    },
    {
      "epoch": 0.792202803917803,
      "grad_norm": 0.00304567557759583,
      "learning_rate": 0.0953403958042264,
      "loss": 0.4151,
      "num_input_tokens_seen": 34036976,
      "step": 24750
    },
    {
      "epoch": 0.7923628448882915,
      "grad_norm": 0.006323827896267176,
      "learning_rate": 0.09528554524676484,
      "loss": 0.4706,
      "num_input_tokens_seen": 34044032,
      "step": 24755
    },
    {
      "epoch": 0.7925228858587798,
      "grad_norm": 0.0063065108843147755,
      "learning_rate": 0.09523070312695978,
      "loss": 0.5916,
      "num_input_tokens_seen": 34050720,
      "step": 24760
    },
    {
      "epoch": 0.7926829268292683,
      "grad_norm": 0.0032736987341195345,
      "learning_rate": 0.09517586945326863,
      "loss": 0.4311,
      "num_input_tokens_seen": 34057648,
      "step": 24765
    },
    {
      "epoch": 0.7928429677997567,
      "grad_norm": 0.0031954257283359766,
      "learning_rate": 0.0951210442341473,
      "loss": 0.4288,
      "num_input_tokens_seen": 34064704,
      "step": 24770
    },
    {
      "epoch": 0.7930030087702452,
      "grad_norm": 0.004051501397043467,
      "learning_rate": 0.09506622747805066,
      "loss": 0.6257,
      "num_input_tokens_seen": 34071200,
      "step": 24775
    },
    {
      "epoch": 0.7931630497407336,
      "grad_norm": 0.004030516371130943,
      "learning_rate": 0.09501141919343203,
      "loss": 0.4529,
      "num_input_tokens_seen": 34077792,
      "step": 24780
    },
    {
      "epoch": 0.7933230907112221,
      "grad_norm": 0.0027228647377341986,
      "learning_rate": 0.09495661938874361,
      "loss": 0.5015,
      "num_input_tokens_seen": 34084432,
      "step": 24785
    },
    {
      "epoch": 0.7934831316817105,
      "grad_norm": 0.003983692731708288,
      "learning_rate": 0.0949018280724362,
      "loss": 0.5264,
      "num_input_tokens_seen": 34091232,
      "step": 24790
    },
    {
      "epoch": 0.793643172652199,
      "grad_norm": 0.004626086447387934,
      "learning_rate": 0.09484704525295934,
      "loss": 0.6355,
      "num_input_tokens_seen": 34098256,
      "step": 24795
    },
    {
      "epoch": 0.7938032136226874,
      "grad_norm": 0.004020712338387966,
      "learning_rate": 0.09479227093876112,
      "loss": 0.5876,
      "num_input_tokens_seen": 34104944,
      "step": 24800
    },
    {
      "epoch": 0.7938032136226874,
      "eval_loss": 0.5242758989334106,
      "eval_runtime": 332.4385,
      "eval_samples_per_second": 41.77,
      "eval_steps_per_second": 20.885,
      "num_input_tokens_seen": 34104944,
      "step": 24800
    },
    {
      "epoch": 0.7939632545931758,
      "grad_norm": 0.00348527729511261,
      "learning_rate": 0.0947375051382886,
      "loss": 0.4709,
      "num_input_tokens_seen": 34111952,
      "step": 24805
    },
    {
      "epoch": 0.7941232955636643,
      "grad_norm": 0.005846159532666206,
      "learning_rate": 0.09468274785998718,
      "loss": 0.4601,
      "num_input_tokens_seen": 34118464,
      "step": 24810
    },
    {
      "epoch": 0.7942833365341527,
      "grad_norm": 0.004809462931007147,
      "learning_rate": 0.09462799911230127,
      "loss": 0.5831,
      "num_input_tokens_seen": 34125136,
      "step": 24815
    },
    {
      "epoch": 0.7944433775046412,
      "grad_norm": 0.002605619840323925,
      "learning_rate": 0.0945732589036737,
      "loss": 0.7396,
      "num_input_tokens_seen": 34132256,
      "step": 24820
    },
    {
      "epoch": 0.7946034184751296,
      "grad_norm": 0.004239069763571024,
      "learning_rate": 0.09451852724254614,
      "loss": 0.6192,
      "num_input_tokens_seen": 34139040,
      "step": 24825
    },
    {
      "epoch": 0.7947634594456181,
      "grad_norm": 0.008089074864983559,
      "learning_rate": 0.09446380413735894,
      "loss": 0.6533,
      "num_input_tokens_seen": 34146336,
      "step": 24830
    },
    {
      "epoch": 0.7949235004161065,
      "grad_norm": 0.011484620161354542,
      "learning_rate": 0.09440908959655099,
      "loss": 0.63,
      "num_input_tokens_seen": 34153200,
      "step": 24835
    },
    {
      "epoch": 0.795083541386595,
      "grad_norm": 0.006741833407431841,
      "learning_rate": 0.09435438362856004,
      "loss": 0.5303,
      "num_input_tokens_seen": 34160096,
      "step": 24840
    },
    {
      "epoch": 0.7952435823570834,
      "grad_norm": 0.004505069460719824,
      "learning_rate": 0.0942996862418225,
      "loss": 0.66,
      "num_input_tokens_seen": 34167120,
      "step": 24845
    },
    {
      "epoch": 0.7954036233275719,
      "grad_norm": 0.0038571490440517664,
      "learning_rate": 0.09424499744477322,
      "loss": 0.5101,
      "num_input_tokens_seen": 34174288,
      "step": 24850
    },
    {
      "epoch": 0.7955636642980604,
      "grad_norm": 0.003709087846800685,
      "learning_rate": 0.09419031724584608,
      "loss": 0.558,
      "num_input_tokens_seen": 34180816,
      "step": 24855
    },
    {
      "epoch": 0.7957237052685487,
      "grad_norm": 0.00614139111712575,
      "learning_rate": 0.09413564565347331,
      "loss": 0.4315,
      "num_input_tokens_seen": 34188048,
      "step": 24860
    },
    {
      "epoch": 0.7958837462390372,
      "grad_norm": 0.005838334094733,
      "learning_rate": 0.094080982676086,
      "loss": 0.4726,
      "num_input_tokens_seen": 34195168,
      "step": 24865
    },
    {
      "epoch": 0.7960437872095256,
      "grad_norm": 0.007329701911658049,
      "learning_rate": 0.09402632832211395,
      "loss": 0.5359,
      "num_input_tokens_seen": 34202336,
      "step": 24870
    },
    {
      "epoch": 0.7962038281800141,
      "grad_norm": 0.002745306584984064,
      "learning_rate": 0.09397168259998541,
      "loss": 0.4229,
      "num_input_tokens_seen": 34209024,
      "step": 24875
    },
    {
      "epoch": 0.7963638691505025,
      "grad_norm": 0.004992759320884943,
      "learning_rate": 0.09391704551812759,
      "loss": 0.5142,
      "num_input_tokens_seen": 34215760,
      "step": 24880
    },
    {
      "epoch": 0.796523910120991,
      "grad_norm": 0.007451056968420744,
      "learning_rate": 0.09386241708496605,
      "loss": 0.5472,
      "num_input_tokens_seen": 34222608,
      "step": 24885
    },
    {
      "epoch": 0.7966839510914794,
      "grad_norm": 0.004596589133143425,
      "learning_rate": 0.09380779730892527,
      "loss": 0.4998,
      "num_input_tokens_seen": 34229712,
      "step": 24890
    },
    {
      "epoch": 0.7968439920619679,
      "grad_norm": 0.0029690461233258247,
      "learning_rate": 0.09375318619842836,
      "loss": 0.4468,
      "num_input_tokens_seen": 34236512,
      "step": 24895
    },
    {
      "epoch": 0.7970040330324563,
      "grad_norm": 0.004832712467759848,
      "learning_rate": 0.09369858376189696,
      "loss": 0.5187,
      "num_input_tokens_seen": 34243120,
      "step": 24900
    },
    {
      "epoch": 0.7971640740029448,
      "grad_norm": 0.005845354404300451,
      "learning_rate": 0.09364399000775143,
      "loss": 0.544,
      "num_input_tokens_seen": 34250208,
      "step": 24905
    },
    {
      "epoch": 0.7973241149734333,
      "grad_norm": 0.0048049865290522575,
      "learning_rate": 0.09358940494441093,
      "loss": 0.4227,
      "num_input_tokens_seen": 34256800,
      "step": 24910
    },
    {
      "epoch": 0.7974841559439216,
      "grad_norm": 0.0063987188041210175,
      "learning_rate": 0.09353482858029301,
      "loss": 0.6596,
      "num_input_tokens_seen": 34263664,
      "step": 24915
    },
    {
      "epoch": 0.7976441969144101,
      "grad_norm": 0.005537652876228094,
      "learning_rate": 0.09348026092381419,
      "loss": 0.4674,
      "num_input_tokens_seen": 34270592,
      "step": 24920
    },
    {
      "epoch": 0.7978042378848985,
      "grad_norm": 0.0031367738265544176,
      "learning_rate": 0.09342570198338931,
      "loss": 0.5466,
      "num_input_tokens_seen": 34277504,
      "step": 24925
    },
    {
      "epoch": 0.797964278855387,
      "grad_norm": 0.0029270597733557224,
      "learning_rate": 0.0933711517674322,
      "loss": 0.4436,
      "num_input_tokens_seen": 34284720,
      "step": 24930
    },
    {
      "epoch": 0.7981243198258754,
      "grad_norm": 0.0041513401083648205,
      "learning_rate": 0.09331661028435513,
      "loss": 0.6199,
      "num_input_tokens_seen": 34291792,
      "step": 24935
    },
    {
      "epoch": 0.7982843607963639,
      "grad_norm": 0.005020825657993555,
      "learning_rate": 0.09326207754256909,
      "loss": 0.706,
      "num_input_tokens_seen": 34299632,
      "step": 24940
    },
    {
      "epoch": 0.7984444017668523,
      "grad_norm": 0.0034602079540491104,
      "learning_rate": 0.09320755355048366,
      "loss": 0.405,
      "num_input_tokens_seen": 34306336,
      "step": 24945
    },
    {
      "epoch": 0.7986044427373408,
      "grad_norm": 0.0030439388938248158,
      "learning_rate": 0.09315303831650722,
      "loss": 0.4328,
      "num_input_tokens_seen": 34313120,
      "step": 24950
    },
    {
      "epoch": 0.7987644837078292,
      "grad_norm": 0.004573076963424683,
      "learning_rate": 0.09309853184904661,
      "loss": 0.5035,
      "num_input_tokens_seen": 34319840,
      "step": 24955
    },
    {
      "epoch": 0.7989245246783176,
      "grad_norm": 0.0034295564983040094,
      "learning_rate": 0.09304403415650753,
      "loss": 0.5134,
      "num_input_tokens_seen": 34326544,
      "step": 24960
    },
    {
      "epoch": 0.7990845656488061,
      "grad_norm": 0.0023948997259140015,
      "learning_rate": 0.09298954524729405,
      "loss": 0.3009,
      "num_input_tokens_seen": 34333616,
      "step": 24965
    },
    {
      "epoch": 0.7992446066192945,
      "grad_norm": 0.0018274864414706826,
      "learning_rate": 0.09293506512980916,
      "loss": 0.4446,
      "num_input_tokens_seen": 34340496,
      "step": 24970
    },
    {
      "epoch": 0.799404647589783,
      "grad_norm": 0.003443835536018014,
      "learning_rate": 0.0928805938124544,
      "loss": 0.5348,
      "num_input_tokens_seen": 34347424,
      "step": 24975
    },
    {
      "epoch": 0.7995646885602714,
      "grad_norm": 0.004618295002728701,
      "learning_rate": 0.09282613130362982,
      "loss": 0.4001,
      "num_input_tokens_seen": 34354464,
      "step": 24980
    },
    {
      "epoch": 0.7997247295307599,
      "grad_norm": 0.002548895077779889,
      "learning_rate": 0.09277167761173427,
      "loss": 0.3952,
      "num_input_tokens_seen": 34361152,
      "step": 24985
    },
    {
      "epoch": 0.7998847705012483,
      "grad_norm": 0.006351415999233723,
      "learning_rate": 0.0927172327451653,
      "loss": 0.7176,
      "num_input_tokens_seen": 34368096,
      "step": 24990
    },
    {
      "epoch": 0.8000448114717368,
      "grad_norm": 0.006313906982541084,
      "learning_rate": 0.09266279671231882,
      "loss": 0.4853,
      "num_input_tokens_seen": 34374880,
      "step": 24995
    },
    {
      "epoch": 0.8002048524422252,
      "grad_norm": 0.005154727026820183,
      "learning_rate": 0.09260836952158967,
      "loss": 0.5314,
      "num_input_tokens_seen": 34381536,
      "step": 25000
    },
    {
      "epoch": 0.8002048524422252,
      "eval_loss": 0.521657407283783,
      "eval_runtime": 332.2814,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 34381536,
      "step": 25000
    },
    {
      "epoch": 0.8003648934127137,
      "grad_norm": 0.0058656660839915276,
      "learning_rate": 0.09255395118137114,
      "loss": 0.6183,
      "num_input_tokens_seen": 34388816,
      "step": 25005
    },
    {
      "epoch": 0.800524934383202,
      "grad_norm": 0.005100387614220381,
      "learning_rate": 0.09249954170005527,
      "loss": 0.3935,
      "num_input_tokens_seen": 34395680,
      "step": 25010
    },
    {
      "epoch": 0.8006849753536905,
      "grad_norm": 0.002987101674079895,
      "learning_rate": 0.0924451410860327,
      "loss": 0.5887,
      "num_input_tokens_seen": 34402512,
      "step": 25015
    },
    {
      "epoch": 0.800845016324179,
      "grad_norm": 0.0037725011352449656,
      "learning_rate": 0.09239074934769258,
      "loss": 0.4186,
      "num_input_tokens_seen": 34409424,
      "step": 25020
    },
    {
      "epoch": 0.8010050572946674,
      "grad_norm": 0.004355670418590307,
      "learning_rate": 0.09233636649342288,
      "loss": 0.6435,
      "num_input_tokens_seen": 34416192,
      "step": 25025
    },
    {
      "epoch": 0.8011650982651559,
      "grad_norm": 0.003722853260114789,
      "learning_rate": 0.09228199253161017,
      "loss": 0.5331,
      "num_input_tokens_seen": 34423440,
      "step": 25030
    },
    {
      "epoch": 0.8013251392356443,
      "grad_norm": 0.0047199539840221405,
      "learning_rate": 0.09222762747063949,
      "loss": 0.5331,
      "num_input_tokens_seen": 34430176,
      "step": 25035
    },
    {
      "epoch": 0.8014851802061328,
      "grad_norm": 0.004058988764882088,
      "learning_rate": 0.09217327131889473,
      "loss": 0.5263,
      "num_input_tokens_seen": 34437136,
      "step": 25040
    },
    {
      "epoch": 0.8016452211766212,
      "grad_norm": 0.002864847192540765,
      "learning_rate": 0.09211892408475818,
      "loss": 0.4674,
      "num_input_tokens_seen": 34444048,
      "step": 25045
    },
    {
      "epoch": 0.8018052621471097,
      "grad_norm": 0.0017419463256374002,
      "learning_rate": 0.09206458577661089,
      "loss": 0.4436,
      "num_input_tokens_seen": 34450496,
      "step": 25050
    },
    {
      "epoch": 0.8019653031175981,
      "grad_norm": 0.0045301904901862144,
      "learning_rate": 0.09201025640283263,
      "loss": 0.4542,
      "num_input_tokens_seen": 34456976,
      "step": 25055
    },
    {
      "epoch": 0.8021253440880866,
      "grad_norm": 0.0043439180590212345,
      "learning_rate": 0.09195593597180148,
      "loss": 0.5772,
      "num_input_tokens_seen": 34463968,
      "step": 25060
    },
    {
      "epoch": 0.8022853850585749,
      "grad_norm": 0.003634899388998747,
      "learning_rate": 0.09190162449189444,
      "loss": 0.5399,
      "num_input_tokens_seen": 34471248,
      "step": 25065
    },
    {
      "epoch": 0.8024454260290634,
      "grad_norm": 0.007683577015995979,
      "learning_rate": 0.09184732197148705,
      "loss": 0.5647,
      "num_input_tokens_seen": 34477744,
      "step": 25070
    },
    {
      "epoch": 0.8026054669995519,
      "grad_norm": 0.0025230776518583298,
      "learning_rate": 0.09179302841895343,
      "loss": 0.4788,
      "num_input_tokens_seen": 34484112,
      "step": 25075
    },
    {
      "epoch": 0.8027655079700403,
      "grad_norm": 0.003817666554823518,
      "learning_rate": 0.09173874384266625,
      "loss": 0.5091,
      "num_input_tokens_seen": 34490896,
      "step": 25080
    },
    {
      "epoch": 0.8029255489405288,
      "grad_norm": 0.002041633939370513,
      "learning_rate": 0.09168446825099695,
      "loss": 0.6343,
      "num_input_tokens_seen": 34497776,
      "step": 25085
    },
    {
      "epoch": 0.8030855899110172,
      "grad_norm": 0.0019841266330331564,
      "learning_rate": 0.09163020165231545,
      "loss": 0.3387,
      "num_input_tokens_seen": 34504720,
      "step": 25090
    },
    {
      "epoch": 0.8032456308815057,
      "grad_norm": 0.004966434091329575,
      "learning_rate": 0.09157594405499044,
      "loss": 0.659,
      "num_input_tokens_seen": 34511152,
      "step": 25095
    },
    {
      "epoch": 0.8034056718519941,
      "grad_norm": 0.00387412472628057,
      "learning_rate": 0.09152169546738899,
      "loss": 0.5569,
      "num_input_tokens_seen": 34518080,
      "step": 25100
    },
    {
      "epoch": 0.8035657128224826,
      "grad_norm": 0.010261587798595428,
      "learning_rate": 0.09146745589787698,
      "loss": 0.4374,
      "num_input_tokens_seen": 34524544,
      "step": 25105
    },
    {
      "epoch": 0.803725753792971,
      "grad_norm": 0.005928184371441603,
      "learning_rate": 0.09141322535481891,
      "loss": 0.5064,
      "num_input_tokens_seen": 34531664,
      "step": 25110
    },
    {
      "epoch": 0.8038857947634594,
      "grad_norm": 0.0018291767919436097,
      "learning_rate": 0.0913590038465777,
      "loss": 0.4166,
      "num_input_tokens_seen": 34538384,
      "step": 25115
    },
    {
      "epoch": 0.804045835733948,
      "grad_norm": 0.004491131752729416,
      "learning_rate": 0.09130479138151505,
      "loss": 0.4801,
      "num_input_tokens_seen": 34545040,
      "step": 25120
    },
    {
      "epoch": 0.8042058767044363,
      "grad_norm": 0.0038028971757739782,
      "learning_rate": 0.09125058796799114,
      "loss": 0.5021,
      "num_input_tokens_seen": 34551872,
      "step": 25125
    },
    {
      "epoch": 0.8043659176749248,
      "grad_norm": 0.003166337264701724,
      "learning_rate": 0.09119639361436485,
      "loss": 0.4884,
      "num_input_tokens_seen": 34558752,
      "step": 25130
    },
    {
      "epoch": 0.8045259586454132,
      "grad_norm": 0.004568733274936676,
      "learning_rate": 0.09114220832899368,
      "loss": 0.4633,
      "num_input_tokens_seen": 34565280,
      "step": 25135
    },
    {
      "epoch": 0.8046859996159017,
      "grad_norm": 0.004122201818972826,
      "learning_rate": 0.0910880321202336,
      "loss": 0.5322,
      "num_input_tokens_seen": 34572320,
      "step": 25140
    },
    {
      "epoch": 0.8048460405863901,
      "grad_norm": 0.006650695577263832,
      "learning_rate": 0.09103386499643933,
      "loss": 0.5319,
      "num_input_tokens_seen": 34579184,
      "step": 25145
    },
    {
      "epoch": 0.8050060815568786,
      "grad_norm": 0.003983411937952042,
      "learning_rate": 0.09097970696596407,
      "loss": 0.512,
      "num_input_tokens_seen": 34585888,
      "step": 25150
    },
    {
      "epoch": 0.805166122527367,
      "grad_norm": 0.006579211913049221,
      "learning_rate": 0.09092555803715971,
      "loss": 0.556,
      "num_input_tokens_seen": 34593040,
      "step": 25155
    },
    {
      "epoch": 0.8053261634978555,
      "grad_norm": 0.0030830686446279287,
      "learning_rate": 0.0908714182183767,
      "loss": 0.3788,
      "num_input_tokens_seen": 34599872,
      "step": 25160
    },
    {
      "epoch": 0.8054862044683438,
      "grad_norm": 0.003872999921441078,
      "learning_rate": 0.090817287517964,
      "loss": 0.477,
      "num_input_tokens_seen": 34606800,
      "step": 25165
    },
    {
      "epoch": 0.8056462454388323,
      "grad_norm": 0.005070657003670931,
      "learning_rate": 0.09076316594426931,
      "loss": 0.485,
      "num_input_tokens_seen": 34613808,
      "step": 25170
    },
    {
      "epoch": 0.8058062864093208,
      "grad_norm": 0.004387910943478346,
      "learning_rate": 0.09070905350563888,
      "loss": 0.5089,
      "num_input_tokens_seen": 34620752,
      "step": 25175
    },
    {
      "epoch": 0.8059663273798092,
      "grad_norm": 0.003874995745718479,
      "learning_rate": 0.09065495021041745,
      "loss": 0.7072,
      "num_input_tokens_seen": 34627456,
      "step": 25180
    },
    {
      "epoch": 0.8061263683502977,
      "grad_norm": 0.007612846326082945,
      "learning_rate": 0.09060085606694851,
      "loss": 0.5529,
      "num_input_tokens_seen": 34634288,
      "step": 25185
    },
    {
      "epoch": 0.8062864093207861,
      "grad_norm": 0.004358894657343626,
      "learning_rate": 0.09054677108357405,
      "loss": 0.3506,
      "num_input_tokens_seen": 34641008,
      "step": 25190
    },
    {
      "epoch": 0.8064464502912746,
      "grad_norm": 0.00241135829128325,
      "learning_rate": 0.09049269526863457,
      "loss": 0.4106,
      "num_input_tokens_seen": 34647600,
      "step": 25195
    },
    {
      "epoch": 0.806606491261763,
      "grad_norm": 0.004093651659786701,
      "learning_rate": 0.09043862863046935,
      "loss": 0.5158,
      "num_input_tokens_seen": 34654672,
      "step": 25200
    },
    {
      "epoch": 0.806606491261763,
      "eval_loss": 0.5267176032066345,
      "eval_runtime": 332.2325,
      "eval_samples_per_second": 41.796,
      "eval_steps_per_second": 20.898,
      "num_input_tokens_seen": 34654672,
      "step": 25200
    },
    {
      "epoch": 0.8067665322322515,
      "grad_norm": 0.004222468473017216,
      "learning_rate": 0.09038457117741602,
      "loss": 0.4184,
      "num_input_tokens_seen": 34661312,
      "step": 25205
    },
    {
      "epoch": 0.8069265732027399,
      "grad_norm": 0.0041554090566933155,
      "learning_rate": 0.09033052291781099,
      "loss": 0.3926,
      "num_input_tokens_seen": 34668176,
      "step": 25210
    },
    {
      "epoch": 0.8070866141732284,
      "grad_norm": 0.0035552841145545244,
      "learning_rate": 0.09027648385998926,
      "loss": 0.4242,
      "num_input_tokens_seen": 34675344,
      "step": 25215
    },
    {
      "epoch": 0.8072466551437167,
      "grad_norm": 0.005699940491467714,
      "learning_rate": 0.09022245401228417,
      "loss": 0.4984,
      "num_input_tokens_seen": 34682560,
      "step": 25220
    },
    {
      "epoch": 0.8074066961142052,
      "grad_norm": 0.0033075844403356314,
      "learning_rate": 0.09016843338302792,
      "loss": 0.6199,
      "num_input_tokens_seen": 34689216,
      "step": 25225
    },
    {
      "epoch": 0.8075667370846937,
      "grad_norm": 0.00478717777878046,
      "learning_rate": 0.09011442198055115,
      "loss": 0.556,
      "num_input_tokens_seen": 34696096,
      "step": 25230
    },
    {
      "epoch": 0.8077267780551821,
      "grad_norm": 0.006529664620757103,
      "learning_rate": 0.09006041981318305,
      "loss": 0.7967,
      "num_input_tokens_seen": 34702912,
      "step": 25235
    },
    {
      "epoch": 0.8078868190256706,
      "grad_norm": 0.005568352527916431,
      "learning_rate": 0.09000642688925149,
      "loss": 0.556,
      "num_input_tokens_seen": 34709728,
      "step": 25240
    },
    {
      "epoch": 0.808046859996159,
      "grad_norm": 0.0029051858000457287,
      "learning_rate": 0.0899524432170828,
      "loss": 0.4679,
      "num_input_tokens_seen": 34716944,
      "step": 25245
    },
    {
      "epoch": 0.8082069009666475,
      "grad_norm": 0.0037167323753237724,
      "learning_rate": 0.08989846880500196,
      "loss": 0.3295,
      "num_input_tokens_seen": 34723616,
      "step": 25250
    },
    {
      "epoch": 0.8083669419371359,
      "grad_norm": 0.0035602550487965345,
      "learning_rate": 0.08984450366133256,
      "loss": 0.3467,
      "num_input_tokens_seen": 34730480,
      "step": 25255
    },
    {
      "epoch": 0.8085269829076244,
      "grad_norm": 0.0018410736229270697,
      "learning_rate": 0.08979054779439664,
      "loss": 0.4869,
      "num_input_tokens_seen": 34737616,
      "step": 25260
    },
    {
      "epoch": 0.8086870238781128,
      "grad_norm": 0.007691208738833666,
      "learning_rate": 0.08973660121251485,
      "loss": 0.4856,
      "num_input_tokens_seen": 34744368,
      "step": 25265
    },
    {
      "epoch": 0.8088470648486012,
      "grad_norm": 0.003476890502497554,
      "learning_rate": 0.08968266392400655,
      "loss": 0.5317,
      "num_input_tokens_seen": 34750992,
      "step": 25270
    },
    {
      "epoch": 0.8090071058190896,
      "grad_norm": 0.006955373100936413,
      "learning_rate": 0.0896287359371894,
      "loss": 0.4847,
      "num_input_tokens_seen": 34758672,
      "step": 25275
    },
    {
      "epoch": 0.8091671467895781,
      "grad_norm": 0.005999623332172632,
      "learning_rate": 0.08957481726037989,
      "loss": 0.5989,
      "num_input_tokens_seen": 34765808,
      "step": 25280
    },
    {
      "epoch": 0.8093271877600666,
      "grad_norm": 0.009339713491499424,
      "learning_rate": 0.08952090790189286,
      "loss": 0.2874,
      "num_input_tokens_seen": 34772832,
      "step": 25285
    },
    {
      "epoch": 0.809487228730555,
      "grad_norm": 0.007122676353901625,
      "learning_rate": 0.08946700787004187,
      "loss": 0.5473,
      "num_input_tokens_seen": 34779760,
      "step": 25290
    },
    {
      "epoch": 0.8096472697010435,
      "grad_norm": 0.006607082672417164,
      "learning_rate": 0.08941311717313899,
      "loss": 0.5551,
      "num_input_tokens_seen": 34786224,
      "step": 25295
    },
    {
      "epoch": 0.8098073106715319,
      "grad_norm": 0.006446785759180784,
      "learning_rate": 0.08935923581949483,
      "loss": 0.4489,
      "num_input_tokens_seen": 34792768,
      "step": 25300
    },
    {
      "epoch": 0.8099673516420204,
      "grad_norm": 0.008130413480103016,
      "learning_rate": 0.0893053638174185,
      "loss": 0.3851,
      "num_input_tokens_seen": 34799408,
      "step": 25305
    },
    {
      "epoch": 0.8101273926125088,
      "grad_norm": 0.0046351454220712185,
      "learning_rate": 0.0892515011752179,
      "loss": 0.4535,
      "num_input_tokens_seen": 34806160,
      "step": 25310
    },
    {
      "epoch": 0.8102874335829973,
      "grad_norm": 0.0037176506593823433,
      "learning_rate": 0.08919764790119918,
      "loss": 0.3651,
      "num_input_tokens_seen": 34812896,
      "step": 25315
    },
    {
      "epoch": 0.8104474745534856,
      "grad_norm": 0.005192394834011793,
      "learning_rate": 0.08914380400366727,
      "loss": 0.6297,
      "num_input_tokens_seen": 34819664,
      "step": 25320
    },
    {
      "epoch": 0.8106075155239741,
      "grad_norm": 0.005814188160002232,
      "learning_rate": 0.08908996949092551,
      "loss": 0.62,
      "num_input_tokens_seen": 34826368,
      "step": 25325
    },
    {
      "epoch": 0.8107675564944626,
      "grad_norm": 0.003955926280468702,
      "learning_rate": 0.08903614437127592,
      "loss": 0.5183,
      "num_input_tokens_seen": 34833408,
      "step": 25330
    },
    {
      "epoch": 0.810927597464951,
      "grad_norm": 0.002513448940590024,
      "learning_rate": 0.088982328653019,
      "loss": 0.5342,
      "num_input_tokens_seen": 34840432,
      "step": 25335
    },
    {
      "epoch": 0.8110876384354395,
      "grad_norm": 0.006346741691231728,
      "learning_rate": 0.0889285223444538,
      "loss": 0.5682,
      "num_input_tokens_seen": 34847440,
      "step": 25340
    },
    {
      "epoch": 0.8112476794059279,
      "grad_norm": 0.006489694118499756,
      "learning_rate": 0.08887472545387787,
      "loss": 0.4581,
      "num_input_tokens_seen": 34854224,
      "step": 25345
    },
    {
      "epoch": 0.8114077203764164,
      "grad_norm": 0.0028866094071418047,
      "learning_rate": 0.08882093798958751,
      "loss": 0.4065,
      "num_input_tokens_seen": 34861040,
      "step": 25350
    },
    {
      "epoch": 0.8115677613469048,
      "grad_norm": 0.003933262545615435,
      "learning_rate": 0.08876715995987726,
      "loss": 0.4215,
      "num_input_tokens_seen": 34867984,
      "step": 25355
    },
    {
      "epoch": 0.8117278023173933,
      "grad_norm": 0.006333711091428995,
      "learning_rate": 0.08871339137304052,
      "loss": 0.5051,
      "num_input_tokens_seen": 34875056,
      "step": 25360
    },
    {
      "epoch": 0.8118878432878817,
      "grad_norm": 0.005284439772367477,
      "learning_rate": 0.0886596322373689,
      "loss": 0.6231,
      "num_input_tokens_seen": 34882336,
      "step": 25365
    },
    {
      "epoch": 0.8120478842583702,
      "grad_norm": 0.00472482293844223,
      "learning_rate": 0.08860588256115293,
      "loss": 0.4864,
      "num_input_tokens_seen": 34888912,
      "step": 25370
    },
    {
      "epoch": 0.8122079252288585,
      "grad_norm": 0.008310808800160885,
      "learning_rate": 0.0885521423526814,
      "loss": 0.4648,
      "num_input_tokens_seen": 34896272,
      "step": 25375
    },
    {
      "epoch": 0.812367966199347,
      "grad_norm": 0.004755984991788864,
      "learning_rate": 0.08849841162024165,
      "loss": 0.5076,
      "num_input_tokens_seen": 34903232,
      "step": 25380
    },
    {
      "epoch": 0.8125280071698355,
      "grad_norm": 0.003500965191051364,
      "learning_rate": 0.08844469037211973,
      "loss": 0.4811,
      "num_input_tokens_seen": 34909888,
      "step": 25385
    },
    {
      "epoch": 0.8126880481403239,
      "grad_norm": 0.003700133180245757,
      "learning_rate": 0.08839097861660014,
      "loss": 0.5812,
      "num_input_tokens_seen": 34917104,
      "step": 25390
    },
    {
      "epoch": 0.8128480891108124,
      "grad_norm": 0.006093580741435289,
      "learning_rate": 0.08833727636196585,
      "loss": 0.5523,
      "num_input_tokens_seen": 34924128,
      "step": 25395
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 0.0056154923513531685,
      "learning_rate": 0.08828358361649848,
      "loss": 0.7053,
      "num_input_tokens_seen": 34931520,
      "step": 25400
    },
    {
      "epoch": 0.8130081300813008,
      "eval_loss": 0.5216692686080933,
      "eval_runtime": 332.2518,
      "eval_samples_per_second": 41.794,
      "eval_steps_per_second": 20.897,
      "num_input_tokens_seen": 34931520,
      "step": 25400
    },
    {
      "epoch": 0.8131681710517893,
      "grad_norm": 0.0048365662805736065,
      "learning_rate": 0.08822990038847807,
      "loss": 0.4242,
      "num_input_tokens_seen": 34938432,
      "step": 25405
    },
    {
      "epoch": 0.8133282120222777,
      "grad_norm": 0.006580015644431114,
      "learning_rate": 0.08817622668618325,
      "loss": 0.6142,
      "num_input_tokens_seen": 34945296,
      "step": 25410
    },
    {
      "epoch": 0.8134882529927662,
      "grad_norm": 0.00615040585398674,
      "learning_rate": 0.08812256251789125,
      "loss": 0.527,
      "num_input_tokens_seen": 34952256,
      "step": 25415
    },
    {
      "epoch": 0.8136482939632546,
      "grad_norm": 0.003650890663266182,
      "learning_rate": 0.08806890789187766,
      "loss": 0.3946,
      "num_input_tokens_seen": 34959232,
      "step": 25420
    },
    {
      "epoch": 0.813808334933743,
      "grad_norm": 0.003657127497717738,
      "learning_rate": 0.08801526281641672,
      "loss": 0.4413,
      "num_input_tokens_seen": 34966208,
      "step": 25425
    },
    {
      "epoch": 0.8139683759042314,
      "grad_norm": 0.007336874958127737,
      "learning_rate": 0.0879616272997813,
      "loss": 0.5868,
      "num_input_tokens_seen": 34973584,
      "step": 25430
    },
    {
      "epoch": 0.8141284168747199,
      "grad_norm": 0.00952393189072609,
      "learning_rate": 0.08790800135024247,
      "loss": 0.5242,
      "num_input_tokens_seen": 34980656,
      "step": 25435
    },
    {
      "epoch": 0.8142884578452084,
      "grad_norm": 0.00254620797932148,
      "learning_rate": 0.08785438497607023,
      "loss": 0.4145,
      "num_input_tokens_seen": 34987616,
      "step": 25440
    },
    {
      "epoch": 0.8144484988156968,
      "grad_norm": 0.004759324248880148,
      "learning_rate": 0.08780077818553277,
      "loss": 0.7411,
      "num_input_tokens_seen": 34994304,
      "step": 25445
    },
    {
      "epoch": 0.8146085397861853,
      "grad_norm": 0.006119485944509506,
      "learning_rate": 0.0877471809868969,
      "loss": 0.5148,
      "num_input_tokens_seen": 35001408,
      "step": 25450
    },
    {
      "epoch": 0.8147685807566737,
      "grad_norm": 0.004147745668888092,
      "learning_rate": 0.08769359338842811,
      "loss": 0.5636,
      "num_input_tokens_seen": 35008224,
      "step": 25455
    },
    {
      "epoch": 0.8149286217271622,
      "grad_norm": 0.003269667038694024,
      "learning_rate": 0.08764001539839016,
      "loss": 0.4884,
      "num_input_tokens_seen": 35015200,
      "step": 25460
    },
    {
      "epoch": 0.8150886626976506,
      "grad_norm": 0.0026226132176816463,
      "learning_rate": 0.08758644702504548,
      "loss": 0.3882,
      "num_input_tokens_seen": 35022000,
      "step": 25465
    },
    {
      "epoch": 0.8152487036681391,
      "grad_norm": 0.0036022996064275503,
      "learning_rate": 0.0875328882766551,
      "loss": 0.4944,
      "num_input_tokens_seen": 35028928,
      "step": 25470
    },
    {
      "epoch": 0.8154087446386274,
      "grad_norm": 0.01811184175312519,
      "learning_rate": 0.08747933916147828,
      "loss": 0.4088,
      "num_input_tokens_seen": 35035760,
      "step": 25475
    },
    {
      "epoch": 0.8155687856091159,
      "grad_norm": 0.003768468741327524,
      "learning_rate": 0.0874257996877731,
      "loss": 0.5,
      "num_input_tokens_seen": 35042240,
      "step": 25480
    },
    {
      "epoch": 0.8157288265796043,
      "grad_norm": 0.002025468507781625,
      "learning_rate": 0.08737226986379593,
      "loss": 0.5356,
      "num_input_tokens_seen": 35049024,
      "step": 25485
    },
    {
      "epoch": 0.8158888675500928,
      "grad_norm": 0.004124523606151342,
      "learning_rate": 0.08731874969780173,
      "loss": 0.5672,
      "num_input_tokens_seen": 35055984,
      "step": 25490
    },
    {
      "epoch": 0.8160489085205813,
      "grad_norm": 0.006454262416809797,
      "learning_rate": 0.08726523919804412,
      "loss": 0.5801,
      "num_input_tokens_seen": 35062688,
      "step": 25495
    },
    {
      "epoch": 0.8162089494910697,
      "grad_norm": 0.008967505767941475,
      "learning_rate": 0.08721173837277492,
      "loss": 0.375,
      "num_input_tokens_seen": 35069872,
      "step": 25500
    },
    {
      "epoch": 0.8163689904615582,
      "grad_norm": 0.0032460305374115705,
      "learning_rate": 0.08715824723024479,
      "loss": 0.5252,
      "num_input_tokens_seen": 35076496,
      "step": 25505
    },
    {
      "epoch": 0.8165290314320466,
      "grad_norm": 0.009215171448886395,
      "learning_rate": 0.08710476577870258,
      "loss": 0.5438,
      "num_input_tokens_seen": 35083632,
      "step": 25510
    },
    {
      "epoch": 0.8166890724025351,
      "grad_norm": 0.0027321134693920612,
      "learning_rate": 0.08705129402639587,
      "loss": 0.6673,
      "num_input_tokens_seen": 35090352,
      "step": 25515
    },
    {
      "epoch": 0.8168491133730235,
      "grad_norm": 0.004857539664953947,
      "learning_rate": 0.08699783198157078,
      "loss": 0.5352,
      "num_input_tokens_seen": 35097280,
      "step": 25520
    },
    {
      "epoch": 0.817009154343512,
      "grad_norm": 0.007319565396755934,
      "learning_rate": 0.08694437965247163,
      "loss": 0.4541,
      "num_input_tokens_seen": 35103888,
      "step": 25525
    },
    {
      "epoch": 0.8171691953140003,
      "grad_norm": 0.003239505225792527,
      "learning_rate": 0.08689093704734165,
      "loss": 0.4943,
      "num_input_tokens_seen": 35110272,
      "step": 25530
    },
    {
      "epoch": 0.8173292362844888,
      "grad_norm": 0.0026971192564815283,
      "learning_rate": 0.08683750417442222,
      "loss": 0.4735,
      "num_input_tokens_seen": 35116928,
      "step": 25535
    },
    {
      "epoch": 0.8174892772549772,
      "grad_norm": 0.005906062666326761,
      "learning_rate": 0.08678408104195334,
      "loss": 0.56,
      "num_input_tokens_seen": 35123904,
      "step": 25540
    },
    {
      "epoch": 0.8176493182254657,
      "grad_norm": 0.003988226875662804,
      "learning_rate": 0.08673066765817365,
      "loss": 0.4926,
      "num_input_tokens_seen": 35131152,
      "step": 25545
    },
    {
      "epoch": 0.8178093591959542,
      "grad_norm": 0.0027064571622759104,
      "learning_rate": 0.08667726403132005,
      "loss": 0.4417,
      "num_input_tokens_seen": 35138064,
      "step": 25550
    },
    {
      "epoch": 0.8179694001664426,
      "grad_norm": 0.003221881343051791,
      "learning_rate": 0.0866238701696281,
      "loss": 0.4244,
      "num_input_tokens_seen": 35144720,
      "step": 25555
    },
    {
      "epoch": 0.8181294411369311,
      "grad_norm": 0.005873254034668207,
      "learning_rate": 0.08657048608133185,
      "loss": 0.4196,
      "num_input_tokens_seen": 35151344,
      "step": 25560
    },
    {
      "epoch": 0.8182894821074195,
      "grad_norm": 0.00383500917814672,
      "learning_rate": 0.08651711177466369,
      "loss": 0.5152,
      "num_input_tokens_seen": 35158080,
      "step": 25565
    },
    {
      "epoch": 0.818449523077908,
      "grad_norm": 0.003189302980899811,
      "learning_rate": 0.08646374725785466,
      "loss": 0.5424,
      "num_input_tokens_seen": 35165616,
      "step": 25570
    },
    {
      "epoch": 0.8186095640483964,
      "grad_norm": 0.003971234895288944,
      "learning_rate": 0.08641039253913434,
      "loss": 0.6035,
      "num_input_tokens_seen": 35172192,
      "step": 25575
    },
    {
      "epoch": 0.8187696050188848,
      "grad_norm": 0.00653417780995369,
      "learning_rate": 0.08635704762673052,
      "loss": 0.5897,
      "num_input_tokens_seen": 35178784,
      "step": 25580
    },
    {
      "epoch": 0.8189296459893732,
      "grad_norm": 0.015298324637115002,
      "learning_rate": 0.08630371252886981,
      "loss": 0.5838,
      "num_input_tokens_seen": 35186016,
      "step": 25585
    },
    {
      "epoch": 0.8190896869598617,
      "grad_norm": 0.0032497933134436607,
      "learning_rate": 0.08625038725377704,
      "loss": 0.4494,
      "num_input_tokens_seen": 35192656,
      "step": 25590
    },
    {
      "epoch": 0.8192497279303502,
      "grad_norm": 0.004387675318866968,
      "learning_rate": 0.08619707180967566,
      "loss": 0.5374,
      "num_input_tokens_seen": 35199344,
      "step": 25595
    },
    {
      "epoch": 0.8194097689008386,
      "grad_norm": 0.005866920109838247,
      "learning_rate": 0.08614376620478768,
      "loss": 0.4934,
      "num_input_tokens_seen": 35206448,
      "step": 25600
    },
    {
      "epoch": 0.8194097689008386,
      "eval_loss": 0.5182533264160156,
      "eval_runtime": 331.9906,
      "eval_samples_per_second": 41.826,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 35206448,
      "step": 25600
    },
    {
      "epoch": 0.8195698098713271,
      "grad_norm": 0.006403959356248379,
      "learning_rate": 0.08609047044733344,
      "loss": 0.7146,
      "num_input_tokens_seen": 35213584,
      "step": 25605
    },
    {
      "epoch": 0.8197298508418155,
      "grad_norm": 0.0062397196888923645,
      "learning_rate": 0.08603718454553168,
      "loss": 0.4918,
      "num_input_tokens_seen": 35220928,
      "step": 25610
    },
    {
      "epoch": 0.819889891812304,
      "grad_norm": 0.008242450654506683,
      "learning_rate": 0.08598390850759997,
      "loss": 0.5558,
      "num_input_tokens_seen": 35227936,
      "step": 25615
    },
    {
      "epoch": 0.8200499327827924,
      "grad_norm": 0.005037902854382992,
      "learning_rate": 0.08593064234175397,
      "loss": 0.5467,
      "num_input_tokens_seen": 35234352,
      "step": 25620
    },
    {
      "epoch": 0.8202099737532809,
      "grad_norm": 0.004915071185678244,
      "learning_rate": 0.08587738605620815,
      "loss": 0.5039,
      "num_input_tokens_seen": 35241040,
      "step": 25625
    },
    {
      "epoch": 0.8203700147237692,
      "grad_norm": 0.006262078415602446,
      "learning_rate": 0.08582413965917512,
      "loss": 0.5152,
      "num_input_tokens_seen": 35247824,
      "step": 25630
    },
    {
      "epoch": 0.8205300556942577,
      "grad_norm": 0.004325537476688623,
      "learning_rate": 0.08577090315886628,
      "loss": 0.4855,
      "num_input_tokens_seen": 35254816,
      "step": 25635
    },
    {
      "epoch": 0.8206900966647461,
      "grad_norm": 0.0038119256496429443,
      "learning_rate": 0.08571767656349136,
      "loss": 0.5996,
      "num_input_tokens_seen": 35261904,
      "step": 25640
    },
    {
      "epoch": 0.8208501376352346,
      "grad_norm": 0.0064861667342484,
      "learning_rate": 0.08566445988125847,
      "loss": 0.3901,
      "num_input_tokens_seen": 35269248,
      "step": 25645
    },
    {
      "epoch": 0.8210101786057231,
      "grad_norm": 0.0025264397263526917,
      "learning_rate": 0.08561125312037436,
      "loss": 0.491,
      "num_input_tokens_seen": 35276064,
      "step": 25650
    },
    {
      "epoch": 0.8211702195762115,
      "grad_norm": 0.004076038487255573,
      "learning_rate": 0.08555805628904424,
      "loss": 0.6288,
      "num_input_tokens_seen": 35282640,
      "step": 25655
    },
    {
      "epoch": 0.8213302605467,
      "grad_norm": 0.0061716255731880665,
      "learning_rate": 0.08550486939547161,
      "loss": 0.5778,
      "num_input_tokens_seen": 35289216,
      "step": 25660
    },
    {
      "epoch": 0.8214903015171884,
      "grad_norm": 0.0032985566649585962,
      "learning_rate": 0.08545169244785869,
      "loss": 0.5057,
      "num_input_tokens_seen": 35296112,
      "step": 25665
    },
    {
      "epoch": 0.8216503424876769,
      "grad_norm": 0.003814979689195752,
      "learning_rate": 0.08539852545440589,
      "loss": 0.3637,
      "num_input_tokens_seen": 35302624,
      "step": 25670
    },
    {
      "epoch": 0.8218103834581653,
      "grad_norm": 0.005345033947378397,
      "learning_rate": 0.08534536842331235,
      "loss": 0.3968,
      "num_input_tokens_seen": 35309632,
      "step": 25675
    },
    {
      "epoch": 0.8219704244286538,
      "grad_norm": 0.005427381955087185,
      "learning_rate": 0.08529222136277545,
      "loss": 0.5323,
      "num_input_tokens_seen": 35316592,
      "step": 25680
    },
    {
      "epoch": 0.8221304653991421,
      "grad_norm": 0.001696236664429307,
      "learning_rate": 0.08523908428099125,
      "loss": 0.5799,
      "num_input_tokens_seen": 35323520,
      "step": 25685
    },
    {
      "epoch": 0.8222905063696306,
      "grad_norm": 0.003423895686864853,
      "learning_rate": 0.08518595718615402,
      "loss": 0.5051,
      "num_input_tokens_seen": 35330304,
      "step": 25690
    },
    {
      "epoch": 0.822450547340119,
      "grad_norm": 0.006009547971189022,
      "learning_rate": 0.08513284008645675,
      "loss": 0.5738,
      "num_input_tokens_seen": 35337072,
      "step": 25695
    },
    {
      "epoch": 0.8226105883106075,
      "grad_norm": 0.005179218016564846,
      "learning_rate": 0.08507973299009065,
      "loss": 0.4495,
      "num_input_tokens_seen": 35343728,
      "step": 25700
    },
    {
      "epoch": 0.822770629281096,
      "grad_norm": 0.0056747812777757645,
      "learning_rate": 0.08502663590524563,
      "loss": 0.6526,
      "num_input_tokens_seen": 35350592,
      "step": 25705
    },
    {
      "epoch": 0.8229306702515844,
      "grad_norm": 0.00506596127524972,
      "learning_rate": 0.08497354884010981,
      "loss": 0.4463,
      "num_input_tokens_seen": 35357568,
      "step": 25710
    },
    {
      "epoch": 0.8230907112220729,
      "grad_norm": 0.00927469227463007,
      "learning_rate": 0.0849204718028699,
      "loss": 0.5634,
      "num_input_tokens_seen": 35364720,
      "step": 25715
    },
    {
      "epoch": 0.8232507521925613,
      "grad_norm": 0.005844544619321823,
      "learning_rate": 0.08486740480171118,
      "loss": 0.4047,
      "num_input_tokens_seen": 35371280,
      "step": 25720
    },
    {
      "epoch": 0.8234107931630498,
      "grad_norm": 0.005193536635488272,
      "learning_rate": 0.08481434784481706,
      "loss": 0.6059,
      "num_input_tokens_seen": 35378432,
      "step": 25725
    },
    {
      "epoch": 0.8235708341335382,
      "grad_norm": 0.006876097526401281,
      "learning_rate": 0.08476130094036968,
      "loss": 0.6329,
      "num_input_tokens_seen": 35385216,
      "step": 25730
    },
    {
      "epoch": 0.8237308751040266,
      "grad_norm": 0.0036810743622481823,
      "learning_rate": 0.08470826409654961,
      "loss": 0.6639,
      "num_input_tokens_seen": 35391728,
      "step": 25735
    },
    {
      "epoch": 0.823890916074515,
      "grad_norm": 0.004545011557638645,
      "learning_rate": 0.08465523732153564,
      "loss": 0.6464,
      "num_input_tokens_seen": 35398608,
      "step": 25740
    },
    {
      "epoch": 0.8240509570450035,
      "grad_norm": 0.0015843308065086603,
      "learning_rate": 0.08460222062350532,
      "loss": 0.4396,
      "num_input_tokens_seen": 35405664,
      "step": 25745
    },
    {
      "epoch": 0.8242109980154919,
      "grad_norm": 0.00528285838663578,
      "learning_rate": 0.08454921401063442,
      "loss": 0.5261,
      "num_input_tokens_seen": 35412368,
      "step": 25750
    },
    {
      "epoch": 0.8243710389859804,
      "grad_norm": 0.0052243685349822044,
      "learning_rate": 0.08449621749109716,
      "loss": 0.3651,
      "num_input_tokens_seen": 35419536,
      "step": 25755
    },
    {
      "epoch": 0.8245310799564689,
      "grad_norm": 0.005252052564173937,
      "learning_rate": 0.08444323107306641,
      "loss": 0.8823,
      "num_input_tokens_seen": 35426288,
      "step": 25760
    },
    {
      "epoch": 0.8246911209269573,
      "grad_norm": 0.004766441881656647,
      "learning_rate": 0.0843902547647132,
      "loss": 0.5771,
      "num_input_tokens_seen": 35433392,
      "step": 25765
    },
    {
      "epoch": 0.8248511618974458,
      "grad_norm": 0.005193920806050301,
      "learning_rate": 0.0843372885742072,
      "loss": 0.4738,
      "num_input_tokens_seen": 35440448,
      "step": 25770
    },
    {
      "epoch": 0.8250112028679342,
      "grad_norm": 0.0031509120017290115,
      "learning_rate": 0.08428433250971652,
      "loss": 0.4496,
      "num_input_tokens_seen": 35447056,
      "step": 25775
    },
    {
      "epoch": 0.8251712438384227,
      "grad_norm": 0.003504253923892975,
      "learning_rate": 0.08423138657940757,
      "loss": 0.402,
      "num_input_tokens_seen": 35453872,
      "step": 25780
    },
    {
      "epoch": 0.825331284808911,
      "grad_norm": 0.0028185457922518253,
      "learning_rate": 0.08417845079144536,
      "loss": 0.4627,
      "num_input_tokens_seen": 35460992,
      "step": 25785
    },
    {
      "epoch": 0.8254913257793995,
      "grad_norm": 0.00296256341971457,
      "learning_rate": 0.08412552515399314,
      "loss": 0.3309,
      "num_input_tokens_seen": 35467856,
      "step": 25790
    },
    {
      "epoch": 0.8256513667498879,
      "grad_norm": 0.00652862386777997,
      "learning_rate": 0.08407260967521278,
      "loss": 0.5141,
      "num_input_tokens_seen": 35475088,
      "step": 25795
    },
    {
      "epoch": 0.8258114077203764,
      "grad_norm": 0.004843416623771191,
      "learning_rate": 0.08401970436326454,
      "loss": 0.4797,
      "num_input_tokens_seen": 35482800,
      "step": 25800
    },
    {
      "epoch": 0.8258114077203764,
      "eval_loss": 0.5184707045555115,
      "eval_runtime": 332.1411,
      "eval_samples_per_second": 41.808,
      "eval_steps_per_second": 20.904,
      "num_input_tokens_seen": 35482800,
      "step": 25800
    },
    {
      "epoch": 0.8259714486908648,
      "grad_norm": 0.006374089978635311,
      "learning_rate": 0.08396680922630702,
      "loss": 0.5911,
      "num_input_tokens_seen": 35489520,
      "step": 25805
    },
    {
      "epoch": 0.8261314896613533,
      "grad_norm": 0.003909167367964983,
      "learning_rate": 0.08391392427249732,
      "loss": 0.5763,
      "num_input_tokens_seen": 35496304,
      "step": 25810
    },
    {
      "epoch": 0.8262915306318418,
      "grad_norm": 0.01638675108551979,
      "learning_rate": 0.08386104950999107,
      "loss": 0.753,
      "num_input_tokens_seen": 35502928,
      "step": 25815
    },
    {
      "epoch": 0.8264515716023302,
      "grad_norm": 0.004372297320514917,
      "learning_rate": 0.0838081849469421,
      "loss": 0.503,
      "num_input_tokens_seen": 35509664,
      "step": 25820
    },
    {
      "epoch": 0.8266116125728187,
      "grad_norm": 0.004373301286250353,
      "learning_rate": 0.08375533059150281,
      "loss": 0.3955,
      "num_input_tokens_seen": 35516336,
      "step": 25825
    },
    {
      "epoch": 0.8267716535433071,
      "grad_norm": 0.005332795903086662,
      "learning_rate": 0.08370248645182406,
      "loss": 0.473,
      "num_input_tokens_seen": 35523456,
      "step": 25830
    },
    {
      "epoch": 0.8269316945137956,
      "grad_norm": 0.005518487188965082,
      "learning_rate": 0.083649652536055,
      "loss": 0.3554,
      "num_input_tokens_seen": 35529808,
      "step": 25835
    },
    {
      "epoch": 0.8270917354842839,
      "grad_norm": 0.005473607685416937,
      "learning_rate": 0.08359682885234339,
      "loss": 0.5731,
      "num_input_tokens_seen": 35536448,
      "step": 25840
    },
    {
      "epoch": 0.8272517764547724,
      "grad_norm": 0.004844644106924534,
      "learning_rate": 0.08354401540883516,
      "loss": 0.6943,
      "num_input_tokens_seen": 35542784,
      "step": 25845
    },
    {
      "epoch": 0.8274118174252608,
      "grad_norm": 0.004221132956445217,
      "learning_rate": 0.0834912122136749,
      "loss": 0.6159,
      "num_input_tokens_seen": 35549520,
      "step": 25850
    },
    {
      "epoch": 0.8275718583957493,
      "grad_norm": 0.003685489995405078,
      "learning_rate": 0.0834384192750056,
      "loss": 0.5897,
      "num_input_tokens_seen": 35556320,
      "step": 25855
    },
    {
      "epoch": 0.8277318993662378,
      "grad_norm": 0.004640358965843916,
      "learning_rate": 0.08338563660096844,
      "loss": 0.4157,
      "num_input_tokens_seen": 35563168,
      "step": 25860
    },
    {
      "epoch": 0.8278919403367262,
      "grad_norm": 0.0039308397099375725,
      "learning_rate": 0.08333286419970329,
      "loss": 0.6089,
      "num_input_tokens_seen": 35570192,
      "step": 25865
    },
    {
      "epoch": 0.8280519813072147,
      "grad_norm": 0.003422453301027417,
      "learning_rate": 0.08328010207934824,
      "loss": 0.6935,
      "num_input_tokens_seen": 35577936,
      "step": 25870
    },
    {
      "epoch": 0.8282120222777031,
      "grad_norm": 0.006191405933350325,
      "learning_rate": 0.08322735024803989,
      "loss": 0.5659,
      "num_input_tokens_seen": 35585168,
      "step": 25875
    },
    {
      "epoch": 0.8283720632481916,
      "grad_norm": 0.003433545585721731,
      "learning_rate": 0.08317460871391331,
      "loss": 0.4562,
      "num_input_tokens_seen": 35591248,
      "step": 25880
    },
    {
      "epoch": 0.82853210421868,
      "grad_norm": 0.01538124680519104,
      "learning_rate": 0.08312187748510179,
      "loss": 0.4982,
      "num_input_tokens_seen": 35598448,
      "step": 25885
    },
    {
      "epoch": 0.8286921451891684,
      "grad_norm": 0.003932721447199583,
      "learning_rate": 0.08306915656973726,
      "loss": 0.4047,
      "num_input_tokens_seen": 35605632,
      "step": 25890
    },
    {
      "epoch": 0.8288521861596568,
      "grad_norm": 0.0020257136784493923,
      "learning_rate": 0.08301644597594988,
      "loss": 0.3414,
      "num_input_tokens_seen": 35612096,
      "step": 25895
    },
    {
      "epoch": 0.8290122271301453,
      "grad_norm": 0.0027559157460927963,
      "learning_rate": 0.08296374571186826,
      "loss": 0.5511,
      "num_input_tokens_seen": 35618864,
      "step": 25900
    },
    {
      "epoch": 0.8291722681006337,
      "grad_norm": 0.003152458928525448,
      "learning_rate": 0.08291105578561955,
      "loss": 0.5196,
      "num_input_tokens_seen": 35626160,
      "step": 25905
    },
    {
      "epoch": 0.8293323090711222,
      "grad_norm": 0.002652376191690564,
      "learning_rate": 0.08285837620532904,
      "loss": 0.4367,
      "num_input_tokens_seen": 35633312,
      "step": 25910
    },
    {
      "epoch": 0.8294923500416107,
      "grad_norm": 0.002029903931543231,
      "learning_rate": 0.0828057069791207,
      "loss": 0.4193,
      "num_input_tokens_seen": 35640192,
      "step": 25915
    },
    {
      "epoch": 0.8296523910120991,
      "grad_norm": 0.005238483659923077,
      "learning_rate": 0.0827530481151168,
      "loss": 0.3119,
      "num_input_tokens_seen": 35647104,
      "step": 25920
    },
    {
      "epoch": 0.8298124319825876,
      "grad_norm": 0.005686548538506031,
      "learning_rate": 0.08270039962143792,
      "loss": 0.5764,
      "num_input_tokens_seen": 35653808,
      "step": 25925
    },
    {
      "epoch": 0.829972472953076,
      "grad_norm": 0.00689256377518177,
      "learning_rate": 0.08264776150620314,
      "loss": 0.541,
      "num_input_tokens_seen": 35660736,
      "step": 25930
    },
    {
      "epoch": 0.8301325139235645,
      "grad_norm": 0.0031606548000127077,
      "learning_rate": 0.08259513377753,
      "loss": 0.4512,
      "num_input_tokens_seen": 35667344,
      "step": 25935
    },
    {
      "epoch": 0.8302925548940528,
      "grad_norm": 0.0026892193127423525,
      "learning_rate": 0.08254251644353423,
      "loss": 0.5636,
      "num_input_tokens_seen": 35674240,
      "step": 25940
    },
    {
      "epoch": 0.8304525958645413,
      "grad_norm": 0.002968753455206752,
      "learning_rate": 0.08248990951233022,
      "loss": 0.5288,
      "num_input_tokens_seen": 35681680,
      "step": 25945
    },
    {
      "epoch": 0.8306126368350297,
      "grad_norm": 0.0029447174165397882,
      "learning_rate": 0.08243731299203048,
      "loss": 0.4077,
      "num_input_tokens_seen": 35688160,
      "step": 25950
    },
    {
      "epoch": 0.8307726778055182,
      "grad_norm": 0.004103557206690311,
      "learning_rate": 0.08238472689074612,
      "loss": 0.6856,
      "num_input_tokens_seen": 35694960,
      "step": 25955
    },
    {
      "epoch": 0.8309327187760066,
      "grad_norm": 0.0034687789157032967,
      "learning_rate": 0.08233215121658666,
      "loss": 0.5022,
      "num_input_tokens_seen": 35702368,
      "step": 25960
    },
    {
      "epoch": 0.8310927597464951,
      "grad_norm": 0.006532520521432161,
      "learning_rate": 0.08227958597765982,
      "loss": 0.6208,
      "num_input_tokens_seen": 35709136,
      "step": 25965
    },
    {
      "epoch": 0.8312528007169836,
      "grad_norm": 0.0035738416481763124,
      "learning_rate": 0.08222703118207181,
      "loss": 0.6402,
      "num_input_tokens_seen": 35716096,
      "step": 25970
    },
    {
      "epoch": 0.831412841687472,
      "grad_norm": 0.004390406887978315,
      "learning_rate": 0.08217448683792734,
      "loss": 0.4139,
      "num_input_tokens_seen": 35722960,
      "step": 25975
    },
    {
      "epoch": 0.8315728826579605,
      "grad_norm": 0.003668200457468629,
      "learning_rate": 0.08212195295332926,
      "loss": 0.5249,
      "num_input_tokens_seen": 35729904,
      "step": 25980
    },
    {
      "epoch": 0.8317329236284489,
      "grad_norm": 0.005272912792861462,
      "learning_rate": 0.08206942953637915,
      "loss": 0.8034,
      "num_input_tokens_seen": 35736640,
      "step": 25985
    },
    {
      "epoch": 0.8318929645989374,
      "grad_norm": 0.0022285261657088995,
      "learning_rate": 0.08201691659517658,
      "loss": 0.6197,
      "num_input_tokens_seen": 35743696,
      "step": 25990
    },
    {
      "epoch": 0.8320530055694257,
      "grad_norm": 0.0029169381596148014,
      "learning_rate": 0.08196441413781981,
      "loss": 0.4997,
      "num_input_tokens_seen": 35750416,
      "step": 25995
    },
    {
      "epoch": 0.8322130465399142,
      "grad_norm": 0.004448354709893465,
      "learning_rate": 0.08191192217240544,
      "loss": 0.5238,
      "num_input_tokens_seen": 35756816,
      "step": 26000
    },
    {
      "epoch": 0.8322130465399142,
      "eval_loss": 0.5198890566825867,
      "eval_runtime": 333.3486,
      "eval_samples_per_second": 41.656,
      "eval_steps_per_second": 20.828,
      "num_input_tokens_seen": 35756816,
      "step": 26000
    },
    {
      "epoch": 0.8323730875104026,
      "grad_norm": 0.003189150942489505,
      "learning_rate": 0.08185944070702823,
      "loss": 0.4003,
      "num_input_tokens_seen": 35763136,
      "step": 26005
    },
    {
      "epoch": 0.8325331284808911,
      "grad_norm": 0.005674141924828291,
      "learning_rate": 0.08180696974978159,
      "loss": 0.6464,
      "num_input_tokens_seen": 35769632,
      "step": 26010
    },
    {
      "epoch": 0.8326931694513795,
      "grad_norm": 0.0014576894463971257,
      "learning_rate": 0.08175450930875724,
      "loss": 0.4187,
      "num_input_tokens_seen": 35776496,
      "step": 26015
    },
    {
      "epoch": 0.832853210421868,
      "grad_norm": 0.004188659135252237,
      "learning_rate": 0.08170205939204513,
      "loss": 0.5616,
      "num_input_tokens_seen": 35783328,
      "step": 26020
    },
    {
      "epoch": 0.8330132513923565,
      "grad_norm": 0.003506055334582925,
      "learning_rate": 0.08164962000773379,
      "loss": 0.5089,
      "num_input_tokens_seen": 35790336,
      "step": 26025
    },
    {
      "epoch": 0.8331732923628449,
      "grad_norm": 0.004665244370698929,
      "learning_rate": 0.08159719116390995,
      "loss": 0.5608,
      "num_input_tokens_seen": 35797344,
      "step": 26030
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.005485698580741882,
      "learning_rate": 0.08154477286865887,
      "loss": 0.443,
      "num_input_tokens_seen": 35804048,
      "step": 26035
    },
    {
      "epoch": 0.8334933743038218,
      "grad_norm": 0.0027829622849822044,
      "learning_rate": 0.08149236513006404,
      "loss": 0.49,
      "num_input_tokens_seen": 35811120,
      "step": 26040
    },
    {
      "epoch": 0.8336534152743102,
      "grad_norm": 0.003452188102528453,
      "learning_rate": 0.08143996795620746,
      "loss": 0.361,
      "num_input_tokens_seen": 35817616,
      "step": 26045
    },
    {
      "epoch": 0.8338134562447986,
      "grad_norm": 0.005440519656985998,
      "learning_rate": 0.08138758135516938,
      "loss": 0.3659,
      "num_input_tokens_seen": 35824384,
      "step": 26050
    },
    {
      "epoch": 0.8339734972152871,
      "grad_norm": 0.0035370986443012953,
      "learning_rate": 0.08133520533502851,
      "loss": 0.4011,
      "num_input_tokens_seen": 35831216,
      "step": 26055
    },
    {
      "epoch": 0.8341335381857755,
      "grad_norm": 0.0035750633105635643,
      "learning_rate": 0.08128283990386184,
      "loss": 0.4342,
      "num_input_tokens_seen": 35838416,
      "step": 26060
    },
    {
      "epoch": 0.834293579156264,
      "grad_norm": 0.00460593169555068,
      "learning_rate": 0.08123048506974488,
      "loss": 0.5684,
      "num_input_tokens_seen": 35845472,
      "step": 26065
    },
    {
      "epoch": 0.8344536201267524,
      "grad_norm": 0.0037240313831716776,
      "learning_rate": 0.08117814084075124,
      "loss": 0.4355,
      "num_input_tokens_seen": 35852160,
      "step": 26070
    },
    {
      "epoch": 0.8346136610972409,
      "grad_norm": 0.0033752534072846174,
      "learning_rate": 0.08112580722495318,
      "loss": 0.5776,
      "num_input_tokens_seen": 35859056,
      "step": 26075
    },
    {
      "epoch": 0.8347737020677294,
      "grad_norm": 0.005838749930262566,
      "learning_rate": 0.08107348423042122,
      "loss": 0.606,
      "num_input_tokens_seen": 35866208,
      "step": 26080
    },
    {
      "epoch": 0.8349337430382178,
      "grad_norm": 0.0042215753346681595,
      "learning_rate": 0.08102117186522413,
      "loss": 0.4651,
      "num_input_tokens_seen": 35872832,
      "step": 26085
    },
    {
      "epoch": 0.8350937840087063,
      "grad_norm": 0.005922353360801935,
      "learning_rate": 0.08096887013742916,
      "loss": 0.5234,
      "num_input_tokens_seen": 35879952,
      "step": 26090
    },
    {
      "epoch": 0.8352538249791946,
      "grad_norm": 0.0036135707050561905,
      "learning_rate": 0.08091657905510198,
      "loss": 0.6463,
      "num_input_tokens_seen": 35886480,
      "step": 26095
    },
    {
      "epoch": 0.8354138659496831,
      "grad_norm": 0.0028410586528480053,
      "learning_rate": 0.08086429862630642,
      "loss": 0.6062,
      "num_input_tokens_seen": 35893824,
      "step": 26100
    },
    {
      "epoch": 0.8355739069201715,
      "grad_norm": 0.002536978805437684,
      "learning_rate": 0.08081202885910488,
      "loss": 0.4204,
      "num_input_tokens_seen": 35900944,
      "step": 26105
    },
    {
      "epoch": 0.83573394789066,
      "grad_norm": 0.00497179152444005,
      "learning_rate": 0.08075976976155795,
      "loss": 0.5768,
      "num_input_tokens_seen": 35907744,
      "step": 26110
    },
    {
      "epoch": 0.8358939888611484,
      "grad_norm": 0.0032778913155198097,
      "learning_rate": 0.08070752134172461,
      "loss": 0.3943,
      "num_input_tokens_seen": 35915232,
      "step": 26115
    },
    {
      "epoch": 0.8360540298316369,
      "grad_norm": 0.002180648036301136,
      "learning_rate": 0.08065528360766229,
      "loss": 0.4669,
      "num_input_tokens_seen": 35922064,
      "step": 26120
    },
    {
      "epoch": 0.8362140708021254,
      "grad_norm": 0.004063660278916359,
      "learning_rate": 0.08060305656742664,
      "loss": 0.5028,
      "num_input_tokens_seen": 35929040,
      "step": 26125
    },
    {
      "epoch": 0.8363741117726138,
      "grad_norm": 0.0024299873039126396,
      "learning_rate": 0.08055084022907182,
      "loss": 0.3976,
      "num_input_tokens_seen": 35935728,
      "step": 26130
    },
    {
      "epoch": 0.8365341527431023,
      "grad_norm": 0.005371626932173967,
      "learning_rate": 0.08049863460065014,
      "loss": 0.4512,
      "num_input_tokens_seen": 35942208,
      "step": 26135
    },
    {
      "epoch": 0.8366941937135907,
      "grad_norm": 0.0036304963286966085,
      "learning_rate": 0.0804464396902124,
      "loss": 0.461,
      "num_input_tokens_seen": 35949184,
      "step": 26140
    },
    {
      "epoch": 0.8368542346840792,
      "grad_norm": 0.0035036008339375257,
      "learning_rate": 0.08039425550580777,
      "loss": 0.4169,
      "num_input_tokens_seen": 35956032,
      "step": 26145
    },
    {
      "epoch": 0.8370142756545675,
      "grad_norm": 0.0036709613632410765,
      "learning_rate": 0.08034208205548363,
      "loss": 0.5525,
      "num_input_tokens_seen": 35962608,
      "step": 26150
    },
    {
      "epoch": 0.837174316625056,
      "grad_norm": 0.004320255946367979,
      "learning_rate": 0.08028991934728581,
      "loss": 0.6001,
      "num_input_tokens_seen": 35970352,
      "step": 26155
    },
    {
      "epoch": 0.8373343575955444,
      "grad_norm": 0.0037343387957662344,
      "learning_rate": 0.0802377673892585,
      "loss": 0.4256,
      "num_input_tokens_seen": 35976784,
      "step": 26160
    },
    {
      "epoch": 0.8374943985660329,
      "grad_norm": 0.005303704179823399,
      "learning_rate": 0.0801856261894441,
      "loss": 0.5837,
      "num_input_tokens_seen": 35983712,
      "step": 26165
    },
    {
      "epoch": 0.8376544395365213,
      "grad_norm": 0.0024630401749163866,
      "learning_rate": 0.08013349575588354,
      "loss": 0.461,
      "num_input_tokens_seen": 35990560,
      "step": 26170
    },
    {
      "epoch": 0.8378144805070098,
      "grad_norm": 0.003015598515048623,
      "learning_rate": 0.08008137609661586,
      "loss": 0.6723,
      "num_input_tokens_seen": 35997520,
      "step": 26175
    },
    {
      "epoch": 0.8379745214774983,
      "grad_norm": 0.004372558556497097,
      "learning_rate": 0.08002926721967872,
      "loss": 0.4181,
      "num_input_tokens_seen": 36004064,
      "step": 26180
    },
    {
      "epoch": 0.8381345624479867,
      "grad_norm": 0.004158422350883484,
      "learning_rate": 0.07997716913310782,
      "loss": 0.4822,
      "num_input_tokens_seen": 36010944,
      "step": 26185
    },
    {
      "epoch": 0.8382946034184752,
      "grad_norm": 0.003142266534268856,
      "learning_rate": 0.07992508184493745,
      "loss": 0.4628,
      "num_input_tokens_seen": 36017488,
      "step": 26190
    },
    {
      "epoch": 0.8384546443889636,
      "grad_norm": 0.0037277606315910816,
      "learning_rate": 0.07987300536320001,
      "loss": 0.5525,
      "num_input_tokens_seen": 36024416,
      "step": 26195
    },
    {
      "epoch": 0.838614685359452,
      "grad_norm": 0.004690831992775202,
      "learning_rate": 0.07982093969592649,
      "loss": 0.4021,
      "num_input_tokens_seen": 36031296,
      "step": 26200
    },
    {
      "epoch": 0.838614685359452,
      "eval_loss": 0.5175352692604065,
      "eval_runtime": 331.707,
      "eval_samples_per_second": 41.862,
      "eval_steps_per_second": 20.931,
      "num_input_tokens_seen": 36031296,
      "step": 26200
    },
    {
      "epoch": 0.8387747263299404,
      "grad_norm": 0.007852542214095592,
      "learning_rate": 0.07976888485114592,
      "loss": 0.4695,
      "num_input_tokens_seen": 36038224,
      "step": 26205
    },
    {
      "epoch": 0.8389347673004289,
      "grad_norm": 0.005034526344388723,
      "learning_rate": 0.07971684083688595,
      "loss": 0.438,
      "num_input_tokens_seen": 36045344,
      "step": 26210
    },
    {
      "epoch": 0.8390948082709173,
      "grad_norm": 0.004475869704037905,
      "learning_rate": 0.0796648076611723,
      "loss": 0.4441,
      "num_input_tokens_seen": 36052096,
      "step": 26215
    },
    {
      "epoch": 0.8392548492414058,
      "grad_norm": 0.004348197486251593,
      "learning_rate": 0.07961278533202922,
      "loss": 0.4436,
      "num_input_tokens_seen": 36059152,
      "step": 26220
    },
    {
      "epoch": 0.8394148902118942,
      "grad_norm": 0.00587256345897913,
      "learning_rate": 0.07956077385747919,
      "loss": 0.5341,
      "num_input_tokens_seen": 36065840,
      "step": 26225
    },
    {
      "epoch": 0.8395749311823827,
      "grad_norm": 0.009400506503880024,
      "learning_rate": 0.079508773245543,
      "loss": 0.6739,
      "num_input_tokens_seen": 36073024,
      "step": 26230
    },
    {
      "epoch": 0.8397349721528712,
      "grad_norm": 0.00627184147015214,
      "learning_rate": 0.07945678350423982,
      "loss": 0.3579,
      "num_input_tokens_seen": 36079808,
      "step": 26235
    },
    {
      "epoch": 0.8398950131233596,
      "grad_norm": 0.004075394943356514,
      "learning_rate": 0.07940480464158717,
      "loss": 0.5374,
      "num_input_tokens_seen": 36086400,
      "step": 26240
    },
    {
      "epoch": 0.8400550540938481,
      "grad_norm": 0.006969534792006016,
      "learning_rate": 0.07935283666560076,
      "loss": 0.6072,
      "num_input_tokens_seen": 36093584,
      "step": 26245
    },
    {
      "epoch": 0.8402150950643364,
      "grad_norm": 0.0024359349627047777,
      "learning_rate": 0.07930087958429478,
      "loss": 0.5371,
      "num_input_tokens_seen": 36100368,
      "step": 26250
    },
    {
      "epoch": 0.8403751360348249,
      "grad_norm": 0.006562584545463324,
      "learning_rate": 0.07924893340568159,
      "loss": 0.4578,
      "num_input_tokens_seen": 36107200,
      "step": 26255
    },
    {
      "epoch": 0.8405351770053133,
      "grad_norm": 0.009027177467942238,
      "learning_rate": 0.07919699813777205,
      "loss": 0.5155,
      "num_input_tokens_seen": 36114048,
      "step": 26260
    },
    {
      "epoch": 0.8406952179758018,
      "grad_norm": 0.007234485819935799,
      "learning_rate": 0.07914507378857515,
      "loss": 0.4995,
      "num_input_tokens_seen": 36121600,
      "step": 26265
    },
    {
      "epoch": 0.8408552589462902,
      "grad_norm": 0.0029030886944383383,
      "learning_rate": 0.07909316036609822,
      "loss": 0.3091,
      "num_input_tokens_seen": 36128288,
      "step": 26270
    },
    {
      "epoch": 0.8410152999167787,
      "grad_norm": 0.004157917574048042,
      "learning_rate": 0.07904125787834704,
      "loss": 0.4834,
      "num_input_tokens_seen": 36135264,
      "step": 26275
    },
    {
      "epoch": 0.8411753408872671,
      "grad_norm": 0.005018006544560194,
      "learning_rate": 0.07898936633332569,
      "loss": 0.4464,
      "num_input_tokens_seen": 36142224,
      "step": 26280
    },
    {
      "epoch": 0.8413353818577556,
      "grad_norm": 0.006967227440327406,
      "learning_rate": 0.07893748573903635,
      "loss": 0.5101,
      "num_input_tokens_seen": 36148752,
      "step": 26285
    },
    {
      "epoch": 0.8414954228282441,
      "grad_norm": 0.005640960764139891,
      "learning_rate": 0.0788856161034798,
      "loss": 0.6661,
      "num_input_tokens_seen": 36155344,
      "step": 26290
    },
    {
      "epoch": 0.8416554637987325,
      "grad_norm": 0.005904954392462969,
      "learning_rate": 0.07883375743465487,
      "loss": 0.4587,
      "num_input_tokens_seen": 36162016,
      "step": 26295
    },
    {
      "epoch": 0.841815504769221,
      "grad_norm": 0.004132525064051151,
      "learning_rate": 0.07878190974055888,
      "loss": 0.4691,
      "num_input_tokens_seen": 36168912,
      "step": 26300
    },
    {
      "epoch": 0.8419755457397093,
      "grad_norm": 0.00471544312313199,
      "learning_rate": 0.07873007302918746,
      "loss": 0.4901,
      "num_input_tokens_seen": 36175584,
      "step": 26305
    },
    {
      "epoch": 0.8421355867101978,
      "grad_norm": 0.0037973071448504925,
      "learning_rate": 0.07867824730853433,
      "loss": 0.5159,
      "num_input_tokens_seen": 36182720,
      "step": 26310
    },
    {
      "epoch": 0.8422956276806862,
      "grad_norm": 0.007064228877425194,
      "learning_rate": 0.07862643258659176,
      "loss": 0.6655,
      "num_input_tokens_seen": 36189520,
      "step": 26315
    },
    {
      "epoch": 0.8424556686511747,
      "grad_norm": 0.006026947870850563,
      "learning_rate": 0.07857462887135026,
      "loss": 0.429,
      "num_input_tokens_seen": 36196288,
      "step": 26320
    },
    {
      "epoch": 0.8426157096216631,
      "grad_norm": 0.007061505224555731,
      "learning_rate": 0.0785228361707986,
      "loss": 0.4434,
      "num_input_tokens_seen": 36203168,
      "step": 26325
    },
    {
      "epoch": 0.8427757505921516,
      "grad_norm": 0.011255883611738682,
      "learning_rate": 0.07847105449292378,
      "loss": 0.5156,
      "num_input_tokens_seen": 36210256,
      "step": 26330
    },
    {
      "epoch": 0.8429357915626401,
      "grad_norm": 0.005259189289063215,
      "learning_rate": 0.0784192838457113,
      "loss": 0.5499,
      "num_input_tokens_seen": 36216992,
      "step": 26335
    },
    {
      "epoch": 0.8430958325331285,
      "grad_norm": 0.005552845541387796,
      "learning_rate": 0.07836752423714473,
      "loss": 0.5185,
      "num_input_tokens_seen": 36223648,
      "step": 26340
    },
    {
      "epoch": 0.843255873503617,
      "grad_norm": 0.004316482227295637,
      "learning_rate": 0.07831577567520616,
      "loss": 0.4061,
      "num_input_tokens_seen": 36230640,
      "step": 26345
    },
    {
      "epoch": 0.8434159144741054,
      "grad_norm": 0.014837580733001232,
      "learning_rate": 0.07826403816787579,
      "loss": 0.3966,
      "num_input_tokens_seen": 36237568,
      "step": 26350
    },
    {
      "epoch": 0.8435759554445939,
      "grad_norm": 0.005095233675092459,
      "learning_rate": 0.0782123117231322,
      "loss": 0.5915,
      "num_input_tokens_seen": 36244432,
      "step": 26355
    },
    {
      "epoch": 0.8437359964150822,
      "grad_norm": 0.0030633099377155304,
      "learning_rate": 0.07816059634895237,
      "loss": 0.4459,
      "num_input_tokens_seen": 36251344,
      "step": 26360
    },
    {
      "epoch": 0.8438960373855707,
      "grad_norm": 0.004756886512041092,
      "learning_rate": 0.0781088920533113,
      "loss": 0.6359,
      "num_input_tokens_seen": 36259216,
      "step": 26365
    },
    {
      "epoch": 0.8440560783560591,
      "grad_norm": 0.003631259547546506,
      "learning_rate": 0.07805719884418257,
      "loss": 0.5109,
      "num_input_tokens_seen": 36266368,
      "step": 26370
    },
    {
      "epoch": 0.8442161193265476,
      "grad_norm": 0.005262619815766811,
      "learning_rate": 0.07800551672953779,
      "loss": 0.6258,
      "num_input_tokens_seen": 36273136,
      "step": 26375
    },
    {
      "epoch": 0.844376160297036,
      "grad_norm": 0.008118929341435432,
      "learning_rate": 0.07795384571734709,
      "loss": 0.543,
      "num_input_tokens_seen": 36280336,
      "step": 26380
    },
    {
      "epoch": 0.8445362012675245,
      "grad_norm": 0.003905773628503084,
      "learning_rate": 0.07790218581557883,
      "loss": 0.569,
      "num_input_tokens_seen": 36287488,
      "step": 26385
    },
    {
      "epoch": 0.844696242238013,
      "grad_norm": 0.011734245344996452,
      "learning_rate": 0.07785053703219949,
      "loss": 0.5434,
      "num_input_tokens_seen": 36294656,
      "step": 26390
    },
    {
      "epoch": 0.8448562832085014,
      "grad_norm": 0.006380126345902681,
      "learning_rate": 0.07779889937517409,
      "loss": 0.6725,
      "num_input_tokens_seen": 36301440,
      "step": 26395
    },
    {
      "epoch": 0.8450163241789899,
      "grad_norm": 0.00522440392524004,
      "learning_rate": 0.0777472728524657,
      "loss": 0.7042,
      "num_input_tokens_seen": 36307968,
      "step": 26400
    },
    {
      "epoch": 0.8450163241789899,
      "eval_loss": 0.52250075340271,
      "eval_runtime": 331.7079,
      "eval_samples_per_second": 41.862,
      "eval_steps_per_second": 20.931,
      "num_input_tokens_seen": 36307968,
      "step": 26400
    },
    {
      "epoch": 0.8451763651494782,
      "grad_norm": 0.0033724240493029356,
      "learning_rate": 0.07769565747203584,
      "loss": 0.4786,
      "num_input_tokens_seen": 36314560,
      "step": 26405
    },
    {
      "epoch": 0.8453364061199667,
      "grad_norm": 0.00938517414033413,
      "learning_rate": 0.07764405324184427,
      "loss": 0.4672,
      "num_input_tokens_seen": 36321168,
      "step": 26410
    },
    {
      "epoch": 0.8454964470904551,
      "grad_norm": 0.004416842479258776,
      "learning_rate": 0.07759246016984889,
      "loss": 0.4965,
      "num_input_tokens_seen": 36328304,
      "step": 26415
    },
    {
      "epoch": 0.8456564880609436,
      "grad_norm": 0.004702616948634386,
      "learning_rate": 0.07754087826400609,
      "loss": 0.4814,
      "num_input_tokens_seen": 36335072,
      "step": 26420
    },
    {
      "epoch": 0.845816529031432,
      "grad_norm": 0.004716373980045319,
      "learning_rate": 0.0774893075322705,
      "loss": 0.4493,
      "num_input_tokens_seen": 36341808,
      "step": 26425
    },
    {
      "epoch": 0.8459765700019205,
      "grad_norm": 0.005262926686555147,
      "learning_rate": 0.07743774798259484,
      "loss": 0.5661,
      "num_input_tokens_seen": 36348224,
      "step": 26430
    },
    {
      "epoch": 0.8461366109724089,
      "grad_norm": 0.004729955457150936,
      "learning_rate": 0.07738619962293032,
      "loss": 0.5601,
      "num_input_tokens_seen": 36354896,
      "step": 26435
    },
    {
      "epoch": 0.8462966519428974,
      "grad_norm": 0.003323962911963463,
      "learning_rate": 0.0773346624612264,
      "loss": 0.4096,
      "num_input_tokens_seen": 36361792,
      "step": 26440
    },
    {
      "epoch": 0.8464566929133859,
      "grad_norm": 0.0038144111167639494,
      "learning_rate": 0.07728313650543066,
      "loss": 0.4946,
      "num_input_tokens_seen": 36368832,
      "step": 26445
    },
    {
      "epoch": 0.8466167338838743,
      "grad_norm": 0.0034150974825024605,
      "learning_rate": 0.07723162176348913,
      "loss": 0.5606,
      "num_input_tokens_seen": 36375536,
      "step": 26450
    },
    {
      "epoch": 0.8467767748543628,
      "grad_norm": 0.008517473004758358,
      "learning_rate": 0.07718011824334593,
      "loss": 0.4212,
      "num_input_tokens_seen": 36381968,
      "step": 26455
    },
    {
      "epoch": 0.8469368158248511,
      "grad_norm": 0.0092539107427001,
      "learning_rate": 0.07712862595294363,
      "loss": 0.6524,
      "num_input_tokens_seen": 36388704,
      "step": 26460
    },
    {
      "epoch": 0.8470968567953396,
      "grad_norm": 0.005959936883300543,
      "learning_rate": 0.07707714490022301,
      "loss": 0.654,
      "num_input_tokens_seen": 36395216,
      "step": 26465
    },
    {
      "epoch": 0.847256897765828,
      "grad_norm": 0.0032486054114997387,
      "learning_rate": 0.07702567509312298,
      "loss": 0.4884,
      "num_input_tokens_seen": 36401792,
      "step": 26470
    },
    {
      "epoch": 0.8474169387363165,
      "grad_norm": 0.004956191871315241,
      "learning_rate": 0.07697421653958098,
      "loss": 0.6135,
      "num_input_tokens_seen": 36409232,
      "step": 26475
    },
    {
      "epoch": 0.8475769797068049,
      "grad_norm": 0.004569231998175383,
      "learning_rate": 0.07692276924753247,
      "loss": 0.6254,
      "num_input_tokens_seen": 36416128,
      "step": 26480
    },
    {
      "epoch": 0.8477370206772934,
      "grad_norm": 0.008201093412935734,
      "learning_rate": 0.07687133322491124,
      "loss": 0.5854,
      "num_input_tokens_seen": 36422976,
      "step": 26485
    },
    {
      "epoch": 0.8478970616477818,
      "grad_norm": 0.004750742577016354,
      "learning_rate": 0.07681990847964948,
      "loss": 0.433,
      "num_input_tokens_seen": 36429792,
      "step": 26490
    },
    {
      "epoch": 0.8480571026182703,
      "grad_norm": 0.0037916398141533136,
      "learning_rate": 0.0767684950196774,
      "loss": 0.5745,
      "num_input_tokens_seen": 36436528,
      "step": 26495
    },
    {
      "epoch": 0.8482171435887588,
      "grad_norm": 0.005433341953903437,
      "learning_rate": 0.0767170928529237,
      "loss": 0.4949,
      "num_input_tokens_seen": 36443120,
      "step": 26500
    },
    {
      "epoch": 0.8483771845592472,
      "grad_norm": 0.002968896646052599,
      "learning_rate": 0.07666570198731526,
      "loss": 0.3652,
      "num_input_tokens_seen": 36450016,
      "step": 26505
    },
    {
      "epoch": 0.8485372255297357,
      "grad_norm": 0.006167560815811157,
      "learning_rate": 0.07661432243077708,
      "loss": 0.6098,
      "num_input_tokens_seen": 36457104,
      "step": 26510
    },
    {
      "epoch": 0.848697266500224,
      "grad_norm": 0.004008921794593334,
      "learning_rate": 0.0765629541912326,
      "loss": 0.5605,
      "num_input_tokens_seen": 36463952,
      "step": 26515
    },
    {
      "epoch": 0.8488573074707125,
      "grad_norm": 0.003632513340562582,
      "learning_rate": 0.07651159727660352,
      "loss": 0.5713,
      "num_input_tokens_seen": 36470688,
      "step": 26520
    },
    {
      "epoch": 0.8490173484412009,
      "grad_norm": 0.0043215518817305565,
      "learning_rate": 0.07646025169480959,
      "loss": 0.6751,
      "num_input_tokens_seen": 36477584,
      "step": 26525
    },
    {
      "epoch": 0.8491773894116894,
      "grad_norm": 0.0056101419031620026,
      "learning_rate": 0.07640891745376908,
      "loss": 0.4517,
      "num_input_tokens_seen": 36484208,
      "step": 26530
    },
    {
      "epoch": 0.8493374303821778,
      "grad_norm": 0.007596283219754696,
      "learning_rate": 0.07635759456139822,
      "loss": 0.6528,
      "num_input_tokens_seen": 36491024,
      "step": 26535
    },
    {
      "epoch": 0.8494974713526663,
      "grad_norm": 0.004045348148792982,
      "learning_rate": 0.0763062830256118,
      "loss": 0.5473,
      "num_input_tokens_seen": 36498080,
      "step": 26540
    },
    {
      "epoch": 0.8496575123231547,
      "grad_norm": 0.00393324950709939,
      "learning_rate": 0.07625498285432258,
      "loss": 0.6284,
      "num_input_tokens_seen": 36504928,
      "step": 26545
    },
    {
      "epoch": 0.8498175532936432,
      "grad_norm": 0.0022350468207150698,
      "learning_rate": 0.07620369405544176,
      "loss": 0.3469,
      "num_input_tokens_seen": 36512112,
      "step": 26550
    },
    {
      "epoch": 0.8499775942641317,
      "grad_norm": 0.00593554088845849,
      "learning_rate": 0.07615241663687868,
      "loss": 0.5913,
      "num_input_tokens_seen": 36519056,
      "step": 26555
    },
    {
      "epoch": 0.85013763523462,
      "grad_norm": 0.005167112220078707,
      "learning_rate": 0.07610115060654106,
      "loss": 0.5105,
      "num_input_tokens_seen": 36525712,
      "step": 26560
    },
    {
      "epoch": 0.8502976762051085,
      "grad_norm": 0.0045301965437829494,
      "learning_rate": 0.07604989597233458,
      "loss": 0.4882,
      "num_input_tokens_seen": 36532800,
      "step": 26565
    },
    {
      "epoch": 0.8504577171755969,
      "grad_norm": 0.005330171436071396,
      "learning_rate": 0.07599865274216352,
      "loss": 0.5422,
      "num_input_tokens_seen": 36539504,
      "step": 26570
    },
    {
      "epoch": 0.8506177581460854,
      "grad_norm": 0.005567891523241997,
      "learning_rate": 0.07594742092393013,
      "loss": 0.3874,
      "num_input_tokens_seen": 36546640,
      "step": 26575
    },
    {
      "epoch": 0.8507777991165738,
      "grad_norm": 0.004925946705043316,
      "learning_rate": 0.07589620052553503,
      "loss": 0.44,
      "num_input_tokens_seen": 36553136,
      "step": 26580
    },
    {
      "epoch": 0.8509378400870623,
      "grad_norm": 0.0060883574187755585,
      "learning_rate": 0.0758449915548771,
      "loss": 0.5947,
      "num_input_tokens_seen": 36559952,
      "step": 26585
    },
    {
      "epoch": 0.8510978810575507,
      "grad_norm": 0.009053162299096584,
      "learning_rate": 0.07579379401985332,
      "loss": 0.7901,
      "num_input_tokens_seen": 36566800,
      "step": 26590
    },
    {
      "epoch": 0.8512579220280392,
      "grad_norm": 0.009042750112712383,
      "learning_rate": 0.07574260792835905,
      "loss": 0.3565,
      "num_input_tokens_seen": 36573504,
      "step": 26595
    },
    {
      "epoch": 0.8514179629985277,
      "grad_norm": 0.008874189108610153,
      "learning_rate": 0.07569143328828784,
      "loss": 0.6731,
      "num_input_tokens_seen": 36580432,
      "step": 26600
    },
    {
      "epoch": 0.8514179629985277,
      "eval_loss": 0.5186812281608582,
      "eval_runtime": 332.0097,
      "eval_samples_per_second": 41.824,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 36580432,
      "step": 26600
    },
    {
      "epoch": 0.8515780039690161,
      "grad_norm": 0.00312195997685194,
      "learning_rate": 0.0756402701075314,
      "loss": 0.6462,
      "num_input_tokens_seen": 36587424,
      "step": 26605
    },
    {
      "epoch": 0.8517380449395046,
      "grad_norm": 0.003125071059912443,
      "learning_rate": 0.07558911839397982,
      "loss": 0.4326,
      "num_input_tokens_seen": 36594048,
      "step": 26610
    },
    {
      "epoch": 0.8518980859099929,
      "grad_norm": 0.009861161932349205,
      "learning_rate": 0.07553797815552123,
      "loss": 0.567,
      "num_input_tokens_seen": 36600640,
      "step": 26615
    },
    {
      "epoch": 0.8520581268804814,
      "grad_norm": 0.006775704212486744,
      "learning_rate": 0.07548684940004222,
      "loss": 0.6251,
      "num_input_tokens_seen": 36607376,
      "step": 26620
    },
    {
      "epoch": 0.8522181678509698,
      "grad_norm": 0.003942094277590513,
      "learning_rate": 0.07543573213542744,
      "loss": 0.4561,
      "num_input_tokens_seen": 36614288,
      "step": 26625
    },
    {
      "epoch": 0.8523782088214583,
      "grad_norm": 0.0019562356173992157,
      "learning_rate": 0.0753846263695597,
      "loss": 0.672,
      "num_input_tokens_seen": 36621600,
      "step": 26630
    },
    {
      "epoch": 0.8525382497919467,
      "grad_norm": 0.007594516966491938,
      "learning_rate": 0.07533353211032029,
      "loss": 0.8608,
      "num_input_tokens_seen": 36628368,
      "step": 26635
    },
    {
      "epoch": 0.8526982907624352,
      "grad_norm": 0.00456826901063323,
      "learning_rate": 0.07528244936558857,
      "loss": 0.6381,
      "num_input_tokens_seen": 36635840,
      "step": 26640
    },
    {
      "epoch": 0.8528583317329236,
      "grad_norm": 0.006902114488184452,
      "learning_rate": 0.07523137814324206,
      "loss": 0.4526,
      "num_input_tokens_seen": 36642688,
      "step": 26645
    },
    {
      "epoch": 0.8530183727034121,
      "grad_norm": 0.004157486837357283,
      "learning_rate": 0.07518031845115672,
      "loss": 0.6068,
      "num_input_tokens_seen": 36649600,
      "step": 26650
    },
    {
      "epoch": 0.8531784136739006,
      "grad_norm": 0.004658672027289867,
      "learning_rate": 0.07512927029720647,
      "loss": 0.8209,
      "num_input_tokens_seen": 36656432,
      "step": 26655
    },
    {
      "epoch": 0.853338454644389,
      "grad_norm": 0.003241476835682988,
      "learning_rate": 0.0750782336892636,
      "loss": 0.4017,
      "num_input_tokens_seen": 36663360,
      "step": 26660
    },
    {
      "epoch": 0.8534984956148775,
      "grad_norm": 0.004577235318720341,
      "learning_rate": 0.0750272086351987,
      "loss": 0.467,
      "num_input_tokens_seen": 36670032,
      "step": 26665
    },
    {
      "epoch": 0.8536585365853658,
      "grad_norm": 0.0024691689759492874,
      "learning_rate": 0.07497619514288031,
      "loss": 0.4831,
      "num_input_tokens_seen": 36676704,
      "step": 26670
    },
    {
      "epoch": 0.8538185775558543,
      "grad_norm": 0.005253626499325037,
      "learning_rate": 0.07492519322017545,
      "loss": 1.0358,
      "num_input_tokens_seen": 36683472,
      "step": 26675
    },
    {
      "epoch": 0.8539786185263427,
      "grad_norm": 0.004074752330780029,
      "learning_rate": 0.0748742028749493,
      "loss": 0.4852,
      "num_input_tokens_seen": 36690496,
      "step": 26680
    },
    {
      "epoch": 0.8541386594968312,
      "grad_norm": 0.006430268753319979,
      "learning_rate": 0.0748232241150651,
      "loss": 0.5072,
      "num_input_tokens_seen": 36697696,
      "step": 26685
    },
    {
      "epoch": 0.8542987004673196,
      "grad_norm": 0.003662393894046545,
      "learning_rate": 0.07477225694838453,
      "loss": 0.4414,
      "num_input_tokens_seen": 36704384,
      "step": 26690
    },
    {
      "epoch": 0.8544587414378081,
      "grad_norm": 0.0038944475818425417,
      "learning_rate": 0.07472130138276731,
      "loss": 0.5123,
      "num_input_tokens_seen": 36711424,
      "step": 26695
    },
    {
      "epoch": 0.8546187824082965,
      "grad_norm": 0.004140603821724653,
      "learning_rate": 0.07467035742607138,
      "loss": 0.3554,
      "num_input_tokens_seen": 36718304,
      "step": 26700
    },
    {
      "epoch": 0.854778823378785,
      "grad_norm": 0.008415115997195244,
      "learning_rate": 0.07461942508615303,
      "loss": 0.4814,
      "num_input_tokens_seen": 36725104,
      "step": 26705
    },
    {
      "epoch": 0.8549388643492735,
      "grad_norm": 0.0035856233444064856,
      "learning_rate": 0.07456850437086657,
      "loss": 0.4804,
      "num_input_tokens_seen": 36731712,
      "step": 26710
    },
    {
      "epoch": 0.8550989053197618,
      "grad_norm": 0.004682404920458794,
      "learning_rate": 0.07451759528806468,
      "loss": 0.5085,
      "num_input_tokens_seen": 36738640,
      "step": 26715
    },
    {
      "epoch": 0.8552589462902503,
      "grad_norm": 0.004628075286746025,
      "learning_rate": 0.0744666978455982,
      "loss": 0.6271,
      "num_input_tokens_seen": 36745232,
      "step": 26720
    },
    {
      "epoch": 0.8554189872607387,
      "grad_norm": 0.011660770513117313,
      "learning_rate": 0.07441581205131609,
      "loss": 0.5433,
      "num_input_tokens_seen": 36752432,
      "step": 26725
    },
    {
      "epoch": 0.8555790282312272,
      "grad_norm": 0.006264521740376949,
      "learning_rate": 0.07436493791306566,
      "loss": 0.3956,
      "num_input_tokens_seen": 36759328,
      "step": 26730
    },
    {
      "epoch": 0.8557390692017156,
      "grad_norm": 0.006673912517726421,
      "learning_rate": 0.07431407543869223,
      "loss": 0.3969,
      "num_input_tokens_seen": 36765696,
      "step": 26735
    },
    {
      "epoch": 0.8558991101722041,
      "grad_norm": 0.005838002543896437,
      "learning_rate": 0.0742632246360395,
      "loss": 0.4437,
      "num_input_tokens_seen": 36772752,
      "step": 26740
    },
    {
      "epoch": 0.8560591511426925,
      "grad_norm": 0.0034939113538712263,
      "learning_rate": 0.07421238551294934,
      "loss": 0.4524,
      "num_input_tokens_seen": 36779568,
      "step": 26745
    },
    {
      "epoch": 0.856219192113181,
      "grad_norm": 0.0031540498603135347,
      "learning_rate": 0.07416155807726171,
      "loss": 0.4866,
      "num_input_tokens_seen": 36787040,
      "step": 26750
    },
    {
      "epoch": 0.8563792330836694,
      "grad_norm": 0.007299327757209539,
      "learning_rate": 0.07411074233681492,
      "loss": 0.5252,
      "num_input_tokens_seen": 36793632,
      "step": 26755
    },
    {
      "epoch": 0.8565392740541579,
      "grad_norm": 0.009273015893995762,
      "learning_rate": 0.07405993829944528,
      "loss": 0.5337,
      "num_input_tokens_seen": 36800032,
      "step": 26760
    },
    {
      "epoch": 0.8566993150246464,
      "grad_norm": 0.004101223312318325,
      "learning_rate": 0.07400914597298755,
      "loss": 0.5383,
      "num_input_tokens_seen": 36807024,
      "step": 26765
    },
    {
      "epoch": 0.8568593559951347,
      "grad_norm": 0.00369862443767488,
      "learning_rate": 0.07395836536527445,
      "loss": 0.5372,
      "num_input_tokens_seen": 36813696,
      "step": 26770
    },
    {
      "epoch": 0.8570193969656232,
      "grad_norm": 0.003290789667516947,
      "learning_rate": 0.07390759648413696,
      "loss": 0.545,
      "num_input_tokens_seen": 36820752,
      "step": 26775
    },
    {
      "epoch": 0.8571794379361116,
      "grad_norm": 0.004367134999483824,
      "learning_rate": 0.07385683933740435,
      "loss": 0.4333,
      "num_input_tokens_seen": 36827648,
      "step": 26780
    },
    {
      "epoch": 0.8573394789066001,
      "grad_norm": 0.002879665931686759,
      "learning_rate": 0.07380609393290402,
      "loss": 0.5366,
      "num_input_tokens_seen": 36834320,
      "step": 26785
    },
    {
      "epoch": 0.8574995198770885,
      "grad_norm": 0.003566778264939785,
      "learning_rate": 0.07375536027846147,
      "loss": 0.4753,
      "num_input_tokens_seen": 36841600,
      "step": 26790
    },
    {
      "epoch": 0.857659560847577,
      "grad_norm": 0.005380814429372549,
      "learning_rate": 0.07370463838190057,
      "loss": 0.5749,
      "num_input_tokens_seen": 36848608,
      "step": 26795
    },
    {
      "epoch": 0.8578196018180654,
      "grad_norm": 0.004229803569614887,
      "learning_rate": 0.07365392825104317,
      "loss": 0.5761,
      "num_input_tokens_seen": 36855328,
      "step": 26800
    },
    {
      "epoch": 0.8578196018180654,
      "eval_loss": 0.5182088613510132,
      "eval_runtime": 331.8558,
      "eval_samples_per_second": 41.843,
      "eval_steps_per_second": 20.922,
      "num_input_tokens_seen": 36855328,
      "step": 26800
    },
    {
      "epoch": 0.8579796427885539,
      "grad_norm": 0.005755239631980658,
      "learning_rate": 0.07360322989370945,
      "loss": 0.6067,
      "num_input_tokens_seen": 36861984,
      "step": 26805
    },
    {
      "epoch": 0.8581396837590423,
      "grad_norm": 0.00501808850094676,
      "learning_rate": 0.07355254331771781,
      "loss": 0.5372,
      "num_input_tokens_seen": 36868848,
      "step": 26810
    },
    {
      "epoch": 0.8582997247295308,
      "grad_norm": 0.0045625739730894566,
      "learning_rate": 0.07350186853088461,
      "loss": 0.497,
      "num_input_tokens_seen": 36876192,
      "step": 26815
    },
    {
      "epoch": 0.8584597657000193,
      "grad_norm": 0.004448394291102886,
      "learning_rate": 0.07345120554102462,
      "loss": 0.3829,
      "num_input_tokens_seen": 36883024,
      "step": 26820
    },
    {
      "epoch": 0.8586198066705076,
      "grad_norm": 0.0064229522831737995,
      "learning_rate": 0.07340055435595079,
      "loss": 0.7723,
      "num_input_tokens_seen": 36890208,
      "step": 26825
    },
    {
      "epoch": 0.8587798476409961,
      "grad_norm": 0.004798680543899536,
      "learning_rate": 0.07334991498347401,
      "loss": 0.6217,
      "num_input_tokens_seen": 36897056,
      "step": 26830
    },
    {
      "epoch": 0.8589398886114845,
      "grad_norm": 0.005284983664751053,
      "learning_rate": 0.07329928743140365,
      "loss": 0.8272,
      "num_input_tokens_seen": 36903712,
      "step": 26835
    },
    {
      "epoch": 0.859099929581973,
      "grad_norm": 0.005384691059589386,
      "learning_rate": 0.07324867170754705,
      "loss": 0.4382,
      "num_input_tokens_seen": 36910848,
      "step": 26840
    },
    {
      "epoch": 0.8592599705524614,
      "grad_norm": 0.006283293478190899,
      "learning_rate": 0.07319806781970974,
      "loss": 0.5607,
      "num_input_tokens_seen": 36918192,
      "step": 26845
    },
    {
      "epoch": 0.8594200115229499,
      "grad_norm": 0.005267408210784197,
      "learning_rate": 0.07314747577569555,
      "loss": 0.535,
      "num_input_tokens_seen": 36925056,
      "step": 26850
    },
    {
      "epoch": 0.8595800524934383,
      "grad_norm": 0.0036873859353363514,
      "learning_rate": 0.07309689558330636,
      "loss": 0.4319,
      "num_input_tokens_seen": 36932064,
      "step": 26855
    },
    {
      "epoch": 0.8597400934639268,
      "grad_norm": 0.008665517903864384,
      "learning_rate": 0.0730463272503423,
      "loss": 0.6112,
      "num_input_tokens_seen": 36938592,
      "step": 26860
    },
    {
      "epoch": 0.8599001344344153,
      "grad_norm": 0.008345507085323334,
      "learning_rate": 0.07299577078460168,
      "loss": 0.3951,
      "num_input_tokens_seen": 36945600,
      "step": 26865
    },
    {
      "epoch": 0.8600601754049036,
      "grad_norm": 0.005310241598635912,
      "learning_rate": 0.07294522619388083,
      "loss": 0.6699,
      "num_input_tokens_seen": 36952144,
      "step": 26870
    },
    {
      "epoch": 0.8602202163753921,
      "grad_norm": 0.003512091003358364,
      "learning_rate": 0.07289469348597452,
      "loss": 0.5395,
      "num_input_tokens_seen": 36958928,
      "step": 26875
    },
    {
      "epoch": 0.8603802573458805,
      "grad_norm": 0.0038135903887450695,
      "learning_rate": 0.07284417266867535,
      "loss": 0.4925,
      "num_input_tokens_seen": 36965952,
      "step": 26880
    },
    {
      "epoch": 0.860540298316369,
      "grad_norm": 0.0033417842350900173,
      "learning_rate": 0.07279366374977439,
      "loss": 0.4206,
      "num_input_tokens_seen": 36973152,
      "step": 26885
    },
    {
      "epoch": 0.8607003392868574,
      "grad_norm": 0.006203772034496069,
      "learning_rate": 0.07274316673706074,
      "loss": 0.5987,
      "num_input_tokens_seen": 36980192,
      "step": 26890
    },
    {
      "epoch": 0.8608603802573459,
      "grad_norm": 0.004366415552794933,
      "learning_rate": 0.07269268163832161,
      "loss": 0.4615,
      "num_input_tokens_seen": 36987392,
      "step": 26895
    },
    {
      "epoch": 0.8610204212278343,
      "grad_norm": 0.004639393649995327,
      "learning_rate": 0.07264220846134248,
      "loss": 0.3676,
      "num_input_tokens_seen": 36994176,
      "step": 26900
    },
    {
      "epoch": 0.8611804621983228,
      "grad_norm": 0.005171205382794142,
      "learning_rate": 0.07259174721390699,
      "loss": 0.4645,
      "num_input_tokens_seen": 37000784,
      "step": 26905
    },
    {
      "epoch": 0.8613405031688112,
      "grad_norm": 0.0036650034599006176,
      "learning_rate": 0.07254129790379686,
      "loss": 0.575,
      "num_input_tokens_seen": 37007680,
      "step": 26910
    },
    {
      "epoch": 0.8615005441392997,
      "grad_norm": 0.004900180269032717,
      "learning_rate": 0.072490860538792,
      "loss": 0.7331,
      "num_input_tokens_seen": 37014384,
      "step": 26915
    },
    {
      "epoch": 0.8616605851097882,
      "grad_norm": 0.00833551399409771,
      "learning_rate": 0.07244043512667042,
      "loss": 0.5457,
      "num_input_tokens_seen": 37021408,
      "step": 26920
    },
    {
      "epoch": 0.8618206260802765,
      "grad_norm": 0.004272268619388342,
      "learning_rate": 0.07239002167520843,
      "loss": 0.4524,
      "num_input_tokens_seen": 37028176,
      "step": 26925
    },
    {
      "epoch": 0.861980667050765,
      "grad_norm": 0.003507244400680065,
      "learning_rate": 0.07233962019218045,
      "loss": 0.5795,
      "num_input_tokens_seen": 37034848,
      "step": 26930
    },
    {
      "epoch": 0.8621407080212534,
      "grad_norm": 0.002444981364533305,
      "learning_rate": 0.07228923068535892,
      "loss": 0.5123,
      "num_input_tokens_seen": 37042080,
      "step": 26935
    },
    {
      "epoch": 0.8623007489917419,
      "grad_norm": 0.005451084114611149,
      "learning_rate": 0.0722388531625146,
      "loss": 0.4681,
      "num_input_tokens_seen": 37049376,
      "step": 26940
    },
    {
      "epoch": 0.8624607899622303,
      "grad_norm": 0.0034254577476531267,
      "learning_rate": 0.07218848763141639,
      "loss": 0.5392,
      "num_input_tokens_seen": 37056576,
      "step": 26945
    },
    {
      "epoch": 0.8626208309327188,
      "grad_norm": 0.004493013955652714,
      "learning_rate": 0.07213813409983118,
      "loss": 0.5514,
      "num_input_tokens_seen": 37063520,
      "step": 26950
    },
    {
      "epoch": 0.8627808719032072,
      "grad_norm": 0.0029111127369105816,
      "learning_rate": 0.0720877925755242,
      "loss": 0.5104,
      "num_input_tokens_seen": 37070560,
      "step": 26955
    },
    {
      "epoch": 0.8629409128736957,
      "grad_norm": 0.005492005962878466,
      "learning_rate": 0.07203746306625866,
      "loss": 0.4839,
      "num_input_tokens_seen": 37077200,
      "step": 26960
    },
    {
      "epoch": 0.8631009538441841,
      "grad_norm": 0.007329805754125118,
      "learning_rate": 0.07198714557979606,
      "loss": 0.5813,
      "num_input_tokens_seen": 37084272,
      "step": 26965
    },
    {
      "epoch": 0.8632609948146726,
      "grad_norm": 0.003850972978398204,
      "learning_rate": 0.07193684012389602,
      "loss": 0.4654,
      "num_input_tokens_seen": 37090864,
      "step": 26970
    },
    {
      "epoch": 0.863421035785161,
      "grad_norm": 0.005505235865712166,
      "learning_rate": 0.07188654670631621,
      "loss": 0.5163,
      "num_input_tokens_seen": 37097712,
      "step": 26975
    },
    {
      "epoch": 0.8635810767556494,
      "grad_norm": 0.003818335710093379,
      "learning_rate": 0.07183626533481258,
      "loss": 0.4684,
      "num_input_tokens_seen": 37104736,
      "step": 26980
    },
    {
      "epoch": 0.8637411177261379,
      "grad_norm": 0.004146234132349491,
      "learning_rate": 0.07178599601713909,
      "loss": 0.5316,
      "num_input_tokens_seen": 37112064,
      "step": 26985
    },
    {
      "epoch": 0.8639011586966263,
      "grad_norm": 0.004166514612734318,
      "learning_rate": 0.07173573876104786,
      "loss": 0.5441,
      "num_input_tokens_seen": 37118976,
      "step": 26990
    },
    {
      "epoch": 0.8640611996671148,
      "grad_norm": 0.007706628646701574,
      "learning_rate": 0.0716854935742893,
      "loss": 0.5542,
      "num_input_tokens_seen": 37125920,
      "step": 26995
    },
    {
      "epoch": 0.8642212406376032,
      "grad_norm": 0.0033014409709721804,
      "learning_rate": 0.07163526046461174,
      "loss": 0.616,
      "num_input_tokens_seen": 37133072,
      "step": 27000
    },
    {
      "epoch": 0.8642212406376032,
      "eval_loss": 0.5133365392684937,
      "eval_runtime": 331.8996,
      "eval_samples_per_second": 41.838,
      "eval_steps_per_second": 20.919,
      "num_input_tokens_seen": 37133072,
      "step": 27000
    },
    {
      "epoch": 0.8643812816080917,
      "grad_norm": 0.00444735586643219,
      "learning_rate": 0.07158503943976181,
      "loss": 0.595,
      "num_input_tokens_seen": 37139632,
      "step": 27005
    },
    {
      "epoch": 0.8645413225785801,
      "grad_norm": 0.006885899230837822,
      "learning_rate": 0.07153483050748427,
      "loss": 0.483,
      "num_input_tokens_seen": 37146080,
      "step": 27010
    },
    {
      "epoch": 0.8647013635490686,
      "grad_norm": 0.004089316353201866,
      "learning_rate": 0.07148463367552188,
      "loss": 0.4232,
      "num_input_tokens_seen": 37152576,
      "step": 27015
    },
    {
      "epoch": 0.864861404519557,
      "grad_norm": 0.0037164429668337107,
      "learning_rate": 0.07143444895161565,
      "loss": 0.4583,
      "num_input_tokens_seen": 37159408,
      "step": 27020
    },
    {
      "epoch": 0.8650214454900454,
      "grad_norm": 0.004684323910623789,
      "learning_rate": 0.07138427634350476,
      "loss": 0.3908,
      "num_input_tokens_seen": 37166432,
      "step": 27025
    },
    {
      "epoch": 0.8651814864605339,
      "grad_norm": 0.0015872186049818993,
      "learning_rate": 0.07133411585892636,
      "loss": 0.4044,
      "num_input_tokens_seen": 37173392,
      "step": 27030
    },
    {
      "epoch": 0.8653415274310223,
      "grad_norm": 0.003451008815318346,
      "learning_rate": 0.07128396750561593,
      "loss": 0.3878,
      "num_input_tokens_seen": 37180320,
      "step": 27035
    },
    {
      "epoch": 0.8655015684015108,
      "grad_norm": 0.004297584295272827,
      "learning_rate": 0.07123383129130685,
      "loss": 0.567,
      "num_input_tokens_seen": 37186816,
      "step": 27040
    },
    {
      "epoch": 0.8656616093719992,
      "grad_norm": 0.004556905943900347,
      "learning_rate": 0.07118370722373084,
      "loss": 0.4425,
      "num_input_tokens_seen": 37193600,
      "step": 27045
    },
    {
      "epoch": 0.8658216503424877,
      "grad_norm": 0.004748350940644741,
      "learning_rate": 0.07113359531061769,
      "loss": 0.4006,
      "num_input_tokens_seen": 37200656,
      "step": 27050
    },
    {
      "epoch": 0.8659816913129761,
      "grad_norm": 0.003034368623048067,
      "learning_rate": 0.07108349555969525,
      "loss": 0.468,
      "num_input_tokens_seen": 37207424,
      "step": 27055
    },
    {
      "epoch": 0.8661417322834646,
      "grad_norm": 0.005233824718743563,
      "learning_rate": 0.07103340797868944,
      "loss": 0.4688,
      "num_input_tokens_seen": 37214208,
      "step": 27060
    },
    {
      "epoch": 0.866301773253953,
      "grad_norm": 0.0035434553865343332,
      "learning_rate": 0.07098333257532453,
      "loss": 0.4603,
      "num_input_tokens_seen": 37220752,
      "step": 27065
    },
    {
      "epoch": 0.8664618142244415,
      "grad_norm": 0.003581432392820716,
      "learning_rate": 0.07093326935732269,
      "loss": 0.3996,
      "num_input_tokens_seen": 37227424,
      "step": 27070
    },
    {
      "epoch": 0.86662185519493,
      "grad_norm": 0.023144571110606194,
      "learning_rate": 0.0708832183324044,
      "loss": 0.3877,
      "num_input_tokens_seen": 37234336,
      "step": 27075
    },
    {
      "epoch": 0.8667818961654183,
      "grad_norm": 0.004684719257056713,
      "learning_rate": 0.07083317950828799,
      "loss": 0.513,
      "num_input_tokens_seen": 37241024,
      "step": 27080
    },
    {
      "epoch": 0.8669419371359068,
      "grad_norm": 0.005930344574153423,
      "learning_rate": 0.0707831528926902,
      "loss": 0.4158,
      "num_input_tokens_seen": 37247632,
      "step": 27085
    },
    {
      "epoch": 0.8671019781063952,
      "grad_norm": 0.005437242332845926,
      "learning_rate": 0.07073313849332578,
      "loss": 0.4751,
      "num_input_tokens_seen": 37254320,
      "step": 27090
    },
    {
      "epoch": 0.8672620190768837,
      "grad_norm": 0.0025772363878786564,
      "learning_rate": 0.07068313631790749,
      "loss": 0.4047,
      "num_input_tokens_seen": 37261712,
      "step": 27095
    },
    {
      "epoch": 0.8674220600473721,
      "grad_norm": 0.004204567521810532,
      "learning_rate": 0.07063314637414632,
      "loss": 0.529,
      "num_input_tokens_seen": 37268320,
      "step": 27100
    },
    {
      "epoch": 0.8675821010178606,
      "grad_norm": 0.005543051287531853,
      "learning_rate": 0.07058316866975144,
      "loss": 0.3731,
      "num_input_tokens_seen": 37274896,
      "step": 27105
    },
    {
      "epoch": 0.867742141988349,
      "grad_norm": 0.006475256755948067,
      "learning_rate": 0.0705332032124299,
      "loss": 0.4901,
      "num_input_tokens_seen": 37281984,
      "step": 27110
    },
    {
      "epoch": 0.8679021829588375,
      "grad_norm": 0.0018504104809835553,
      "learning_rate": 0.0704832500098871,
      "loss": 0.3338,
      "num_input_tokens_seen": 37288800,
      "step": 27115
    },
    {
      "epoch": 0.8680622239293259,
      "grad_norm": 0.006184895057231188,
      "learning_rate": 0.07043330906982641,
      "loss": 0.5671,
      "num_input_tokens_seen": 37295520,
      "step": 27120
    },
    {
      "epoch": 0.8682222648998144,
      "grad_norm": 0.006020343862473965,
      "learning_rate": 0.07038338039994936,
      "loss": 0.4625,
      "num_input_tokens_seen": 37302384,
      "step": 27125
    },
    {
      "epoch": 0.8683823058703029,
      "grad_norm": 0.005874623078852892,
      "learning_rate": 0.07033346400795562,
      "loss": 0.658,
      "num_input_tokens_seen": 37309680,
      "step": 27130
    },
    {
      "epoch": 0.8685423468407912,
      "grad_norm": 0.006423109211027622,
      "learning_rate": 0.07028355990154282,
      "loss": 0.4004,
      "num_input_tokens_seen": 37316000,
      "step": 27135
    },
    {
      "epoch": 0.8687023878112797,
      "grad_norm": 0.004387808497995138,
      "learning_rate": 0.07023366808840685,
      "loss": 0.4747,
      "num_input_tokens_seen": 37323184,
      "step": 27140
    },
    {
      "epoch": 0.8688624287817681,
      "grad_norm": 0.005384811200201511,
      "learning_rate": 0.07018378857624172,
      "loss": 0.5079,
      "num_input_tokens_seen": 37329920,
      "step": 27145
    },
    {
      "epoch": 0.8690224697522566,
      "grad_norm": 0.003010452026501298,
      "learning_rate": 0.0701339213727394,
      "loss": 0.4818,
      "num_input_tokens_seen": 37336736,
      "step": 27150
    },
    {
      "epoch": 0.869182510722745,
      "grad_norm": 0.004632956814020872,
      "learning_rate": 0.07008406648559008,
      "loss": 0.4996,
      "num_input_tokens_seen": 37343488,
      "step": 27155
    },
    {
      "epoch": 0.8693425516932335,
      "grad_norm": 0.00562205258756876,
      "learning_rate": 0.07003422392248196,
      "loss": 0.5182,
      "num_input_tokens_seen": 37350608,
      "step": 27160
    },
    {
      "epoch": 0.8695025926637219,
      "grad_norm": 0.004697566386312246,
      "learning_rate": 0.06998439369110142,
      "loss": 0.4281,
      "num_input_tokens_seen": 37357056,
      "step": 27165
    },
    {
      "epoch": 0.8696626336342104,
      "grad_norm": 0.006452847272157669,
      "learning_rate": 0.06993457579913295,
      "loss": 0.6659,
      "num_input_tokens_seen": 37363632,
      "step": 27170
    },
    {
      "epoch": 0.8698226746046988,
      "grad_norm": 0.004673555493354797,
      "learning_rate": 0.06988477025425903,
      "loss": 0.5245,
      "num_input_tokens_seen": 37370832,
      "step": 27175
    },
    {
      "epoch": 0.8699827155751872,
      "grad_norm": 0.0020568997133523226,
      "learning_rate": 0.06983497706416032,
      "loss": 0.408,
      "num_input_tokens_seen": 37377776,
      "step": 27180
    },
    {
      "epoch": 0.8701427565456757,
      "grad_norm": 0.003352740313857794,
      "learning_rate": 0.0697851962365156,
      "loss": 0.4197,
      "num_input_tokens_seen": 37384288,
      "step": 27185
    },
    {
      "epoch": 0.8703027975161641,
      "grad_norm": 0.007083093747496605,
      "learning_rate": 0.06973542777900163,
      "loss": 0.4987,
      "num_input_tokens_seen": 37391056,
      "step": 27190
    },
    {
      "epoch": 0.8704628384866526,
      "grad_norm": 0.003873917507007718,
      "learning_rate": 0.06968567169929342,
      "loss": 0.4853,
      "num_input_tokens_seen": 37397792,
      "step": 27195
    },
    {
      "epoch": 0.870622879457141,
      "grad_norm": 0.0032700931187719107,
      "learning_rate": 0.06963592800506392,
      "loss": 0.3624,
      "num_input_tokens_seen": 37404464,
      "step": 27200
    },
    {
      "epoch": 0.870622879457141,
      "eval_loss": 0.5140663981437683,
      "eval_runtime": 331.6007,
      "eval_samples_per_second": 41.876,
      "eval_steps_per_second": 20.938,
      "num_input_tokens_seen": 37404464,
      "step": 27200
    },
    {
      "epoch": 0.8707829204276295,
      "grad_norm": 0.004483099095523357,
      "learning_rate": 0.06958619670398417,
      "loss": 0.4266,
      "num_input_tokens_seen": 37411152,
      "step": 27205
    },
    {
      "epoch": 0.8709429613981179,
      "grad_norm": 0.005353831220418215,
      "learning_rate": 0.0695364778037235,
      "loss": 0.4542,
      "num_input_tokens_seen": 37418016,
      "step": 27210
    },
    {
      "epoch": 0.8711030023686064,
      "grad_norm": 0.008320700377225876,
      "learning_rate": 0.06948677131194907,
      "loss": 0.5169,
      "num_input_tokens_seen": 37425024,
      "step": 27215
    },
    {
      "epoch": 0.8712630433390948,
      "grad_norm": 0.00773592246696353,
      "learning_rate": 0.06943707723632629,
      "loss": 0.4378,
      "num_input_tokens_seen": 37432144,
      "step": 27220
    },
    {
      "epoch": 0.8714230843095833,
      "grad_norm": 0.007048585452139378,
      "learning_rate": 0.06938739558451867,
      "loss": 0.544,
      "num_input_tokens_seen": 37438640,
      "step": 27225
    },
    {
      "epoch": 0.8715831252800716,
      "grad_norm": 0.004302774555981159,
      "learning_rate": 0.06933772636418763,
      "loss": 0.4635,
      "num_input_tokens_seen": 37445808,
      "step": 27230
    },
    {
      "epoch": 0.8717431662505601,
      "grad_norm": 0.007167422212660313,
      "learning_rate": 0.06928806958299293,
      "loss": 0.5798,
      "num_input_tokens_seen": 37453040,
      "step": 27235
    },
    {
      "epoch": 0.8719032072210486,
      "grad_norm": 0.006547635421156883,
      "learning_rate": 0.06923842524859211,
      "loss": 0.5053,
      "num_input_tokens_seen": 37459920,
      "step": 27240
    },
    {
      "epoch": 0.872063248191537,
      "grad_norm": 0.004767126403748989,
      "learning_rate": 0.06918879336864105,
      "loss": 0.557,
      "num_input_tokens_seen": 37466720,
      "step": 27245
    },
    {
      "epoch": 0.8722232891620255,
      "grad_norm": 0.00431924918666482,
      "learning_rate": 0.06913917395079362,
      "loss": 0.3476,
      "num_input_tokens_seen": 37473296,
      "step": 27250
    },
    {
      "epoch": 0.8723833301325139,
      "grad_norm": 0.010228690691292286,
      "learning_rate": 0.0690895670027017,
      "loss": 0.6272,
      "num_input_tokens_seen": 37479856,
      "step": 27255
    },
    {
      "epoch": 0.8725433711030024,
      "grad_norm": 0.0037479365710169077,
      "learning_rate": 0.06903997253201531,
      "loss": 0.4963,
      "num_input_tokens_seen": 37486624,
      "step": 27260
    },
    {
      "epoch": 0.8727034120734908,
      "grad_norm": 0.005982297472655773,
      "learning_rate": 0.06899039054638263,
      "loss": 0.4947,
      "num_input_tokens_seen": 37493200,
      "step": 27265
    },
    {
      "epoch": 0.8728634530439793,
      "grad_norm": 0.004585164599120617,
      "learning_rate": 0.06894082105344976,
      "loss": 0.8008,
      "num_input_tokens_seen": 37499872,
      "step": 27270
    },
    {
      "epoch": 0.8730234940144677,
      "grad_norm": 0.0052345250733196735,
      "learning_rate": 0.06889126406086087,
      "loss": 0.5928,
      "num_input_tokens_seen": 37506768,
      "step": 27275
    },
    {
      "epoch": 0.8731835349849562,
      "grad_norm": 0.002731738146394491,
      "learning_rate": 0.0688417195762584,
      "loss": 0.5358,
      "num_input_tokens_seen": 37513712,
      "step": 27280
    },
    {
      "epoch": 0.8733435759554445,
      "grad_norm": 0.00569634884595871,
      "learning_rate": 0.06879218760728262,
      "loss": 0.4046,
      "num_input_tokens_seen": 37520560,
      "step": 27285
    },
    {
      "epoch": 0.873503616925933,
      "grad_norm": 0.0030761733651161194,
      "learning_rate": 0.06874266816157207,
      "loss": 0.4368,
      "num_input_tokens_seen": 37527184,
      "step": 27290
    },
    {
      "epoch": 0.8736636578964215,
      "grad_norm": 0.003437973791733384,
      "learning_rate": 0.06869316124676321,
      "loss": 0.4631,
      "num_input_tokens_seen": 37533840,
      "step": 27295
    },
    {
      "epoch": 0.8738236988669099,
      "grad_norm": 0.0033812453038990498,
      "learning_rate": 0.06864366687049062,
      "loss": 0.5063,
      "num_input_tokens_seen": 37540736,
      "step": 27300
    },
    {
      "epoch": 0.8739837398373984,
      "grad_norm": 0.005387371405959129,
      "learning_rate": 0.06859418504038704,
      "loss": 0.5281,
      "num_input_tokens_seen": 37547616,
      "step": 27305
    },
    {
      "epoch": 0.8741437808078868,
      "grad_norm": 0.004942051135003567,
      "learning_rate": 0.06854471576408311,
      "loss": 0.5418,
      "num_input_tokens_seen": 37554336,
      "step": 27310
    },
    {
      "epoch": 0.8743038217783753,
      "grad_norm": 0.0030679339542984962,
      "learning_rate": 0.06849525904920767,
      "loss": 0.4692,
      "num_input_tokens_seen": 37561104,
      "step": 27315
    },
    {
      "epoch": 0.8744638627488637,
      "grad_norm": 0.006992787588387728,
      "learning_rate": 0.06844581490338748,
      "loss": 0.5761,
      "num_input_tokens_seen": 37568016,
      "step": 27320
    },
    {
      "epoch": 0.8746239037193522,
      "grad_norm": 0.005636157002300024,
      "learning_rate": 0.06839638333424752,
      "loss": 0.5096,
      "num_input_tokens_seen": 37575152,
      "step": 27325
    },
    {
      "epoch": 0.8747839446898406,
      "grad_norm": 0.0059483470395207405,
      "learning_rate": 0.06834696434941082,
      "loss": 0.4211,
      "num_input_tokens_seen": 37581664,
      "step": 27330
    },
    {
      "epoch": 0.874943985660329,
      "grad_norm": 0.0032917982898652554,
      "learning_rate": 0.06829755795649824,
      "loss": 0.4418,
      "num_input_tokens_seen": 37588304,
      "step": 27335
    },
    {
      "epoch": 0.8751040266308175,
      "grad_norm": 0.00325089693069458,
      "learning_rate": 0.06824816416312904,
      "loss": 0.3776,
      "num_input_tokens_seen": 37595200,
      "step": 27340
    },
    {
      "epoch": 0.8752640676013059,
      "grad_norm": 0.0052818357944488525,
      "learning_rate": 0.06819878297692027,
      "loss": 0.5987,
      "num_input_tokens_seen": 37601792,
      "step": 27345
    },
    {
      "epoch": 0.8754241085717944,
      "grad_norm": 0.0032723662443459034,
      "learning_rate": 0.0681494144054871,
      "loss": 0.4171,
      "num_input_tokens_seen": 37608544,
      "step": 27350
    },
    {
      "epoch": 0.8755841495422828,
      "grad_norm": 0.003611738560721278,
      "learning_rate": 0.06810005845644286,
      "loss": 0.5488,
      "num_input_tokens_seen": 37615264,
      "step": 27355
    },
    {
      "epoch": 0.8757441905127713,
      "grad_norm": 0.0062732701189816,
      "learning_rate": 0.06805071513739878,
      "loss": 0.4961,
      "num_input_tokens_seen": 37621808,
      "step": 27360
    },
    {
      "epoch": 0.8759042314832597,
      "grad_norm": 0.004675328731536865,
      "learning_rate": 0.06800138445596428,
      "loss": 0.4303,
      "num_input_tokens_seen": 37628368,
      "step": 27365
    },
    {
      "epoch": 0.8760642724537482,
      "grad_norm": 0.0059192366898059845,
      "learning_rate": 0.06795206641974678,
      "loss": 0.581,
      "num_input_tokens_seen": 37635200,
      "step": 27370
    },
    {
      "epoch": 0.8762243134242366,
      "grad_norm": 0.005811857059597969,
      "learning_rate": 0.06790276103635169,
      "loss": 0.6133,
      "num_input_tokens_seen": 37641648,
      "step": 27375
    },
    {
      "epoch": 0.8763843543947251,
      "grad_norm": 0.004081714432686567,
      "learning_rate": 0.0678534683133826,
      "loss": 0.4867,
      "num_input_tokens_seen": 37648640,
      "step": 27380
    },
    {
      "epoch": 0.8765443953652134,
      "grad_norm": 0.0042199199087917805,
      "learning_rate": 0.06780418825844095,
      "loss": 0.6243,
      "num_input_tokens_seen": 37655296,
      "step": 27385
    },
    {
      "epoch": 0.8767044363357019,
      "grad_norm": 0.0032463879324495792,
      "learning_rate": 0.0677549208791264,
      "loss": 0.4838,
      "num_input_tokens_seen": 37662032,
      "step": 27390
    },
    {
      "epoch": 0.8768644773061904,
      "grad_norm": 0.011012490838766098,
      "learning_rate": 0.06770566618303668,
      "loss": 0.6002,
      "num_input_tokens_seen": 37668672,
      "step": 27395
    },
    {
      "epoch": 0.8770245182766788,
      "grad_norm": 0.006612598430365324,
      "learning_rate": 0.06765642417776736,
      "loss": 0.5559,
      "num_input_tokens_seen": 37675456,
      "step": 27400
    },
    {
      "epoch": 0.8770245182766788,
      "eval_loss": 0.5117436051368713,
      "eval_runtime": 332.2422,
      "eval_samples_per_second": 41.795,
      "eval_steps_per_second": 20.897,
      "num_input_tokens_seen": 37675456,
      "step": 27400
    },
    {
      "epoch": 0.8771845592471673,
      "grad_norm": 0.003159747924655676,
      "learning_rate": 0.0676071948709122,
      "loss": 0.5227,
      "num_input_tokens_seen": 37682624,
      "step": 27405
    },
    {
      "epoch": 0.8773446002176557,
      "grad_norm": 0.0049636284820735455,
      "learning_rate": 0.06755797827006307,
      "loss": 0.6191,
      "num_input_tokens_seen": 37689552,
      "step": 27410
    },
    {
      "epoch": 0.8775046411881442,
      "grad_norm": 0.0030485268216580153,
      "learning_rate": 0.06750877438280974,
      "loss": 0.3599,
      "num_input_tokens_seen": 37696400,
      "step": 27415
    },
    {
      "epoch": 0.8776646821586326,
      "grad_norm": 0.005464312620460987,
      "learning_rate": 0.06745958321673998,
      "loss": 0.6722,
      "num_input_tokens_seen": 37703168,
      "step": 27420
    },
    {
      "epoch": 0.8778247231291211,
      "grad_norm": 0.0035258913412690163,
      "learning_rate": 0.0674104047794398,
      "loss": 0.4418,
      "num_input_tokens_seen": 37710160,
      "step": 27425
    },
    {
      "epoch": 0.8779847640996095,
      "grad_norm": 0.008255787193775177,
      "learning_rate": 0.06736123907849303,
      "loss": 0.7185,
      "num_input_tokens_seen": 37717200,
      "step": 27430
    },
    {
      "epoch": 0.878144805070098,
      "grad_norm": 0.0042372578755021095,
      "learning_rate": 0.06731208612148178,
      "loss": 0.5181,
      "num_input_tokens_seen": 37724240,
      "step": 27435
    },
    {
      "epoch": 0.8783048460405863,
      "grad_norm": 0.0029720868915319443,
      "learning_rate": 0.0672629459159859,
      "loss": 0.6973,
      "num_input_tokens_seen": 37731376,
      "step": 27440
    },
    {
      "epoch": 0.8784648870110748,
      "grad_norm": 0.003933004569262266,
      "learning_rate": 0.0672138184695835,
      "loss": 0.4099,
      "num_input_tokens_seen": 37738384,
      "step": 27445
    },
    {
      "epoch": 0.8786249279815633,
      "grad_norm": 0.0030778569635003805,
      "learning_rate": 0.0671647037898507,
      "loss": 0.4366,
      "num_input_tokens_seen": 37745136,
      "step": 27450
    },
    {
      "epoch": 0.8787849689520517,
      "grad_norm": 0.004074443131685257,
      "learning_rate": 0.0671156018843615,
      "loss": 0.609,
      "num_input_tokens_seen": 37752064,
      "step": 27455
    },
    {
      "epoch": 0.8789450099225402,
      "grad_norm": 0.0028876117430627346,
      "learning_rate": 0.06706651276068812,
      "loss": 0.521,
      "num_input_tokens_seen": 37758848,
      "step": 27460
    },
    {
      "epoch": 0.8791050508930286,
      "grad_norm": 0.007112839724868536,
      "learning_rate": 0.06701743642640064,
      "loss": 0.5287,
      "num_input_tokens_seen": 37765904,
      "step": 27465
    },
    {
      "epoch": 0.8792650918635171,
      "grad_norm": 0.0027936489786952734,
      "learning_rate": 0.06696837288906729,
      "loss": 0.4116,
      "num_input_tokens_seen": 37772608,
      "step": 27470
    },
    {
      "epoch": 0.8794251328340055,
      "grad_norm": 0.01729479804635048,
      "learning_rate": 0.06691932215625432,
      "loss": 0.5019,
      "num_input_tokens_seen": 37779792,
      "step": 27475
    },
    {
      "epoch": 0.879585173804494,
      "grad_norm": 0.002654806710779667,
      "learning_rate": 0.06687028423552589,
      "loss": 0.6105,
      "num_input_tokens_seen": 37786864,
      "step": 27480
    },
    {
      "epoch": 0.8797452147749824,
      "grad_norm": 0.002724181627854705,
      "learning_rate": 0.06682125913444435,
      "loss": 0.5846,
      "num_input_tokens_seen": 37793680,
      "step": 27485
    },
    {
      "epoch": 0.8799052557454708,
      "grad_norm": 0.0030355043709278107,
      "learning_rate": 0.0667722468605699,
      "loss": 0.5087,
      "num_input_tokens_seen": 37800192,
      "step": 27490
    },
    {
      "epoch": 0.8800652967159592,
      "grad_norm": 0.0020034699700772762,
      "learning_rate": 0.06672324742146094,
      "loss": 0.4237,
      "num_input_tokens_seen": 37806736,
      "step": 27495
    },
    {
      "epoch": 0.8802253376864477,
      "grad_norm": 0.00833948701620102,
      "learning_rate": 0.06667426082467373,
      "loss": 0.4749,
      "num_input_tokens_seen": 37813168,
      "step": 27500
    },
    {
      "epoch": 0.8803853786569362,
      "grad_norm": 0.00311453640460968,
      "learning_rate": 0.0666252870777626,
      "loss": 0.3747,
      "num_input_tokens_seen": 37820352,
      "step": 27505
    },
    {
      "epoch": 0.8805454196274246,
      "grad_norm": 0.003652189625427127,
      "learning_rate": 0.06657632618827995,
      "loss": 0.4326,
      "num_input_tokens_seen": 37827184,
      "step": 27510
    },
    {
      "epoch": 0.8807054605979131,
      "grad_norm": 0.0034327609464526176,
      "learning_rate": 0.06652737816377623,
      "loss": 0.4607,
      "num_input_tokens_seen": 37834208,
      "step": 27515
    },
    {
      "epoch": 0.8808655015684015,
      "grad_norm": 0.005394717678427696,
      "learning_rate": 0.06647844301179971,
      "loss": 0.4828,
      "num_input_tokens_seen": 37840928,
      "step": 27520
    },
    {
      "epoch": 0.88102554253889,
      "grad_norm": 0.0047437893226742744,
      "learning_rate": 0.06642952073989689,
      "loss": 0.5274,
      "num_input_tokens_seen": 37847696,
      "step": 27525
    },
    {
      "epoch": 0.8811855835093784,
      "grad_norm": 0.0036170959938317537,
      "learning_rate": 0.06638061135561223,
      "loss": 0.3545,
      "num_input_tokens_seen": 37854656,
      "step": 27530
    },
    {
      "epoch": 0.8813456244798669,
      "grad_norm": 0.008847699500620365,
      "learning_rate": 0.06633171486648808,
      "loss": 0.5391,
      "num_input_tokens_seen": 37861360,
      "step": 27535
    },
    {
      "epoch": 0.8815056654503552,
      "grad_norm": 0.006604545284062624,
      "learning_rate": 0.06628283128006499,
      "loss": 0.4266,
      "num_input_tokens_seen": 37869264,
      "step": 27540
    },
    {
      "epoch": 0.8816657064208437,
      "grad_norm": 0.008438008837401867,
      "learning_rate": 0.0662339606038813,
      "loss": 0.5285,
      "num_input_tokens_seen": 37876128,
      "step": 27545
    },
    {
      "epoch": 0.8818257473913321,
      "grad_norm": 0.004358770791441202,
      "learning_rate": 0.06618510284547358,
      "loss": 0.4724,
      "num_input_tokens_seen": 37882592,
      "step": 27550
    },
    {
      "epoch": 0.8819857883618206,
      "grad_norm": 0.0029081387910991907,
      "learning_rate": 0.06613625801237633,
      "loss": 0.5181,
      "num_input_tokens_seen": 37889664,
      "step": 27555
    },
    {
      "epoch": 0.8821458293323091,
      "grad_norm": 0.004463063552975655,
      "learning_rate": 0.066087426112122,
      "loss": 0.4766,
      "num_input_tokens_seen": 37896464,
      "step": 27560
    },
    {
      "epoch": 0.8823058703027975,
      "grad_norm": 0.011500423774123192,
      "learning_rate": 0.06603860715224101,
      "loss": 0.4789,
      "num_input_tokens_seen": 37903712,
      "step": 27565
    },
    {
      "epoch": 0.882465911273286,
      "grad_norm": 0.00471881777048111,
      "learning_rate": 0.06598980114026198,
      "loss": 0.6951,
      "num_input_tokens_seen": 37910640,
      "step": 27570
    },
    {
      "epoch": 0.8826259522437744,
      "grad_norm": 0.010056703351438046,
      "learning_rate": 0.06594100808371128,
      "loss": 0.5099,
      "num_input_tokens_seen": 37917536,
      "step": 27575
    },
    {
      "epoch": 0.8827859932142629,
      "grad_norm": 0.005126235540956259,
      "learning_rate": 0.06589222799011357,
      "loss": 0.62,
      "num_input_tokens_seen": 37924128,
      "step": 27580
    },
    {
      "epoch": 0.8829460341847513,
      "grad_norm": 0.0036950288340449333,
      "learning_rate": 0.0658434608669912,
      "loss": 0.5182,
      "num_input_tokens_seen": 37930848,
      "step": 27585
    },
    {
      "epoch": 0.8831060751552398,
      "grad_norm": 0.003752041608095169,
      "learning_rate": 0.06579470672186473,
      "loss": 0.5652,
      "num_input_tokens_seen": 37937792,
      "step": 27590
    },
    {
      "epoch": 0.8832661161257281,
      "grad_norm": 0.006341902539134026,
      "learning_rate": 0.06574596556225275,
      "loss": 0.4234,
      "num_input_tokens_seen": 37944720,
      "step": 27595
    },
    {
      "epoch": 0.8834261570962166,
      "grad_norm": 0.0036020560655742884,
      "learning_rate": 0.06569723739567161,
      "loss": 0.4689,
      "num_input_tokens_seen": 37951616,
      "step": 27600
    },
    {
      "epoch": 0.8834261570962166,
      "eval_loss": 0.5122509002685547,
      "eval_runtime": 331.5921,
      "eval_samples_per_second": 41.877,
      "eval_steps_per_second": 20.938,
      "num_input_tokens_seen": 37951616,
      "step": 27600
    },
    {
      "epoch": 0.8835861980667051,
      "grad_norm": 0.00494958832859993,
      "learning_rate": 0.06564852222963588,
      "loss": 0.6366,
      "num_input_tokens_seen": 37958592,
      "step": 27605
    },
    {
      "epoch": 0.8837462390371935,
      "grad_norm": 0.0030083036981523037,
      "learning_rate": 0.06559982007165813,
      "loss": 0.4493,
      "num_input_tokens_seen": 37965488,
      "step": 27610
    },
    {
      "epoch": 0.883906280007682,
      "grad_norm": 0.006387231405824423,
      "learning_rate": 0.06555113092924868,
      "loss": 0.6382,
      "num_input_tokens_seen": 37972304,
      "step": 27615
    },
    {
      "epoch": 0.8840663209781704,
      "grad_norm": 0.008601360023021698,
      "learning_rate": 0.06550245480991615,
      "loss": 0.6048,
      "num_input_tokens_seen": 37979328,
      "step": 27620
    },
    {
      "epoch": 0.8842263619486589,
      "grad_norm": 0.0038087130524218082,
      "learning_rate": 0.0654537917211669,
      "loss": 0.4929,
      "num_input_tokens_seen": 37985696,
      "step": 27625
    },
    {
      "epoch": 0.8843864029191473,
      "grad_norm": 0.003019978292286396,
      "learning_rate": 0.0654051416705055,
      "loss": 0.5414,
      "num_input_tokens_seen": 37992592,
      "step": 27630
    },
    {
      "epoch": 0.8845464438896358,
      "grad_norm": 0.004923567175865173,
      "learning_rate": 0.06535650466543427,
      "loss": 0.531,
      "num_input_tokens_seen": 37998720,
      "step": 27635
    },
    {
      "epoch": 0.8847064848601242,
      "grad_norm": 0.004633017815649509,
      "learning_rate": 0.0653078807134538,
      "loss": 0.4636,
      "num_input_tokens_seen": 38005328,
      "step": 27640
    },
    {
      "epoch": 0.8848665258306126,
      "grad_norm": 0.008350727148354053,
      "learning_rate": 0.06525926982206236,
      "loss": 0.5034,
      "num_input_tokens_seen": 38012512,
      "step": 27645
    },
    {
      "epoch": 0.885026566801101,
      "grad_norm": 0.004563808441162109,
      "learning_rate": 0.06521067199875648,
      "loss": 0.5977,
      "num_input_tokens_seen": 38019248,
      "step": 27650
    },
    {
      "epoch": 0.8851866077715895,
      "grad_norm": 0.005567396059632301,
      "learning_rate": 0.06516208725103047,
      "loss": 0.4957,
      "num_input_tokens_seen": 38026096,
      "step": 27655
    },
    {
      "epoch": 0.885346648742078,
      "grad_norm": 0.006141791120171547,
      "learning_rate": 0.06511351558637678,
      "loss": 0.4394,
      "num_input_tokens_seen": 38032832,
      "step": 27660
    },
    {
      "epoch": 0.8855066897125664,
      "grad_norm": 0.005908170714974403,
      "learning_rate": 0.06506495701228569,
      "loss": 0.5015,
      "num_input_tokens_seen": 38039488,
      "step": 27665
    },
    {
      "epoch": 0.8856667306830549,
      "grad_norm": 0.009814023040235043,
      "learning_rate": 0.06501641153624559,
      "loss": 0.5073,
      "num_input_tokens_seen": 38046752,
      "step": 27670
    },
    {
      "epoch": 0.8858267716535433,
      "grad_norm": 0.0035712621174752712,
      "learning_rate": 0.06496787916574286,
      "loss": 0.5837,
      "num_input_tokens_seen": 38053904,
      "step": 27675
    },
    {
      "epoch": 0.8859868126240318,
      "grad_norm": 0.005328933708369732,
      "learning_rate": 0.06491935990826168,
      "loss": 0.5442,
      "num_input_tokens_seen": 38060992,
      "step": 27680
    },
    {
      "epoch": 0.8861468535945202,
      "grad_norm": 0.003846638137474656,
      "learning_rate": 0.0648708537712844,
      "loss": 0.3915,
      "num_input_tokens_seen": 38068048,
      "step": 27685
    },
    {
      "epoch": 0.8863068945650087,
      "grad_norm": 0.00562066026031971,
      "learning_rate": 0.06482236076229132,
      "loss": 0.5408,
      "num_input_tokens_seen": 38075040,
      "step": 27690
    },
    {
      "epoch": 0.886466935535497,
      "grad_norm": 0.008134579285979271,
      "learning_rate": 0.06477388088876056,
      "loss": 0.4208,
      "num_input_tokens_seen": 38081616,
      "step": 27695
    },
    {
      "epoch": 0.8866269765059855,
      "grad_norm": 0.004628532100468874,
      "learning_rate": 0.06472541415816846,
      "loss": 0.5613,
      "num_input_tokens_seen": 38088848,
      "step": 27700
    },
    {
      "epoch": 0.8867870174764739,
      "grad_norm": 0.005806329660117626,
      "learning_rate": 0.06467696057798909,
      "loss": 0.5163,
      "num_input_tokens_seen": 38095920,
      "step": 27705
    },
    {
      "epoch": 0.8869470584469624,
      "grad_norm": 0.005794523283839226,
      "learning_rate": 0.0646285201556946,
      "loss": 0.5066,
      "num_input_tokens_seen": 38103552,
      "step": 27710
    },
    {
      "epoch": 0.8871070994174509,
      "grad_norm": 0.005768493749201298,
      "learning_rate": 0.06458009289875521,
      "loss": 0.5678,
      "num_input_tokens_seen": 38110336,
      "step": 27715
    },
    {
      "epoch": 0.8872671403879393,
      "grad_norm": 0.0071105994284152985,
      "learning_rate": 0.0645316788146389,
      "loss": 0.678,
      "num_input_tokens_seen": 38116960,
      "step": 27720
    },
    {
      "epoch": 0.8874271813584278,
      "grad_norm": 0.0022880949545651674,
      "learning_rate": 0.06448327791081175,
      "loss": 0.4603,
      "num_input_tokens_seen": 38123744,
      "step": 27725
    },
    {
      "epoch": 0.8875872223289162,
      "grad_norm": 0.0035081906244158745,
      "learning_rate": 0.0644348901947379,
      "loss": 0.5271,
      "num_input_tokens_seen": 38130560,
      "step": 27730
    },
    {
      "epoch": 0.8877472632994047,
      "grad_norm": 0.00429524714127183,
      "learning_rate": 0.06438651567387917,
      "loss": 0.5734,
      "num_input_tokens_seen": 38137584,
      "step": 27735
    },
    {
      "epoch": 0.8879073042698931,
      "grad_norm": 0.004047325812280178,
      "learning_rate": 0.0643381543556957,
      "loss": 0.3595,
      "num_input_tokens_seen": 38144720,
      "step": 27740
    },
    {
      "epoch": 0.8880673452403816,
      "grad_norm": 0.0024583733174949884,
      "learning_rate": 0.06428980624764526,
      "loss": 0.3804,
      "num_input_tokens_seen": 38152032,
      "step": 27745
    },
    {
      "epoch": 0.8882273862108699,
      "grad_norm": 0.006498812232166529,
      "learning_rate": 0.06424147135718378,
      "loss": 0.442,
      "num_input_tokens_seen": 38158848,
      "step": 27750
    },
    {
      "epoch": 0.8883874271813584,
      "grad_norm": 0.006097822915762663,
      "learning_rate": 0.06419314969176519,
      "loss": 0.539,
      "num_input_tokens_seen": 38165568,
      "step": 27755
    },
    {
      "epoch": 0.8885474681518468,
      "grad_norm": 0.0026530008763074875,
      "learning_rate": 0.06414484125884118,
      "loss": 0.4538,
      "num_input_tokens_seen": 38172112,
      "step": 27760
    },
    {
      "epoch": 0.8887075091223353,
      "grad_norm": 0.009077442809939384,
      "learning_rate": 0.06409654606586157,
      "loss": 0.5051,
      "num_input_tokens_seen": 38178576,
      "step": 27765
    },
    {
      "epoch": 0.8888675500928238,
      "grad_norm": 0.011590363457798958,
      "learning_rate": 0.06404826412027415,
      "loss": 0.5121,
      "num_input_tokens_seen": 38185488,
      "step": 27770
    },
    {
      "epoch": 0.8890275910633122,
      "grad_norm": 0.00344660016708076,
      "learning_rate": 0.06399999542952453,
      "loss": 0.4379,
      "num_input_tokens_seen": 38191872,
      "step": 27775
    },
    {
      "epoch": 0.8891876320338007,
      "grad_norm": 0.006874872371554375,
      "learning_rate": 0.0639517400010563,
      "loss": 0.5826,
      "num_input_tokens_seen": 38198320,
      "step": 27780
    },
    {
      "epoch": 0.8893476730042891,
      "grad_norm": 0.00554203474894166,
      "learning_rate": 0.06390349784231118,
      "loss": 0.5373,
      "num_input_tokens_seen": 38205040,
      "step": 27785
    },
    {
      "epoch": 0.8895077139747776,
      "grad_norm": 0.003817951073870063,
      "learning_rate": 0.06385526896072859,
      "loss": 0.6677,
      "num_input_tokens_seen": 38212400,
      "step": 27790
    },
    {
      "epoch": 0.889667754945266,
      "grad_norm": 0.003747663926333189,
      "learning_rate": 0.06380705336374613,
      "loss": 0.6688,
      "num_input_tokens_seen": 38219232,
      "step": 27795
    },
    {
      "epoch": 0.8898277959157544,
      "grad_norm": 0.005224922671914101,
      "learning_rate": 0.06375885105879918,
      "loss": 0.5883,
      "num_input_tokens_seen": 38225840,
      "step": 27800
    },
    {
      "epoch": 0.8898277959157544,
      "eval_loss": 0.509453296661377,
      "eval_runtime": 332.2081,
      "eval_samples_per_second": 41.799,
      "eval_steps_per_second": 20.9,
      "num_input_tokens_seen": 38225840,
      "step": 27800
    },
    {
      "epoch": 0.8899878368862428,
      "grad_norm": 0.0053269341588020325,
      "learning_rate": 0.06371066205332115,
      "loss": 0.3325,
      "num_input_tokens_seen": 38232304,
      "step": 27805
    },
    {
      "epoch": 0.8901478778567313,
      "grad_norm": 0.003673740429803729,
      "learning_rate": 0.06366248635474347,
      "loss": 0.5445,
      "num_input_tokens_seen": 38238848,
      "step": 27810
    },
    {
      "epoch": 0.8903079188272197,
      "grad_norm": 0.010311811231076717,
      "learning_rate": 0.06361432397049532,
      "loss": 0.4792,
      "num_input_tokens_seen": 38246048,
      "step": 27815
    },
    {
      "epoch": 0.8904679597977082,
      "grad_norm": 0.010040816850960255,
      "learning_rate": 0.06356617490800408,
      "loss": 0.627,
      "num_input_tokens_seen": 38252768,
      "step": 27820
    },
    {
      "epoch": 0.8906280007681967,
      "grad_norm": 0.006239621434360743,
      "learning_rate": 0.06351803917469478,
      "loss": 0.8102,
      "num_input_tokens_seen": 38259504,
      "step": 27825
    },
    {
      "epoch": 0.8907880417386851,
      "grad_norm": 0.002413753420114517,
      "learning_rate": 0.06346991677799067,
      "loss": 0.41,
      "num_input_tokens_seen": 38266192,
      "step": 27830
    },
    {
      "epoch": 0.8909480827091736,
      "grad_norm": 0.0068312156945466995,
      "learning_rate": 0.06342180772531283,
      "loss": 0.4261,
      "num_input_tokens_seen": 38272928,
      "step": 27835
    },
    {
      "epoch": 0.891108123679662,
      "grad_norm": 0.005317605100572109,
      "learning_rate": 0.06337371202408021,
      "loss": 0.6423,
      "num_input_tokens_seen": 38279616,
      "step": 27840
    },
    {
      "epoch": 0.8912681646501505,
      "grad_norm": 0.003147427225485444,
      "learning_rate": 0.06332562968170984,
      "loss": 0.4827,
      "num_input_tokens_seen": 38286672,
      "step": 27845
    },
    {
      "epoch": 0.8914282056206388,
      "grad_norm": 0.008170399814844131,
      "learning_rate": 0.06327756070561656,
      "loss": 0.8558,
      "num_input_tokens_seen": 38294016,
      "step": 27850
    },
    {
      "epoch": 0.8915882465911273,
      "grad_norm": 0.005651840008795261,
      "learning_rate": 0.06322950510321329,
      "loss": 0.5348,
      "num_input_tokens_seen": 38300432,
      "step": 27855
    },
    {
      "epoch": 0.8917482875616157,
      "grad_norm": 0.004322609398514032,
      "learning_rate": 0.06318146288191076,
      "loss": 0.5689,
      "num_input_tokens_seen": 38307232,
      "step": 27860
    },
    {
      "epoch": 0.8919083285321042,
      "grad_norm": 0.0052107106894254684,
      "learning_rate": 0.06313343404911763,
      "loss": 0.4857,
      "num_input_tokens_seen": 38314048,
      "step": 27865
    },
    {
      "epoch": 0.8920683695025927,
      "grad_norm": 0.0032943657133728266,
      "learning_rate": 0.0630854186122406,
      "loss": 0.3654,
      "num_input_tokens_seen": 38320480,
      "step": 27870
    },
    {
      "epoch": 0.8922284104730811,
      "grad_norm": 0.005796919111162424,
      "learning_rate": 0.06303741657868431,
      "loss": 0.4219,
      "num_input_tokens_seen": 38327184,
      "step": 27875
    },
    {
      "epoch": 0.8923884514435696,
      "grad_norm": 0.0030904975719749928,
      "learning_rate": 0.06298942795585115,
      "loss": 0.4458,
      "num_input_tokens_seen": 38334128,
      "step": 27880
    },
    {
      "epoch": 0.892548492414058,
      "grad_norm": 0.004323871806263924,
      "learning_rate": 0.06294145275114167,
      "loss": 0.3718,
      "num_input_tokens_seen": 38340896,
      "step": 27885
    },
    {
      "epoch": 0.8927085333845465,
      "grad_norm": 0.003908188082277775,
      "learning_rate": 0.06289349097195428,
      "loss": 0.5139,
      "num_input_tokens_seen": 38347888,
      "step": 27890
    },
    {
      "epoch": 0.8928685743550349,
      "grad_norm": 0.008190911263227463,
      "learning_rate": 0.06284554262568516,
      "loss": 0.4815,
      "num_input_tokens_seen": 38354640,
      "step": 27895
    },
    {
      "epoch": 0.8930286153255234,
      "grad_norm": 0.00502841267734766,
      "learning_rate": 0.06279760771972868,
      "loss": 0.5994,
      "num_input_tokens_seen": 38361568,
      "step": 27900
    },
    {
      "epoch": 0.8931886562960117,
      "grad_norm": 0.0058253249153494835,
      "learning_rate": 0.06274968626147688,
      "loss": 0.5908,
      "num_input_tokens_seen": 38368608,
      "step": 27905
    },
    {
      "epoch": 0.8933486972665002,
      "grad_norm": 0.005438762716948986,
      "learning_rate": 0.06270177825831993,
      "loss": 0.6312,
      "num_input_tokens_seen": 38375424,
      "step": 27910
    },
    {
      "epoch": 0.8935087382369886,
      "grad_norm": 0.009575128555297852,
      "learning_rate": 0.06265388371764587,
      "loss": 0.5835,
      "num_input_tokens_seen": 38382288,
      "step": 27915
    },
    {
      "epoch": 0.8936687792074771,
      "grad_norm": 0.006552362814545631,
      "learning_rate": 0.0626060026468406,
      "loss": 0.5524,
      "num_input_tokens_seen": 38389488,
      "step": 27920
    },
    {
      "epoch": 0.8938288201779656,
      "grad_norm": 0.004644629079848528,
      "learning_rate": 0.06255813505328794,
      "loss": 0.4276,
      "num_input_tokens_seen": 38395936,
      "step": 27925
    },
    {
      "epoch": 0.893988861148454,
      "grad_norm": 0.004783387295901775,
      "learning_rate": 0.06251028094436978,
      "loss": 0.5829,
      "num_input_tokens_seen": 38403216,
      "step": 27930
    },
    {
      "epoch": 0.8941489021189425,
      "grad_norm": 0.0023840349167585373,
      "learning_rate": 0.06246244032746568,
      "loss": 0.4509,
      "num_input_tokens_seen": 38410192,
      "step": 27935
    },
    {
      "epoch": 0.8943089430894309,
      "grad_norm": 0.008180913515388966,
      "learning_rate": 0.06241461320995342,
      "loss": 0.5752,
      "num_input_tokens_seen": 38417168,
      "step": 27940
    },
    {
      "epoch": 0.8944689840599194,
      "grad_norm": 0.0033074195962399244,
      "learning_rate": 0.062366799599208426,
      "loss": 0.6649,
      "num_input_tokens_seen": 38424176,
      "step": 27945
    },
    {
      "epoch": 0.8946290250304078,
      "grad_norm": 0.005105039104819298,
      "learning_rate": 0.06231899950260418,
      "loss": 0.5824,
      "num_input_tokens_seen": 38431056,
      "step": 27950
    },
    {
      "epoch": 0.8947890660008962,
      "grad_norm": 0.005288850981742144,
      "learning_rate": 0.06227121292751214,
      "loss": 0.5931,
      "num_input_tokens_seen": 38437728,
      "step": 27955
    },
    {
      "epoch": 0.8949491069713846,
      "grad_norm": 0.006224769167602062,
      "learning_rate": 0.062223439881301496,
      "loss": 0.5137,
      "num_input_tokens_seen": 38444576,
      "step": 27960
    },
    {
      "epoch": 0.8951091479418731,
      "grad_norm": 0.007253636606037617,
      "learning_rate": 0.06217568037133948,
      "loss": 0.5741,
      "num_input_tokens_seen": 38451312,
      "step": 27965
    },
    {
      "epoch": 0.8952691889123615,
      "grad_norm": 0.0032509458251297474,
      "learning_rate": 0.06212793440499126,
      "loss": 0.539,
      "num_input_tokens_seen": 38458448,
      "step": 27970
    },
    {
      "epoch": 0.89542922988285,
      "grad_norm": 0.002545781899243593,
      "learning_rate": 0.062080201989619783,
      "loss": 0.3349,
      "num_input_tokens_seen": 38464800,
      "step": 27975
    },
    {
      "epoch": 0.8955892708533385,
      "grad_norm": 0.0036242040805518627,
      "learning_rate": 0.062032483132586094,
      "loss": 0.609,
      "num_input_tokens_seen": 38471584,
      "step": 27980
    },
    {
      "epoch": 0.8957493118238269,
      "grad_norm": 0.00545582827180624,
      "learning_rate": 0.0619847778412489,
      "loss": 0.5057,
      "num_input_tokens_seen": 38478256,
      "step": 27985
    },
    {
      "epoch": 0.8959093527943154,
      "grad_norm": 0.0071346815675497055,
      "learning_rate": 0.06193708612296509,
      "loss": 0.5318,
      "num_input_tokens_seen": 38485408,
      "step": 27990
    },
    {
      "epoch": 0.8960693937648038,
      "grad_norm": 0.0033981362357735634,
      "learning_rate": 0.06188940798508923,
      "loss": 0.5766,
      "num_input_tokens_seen": 38492144,
      "step": 27995
    },
    {
      "epoch": 0.8962294347352923,
      "grad_norm": 0.0033152548130601645,
      "learning_rate": 0.06184174343497397,
      "loss": 0.4924,
      "num_input_tokens_seen": 38498736,
      "step": 28000
    },
    {
      "epoch": 0.8962294347352923,
      "eval_loss": 0.5094183683395386,
      "eval_runtime": 332.1232,
      "eval_samples_per_second": 41.81,
      "eval_steps_per_second": 20.905,
      "num_input_tokens_seen": 38498736,
      "step": 28000
    },
    {
      "epoch": 0.8963894757057806,
      "grad_norm": 0.005301818251609802,
      "learning_rate": 0.061794092479969726,
      "loss": 0.6012,
      "num_input_tokens_seen": 38505536,
      "step": 28005
    },
    {
      "epoch": 0.8965495166762691,
      "grad_norm": 0.0036753888707607985,
      "learning_rate": 0.06174645512742485,
      "loss": 0.67,
      "num_input_tokens_seen": 38512416,
      "step": 28010
    },
    {
      "epoch": 0.8967095576467575,
      "grad_norm": 0.005109559278935194,
      "learning_rate": 0.06169883138468565,
      "loss": 0.4838,
      "num_input_tokens_seen": 38519152,
      "step": 28015
    },
    {
      "epoch": 0.896869598617246,
      "grad_norm": 0.005170960910618305,
      "learning_rate": 0.06165122125909637,
      "loss": 0.4937,
      "num_input_tokens_seen": 38526048,
      "step": 28020
    },
    {
      "epoch": 0.8970296395877344,
      "grad_norm": 0.003831163514405489,
      "learning_rate": 0.061603624757998965,
      "loss": 0.4926,
      "num_input_tokens_seen": 38532816,
      "step": 28025
    },
    {
      "epoch": 0.8971896805582229,
      "grad_norm": 0.004438104573637247,
      "learning_rate": 0.0615560418887335,
      "loss": 0.4609,
      "num_input_tokens_seen": 38539216,
      "step": 28030
    },
    {
      "epoch": 0.8973497215287114,
      "grad_norm": 0.00660707987844944,
      "learning_rate": 0.06150847265863787,
      "loss": 0.5754,
      "num_input_tokens_seen": 38546240,
      "step": 28035
    },
    {
      "epoch": 0.8975097624991998,
      "grad_norm": 0.00624790508300066,
      "learning_rate": 0.061460917075047757,
      "loss": 0.5942,
      "num_input_tokens_seen": 38552976,
      "step": 28040
    },
    {
      "epoch": 0.8976698034696883,
      "grad_norm": 0.004375270567834377,
      "learning_rate": 0.06141337514529694,
      "loss": 0.5432,
      "num_input_tokens_seen": 38560320,
      "step": 28045
    },
    {
      "epoch": 0.8978298444401767,
      "grad_norm": 0.007264274172484875,
      "learning_rate": 0.06136584687671687,
      "loss": 0.6748,
      "num_input_tokens_seen": 38567152,
      "step": 28050
    },
    {
      "epoch": 0.8979898854106652,
      "grad_norm": 0.0034570142161101103,
      "learning_rate": 0.061318332276637064,
      "loss": 0.4373,
      "num_input_tokens_seen": 38573456,
      "step": 28055
    },
    {
      "epoch": 0.8981499263811535,
      "grad_norm": 0.005991083104163408,
      "learning_rate": 0.06127083135238491,
      "loss": 0.4555,
      "num_input_tokens_seen": 38580352,
      "step": 28060
    },
    {
      "epoch": 0.898309967351642,
      "grad_norm": 0.006162109784781933,
      "learning_rate": 0.06122334411128555,
      "loss": 0.6593,
      "num_input_tokens_seen": 38586848,
      "step": 28065
    },
    {
      "epoch": 0.8984700083221304,
      "grad_norm": 0.003770333481952548,
      "learning_rate": 0.06117587056066223,
      "loss": 0.4711,
      "num_input_tokens_seen": 38594272,
      "step": 28070
    },
    {
      "epoch": 0.8986300492926189,
      "grad_norm": 0.007549569942057133,
      "learning_rate": 0.06112841070783589,
      "loss": 0.5295,
      "num_input_tokens_seen": 38600992,
      "step": 28075
    },
    {
      "epoch": 0.8987900902631074,
      "grad_norm": 0.0031266186852008104,
      "learning_rate": 0.061080964560125406,
      "loss": 0.4911,
      "num_input_tokens_seen": 38607776,
      "step": 28080
    },
    {
      "epoch": 0.8989501312335958,
      "grad_norm": 0.003066882025450468,
      "learning_rate": 0.06103353212484766,
      "loss": 0.5526,
      "num_input_tokens_seen": 38614560,
      "step": 28085
    },
    {
      "epoch": 0.8991101722040843,
      "grad_norm": 0.01141316071152687,
      "learning_rate": 0.06098611340931722,
      "loss": 0.6944,
      "num_input_tokens_seen": 38621280,
      "step": 28090
    },
    {
      "epoch": 0.8992702131745727,
      "grad_norm": 0.0037618952337652445,
      "learning_rate": 0.06093870842084672,
      "loss": 0.437,
      "num_input_tokens_seen": 38628224,
      "step": 28095
    },
    {
      "epoch": 0.8994302541450612,
      "grad_norm": 0.00536479102447629,
      "learning_rate": 0.06089131716674666,
      "loss": 0.4643,
      "num_input_tokens_seen": 38635552,
      "step": 28100
    },
    {
      "epoch": 0.8995902951155496,
      "grad_norm": 0.004404568579047918,
      "learning_rate": 0.060843939654325226,
      "loss": 0.4573,
      "num_input_tokens_seen": 38642704,
      "step": 28105
    },
    {
      "epoch": 0.899750336086038,
      "grad_norm": 0.0036618891172111034,
      "learning_rate": 0.06079657589088873,
      "loss": 0.4853,
      "num_input_tokens_seen": 38649728,
      "step": 28110
    },
    {
      "epoch": 0.8999103770565264,
      "grad_norm": 0.005412687547504902,
      "learning_rate": 0.06074922588374126,
      "loss": 0.5564,
      "num_input_tokens_seen": 38656688,
      "step": 28115
    },
    {
      "epoch": 0.9000704180270149,
      "grad_norm": 0.004267572890967131,
      "learning_rate": 0.06070188964018472,
      "loss": 0.5355,
      "num_input_tokens_seen": 38663280,
      "step": 28120
    },
    {
      "epoch": 0.9002304589975033,
      "grad_norm": 0.006408481393009424,
      "learning_rate": 0.06065456716751902,
      "loss": 0.5179,
      "num_input_tokens_seen": 38669872,
      "step": 28125
    },
    {
      "epoch": 0.9003904999679918,
      "grad_norm": 0.005408635828644037,
      "learning_rate": 0.06060725847304182,
      "loss": 0.5065,
      "num_input_tokens_seen": 38676720,
      "step": 28130
    },
    {
      "epoch": 0.9005505409384803,
      "grad_norm": 0.005353043787181377,
      "learning_rate": 0.06055996356404877,
      "loss": 0.5064,
      "num_input_tokens_seen": 38683696,
      "step": 28135
    },
    {
      "epoch": 0.9007105819089687,
      "grad_norm": 0.0038113596383482218,
      "learning_rate": 0.06051268244783327,
      "loss": 0.4147,
      "num_input_tokens_seen": 38690288,
      "step": 28140
    },
    {
      "epoch": 0.9008706228794572,
      "grad_norm": 0.004822753369808197,
      "learning_rate": 0.06046541513168676,
      "loss": 0.3847,
      "num_input_tokens_seen": 38696944,
      "step": 28145
    },
    {
      "epoch": 0.9010306638499456,
      "grad_norm": 0.004716976545751095,
      "learning_rate": 0.060418161622898356,
      "loss": 0.5609,
      "num_input_tokens_seen": 38703600,
      "step": 28150
    },
    {
      "epoch": 0.9011907048204341,
      "grad_norm": 0.007305409759283066,
      "learning_rate": 0.06037092192875521,
      "loss": 0.5199,
      "num_input_tokens_seen": 38710448,
      "step": 28155
    },
    {
      "epoch": 0.9013507457909224,
      "grad_norm": 0.0026565822772681713,
      "learning_rate": 0.060323696056542225,
      "loss": 0.3531,
      "num_input_tokens_seen": 38717024,
      "step": 28160
    },
    {
      "epoch": 0.9015107867614109,
      "grad_norm": 0.0031985174864530563,
      "learning_rate": 0.06027648401354229,
      "loss": 0.4514,
      "num_input_tokens_seen": 38724032,
      "step": 28165
    },
    {
      "epoch": 0.9016708277318993,
      "grad_norm": 0.008088653907179832,
      "learning_rate": 0.06022928580703601,
      "loss": 0.4748,
      "num_input_tokens_seen": 38731152,
      "step": 28170
    },
    {
      "epoch": 0.9018308687023878,
      "grad_norm": 0.006095502991229296,
      "learning_rate": 0.060182101444301986,
      "loss": 0.4115,
      "num_input_tokens_seen": 38738016,
      "step": 28175
    },
    {
      "epoch": 0.9019909096728762,
      "grad_norm": 0.003982992842793465,
      "learning_rate": 0.06013493093261669,
      "loss": 0.7293,
      "num_input_tokens_seen": 38744608,
      "step": 28180
    },
    {
      "epoch": 0.9021509506433647,
      "grad_norm": 0.006601257715374231,
      "learning_rate": 0.06008777427925432,
      "loss": 0.6055,
      "num_input_tokens_seen": 38751568,
      "step": 28185
    },
    {
      "epoch": 0.9023109916138532,
      "grad_norm": 0.0030071570072323084,
      "learning_rate": 0.06004063149148705,
      "loss": 0.4666,
      "num_input_tokens_seen": 38758352,
      "step": 28190
    },
    {
      "epoch": 0.9024710325843416,
      "grad_norm": 0.005185949150472879,
      "learning_rate": 0.05999350257658497,
      "loss": 0.3382,
      "num_input_tokens_seen": 38765072,
      "step": 28195
    },
    {
      "epoch": 0.9026310735548301,
      "grad_norm": 0.00703087355941534,
      "learning_rate": 0.05994638754181582,
      "loss": 0.5541,
      "num_input_tokens_seen": 38771760,
      "step": 28200
    },
    {
      "epoch": 0.9026310735548301,
      "eval_loss": 0.5091822743415833,
      "eval_runtime": 332.0133,
      "eval_samples_per_second": 41.824,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 38771760,
      "step": 28200
    },
    {
      "epoch": 0.9027911145253185,
      "grad_norm": 0.0031139766797423363,
      "learning_rate": 0.059899286394445445,
      "loss": 0.4897,
      "num_input_tokens_seen": 38778448,
      "step": 28205
    },
    {
      "epoch": 0.902951155495807,
      "grad_norm": 0.003109917975962162,
      "learning_rate": 0.059852199141737346,
      "loss": 0.495,
      "num_input_tokens_seen": 38785344,
      "step": 28210
    },
    {
      "epoch": 0.9031111964662953,
      "grad_norm": 0.005816150922328234,
      "learning_rate": 0.05980512579095304,
      "loss": 0.4364,
      "num_input_tokens_seen": 38793056,
      "step": 28215
    },
    {
      "epoch": 0.9032712374367838,
      "grad_norm": 0.004020435735583305,
      "learning_rate": 0.05975806634935181,
      "loss": 0.6472,
      "num_input_tokens_seen": 38800032,
      "step": 28220
    },
    {
      "epoch": 0.9034312784072722,
      "grad_norm": 0.0067585003562271595,
      "learning_rate": 0.05971102082419076,
      "loss": 0.4301,
      "num_input_tokens_seen": 38806480,
      "step": 28225
    },
    {
      "epoch": 0.9035913193777607,
      "grad_norm": 0.002270557451993227,
      "learning_rate": 0.05966398922272492,
      "loss": 0.4285,
      "num_input_tokens_seen": 38813232,
      "step": 28230
    },
    {
      "epoch": 0.9037513603482491,
      "grad_norm": 0.004433826077729464,
      "learning_rate": 0.059616971552207236,
      "loss": 0.5039,
      "num_input_tokens_seen": 38819856,
      "step": 28235
    },
    {
      "epoch": 0.9039114013187376,
      "grad_norm": 0.0018012220971286297,
      "learning_rate": 0.059569967819888305,
      "loss": 0.4031,
      "num_input_tokens_seen": 38826624,
      "step": 28240
    },
    {
      "epoch": 0.9040714422892261,
      "grad_norm": 0.00494990823790431,
      "learning_rate": 0.05952297803301681,
      "loss": 0.4617,
      "num_input_tokens_seen": 38833408,
      "step": 28245
    },
    {
      "epoch": 0.9042314832597145,
      "grad_norm": 0.005484606605023146,
      "learning_rate": 0.059476002198839056,
      "loss": 0.4719,
      "num_input_tokens_seen": 38840208,
      "step": 28250
    },
    {
      "epoch": 0.904391524230203,
      "grad_norm": 0.0034718711394816637,
      "learning_rate": 0.05942904032459935,
      "loss": 0.7756,
      "num_input_tokens_seen": 38847744,
      "step": 28255
    },
    {
      "epoch": 0.9045515652006914,
      "grad_norm": 0.004017300438135862,
      "learning_rate": 0.05938209241753987,
      "loss": 0.3509,
      "num_input_tokens_seen": 38854240,
      "step": 28260
    },
    {
      "epoch": 0.9047116061711798,
      "grad_norm": 0.0021270730067044497,
      "learning_rate": 0.05933515848490046,
      "loss": 0.5731,
      "num_input_tokens_seen": 38861728,
      "step": 28265
    },
    {
      "epoch": 0.9048716471416682,
      "grad_norm": 0.007262269034981728,
      "learning_rate": 0.059288238533918985,
      "loss": 0.5878,
      "num_input_tokens_seen": 38868800,
      "step": 28270
    },
    {
      "epoch": 0.9050316881121567,
      "grad_norm": 0.003403706243261695,
      "learning_rate": 0.05924133257183113,
      "loss": 0.4241,
      "num_input_tokens_seen": 38875600,
      "step": 28275
    },
    {
      "epoch": 0.9051917290826451,
      "grad_norm": 0.007776123005896807,
      "learning_rate": 0.059194440605870285,
      "loss": 0.5786,
      "num_input_tokens_seen": 38882400,
      "step": 28280
    },
    {
      "epoch": 0.9053517700531336,
      "grad_norm": 0.004948895424604416,
      "learning_rate": 0.059147562643267884,
      "loss": 0.6107,
      "num_input_tokens_seen": 38889168,
      "step": 28285
    },
    {
      "epoch": 0.905511811023622,
      "grad_norm": 0.009915845468640327,
      "learning_rate": 0.059100698691253055,
      "loss": 0.5445,
      "num_input_tokens_seen": 38895776,
      "step": 28290
    },
    {
      "epoch": 0.9056718519941105,
      "grad_norm": 0.004093893337994814,
      "learning_rate": 0.05905384875705273,
      "loss": 0.4212,
      "num_input_tokens_seen": 38903056,
      "step": 28295
    },
    {
      "epoch": 0.905831892964599,
      "grad_norm": 0.003475519362837076,
      "learning_rate": 0.05900701284789189,
      "loss": 0.4807,
      "num_input_tokens_seen": 38910112,
      "step": 28300
    },
    {
      "epoch": 0.9059919339350874,
      "grad_norm": 0.006791291292756796,
      "learning_rate": 0.058960190970993115,
      "loss": 0.4481,
      "num_input_tokens_seen": 38916832,
      "step": 28305
    },
    {
      "epoch": 0.9061519749055759,
      "grad_norm": 0.005941628944128752,
      "learning_rate": 0.058913383133576955,
      "loss": 0.6647,
      "num_input_tokens_seen": 38923696,
      "step": 28310
    },
    {
      "epoch": 0.9063120158760642,
      "grad_norm": 0.0042356206104159355,
      "learning_rate": 0.05886658934286185,
      "loss": 0.5311,
      "num_input_tokens_seen": 38930784,
      "step": 28315
    },
    {
      "epoch": 0.9064720568465527,
      "grad_norm": 0.0046985577791929245,
      "learning_rate": 0.058819809606063846,
      "loss": 0.543,
      "num_input_tokens_seen": 38937424,
      "step": 28320
    },
    {
      "epoch": 0.9066320978170411,
      "grad_norm": 0.003270609537139535,
      "learning_rate": 0.05877304393039711,
      "loss": 0.392,
      "num_input_tokens_seen": 38943984,
      "step": 28325
    },
    {
      "epoch": 0.9067921387875296,
      "grad_norm": 0.009405034594237804,
      "learning_rate": 0.05872629232307338,
      "loss": 0.6095,
      "num_input_tokens_seen": 38950688,
      "step": 28330
    },
    {
      "epoch": 0.906952179758018,
      "grad_norm": 0.006214601919054985,
      "learning_rate": 0.05867955479130239,
      "loss": 0.6248,
      "num_input_tokens_seen": 38957456,
      "step": 28335
    },
    {
      "epoch": 0.9071122207285065,
      "grad_norm": 0.004442626144737005,
      "learning_rate": 0.058632831342291705,
      "loss": 0.4831,
      "num_input_tokens_seen": 38964208,
      "step": 28340
    },
    {
      "epoch": 0.907272261698995,
      "grad_norm": 0.001888857688754797,
      "learning_rate": 0.05858612198324655,
      "loss": 0.4064,
      "num_input_tokens_seen": 38971152,
      "step": 28345
    },
    {
      "epoch": 0.9074323026694834,
      "grad_norm": 0.007328737061470747,
      "learning_rate": 0.05853942672137025,
      "loss": 0.4499,
      "num_input_tokens_seen": 38978112,
      "step": 28350
    },
    {
      "epoch": 0.9075923436399719,
      "grad_norm": 0.004644669126719236,
      "learning_rate": 0.05849274556386363,
      "loss": 0.3388,
      "num_input_tokens_seen": 38984736,
      "step": 28355
    },
    {
      "epoch": 0.9077523846104603,
      "grad_norm": 0.0037627345882356167,
      "learning_rate": 0.05844607851792567,
      "loss": 0.4291,
      "num_input_tokens_seen": 38991696,
      "step": 28360
    },
    {
      "epoch": 0.9079124255809488,
      "grad_norm": 0.005647262558341026,
      "learning_rate": 0.058399425590752924,
      "loss": 0.3475,
      "num_input_tokens_seen": 38998800,
      "step": 28365
    },
    {
      "epoch": 0.9080724665514371,
      "grad_norm": 0.004888692405074835,
      "learning_rate": 0.05835278678953985,
      "loss": 0.4475,
      "num_input_tokens_seen": 39005536,
      "step": 28370
    },
    {
      "epoch": 0.9082325075219256,
      "grad_norm": 0.004810594022274017,
      "learning_rate": 0.05830616212147874,
      "loss": 0.6179,
      "num_input_tokens_seen": 39012352,
      "step": 28375
    },
    {
      "epoch": 0.908392548492414,
      "grad_norm": 0.005558249540627003,
      "learning_rate": 0.058259551593759784,
      "loss": 0.4144,
      "num_input_tokens_seen": 39019088,
      "step": 28380
    },
    {
      "epoch": 0.9085525894629025,
      "grad_norm": 0.002801172435283661,
      "learning_rate": 0.058212955213570804,
      "loss": 0.5801,
      "num_input_tokens_seen": 39025728,
      "step": 28385
    },
    {
      "epoch": 0.9087126304333909,
      "grad_norm": 0.0029367536772042513,
      "learning_rate": 0.0581663729880976,
      "loss": 0.4618,
      "num_input_tokens_seen": 39032416,
      "step": 28390
    },
    {
      "epoch": 0.9088726714038794,
      "grad_norm": 0.006222155876457691,
      "learning_rate": 0.05811980492452379,
      "loss": 0.4383,
      "num_input_tokens_seen": 39039040,
      "step": 28395
    },
    {
      "epoch": 0.9090327123743679,
      "grad_norm": 0.004764068406075239,
      "learning_rate": 0.058073251030030644,
      "loss": 0.4895,
      "num_input_tokens_seen": 39045824,
      "step": 28400
    },
    {
      "epoch": 0.9090327123743679,
      "eval_loss": 0.5090657472610474,
      "eval_runtime": 331.8717,
      "eval_samples_per_second": 41.841,
      "eval_steps_per_second": 20.921,
      "num_input_tokens_seen": 39045824,
      "step": 28400
    },
    {
      "epoch": 0.9091927533448563,
      "grad_norm": 0.004711235873401165,
      "learning_rate": 0.05802671131179747,
      "loss": 0.3722,
      "num_input_tokens_seen": 39052480,
      "step": 28405
    },
    {
      "epoch": 0.9093527943153448,
      "grad_norm": 0.002775033703073859,
      "learning_rate": 0.057980185777001154,
      "loss": 0.4236,
      "num_input_tokens_seen": 39059024,
      "step": 28410
    },
    {
      "epoch": 0.9095128352858332,
      "grad_norm": 0.007851681672036648,
      "learning_rate": 0.057933674432816606,
      "loss": 0.5701,
      "num_input_tokens_seen": 39066000,
      "step": 28415
    },
    {
      "epoch": 0.9096728762563216,
      "grad_norm": 0.005883705336600542,
      "learning_rate": 0.05788717728641648,
      "loss": 0.5118,
      "num_input_tokens_seen": 39073024,
      "step": 28420
    },
    {
      "epoch": 0.90983291722681,
      "grad_norm": 0.004096079617738724,
      "learning_rate": 0.057840694344971126,
      "loss": 0.5023,
      "num_input_tokens_seen": 39080464,
      "step": 28425
    },
    {
      "epoch": 0.9099929581972985,
      "grad_norm": 0.0038472148589789867,
      "learning_rate": 0.0577942256156489,
      "loss": 0.4649,
      "num_input_tokens_seen": 39087760,
      "step": 28430
    },
    {
      "epoch": 0.9101529991677869,
      "grad_norm": 0.004994986113160849,
      "learning_rate": 0.057747771105615804,
      "loss": 0.4262,
      "num_input_tokens_seen": 39094768,
      "step": 28435
    },
    {
      "epoch": 0.9103130401382754,
      "grad_norm": 0.0065709445625543594,
      "learning_rate": 0.05770133082203568,
      "loss": 0.5743,
      "num_input_tokens_seen": 39101632,
      "step": 28440
    },
    {
      "epoch": 0.9104730811087638,
      "grad_norm": 0.005152306519448757,
      "learning_rate": 0.0576549047720703,
      "loss": 0.4626,
      "num_input_tokens_seen": 39108576,
      "step": 28445
    },
    {
      "epoch": 0.9106331220792523,
      "grad_norm": 0.0036795546766370535,
      "learning_rate": 0.05760849296287902,
      "loss": 0.3793,
      "num_input_tokens_seen": 39115392,
      "step": 28450
    },
    {
      "epoch": 0.9107931630497408,
      "grad_norm": 0.0033041418064385653,
      "learning_rate": 0.05756209540161919,
      "loss": 0.3803,
      "num_input_tokens_seen": 39122288,
      "step": 28455
    },
    {
      "epoch": 0.9109532040202292,
      "grad_norm": 0.004454812500625849,
      "learning_rate": 0.05751571209544595,
      "loss": 0.5323,
      "num_input_tokens_seen": 39129680,
      "step": 28460
    },
    {
      "epoch": 0.9111132449907177,
      "grad_norm": 0.0027283495292067528,
      "learning_rate": 0.057469343051512085,
      "loss": 0.5428,
      "num_input_tokens_seen": 39136096,
      "step": 28465
    },
    {
      "epoch": 0.911273285961206,
      "grad_norm": 0.004823064897209406,
      "learning_rate": 0.057422988276968324,
      "loss": 0.3415,
      "num_input_tokens_seen": 39142608,
      "step": 28470
    },
    {
      "epoch": 0.9114333269316945,
      "grad_norm": 0.006035743281245232,
      "learning_rate": 0.05737664777896323,
      "loss": 0.4021,
      "num_input_tokens_seen": 39149520,
      "step": 28475
    },
    {
      "epoch": 0.9115933679021829,
      "grad_norm": 0.0057624164037406445,
      "learning_rate": 0.057330321564642975,
      "loss": 0.5237,
      "num_input_tokens_seen": 39156384,
      "step": 28480
    },
    {
      "epoch": 0.9117534088726714,
      "grad_norm": 0.004293810110539198,
      "learning_rate": 0.05728400964115174,
      "loss": 0.5297,
      "num_input_tokens_seen": 39162912,
      "step": 28485
    },
    {
      "epoch": 0.9119134498431598,
      "grad_norm": 0.003032421227544546,
      "learning_rate": 0.057237712015631305,
      "loss": 0.4883,
      "num_input_tokens_seen": 39170208,
      "step": 28490
    },
    {
      "epoch": 0.9120734908136483,
      "grad_norm": 0.0039564636535942554,
      "learning_rate": 0.057191428695221425,
      "loss": 0.4953,
      "num_input_tokens_seen": 39176832,
      "step": 28495
    },
    {
      "epoch": 0.9122335317841367,
      "grad_norm": 0.005447630304843187,
      "learning_rate": 0.05714515968705958,
      "loss": 0.5039,
      "num_input_tokens_seen": 39183968,
      "step": 28500
    },
    {
      "epoch": 0.9123935727546252,
      "grad_norm": 0.002739167073741555,
      "learning_rate": 0.05709890499828099,
      "loss": 0.6309,
      "num_input_tokens_seen": 39190624,
      "step": 28505
    },
    {
      "epoch": 0.9125536137251137,
      "grad_norm": 0.009400452487170696,
      "learning_rate": 0.05705266463601868,
      "loss": 0.6617,
      "num_input_tokens_seen": 39197344,
      "step": 28510
    },
    {
      "epoch": 0.9127136546956021,
      "grad_norm": 0.004393095150589943,
      "learning_rate": 0.057006438607403565,
      "loss": 0.4384,
      "num_input_tokens_seen": 39204336,
      "step": 28515
    },
    {
      "epoch": 0.9128736956660906,
      "grad_norm": 0.0031046909280121326,
      "learning_rate": 0.056960226919564205,
      "loss": 0.4388,
      "num_input_tokens_seen": 39210896,
      "step": 28520
    },
    {
      "epoch": 0.9130337366365789,
      "grad_norm": 0.007694361265748739,
      "learning_rate": 0.05691402957962713,
      "loss": 0.5693,
      "num_input_tokens_seen": 39218048,
      "step": 28525
    },
    {
      "epoch": 0.9131937776070674,
      "grad_norm": 0.00285990908741951,
      "learning_rate": 0.05686784659471642,
      "loss": 0.6305,
      "num_input_tokens_seen": 39224896,
      "step": 28530
    },
    {
      "epoch": 0.9133538185775558,
      "grad_norm": 0.004274012520909309,
      "learning_rate": 0.056821677971954136,
      "loss": 0.4917,
      "num_input_tokens_seen": 39231952,
      "step": 28535
    },
    {
      "epoch": 0.9135138595480443,
      "grad_norm": 0.0049010165967047215,
      "learning_rate": 0.05677552371846012,
      "loss": 0.5572,
      "num_input_tokens_seen": 39238736,
      "step": 28540
    },
    {
      "epoch": 0.9136739005185327,
      "grad_norm": 0.00527940271422267,
      "learning_rate": 0.05672938384135182,
      "loss": 0.4585,
      "num_input_tokens_seen": 39245728,
      "step": 28545
    },
    {
      "epoch": 0.9138339414890212,
      "grad_norm": 0.005456010811030865,
      "learning_rate": 0.05668325834774465,
      "loss": 0.4729,
      "num_input_tokens_seen": 39252256,
      "step": 28550
    },
    {
      "epoch": 0.9139939824595096,
      "grad_norm": 0.003864401951432228,
      "learning_rate": 0.05663714724475177,
      "loss": 0.4498,
      "num_input_tokens_seen": 39258928,
      "step": 28555
    },
    {
      "epoch": 0.9141540234299981,
      "grad_norm": 0.006218394264578819,
      "learning_rate": 0.05659105053948403,
      "loss": 0.4079,
      "num_input_tokens_seen": 39266016,
      "step": 28560
    },
    {
      "epoch": 0.9143140644004866,
      "grad_norm": 0.0036058491095900536,
      "learning_rate": 0.056544968239050176,
      "loss": 0.5731,
      "num_input_tokens_seen": 39273424,
      "step": 28565
    },
    {
      "epoch": 0.914474105370975,
      "grad_norm": 0.003793464507907629,
      "learning_rate": 0.056498900350556616,
      "loss": 0.535,
      "num_input_tokens_seen": 39280320,
      "step": 28570
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 0.005509550217539072,
      "learning_rate": 0.05645284688110766,
      "loss": 0.4818,
      "num_input_tokens_seen": 39286912,
      "step": 28575
    },
    {
      "epoch": 0.9147941873119518,
      "grad_norm": 0.005691469646990299,
      "learning_rate": 0.05640680783780532,
      "loss": 0.4959,
      "num_input_tokens_seen": 39293600,
      "step": 28580
    },
    {
      "epoch": 0.9149542282824403,
      "grad_norm": 0.00337778078392148,
      "learning_rate": 0.056360783227749324,
      "loss": 0.3787,
      "num_input_tokens_seen": 39300496,
      "step": 28585
    },
    {
      "epoch": 0.9151142692529287,
      "grad_norm": 0.004301091190427542,
      "learning_rate": 0.05631477305803728,
      "loss": 0.5217,
      "num_input_tokens_seen": 39307152,
      "step": 28590
    },
    {
      "epoch": 0.9152743102234172,
      "grad_norm": 0.005086846649646759,
      "learning_rate": 0.05626877733576462,
      "loss": 0.7908,
      "num_input_tokens_seen": 39313712,
      "step": 28595
    },
    {
      "epoch": 0.9154343511939056,
      "grad_norm": 0.005232734140008688,
      "learning_rate": 0.05622279606802435,
      "loss": 0.3437,
      "num_input_tokens_seen": 39320736,
      "step": 28600
    },
    {
      "epoch": 0.9154343511939056,
      "eval_loss": 0.5097111463546753,
      "eval_runtime": 332.5706,
      "eval_samples_per_second": 41.754,
      "eval_steps_per_second": 20.877,
      "num_input_tokens_seen": 39320736,
      "step": 28600
    },
    {
      "epoch": 0.9155943921643941,
      "grad_norm": 0.004491601604968309,
      "learning_rate": 0.05617682926190744,
      "loss": 0.531,
      "num_input_tokens_seen": 39327648,
      "step": 28605
    },
    {
      "epoch": 0.9157544331348826,
      "grad_norm": 0.007270353846251965,
      "learning_rate": 0.05613087692450248,
      "loss": 0.5726,
      "num_input_tokens_seen": 39334192,
      "step": 28610
    },
    {
      "epoch": 0.915914474105371,
      "grad_norm": 0.004251271486282349,
      "learning_rate": 0.05608493906289592,
      "loss": 0.4769,
      "num_input_tokens_seen": 39341376,
      "step": 28615
    },
    {
      "epoch": 0.9160745150758595,
      "grad_norm": 0.006134895142167807,
      "learning_rate": 0.05603901568417201,
      "loss": 0.5411,
      "num_input_tokens_seen": 39348320,
      "step": 28620
    },
    {
      "epoch": 0.9162345560463478,
      "grad_norm": 0.008080917410552502,
      "learning_rate": 0.055993106795412625,
      "loss": 0.528,
      "num_input_tokens_seen": 39354976,
      "step": 28625
    },
    {
      "epoch": 0.9163945970168363,
      "grad_norm": 0.004916769452393055,
      "learning_rate": 0.05594721240369759,
      "loss": 0.4813,
      "num_input_tokens_seen": 39362032,
      "step": 28630
    },
    {
      "epoch": 0.9165546379873247,
      "grad_norm": 0.005455807317048311,
      "learning_rate": 0.055901332516104296,
      "loss": 0.4105,
      "num_input_tokens_seen": 39368960,
      "step": 28635
    },
    {
      "epoch": 0.9167146789578132,
      "grad_norm": 0.004615337122231722,
      "learning_rate": 0.05585546713970804,
      "loss": 0.4504,
      "num_input_tokens_seen": 39376304,
      "step": 28640
    },
    {
      "epoch": 0.9168747199283016,
      "grad_norm": 0.0044007208198308945,
      "learning_rate": 0.05580961628158189,
      "loss": 0.5243,
      "num_input_tokens_seen": 39383664,
      "step": 28645
    },
    {
      "epoch": 0.9170347608987901,
      "grad_norm": 0.006094267591834068,
      "learning_rate": 0.05576377994879659,
      "loss": 0.4099,
      "num_input_tokens_seen": 39390384,
      "step": 28650
    },
    {
      "epoch": 0.9171948018692785,
      "grad_norm": 0.006137648597359657,
      "learning_rate": 0.05571795814842063,
      "loss": 0.4655,
      "num_input_tokens_seen": 39397152,
      "step": 28655
    },
    {
      "epoch": 0.917354842839767,
      "grad_norm": 0.007224716246128082,
      "learning_rate": 0.05567215088752037,
      "loss": 0.4886,
      "num_input_tokens_seen": 39404048,
      "step": 28660
    },
    {
      "epoch": 0.9175148838102555,
      "grad_norm": 0.004110565409064293,
      "learning_rate": 0.05562635817315981,
      "loss": 0.3018,
      "num_input_tokens_seen": 39410512,
      "step": 28665
    },
    {
      "epoch": 0.9176749247807439,
      "grad_norm": 0.005110644269734621,
      "learning_rate": 0.05558058001240083,
      "loss": 0.4002,
      "num_input_tokens_seen": 39417072,
      "step": 28670
    },
    {
      "epoch": 0.9178349657512324,
      "grad_norm": 0.0055067278444767,
      "learning_rate": 0.055534816412302915,
      "loss": 0.4218,
      "num_input_tokens_seen": 39423936,
      "step": 28675
    },
    {
      "epoch": 0.9179950067217207,
      "grad_norm": 0.00619698828086257,
      "learning_rate": 0.055489067379923436,
      "loss": 0.7135,
      "num_input_tokens_seen": 39430640,
      "step": 28680
    },
    {
      "epoch": 0.9181550476922092,
      "grad_norm": 0.005531177390366793,
      "learning_rate": 0.055443332922317505,
      "loss": 0.5542,
      "num_input_tokens_seen": 39437952,
      "step": 28685
    },
    {
      "epoch": 0.9183150886626976,
      "grad_norm": 0.0033926262985914946,
      "learning_rate": 0.055397613046537876,
      "loss": 0.4303,
      "num_input_tokens_seen": 39444400,
      "step": 28690
    },
    {
      "epoch": 0.9184751296331861,
      "grad_norm": 0.003449621144682169,
      "learning_rate": 0.055351907759635145,
      "loss": 0.4532,
      "num_input_tokens_seen": 39451360,
      "step": 28695
    },
    {
      "epoch": 0.9186351706036745,
      "grad_norm": 0.0031164183747023344,
      "learning_rate": 0.05530621706865772,
      "loss": 0.4153,
      "num_input_tokens_seen": 39457952,
      "step": 28700
    },
    {
      "epoch": 0.918795211574163,
      "grad_norm": 0.002953724004328251,
      "learning_rate": 0.055260540980651564,
      "loss": 0.3788,
      "num_input_tokens_seen": 39464704,
      "step": 28705
    },
    {
      "epoch": 0.9189552525446514,
      "grad_norm": 0.005202647764235735,
      "learning_rate": 0.05521487950266062,
      "loss": 0.5793,
      "num_input_tokens_seen": 39471536,
      "step": 28710
    },
    {
      "epoch": 0.9191152935151399,
      "grad_norm": 0.004488151054829359,
      "learning_rate": 0.055169232641726344,
      "loss": 0.3776,
      "num_input_tokens_seen": 39478704,
      "step": 28715
    },
    {
      "epoch": 0.9192753344856284,
      "grad_norm": 0.0041967895813286304,
      "learning_rate": 0.055123600404888166,
      "loss": 0.4765,
      "num_input_tokens_seen": 39485552,
      "step": 28720
    },
    {
      "epoch": 0.9194353754561168,
      "grad_norm": 0.003344462253153324,
      "learning_rate": 0.05507798279918309,
      "loss": 0.5661,
      "num_input_tokens_seen": 39492560,
      "step": 28725
    },
    {
      "epoch": 0.9195954164266052,
      "grad_norm": 0.0028171709273010492,
      "learning_rate": 0.0550323798316459,
      "loss": 0.3721,
      "num_input_tokens_seen": 39499072,
      "step": 28730
    },
    {
      "epoch": 0.9197554573970936,
      "grad_norm": 0.005877265240997076,
      "learning_rate": 0.05498679150930916,
      "loss": 0.539,
      "num_input_tokens_seen": 39505536,
      "step": 28735
    },
    {
      "epoch": 0.9199154983675821,
      "grad_norm": 0.0036439031828194857,
      "learning_rate": 0.05494121783920323,
      "loss": 0.6004,
      "num_input_tokens_seen": 39512032,
      "step": 28740
    },
    {
      "epoch": 0.9200755393380705,
      "grad_norm": 0.018740009516477585,
      "learning_rate": 0.05489565882835605,
      "loss": 0.4938,
      "num_input_tokens_seen": 39518704,
      "step": 28745
    },
    {
      "epoch": 0.920235580308559,
      "grad_norm": 0.004555624444037676,
      "learning_rate": 0.05485011448379348,
      "loss": 0.399,
      "num_input_tokens_seen": 39525584,
      "step": 28750
    },
    {
      "epoch": 0.9203956212790474,
      "grad_norm": 0.003255107207223773,
      "learning_rate": 0.05480458481253893,
      "loss": 0.4859,
      "num_input_tokens_seen": 39531840,
      "step": 28755
    },
    {
      "epoch": 0.9205556622495359,
      "grad_norm": 0.0031488046515733004,
      "learning_rate": 0.054759069821613715,
      "loss": 0.5666,
      "num_input_tokens_seen": 39538800,
      "step": 28760
    },
    {
      "epoch": 0.9207157032200243,
      "grad_norm": 0.009035496041178703,
      "learning_rate": 0.05471356951803683,
      "loss": 0.6655,
      "num_input_tokens_seen": 39545440,
      "step": 28765
    },
    {
      "epoch": 0.9208757441905128,
      "grad_norm": 0.003330423729494214,
      "learning_rate": 0.054668083908824945,
      "loss": 0.4597,
      "num_input_tokens_seen": 39552272,
      "step": 28770
    },
    {
      "epoch": 0.9210357851610013,
      "grad_norm": 0.0037896581925451756,
      "learning_rate": 0.054622613000992526,
      "loss": 0.4729,
      "num_input_tokens_seen": 39559328,
      "step": 28775
    },
    {
      "epoch": 0.9211958261314896,
      "grad_norm": 0.004960062447935343,
      "learning_rate": 0.05457715680155182,
      "loss": 0.4933,
      "num_input_tokens_seen": 39566688,
      "step": 28780
    },
    {
      "epoch": 0.9213558671019781,
      "grad_norm": 0.003056451678276062,
      "learning_rate": 0.05453171531751265,
      "loss": 0.462,
      "num_input_tokens_seen": 39573536,
      "step": 28785
    },
    {
      "epoch": 0.9215159080724665,
      "grad_norm": 0.004429814871400595,
      "learning_rate": 0.05448628855588276,
      "loss": 0.5728,
      "num_input_tokens_seen": 39580784,
      "step": 28790
    },
    {
      "epoch": 0.921675949042955,
      "grad_norm": 0.0039617447182536125,
      "learning_rate": 0.05444087652366746,
      "loss": 0.5081,
      "num_input_tokens_seen": 39588080,
      "step": 28795
    },
    {
      "epoch": 0.9218359900134434,
      "grad_norm": 0.005924553610384464,
      "learning_rate": 0.05439547922786984,
      "loss": 0.4957,
      "num_input_tokens_seen": 39594816,
      "step": 28800
    },
    {
      "epoch": 0.9218359900134434,
      "eval_loss": 0.5112881660461426,
      "eval_runtime": 332.1159,
      "eval_samples_per_second": 41.811,
      "eval_steps_per_second": 20.905,
      "num_input_tokens_seen": 39594816,
      "step": 28800
    },
    {
      "epoch": 0.9219960309839319,
      "grad_norm": 0.002723233075812459,
      "learning_rate": 0.0543500966754908,
      "loss": 0.3261,
      "num_input_tokens_seen": 39602064,
      "step": 28805
    },
    {
      "epoch": 0.9221560719544203,
      "grad_norm": 0.006309398915618658,
      "learning_rate": 0.05430472887352882,
      "loss": 0.5466,
      "num_input_tokens_seen": 39609088,
      "step": 28810
    },
    {
      "epoch": 0.9223161129249088,
      "grad_norm": 0.00619633961468935,
      "learning_rate": 0.05425937582898023,
      "loss": 0.534,
      "num_input_tokens_seen": 39615952,
      "step": 28815
    },
    {
      "epoch": 0.9224761538953972,
      "grad_norm": 0.003760426538065076,
      "learning_rate": 0.054214037548839085,
      "loss": 0.369,
      "num_input_tokens_seen": 39622720,
      "step": 28820
    },
    {
      "epoch": 0.9226361948658857,
      "grad_norm": 0.004633475095033646,
      "learning_rate": 0.05416871404009703,
      "loss": 0.4624,
      "num_input_tokens_seen": 39629968,
      "step": 28825
    },
    {
      "epoch": 0.9227962358363742,
      "grad_norm": 0.004416496958583593,
      "learning_rate": 0.054123405309743605,
      "loss": 0.5303,
      "num_input_tokens_seen": 39636880,
      "step": 28830
    },
    {
      "epoch": 0.9229562768068625,
      "grad_norm": 0.005037115421146154,
      "learning_rate": 0.0540781113647659,
      "loss": 0.5117,
      "num_input_tokens_seen": 39643792,
      "step": 28835
    },
    {
      "epoch": 0.923116317777351,
      "grad_norm": 0.007320908363908529,
      "learning_rate": 0.054032832212148836,
      "loss": 0.5901,
      "num_input_tokens_seen": 39650896,
      "step": 28840
    },
    {
      "epoch": 0.9232763587478394,
      "grad_norm": 0.0032525379210710526,
      "learning_rate": 0.0539875678588751,
      "loss": 0.2819,
      "num_input_tokens_seen": 39657568,
      "step": 28845
    },
    {
      "epoch": 0.9234363997183279,
      "grad_norm": 0.006659949664026499,
      "learning_rate": 0.05394231831192492,
      "loss": 0.4605,
      "num_input_tokens_seen": 39664352,
      "step": 28850
    },
    {
      "epoch": 0.9235964406888163,
      "grad_norm": 0.005537872202694416,
      "learning_rate": 0.05389708357827639,
      "loss": 0.4515,
      "num_input_tokens_seen": 39671248,
      "step": 28855
    },
    {
      "epoch": 0.9237564816593048,
      "grad_norm": 0.0034565399400889874,
      "learning_rate": 0.05385186366490533,
      "loss": 0.6817,
      "num_input_tokens_seen": 39677856,
      "step": 28860
    },
    {
      "epoch": 0.9239165226297932,
      "grad_norm": 0.008326293900609016,
      "learning_rate": 0.053806658578785166,
      "loss": 0.4255,
      "num_input_tokens_seen": 39685184,
      "step": 28865
    },
    {
      "epoch": 0.9240765636002817,
      "grad_norm": 0.003117835381999612,
      "learning_rate": 0.05376146832688705,
      "loss": 0.3786,
      "num_input_tokens_seen": 39691888,
      "step": 28870
    },
    {
      "epoch": 0.9242366045707702,
      "grad_norm": 0.004993585869669914,
      "learning_rate": 0.053716292916179964,
      "loss": 0.4473,
      "num_input_tokens_seen": 39698928,
      "step": 28875
    },
    {
      "epoch": 0.9243966455412586,
      "grad_norm": 0.0033365634735673666,
      "learning_rate": 0.05367113235363045,
      "loss": 0.3577,
      "num_input_tokens_seen": 39705584,
      "step": 28880
    },
    {
      "epoch": 0.924556686511747,
      "grad_norm": 0.002550778677687049,
      "learning_rate": 0.05362598664620289,
      "loss": 0.5074,
      "num_input_tokens_seen": 39712288,
      "step": 28885
    },
    {
      "epoch": 0.9247167274822354,
      "grad_norm": 0.0033374312333762646,
      "learning_rate": 0.053580855800859285,
      "loss": 0.3563,
      "num_input_tokens_seen": 39719392,
      "step": 28890
    },
    {
      "epoch": 0.9248767684527239,
      "grad_norm": 0.005564319901168346,
      "learning_rate": 0.05353573982455938,
      "loss": 0.4139,
      "num_input_tokens_seen": 39726464,
      "step": 28895
    },
    {
      "epoch": 0.9250368094232123,
      "grad_norm": 0.005771728232502937,
      "learning_rate": 0.053490638724260686,
      "loss": 0.5593,
      "num_input_tokens_seen": 39733696,
      "step": 28900
    },
    {
      "epoch": 0.9251968503937008,
      "grad_norm": 0.0030842344276607037,
      "learning_rate": 0.05344555250691827,
      "loss": 0.4723,
      "num_input_tokens_seen": 39740624,
      "step": 28905
    },
    {
      "epoch": 0.9253568913641892,
      "grad_norm": 0.0034340620040893555,
      "learning_rate": 0.053400481179485086,
      "loss": 0.5317,
      "num_input_tokens_seen": 39747680,
      "step": 28910
    },
    {
      "epoch": 0.9255169323346777,
      "grad_norm": 0.00823709461838007,
      "learning_rate": 0.05335542474891159,
      "loss": 0.5588,
      "num_input_tokens_seen": 39754848,
      "step": 28915
    },
    {
      "epoch": 0.9256769733051661,
      "grad_norm": 0.004088544752448797,
      "learning_rate": 0.053310383222146124,
      "loss": 0.5486,
      "num_input_tokens_seen": 39761664,
      "step": 28920
    },
    {
      "epoch": 0.9258370142756546,
      "grad_norm": 0.00607747957110405,
      "learning_rate": 0.053265356606134684,
      "loss": 0.6345,
      "num_input_tokens_seen": 39768688,
      "step": 28925
    },
    {
      "epoch": 0.9259970552461431,
      "grad_norm": 0.004677519202232361,
      "learning_rate": 0.053220344907820856,
      "loss": 0.4312,
      "num_input_tokens_seen": 39775456,
      "step": 28930
    },
    {
      "epoch": 0.9261570962166314,
      "grad_norm": 0.0026903380639851093,
      "learning_rate": 0.05317534813414608,
      "loss": 0.3375,
      "num_input_tokens_seen": 39782048,
      "step": 28935
    },
    {
      "epoch": 0.9263171371871199,
      "grad_norm": 0.0037966426461935043,
      "learning_rate": 0.05313036629204942,
      "loss": 0.4841,
      "num_input_tokens_seen": 39789024,
      "step": 28940
    },
    {
      "epoch": 0.9264771781576083,
      "grad_norm": 0.00558742880821228,
      "learning_rate": 0.05308539938846756,
      "loss": 0.3956,
      "num_input_tokens_seen": 39795968,
      "step": 28945
    },
    {
      "epoch": 0.9266372191280968,
      "grad_norm": 0.004928117152303457,
      "learning_rate": 0.05304044743033507,
      "loss": 0.5225,
      "num_input_tokens_seen": 39802944,
      "step": 28950
    },
    {
      "epoch": 0.9267972600985852,
      "grad_norm": 0.0054198880679905415,
      "learning_rate": 0.05299551042458401,
      "loss": 0.5491,
      "num_input_tokens_seen": 39809600,
      "step": 28955
    },
    {
      "epoch": 0.9269573010690737,
      "grad_norm": 0.002201333176344633,
      "learning_rate": 0.052950588378144266,
      "loss": 0.4225,
      "num_input_tokens_seen": 39816320,
      "step": 28960
    },
    {
      "epoch": 0.9271173420395621,
      "grad_norm": 0.004452093504369259,
      "learning_rate": 0.052905681297943465,
      "loss": 0.4297,
      "num_input_tokens_seen": 39823056,
      "step": 28965
    },
    {
      "epoch": 0.9272773830100506,
      "grad_norm": 0.004280089866369963,
      "learning_rate": 0.0528607891909067,
      "loss": 0.4596,
      "num_input_tokens_seen": 39830000,
      "step": 28970
    },
    {
      "epoch": 0.927437423980539,
      "grad_norm": 0.005052715074270964,
      "learning_rate": 0.05281591206395697,
      "loss": 0.5066,
      "num_input_tokens_seen": 39836560,
      "step": 28975
    },
    {
      "epoch": 0.9275974649510275,
      "grad_norm": 0.004871147684752941,
      "learning_rate": 0.05277104992401496,
      "loss": 0.5017,
      "num_input_tokens_seen": 39843344,
      "step": 28980
    },
    {
      "epoch": 0.927757505921516,
      "grad_norm": 0.005258605349808931,
      "learning_rate": 0.05272620277799884,
      "loss": 0.6551,
      "num_input_tokens_seen": 39850128,
      "step": 28985
    },
    {
      "epoch": 0.9279175468920043,
      "grad_norm": 0.006909629330039024,
      "learning_rate": 0.05268137063282473,
      "loss": 0.6839,
      "num_input_tokens_seen": 39856736,
      "step": 28990
    },
    {
      "epoch": 0.9280775878624928,
      "grad_norm": 0.0036039832048118114,
      "learning_rate": 0.0526365534954062,
      "loss": 0.4813,
      "num_input_tokens_seen": 39864128,
      "step": 28995
    },
    {
      "epoch": 0.9282376288329812,
      "grad_norm": 0.0027762854006141424,
      "learning_rate": 0.052591751372654656,
      "loss": 0.5137,
      "num_input_tokens_seen": 39870432,
      "step": 29000
    },
    {
      "epoch": 0.9282376288329812,
      "eval_loss": 0.5089755654335022,
      "eval_runtime": 332.4713,
      "eval_samples_per_second": 41.766,
      "eval_steps_per_second": 20.883,
      "num_input_tokens_seen": 39870432,
      "step": 29000
    },
    {
      "epoch": 0.9283976698034697,
      "grad_norm": 0.0016032615676522255,
      "learning_rate": 0.05254696427147921,
      "loss": 0.3839,
      "num_input_tokens_seen": 39876960,
      "step": 29005
    },
    {
      "epoch": 0.9285577107739581,
      "grad_norm": 0.00555607071146369,
      "learning_rate": 0.052502192198786546,
      "loss": 0.4651,
      "num_input_tokens_seen": 39884032,
      "step": 29010
    },
    {
      "epoch": 0.9287177517444466,
      "grad_norm": 0.004399891477078199,
      "learning_rate": 0.05245743516148103,
      "loss": 0.3984,
      "num_input_tokens_seen": 39890640,
      "step": 29015
    },
    {
      "epoch": 0.928877792714935,
      "grad_norm": 0.004674617201089859,
      "learning_rate": 0.05241269316646486,
      "loss": 0.6575,
      "num_input_tokens_seen": 39897568,
      "step": 29020
    },
    {
      "epoch": 0.9290378336854235,
      "grad_norm": 0.004897128324955702,
      "learning_rate": 0.052367966220637725,
      "loss": 0.5781,
      "num_input_tokens_seen": 39904592,
      "step": 29025
    },
    {
      "epoch": 0.9291978746559119,
      "grad_norm": 0.007861691527068615,
      "learning_rate": 0.05232325433089716,
      "loss": 0.535,
      "num_input_tokens_seen": 39911360,
      "step": 29030
    },
    {
      "epoch": 0.9293579156264004,
      "grad_norm": 0.007030309643596411,
      "learning_rate": 0.052278557504138214,
      "loss": 0.5284,
      "num_input_tokens_seen": 39918784,
      "step": 29035
    },
    {
      "epoch": 0.9295179565968889,
      "grad_norm": 0.004978757351636887,
      "learning_rate": 0.05223387574725372,
      "loss": 0.5626,
      "num_input_tokens_seen": 39925712,
      "step": 29040
    },
    {
      "epoch": 0.9296779975673772,
      "grad_norm": 0.005931941792368889,
      "learning_rate": 0.05218920906713428,
      "loss": 0.6117,
      "num_input_tokens_seen": 39932384,
      "step": 29045
    },
    {
      "epoch": 0.9298380385378657,
      "grad_norm": 0.0034695731010288,
      "learning_rate": 0.05214455747066789,
      "loss": 0.5556,
      "num_input_tokens_seen": 39939408,
      "step": 29050
    },
    {
      "epoch": 0.9299980795083541,
      "grad_norm": 0.003842143341898918,
      "learning_rate": 0.05209992096474048,
      "loss": 0.596,
      "num_input_tokens_seen": 39946032,
      "step": 29055
    },
    {
      "epoch": 0.9301581204788426,
      "grad_norm": 0.005205618683248758,
      "learning_rate": 0.05205529955623559,
      "loss": 0.7769,
      "num_input_tokens_seen": 39952544,
      "step": 29060
    },
    {
      "epoch": 0.930318161449331,
      "grad_norm": 0.0036829763557761908,
      "learning_rate": 0.052010693252034314,
      "loss": 0.4051,
      "num_input_tokens_seen": 39959296,
      "step": 29065
    },
    {
      "epoch": 0.9304782024198195,
      "grad_norm": 0.0033700945787131786,
      "learning_rate": 0.0519661020590156,
      "loss": 0.4666,
      "num_input_tokens_seen": 39965968,
      "step": 29070
    },
    {
      "epoch": 0.9306382433903079,
      "grad_norm": 0.0061441948637366295,
      "learning_rate": 0.05192152598405586,
      "loss": 0.3814,
      "num_input_tokens_seen": 39973200,
      "step": 29075
    },
    {
      "epoch": 0.9307982843607964,
      "grad_norm": 0.004597972147166729,
      "learning_rate": 0.05187696503402941,
      "loss": 0.5799,
      "num_input_tokens_seen": 39980112,
      "step": 29080
    },
    {
      "epoch": 0.9309583253312849,
      "grad_norm": 0.00541657255962491,
      "learning_rate": 0.05183241921580798,
      "loss": 0.518,
      "num_input_tokens_seen": 39986880,
      "step": 29085
    },
    {
      "epoch": 0.9311183663017732,
      "grad_norm": 0.001372452941723168,
      "learning_rate": 0.051787888536261206,
      "loss": 0.4822,
      "num_input_tokens_seen": 39993376,
      "step": 29090
    },
    {
      "epoch": 0.9312784072722617,
      "grad_norm": 0.0035070583689957857,
      "learning_rate": 0.051743373002256184,
      "loss": 0.3813,
      "num_input_tokens_seen": 40000544,
      "step": 29095
    },
    {
      "epoch": 0.9314384482427501,
      "grad_norm": 0.0029643927700817585,
      "learning_rate": 0.05169887262065787,
      "loss": 0.3583,
      "num_input_tokens_seen": 40006912,
      "step": 29100
    },
    {
      "epoch": 0.9315984892132386,
      "grad_norm": 0.007218221202492714,
      "learning_rate": 0.051654387398328665,
      "loss": 0.6523,
      "num_input_tokens_seen": 40013456,
      "step": 29105
    },
    {
      "epoch": 0.931758530183727,
      "grad_norm": 0.004574640654027462,
      "learning_rate": 0.05160991734212888,
      "loss": 0.5791,
      "num_input_tokens_seen": 40020112,
      "step": 29110
    },
    {
      "epoch": 0.9319185711542155,
      "grad_norm": 0.003401479683816433,
      "learning_rate": 0.051565462458916224,
      "loss": 0.6138,
      "num_input_tokens_seen": 40026736,
      "step": 29115
    },
    {
      "epoch": 0.9320786121247039,
      "grad_norm": 0.005807660054415464,
      "learning_rate": 0.05152102275554627,
      "loss": 0.5048,
      "num_input_tokens_seen": 40033664,
      "step": 29120
    },
    {
      "epoch": 0.9322386530951924,
      "grad_norm": 0.0027312219608575106,
      "learning_rate": 0.05147659823887222,
      "loss": 0.6795,
      "num_input_tokens_seen": 40040768,
      "step": 29125
    },
    {
      "epoch": 0.9323986940656808,
      "grad_norm": 0.005055158864706755,
      "learning_rate": 0.05143218891574479,
      "loss": 0.6008,
      "num_input_tokens_seen": 40047344,
      "step": 29130
    },
    {
      "epoch": 0.9325587350361693,
      "grad_norm": 0.003775889752432704,
      "learning_rate": 0.0513877947930125,
      "loss": 0.6195,
      "num_input_tokens_seen": 40054336,
      "step": 29135
    },
    {
      "epoch": 0.9327187760066578,
      "grad_norm": 0.006970301736146212,
      "learning_rate": 0.051343415877521566,
      "loss": 0.5681,
      "num_input_tokens_seen": 40061312,
      "step": 29140
    },
    {
      "epoch": 0.9328788169771461,
      "grad_norm": 0.004162691533565521,
      "learning_rate": 0.051299052176115634,
      "loss": 0.6701,
      "num_input_tokens_seen": 40068064,
      "step": 29145
    },
    {
      "epoch": 0.9330388579476346,
      "grad_norm": 0.005836218595504761,
      "learning_rate": 0.051254703695636256,
      "loss": 0.4712,
      "num_input_tokens_seen": 40075328,
      "step": 29150
    },
    {
      "epoch": 0.933198898918123,
      "grad_norm": 0.0033626938238739967,
      "learning_rate": 0.05121037044292249,
      "loss": 0.5668,
      "num_input_tokens_seen": 40082032,
      "step": 29155
    },
    {
      "epoch": 0.9333589398886115,
      "grad_norm": 0.004220439121127129,
      "learning_rate": 0.05116605242481101,
      "loss": 0.424,
      "num_input_tokens_seen": 40088832,
      "step": 29160
    },
    {
      "epoch": 0.9335189808590999,
      "grad_norm": 0.005224033258855343,
      "learning_rate": 0.05112174964813634,
      "loss": 0.6163,
      "num_input_tokens_seen": 40096112,
      "step": 29165
    },
    {
      "epoch": 0.9336790218295884,
      "grad_norm": 0.0033759120851755142,
      "learning_rate": 0.05107746211973038,
      "loss": 0.6512,
      "num_input_tokens_seen": 40103232,
      "step": 29170
    },
    {
      "epoch": 0.9338390628000768,
      "grad_norm": 0.004019856918603182,
      "learning_rate": 0.05103318984642291,
      "loss": 0.503,
      "num_input_tokens_seen": 40110256,
      "step": 29175
    },
    {
      "epoch": 0.9339991037705653,
      "grad_norm": 0.006322816479951143,
      "learning_rate": 0.05098893283504131,
      "loss": 0.438,
      "num_input_tokens_seen": 40117680,
      "step": 29180
    },
    {
      "epoch": 0.9341591447410537,
      "grad_norm": 0.005106933414936066,
      "learning_rate": 0.050944691092410475,
      "loss": 0.5617,
      "num_input_tokens_seen": 40124336,
      "step": 29185
    },
    {
      "epoch": 0.9343191857115422,
      "grad_norm": 0.005221895407885313,
      "learning_rate": 0.05090046462535313,
      "loss": 0.6521,
      "num_input_tokens_seen": 40131136,
      "step": 29190
    },
    {
      "epoch": 0.9344792266820307,
      "grad_norm": 0.005306974519044161,
      "learning_rate": 0.050856253440689454,
      "loss": 0.497,
      "num_input_tokens_seen": 40138064,
      "step": 29195
    },
    {
      "epoch": 0.934639267652519,
      "grad_norm": 0.003942559938877821,
      "learning_rate": 0.050812057545237405,
      "loss": 0.5425,
      "num_input_tokens_seen": 40144672,
      "step": 29200
    },
    {
      "epoch": 0.934639267652519,
      "eval_loss": 0.5090768933296204,
      "eval_runtime": 332.4867,
      "eval_samples_per_second": 41.764,
      "eval_steps_per_second": 20.882,
      "num_input_tokens_seen": 40144672,
      "step": 29200
    },
    {
      "epoch": 0.9347993086230075,
      "grad_norm": 0.002980067627504468,
      "learning_rate": 0.0507678769458126,
      "loss": 0.3869,
      "num_input_tokens_seen": 40151824,
      "step": 29205
    },
    {
      "epoch": 0.9349593495934959,
      "grad_norm": 0.0038325353525578976,
      "learning_rate": 0.050723711649228155,
      "loss": 0.6095,
      "num_input_tokens_seen": 40158784,
      "step": 29210
    },
    {
      "epoch": 0.9351193905639844,
      "grad_norm": 0.004647033754736185,
      "learning_rate": 0.05067956166229496,
      "loss": 0.4895,
      "num_input_tokens_seen": 40165984,
      "step": 29215
    },
    {
      "epoch": 0.9352794315344728,
      "grad_norm": 0.008028712123632431,
      "learning_rate": 0.05063542699182155,
      "loss": 0.5485,
      "num_input_tokens_seen": 40173088,
      "step": 29220
    },
    {
      "epoch": 0.9354394725049613,
      "grad_norm": 0.0028699396643787622,
      "learning_rate": 0.050591307644613996,
      "loss": 0.3883,
      "num_input_tokens_seen": 40179440,
      "step": 29225
    },
    {
      "epoch": 0.9355995134754497,
      "grad_norm": 0.005391864571720362,
      "learning_rate": 0.05054720362747599,
      "loss": 0.5009,
      "num_input_tokens_seen": 40186432,
      "step": 29230
    },
    {
      "epoch": 0.9357595544459382,
      "grad_norm": 0.0034850677475333214,
      "learning_rate": 0.050503114947209035,
      "loss": 0.4926,
      "num_input_tokens_seen": 40193424,
      "step": 29235
    },
    {
      "epoch": 0.9359195954164266,
      "grad_norm": 0.0023230910301208496,
      "learning_rate": 0.05045904161061207,
      "loss": 0.399,
      "num_input_tokens_seen": 40200336,
      "step": 29240
    },
    {
      "epoch": 0.936079636386915,
      "grad_norm": 0.006460570264607668,
      "learning_rate": 0.05041498362448185,
      "loss": 0.4586,
      "num_input_tokens_seen": 40207472,
      "step": 29245
    },
    {
      "epoch": 0.9362396773574035,
      "grad_norm": 0.0038835732266306877,
      "learning_rate": 0.05037094099561256,
      "loss": 0.4567,
      "num_input_tokens_seen": 40214000,
      "step": 29250
    },
    {
      "epoch": 0.9363997183278919,
      "grad_norm": 0.004595678299665451,
      "learning_rate": 0.05032691373079624,
      "loss": 0.5056,
      "num_input_tokens_seen": 40220560,
      "step": 29255
    },
    {
      "epoch": 0.9365597592983804,
      "grad_norm": 0.0033376612700521946,
      "learning_rate": 0.05028290183682234,
      "loss": 0.6471,
      "num_input_tokens_seen": 40227712,
      "step": 29260
    },
    {
      "epoch": 0.9367198002688688,
      "grad_norm": 0.003493408439680934,
      "learning_rate": 0.050238905320478096,
      "loss": 0.5514,
      "num_input_tokens_seen": 40234480,
      "step": 29265
    },
    {
      "epoch": 0.9368798412393573,
      "grad_norm": 0.006550697609782219,
      "learning_rate": 0.05019492418854838,
      "loss": 0.7386,
      "num_input_tokens_seen": 40241856,
      "step": 29270
    },
    {
      "epoch": 0.9370398822098457,
      "grad_norm": 0.004449139349162579,
      "learning_rate": 0.05015095844781554,
      "loss": 0.4983,
      "num_input_tokens_seen": 40248448,
      "step": 29275
    },
    {
      "epoch": 0.9371999231803342,
      "grad_norm": 0.00532534159719944,
      "learning_rate": 0.05010700810505968,
      "loss": 0.3576,
      "num_input_tokens_seen": 40255200,
      "step": 29280
    },
    {
      "epoch": 0.9373599641508226,
      "grad_norm": 0.00463995523750782,
      "learning_rate": 0.05006307316705856,
      "loss": 0.5123,
      "num_input_tokens_seen": 40261872,
      "step": 29285
    },
    {
      "epoch": 0.9375200051213111,
      "grad_norm": 0.003582240315154195,
      "learning_rate": 0.0500191536405874,
      "loss": 0.5707,
      "num_input_tokens_seen": 40268496,
      "step": 29290
    },
    {
      "epoch": 0.9376800460917994,
      "grad_norm": 0.005352701526135206,
      "learning_rate": 0.04997524953241922,
      "loss": 0.4693,
      "num_input_tokens_seen": 40275200,
      "step": 29295
    },
    {
      "epoch": 0.9378400870622879,
      "grad_norm": 0.004411535337567329,
      "learning_rate": 0.049931360849324556,
      "loss": 0.5761,
      "num_input_tokens_seen": 40282320,
      "step": 29300
    },
    {
      "epoch": 0.9380001280327764,
      "grad_norm": 0.0033008940517902374,
      "learning_rate": 0.04988748759807155,
      "loss": 0.3422,
      "num_input_tokens_seen": 40289184,
      "step": 29305
    },
    {
      "epoch": 0.9381601690032648,
      "grad_norm": 0.004365883767604828,
      "learning_rate": 0.0498436297854261,
      "loss": 0.5348,
      "num_input_tokens_seen": 40296000,
      "step": 29310
    },
    {
      "epoch": 0.9383202099737533,
      "grad_norm": 0.004834075458347797,
      "learning_rate": 0.04979978741815152,
      "loss": 0.5605,
      "num_input_tokens_seen": 40302720,
      "step": 29315
    },
    {
      "epoch": 0.9384802509442417,
      "grad_norm": 0.0026593240909278393,
      "learning_rate": 0.04975596050300891,
      "loss": 0.469,
      "num_input_tokens_seen": 40309584,
      "step": 29320
    },
    {
      "epoch": 0.9386402919147302,
      "grad_norm": 0.002661466831341386,
      "learning_rate": 0.049712149046757005,
      "loss": 0.4336,
      "num_input_tokens_seen": 40316576,
      "step": 29325
    },
    {
      "epoch": 0.9388003328852186,
      "grad_norm": 0.0028860156890004873,
      "learning_rate": 0.04966835305615194,
      "loss": 0.5709,
      "num_input_tokens_seen": 40323488,
      "step": 29330
    },
    {
      "epoch": 0.9389603738557071,
      "grad_norm": 0.0041739316657185555,
      "learning_rate": 0.049624572537947755,
      "loss": 0.5349,
      "num_input_tokens_seen": 40330768,
      "step": 29335
    },
    {
      "epoch": 0.9391204148261955,
      "grad_norm": 0.0033079548738896847,
      "learning_rate": 0.04958080749889582,
      "loss": 0.489,
      "num_input_tokens_seen": 40337760,
      "step": 29340
    },
    {
      "epoch": 0.939280455796684,
      "grad_norm": 0.003428333206102252,
      "learning_rate": 0.049537057945745304,
      "loss": 0.5107,
      "num_input_tokens_seen": 40344256,
      "step": 29345
    },
    {
      "epoch": 0.9394404967671725,
      "grad_norm": 0.00736693711951375,
      "learning_rate": 0.049493323885243,
      "loss": 0.6072,
      "num_input_tokens_seen": 40352416,
      "step": 29350
    },
    {
      "epoch": 0.9396005377376608,
      "grad_norm": 0.003685912350192666,
      "learning_rate": 0.04944960532413318,
      "loss": 0.467,
      "num_input_tokens_seen": 40359472,
      "step": 29355
    },
    {
      "epoch": 0.9397605787081493,
      "grad_norm": 0.0034351462963968515,
      "learning_rate": 0.049405902269157774,
      "loss": 0.3763,
      "num_input_tokens_seen": 40366864,
      "step": 29360
    },
    {
      "epoch": 0.9399206196786377,
      "grad_norm": 0.0057803066447377205,
      "learning_rate": 0.04936221472705646,
      "loss": 0.4465,
      "num_input_tokens_seen": 40373520,
      "step": 29365
    },
    {
      "epoch": 0.9400806606491262,
      "grad_norm": 0.009594579227268696,
      "learning_rate": 0.04931854270456632,
      "loss": 0.5736,
      "num_input_tokens_seen": 40380752,
      "step": 29370
    },
    {
      "epoch": 0.9402407016196146,
      "grad_norm": 0.004353852476924658,
      "learning_rate": 0.049274886208422075,
      "loss": 0.7131,
      "num_input_tokens_seen": 40387712,
      "step": 29375
    },
    {
      "epoch": 0.9404007425901031,
      "grad_norm": 0.00624494906514883,
      "learning_rate": 0.049231245245356235,
      "loss": 0.4731,
      "num_input_tokens_seen": 40394320,
      "step": 29380
    },
    {
      "epoch": 0.9405607835605915,
      "grad_norm": 0.007510931231081486,
      "learning_rate": 0.049187619822098655,
      "loss": 0.6718,
      "num_input_tokens_seen": 40400912,
      "step": 29385
    },
    {
      "epoch": 0.94072082453108,
      "grad_norm": 0.004257551860064268,
      "learning_rate": 0.04914400994537705,
      "loss": 0.6878,
      "num_input_tokens_seen": 40407424,
      "step": 29390
    },
    {
      "epoch": 0.9408808655015684,
      "grad_norm": 0.0036328418646007776,
      "learning_rate": 0.049100415621916485,
      "loss": 0.5109,
      "num_input_tokens_seen": 40414240,
      "step": 29395
    },
    {
      "epoch": 0.9410409064720568,
      "grad_norm": 0.005520147737115622,
      "learning_rate": 0.04905683685843981,
      "loss": 0.4358,
      "num_input_tokens_seen": 40420752,
      "step": 29400
    },
    {
      "epoch": 0.9410409064720568,
      "eval_loss": 0.5061813592910767,
      "eval_runtime": 332.4729,
      "eval_samples_per_second": 41.766,
      "eval_steps_per_second": 20.883,
      "num_input_tokens_seen": 40420752,
      "step": 29400
    },
    {
      "epoch": 0.9412009474425453,
      "grad_norm": 0.0032959249801933765,
      "learning_rate": 0.049013273661667495,
      "loss": 0.4781,
      "num_input_tokens_seen": 40427792,
      "step": 29405
    },
    {
      "epoch": 0.9413609884130337,
      "grad_norm": 0.004417938645929098,
      "learning_rate": 0.048969726038317396,
      "loss": 0.4588,
      "num_input_tokens_seen": 40435024,
      "step": 29410
    },
    {
      "epoch": 0.9415210293835222,
      "grad_norm": 0.0033155870623886585,
      "learning_rate": 0.048926193995105206,
      "loss": 0.4934,
      "num_input_tokens_seen": 40441856,
      "step": 29415
    },
    {
      "epoch": 0.9416810703540106,
      "grad_norm": 0.0033673287834972143,
      "learning_rate": 0.048882677538744035,
      "loss": 0.4072,
      "num_input_tokens_seen": 40448672,
      "step": 29420
    },
    {
      "epoch": 0.9418411113244991,
      "grad_norm": 0.0061150845140218735,
      "learning_rate": 0.048839176675944715,
      "loss": 0.6136,
      "num_input_tokens_seen": 40455360,
      "step": 29425
    },
    {
      "epoch": 0.9420011522949875,
      "grad_norm": 0.0037787179462611675,
      "learning_rate": 0.04879569141341566,
      "loss": 0.5853,
      "num_input_tokens_seen": 40462144,
      "step": 29430
    },
    {
      "epoch": 0.942161193265476,
      "grad_norm": 0.004569860175251961,
      "learning_rate": 0.04875222175786274,
      "loss": 0.4528,
      "num_input_tokens_seen": 40469168,
      "step": 29435
    },
    {
      "epoch": 0.9423212342359644,
      "grad_norm": 0.005955499596893787,
      "learning_rate": 0.04870876771598966,
      "loss": 0.4053,
      "num_input_tokens_seen": 40475856,
      "step": 29440
    },
    {
      "epoch": 0.9424812752064529,
      "grad_norm": 0.0037000325974076986,
      "learning_rate": 0.04866532929449744,
      "loss": 0.4921,
      "num_input_tokens_seen": 40482688,
      "step": 29445
    },
    {
      "epoch": 0.9426413161769412,
      "grad_norm": 0.004796482156962156,
      "learning_rate": 0.048621906500084945,
      "loss": 0.6281,
      "num_input_tokens_seen": 40489568,
      "step": 29450
    },
    {
      "epoch": 0.9428013571474297,
      "grad_norm": 0.0036219151224941015,
      "learning_rate": 0.04857849933944845,
      "loss": 0.3823,
      "num_input_tokens_seen": 40496624,
      "step": 29455
    },
    {
      "epoch": 0.9429613981179182,
      "grad_norm": 0.00257282517850399,
      "learning_rate": 0.048535107819281866,
      "loss": 0.2877,
      "num_input_tokens_seen": 40503328,
      "step": 29460
    },
    {
      "epoch": 0.9431214390884066,
      "grad_norm": 0.003024621866643429,
      "learning_rate": 0.04849173194627675,
      "loss": 0.5682,
      "num_input_tokens_seen": 40510208,
      "step": 29465
    },
    {
      "epoch": 0.9432814800588951,
      "grad_norm": 0.005060144700109959,
      "learning_rate": 0.04844837172712223,
      "loss": 0.3543,
      "num_input_tokens_seen": 40517152,
      "step": 29470
    },
    {
      "epoch": 0.9434415210293835,
      "grad_norm": 0.004970639478415251,
      "learning_rate": 0.04840502716850494,
      "loss": 0.6577,
      "num_input_tokens_seen": 40524112,
      "step": 29475
    },
    {
      "epoch": 0.943601561999872,
      "grad_norm": 0.008455494418740273,
      "learning_rate": 0.04836169827710916,
      "loss": 0.5283,
      "num_input_tokens_seen": 40531440,
      "step": 29480
    },
    {
      "epoch": 0.9437616029703604,
      "grad_norm": 0.0027620703913271427,
      "learning_rate": 0.04831838505961684,
      "loss": 0.3,
      "num_input_tokens_seen": 40538368,
      "step": 29485
    },
    {
      "epoch": 0.9439216439408489,
      "grad_norm": 0.008153427392244339,
      "learning_rate": 0.048275087522707295,
      "loss": 0.5125,
      "num_input_tokens_seen": 40545408,
      "step": 29490
    },
    {
      "epoch": 0.9440816849113373,
      "grad_norm": 0.006875257007777691,
      "learning_rate": 0.04823180567305766,
      "loss": 0.6343,
      "num_input_tokens_seen": 40552528,
      "step": 29495
    },
    {
      "epoch": 0.9442417258818258,
      "grad_norm": 0.004938194528222084,
      "learning_rate": 0.04818853951734244,
      "loss": 0.6785,
      "num_input_tokens_seen": 40559824,
      "step": 29500
    },
    {
      "epoch": 0.9444017668523141,
      "grad_norm": 0.0064884377643466,
      "learning_rate": 0.04814528906223387,
      "loss": 0.5466,
      "num_input_tokens_seen": 40566368,
      "step": 29505
    },
    {
      "epoch": 0.9445618078228026,
      "grad_norm": 0.007683447562158108,
      "learning_rate": 0.04810205431440177,
      "loss": 0.4665,
      "num_input_tokens_seen": 40573536,
      "step": 29510
    },
    {
      "epoch": 0.9447218487932911,
      "grad_norm": 0.0027082539163529873,
      "learning_rate": 0.04805883528051341,
      "loss": 0.5004,
      "num_input_tokens_seen": 40580064,
      "step": 29515
    },
    {
      "epoch": 0.9448818897637795,
      "grad_norm": 0.005963844712823629,
      "learning_rate": 0.048015631967233685,
      "loss": 0.6363,
      "num_input_tokens_seen": 40586928,
      "step": 29520
    },
    {
      "epoch": 0.945041930734268,
      "grad_norm": 0.0031938699539750814,
      "learning_rate": 0.04797244438122517,
      "loss": 0.303,
      "num_input_tokens_seen": 40593760,
      "step": 29525
    },
    {
      "epoch": 0.9452019717047564,
      "grad_norm": 0.00262182648293674,
      "learning_rate": 0.04792927252914784,
      "loss": 0.4164,
      "num_input_tokens_seen": 40601264,
      "step": 29530
    },
    {
      "epoch": 0.9453620126752449,
      "grad_norm": 0.004220705013722181,
      "learning_rate": 0.04788611641765944,
      "loss": 0.5153,
      "num_input_tokens_seen": 40608000,
      "step": 29535
    },
    {
      "epoch": 0.9455220536457333,
      "grad_norm": 0.005698947235941887,
      "learning_rate": 0.04784297605341508,
      "loss": 0.4542,
      "num_input_tokens_seen": 40614608,
      "step": 29540
    },
    {
      "epoch": 0.9456820946162218,
      "grad_norm": 0.004496917128562927,
      "learning_rate": 0.04779985144306761,
      "loss": 0.5044,
      "num_input_tokens_seen": 40621152,
      "step": 29545
    },
    {
      "epoch": 0.9458421355867102,
      "grad_norm": 0.006778939161449671,
      "learning_rate": 0.047756742593267405,
      "loss": 0.6322,
      "num_input_tokens_seen": 40628160,
      "step": 29550
    },
    {
      "epoch": 0.9460021765571986,
      "grad_norm": 0.0038803904317319393,
      "learning_rate": 0.047713649510662315,
      "loss": 0.509,
      "num_input_tokens_seen": 40635152,
      "step": 29555
    },
    {
      "epoch": 0.946162217527687,
      "grad_norm": 0.005107265431433916,
      "learning_rate": 0.04767057220189789,
      "loss": 0.4257,
      "num_input_tokens_seen": 40642608,
      "step": 29560
    },
    {
      "epoch": 0.9463222584981755,
      "grad_norm": 0.003394113853573799,
      "learning_rate": 0.04762751067361722,
      "loss": 0.431,
      "num_input_tokens_seen": 40649456,
      "step": 29565
    },
    {
      "epoch": 0.946482299468664,
      "grad_norm": 0.007173788733780384,
      "learning_rate": 0.04758446493246086,
      "loss": 0.5072,
      "num_input_tokens_seen": 40656352,
      "step": 29570
    },
    {
      "epoch": 0.9466423404391524,
      "grad_norm": 0.0042920359410345554,
      "learning_rate": 0.047541434985067084,
      "loss": 0.5148,
      "num_input_tokens_seen": 40663200,
      "step": 29575
    },
    {
      "epoch": 0.9468023814096409,
      "grad_norm": 0.007195353973656893,
      "learning_rate": 0.047498420838071556,
      "loss": 0.5318,
      "num_input_tokens_seen": 40669808,
      "step": 29580
    },
    {
      "epoch": 0.9469624223801293,
      "grad_norm": 0.006300772074609995,
      "learning_rate": 0.04745542249810772,
      "loss": 0.3076,
      "num_input_tokens_seen": 40676368,
      "step": 29585
    },
    {
      "epoch": 0.9471224633506178,
      "grad_norm": 0.005022415891289711,
      "learning_rate": 0.047412439971806324,
      "loss": 0.4916,
      "num_input_tokens_seen": 40683264,
      "step": 29590
    },
    {
      "epoch": 0.9472825043211062,
      "grad_norm": 0.006741049699485302,
      "learning_rate": 0.04736947326579592,
      "loss": 0.7441,
      "num_input_tokens_seen": 40689984,
      "step": 29595
    },
    {
      "epoch": 0.9474425452915947,
      "grad_norm": 0.004296532366424799,
      "learning_rate": 0.04732652238670245,
      "loss": 0.4746,
      "num_input_tokens_seen": 40696672,
      "step": 29600
    },
    {
      "epoch": 0.9474425452915947,
      "eval_loss": 0.5063950419425964,
      "eval_runtime": 332.3351,
      "eval_samples_per_second": 41.783,
      "eval_steps_per_second": 20.892,
      "num_input_tokens_seen": 40696672,
      "step": 29600
    },
    {
      "epoch": 0.947602586262083,
      "grad_norm": 0.003097301349043846,
      "learning_rate": 0.04728358734114952,
      "loss": 0.3787,
      "num_input_tokens_seen": 40703120,
      "step": 29605
    },
    {
      "epoch": 0.9477626272325715,
      "grad_norm": 0.0029811635613441467,
      "learning_rate": 0.04724066813575821,
      "loss": 0.474,
      "num_input_tokens_seen": 40709936,
      "step": 29610
    },
    {
      "epoch": 0.94792266820306,
      "grad_norm": 0.004891373682767153,
      "learning_rate": 0.04719776477714729,
      "loss": 0.5539,
      "num_input_tokens_seen": 40717296,
      "step": 29615
    },
    {
      "epoch": 0.9480827091735484,
      "grad_norm": 0.003446088172495365,
      "learning_rate": 0.047154877271932856,
      "loss": 0.523,
      "num_input_tokens_seen": 40724032,
      "step": 29620
    },
    {
      "epoch": 0.9482427501440369,
      "grad_norm": 0.0036176105495542288,
      "learning_rate": 0.0471120056267288,
      "loss": 0.3933,
      "num_input_tokens_seen": 40730768,
      "step": 29625
    },
    {
      "epoch": 0.9484027911145253,
      "grad_norm": 0.008908878080546856,
      "learning_rate": 0.047069149848146495,
      "loss": 0.6911,
      "num_input_tokens_seen": 40737744,
      "step": 29630
    },
    {
      "epoch": 0.9485628320850138,
      "grad_norm": 0.002764762844890356,
      "learning_rate": 0.04702630994279473,
      "loss": 0.638,
      "num_input_tokens_seen": 40744400,
      "step": 29635
    },
    {
      "epoch": 0.9487228730555022,
      "grad_norm": 0.006911026779562235,
      "learning_rate": 0.046983485917280035,
      "loss": 0.5418,
      "num_input_tokens_seen": 40751024,
      "step": 29640
    },
    {
      "epoch": 0.9488829140259907,
      "grad_norm": 0.005738649982959032,
      "learning_rate": 0.04694067777820644,
      "loss": 0.5721,
      "num_input_tokens_seen": 40757808,
      "step": 29645
    },
    {
      "epoch": 0.9490429549964791,
      "grad_norm": 0.008341649547219276,
      "learning_rate": 0.046897885532175415,
      "loss": 0.7275,
      "num_input_tokens_seen": 40764672,
      "step": 29650
    },
    {
      "epoch": 0.9492029959669676,
      "grad_norm": 0.0050587463192641735,
      "learning_rate": 0.04685510918578613,
      "loss": 0.531,
      "num_input_tokens_seen": 40771264,
      "step": 29655
    },
    {
      "epoch": 0.9493630369374559,
      "grad_norm": 0.0054163504391908646,
      "learning_rate": 0.04681234874563519,
      "loss": 0.7181,
      "num_input_tokens_seen": 40778256,
      "step": 29660
    },
    {
      "epoch": 0.9495230779079444,
      "grad_norm": 0.002678815508261323,
      "learning_rate": 0.046769604218316836,
      "loss": 0.3897,
      "num_input_tokens_seen": 40784912,
      "step": 29665
    },
    {
      "epoch": 0.9496831188784329,
      "grad_norm": 0.0035178051330149174,
      "learning_rate": 0.04672687561042279,
      "loss": 0.5006,
      "num_input_tokens_seen": 40791952,
      "step": 29670
    },
    {
      "epoch": 0.9498431598489213,
      "grad_norm": 0.007781251799315214,
      "learning_rate": 0.046684162928542286,
      "loss": 0.4812,
      "num_input_tokens_seen": 40798944,
      "step": 29675
    },
    {
      "epoch": 0.9500032008194098,
      "grad_norm": 0.0024857702665030956,
      "learning_rate": 0.04664146617926222,
      "loss": 0.3919,
      "num_input_tokens_seen": 40806048,
      "step": 29680
    },
    {
      "epoch": 0.9501632417898982,
      "grad_norm": 0.006528627127408981,
      "learning_rate": 0.046598785369167,
      "loss": 0.7607,
      "num_input_tokens_seen": 40813040,
      "step": 29685
    },
    {
      "epoch": 0.9503232827603867,
      "grad_norm": 0.0044284723699092865,
      "learning_rate": 0.046556120504838434,
      "loss": 0.3853,
      "num_input_tokens_seen": 40819792,
      "step": 29690
    },
    {
      "epoch": 0.9504833237308751,
      "grad_norm": 0.007295149378478527,
      "learning_rate": 0.04651347159285609,
      "loss": 0.4635,
      "num_input_tokens_seen": 40826336,
      "step": 29695
    },
    {
      "epoch": 0.9506433647013636,
      "grad_norm": 0.004241581540554762,
      "learning_rate": 0.04647083863979688,
      "loss": 0.6501,
      "num_input_tokens_seen": 40833264,
      "step": 29700
    },
    {
      "epoch": 0.950803405671852,
      "grad_norm": 0.004131798632442951,
      "learning_rate": 0.04642822165223538,
      "loss": 0.4162,
      "num_input_tokens_seen": 40840000,
      "step": 29705
    },
    {
      "epoch": 0.9509634466423404,
      "grad_norm": 0.0030168145895004272,
      "learning_rate": 0.046385620636743716,
      "loss": 0.547,
      "num_input_tokens_seen": 40846864,
      "step": 29710
    },
    {
      "epoch": 0.9511234876128288,
      "grad_norm": 0.0026786362286657095,
      "learning_rate": 0.04634303559989141,
      "loss": 0.5089,
      "num_input_tokens_seen": 40853504,
      "step": 29715
    },
    {
      "epoch": 0.9512835285833173,
      "grad_norm": 0.0035833101719617844,
      "learning_rate": 0.046300466548245635,
      "loss": 0.4412,
      "num_input_tokens_seen": 40860608,
      "step": 29720
    },
    {
      "epoch": 0.9514435695538058,
      "grad_norm": 0.002966301515698433,
      "learning_rate": 0.04625791348837114,
      "loss": 0.6985,
      "num_input_tokens_seen": 40867328,
      "step": 29725
    },
    {
      "epoch": 0.9516036105242942,
      "grad_norm": 0.0049364278092980385,
      "learning_rate": 0.046215376426830095,
      "loss": 0.4947,
      "num_input_tokens_seen": 40874064,
      "step": 29730
    },
    {
      "epoch": 0.9517636514947827,
      "grad_norm": 0.0031876368448138237,
      "learning_rate": 0.04617285537018219,
      "loss": 0.4828,
      "num_input_tokens_seen": 40880784,
      "step": 29735
    },
    {
      "epoch": 0.9519236924652711,
      "grad_norm": 0.0033525226172059774,
      "learning_rate": 0.046130350324984803,
      "loss": 0.5526,
      "num_input_tokens_seen": 40887744,
      "step": 29740
    },
    {
      "epoch": 0.9520837334357596,
      "grad_norm": 0.0036841505207121372,
      "learning_rate": 0.046087861297792666,
      "loss": 0.4175,
      "num_input_tokens_seen": 40894464,
      "step": 29745
    },
    {
      "epoch": 0.952243774406248,
      "grad_norm": 0.003951814025640488,
      "learning_rate": 0.0460453882951582,
      "loss": 0.3298,
      "num_input_tokens_seen": 40900976,
      "step": 29750
    },
    {
      "epoch": 0.9524038153767365,
      "grad_norm": 0.008926313370466232,
      "learning_rate": 0.04600293132363119,
      "loss": 0.5697,
      "num_input_tokens_seen": 40907824,
      "step": 29755
    },
    {
      "epoch": 0.9525638563472248,
      "grad_norm": 0.0058607663959264755,
      "learning_rate": 0.045960490389759086,
      "loss": 0.7368,
      "num_input_tokens_seen": 40914400,
      "step": 29760
    },
    {
      "epoch": 0.9527238973177133,
      "grad_norm": 0.004375599790364504,
      "learning_rate": 0.04591806550008685,
      "loss": 0.3755,
      "num_input_tokens_seen": 40920976,
      "step": 29765
    },
    {
      "epoch": 0.9528839382882017,
      "grad_norm": 0.0032546864822506905,
      "learning_rate": 0.045875656661156825,
      "loss": 0.5346,
      "num_input_tokens_seen": 40927952,
      "step": 29770
    },
    {
      "epoch": 0.9530439792586902,
      "grad_norm": 0.003987098578363657,
      "learning_rate": 0.04583326387950911,
      "loss": 0.4975,
      "num_input_tokens_seen": 40935168,
      "step": 29775
    },
    {
      "epoch": 0.9532040202291787,
      "grad_norm": 0.0028453231789171696,
      "learning_rate": 0.0457908871616811,
      "loss": 0.4371,
      "num_input_tokens_seen": 40941840,
      "step": 29780
    },
    {
      "epoch": 0.9533640611996671,
      "grad_norm": 0.0034389779902994633,
      "learning_rate": 0.04574852651420786,
      "loss": 0.4486,
      "num_input_tokens_seen": 40949136,
      "step": 29785
    },
    {
      "epoch": 0.9535241021701556,
      "grad_norm": 0.003941282629966736,
      "learning_rate": 0.045706181943621985,
      "loss": 0.5819,
      "num_input_tokens_seen": 40956400,
      "step": 29790
    },
    {
      "epoch": 0.953684143140644,
      "grad_norm": 0.002214349340647459,
      "learning_rate": 0.04566385345645344,
      "loss": 0.4002,
      "num_input_tokens_seen": 40963344,
      "step": 29795
    },
    {
      "epoch": 0.9538441841111325,
      "grad_norm": 0.006053952034562826,
      "learning_rate": 0.04562154105922993,
      "loss": 0.3488,
      "num_input_tokens_seen": 40970096,
      "step": 29800
    },
    {
      "epoch": 0.9538441841111325,
      "eval_loss": 0.5054609179496765,
      "eval_runtime": 331.909,
      "eval_samples_per_second": 41.837,
      "eval_steps_per_second": 20.918,
      "num_input_tokens_seen": 40970096,
      "step": 29800
    },
    {
      "epoch": 0.9540042250816209,
      "grad_norm": 0.004150104243308306,
      "learning_rate": 0.04557924475847642,
      "loss": 0.4916,
      "num_input_tokens_seen": 40977296,
      "step": 29805
    },
    {
      "epoch": 0.9541642660521094,
      "grad_norm": 0.003005371894687414,
      "learning_rate": 0.04553696456071567,
      "loss": 0.4655,
      "num_input_tokens_seen": 40984320,
      "step": 29810
    },
    {
      "epoch": 0.9543243070225977,
      "grad_norm": 0.0027803967241197824,
      "learning_rate": 0.045494700472467724,
      "loss": 0.3726,
      "num_input_tokens_seen": 40991040,
      "step": 29815
    },
    {
      "epoch": 0.9544843479930862,
      "grad_norm": 0.0038955591153353453,
      "learning_rate": 0.04545245250025024,
      "loss": 0.6826,
      "num_input_tokens_seen": 40997776,
      "step": 29820
    },
    {
      "epoch": 0.9546443889635747,
      "grad_norm": 0.0038497643545269966,
      "learning_rate": 0.045410220650578384,
      "loss": 0.6323,
      "num_input_tokens_seen": 41004384,
      "step": 29825
    },
    {
      "epoch": 0.9548044299340631,
      "grad_norm": 0.003972586710005999,
      "learning_rate": 0.04536800492996492,
      "loss": 0.4625,
      "num_input_tokens_seen": 41011072,
      "step": 29830
    },
    {
      "epoch": 0.9549644709045516,
      "grad_norm": 0.002907675690948963,
      "learning_rate": 0.04532580534491994,
      "loss": 0.341,
      "num_input_tokens_seen": 41018032,
      "step": 29835
    },
    {
      "epoch": 0.95512451187504,
      "grad_norm": 0.004245399497449398,
      "learning_rate": 0.045283621901951183,
      "loss": 0.5044,
      "num_input_tokens_seen": 41025056,
      "step": 29840
    },
    {
      "epoch": 0.9552845528455285,
      "grad_norm": 0.006661683786660433,
      "learning_rate": 0.04524145460756393,
      "loss": 0.5541,
      "num_input_tokens_seen": 41031760,
      "step": 29845
    },
    {
      "epoch": 0.9554445938160169,
      "grad_norm": 0.005788080859929323,
      "learning_rate": 0.045199303468260794,
      "loss": 0.3899,
      "num_input_tokens_seen": 41038400,
      "step": 29850
    },
    {
      "epoch": 0.9556046347865054,
      "grad_norm": 0.0060399556532502174,
      "learning_rate": 0.04515716849054214,
      "loss": 0.4246,
      "num_input_tokens_seen": 41045152,
      "step": 29855
    },
    {
      "epoch": 0.9557646757569938,
      "grad_norm": 0.0031920215114951134,
      "learning_rate": 0.04511504968090558,
      "loss": 0.5599,
      "num_input_tokens_seen": 41052208,
      "step": 29860
    },
    {
      "epoch": 0.9559247167274822,
      "grad_norm": 0.003915423061698675,
      "learning_rate": 0.04507294704584644,
      "loss": 0.4897,
      "num_input_tokens_seen": 41058816,
      "step": 29865
    },
    {
      "epoch": 0.9560847576979706,
      "grad_norm": 0.006931809708476067,
      "learning_rate": 0.04503086059185749,
      "loss": 0.5383,
      "num_input_tokens_seen": 41065600,
      "step": 29870
    },
    {
      "epoch": 0.9562447986684591,
      "grad_norm": 0.0041376822628080845,
      "learning_rate": 0.04498879032542893,
      "loss": 0.4494,
      "num_input_tokens_seen": 41072272,
      "step": 29875
    },
    {
      "epoch": 0.9564048396389476,
      "grad_norm": 0.005781719461083412,
      "learning_rate": 0.0449467362530486,
      "loss": 0.4788,
      "num_input_tokens_seen": 41079728,
      "step": 29880
    },
    {
      "epoch": 0.956564880609436,
      "grad_norm": 0.003852982772514224,
      "learning_rate": 0.04490469838120171,
      "loss": 0.4415,
      "num_input_tokens_seen": 41086432,
      "step": 29885
    },
    {
      "epoch": 0.9567249215799245,
      "grad_norm": 0.0031371493823826313,
      "learning_rate": 0.04486267671637101,
      "loss": 0.4433,
      "num_input_tokens_seen": 41093200,
      "step": 29890
    },
    {
      "epoch": 0.9568849625504129,
      "grad_norm": 0.0013497089967131615,
      "learning_rate": 0.04482067126503683,
      "loss": 0.2683,
      "num_input_tokens_seen": 41100352,
      "step": 29895
    },
    {
      "epoch": 0.9570450035209014,
      "grad_norm": 0.004131306894123554,
      "learning_rate": 0.04477868203367687,
      "loss": 0.574,
      "num_input_tokens_seen": 41107376,
      "step": 29900
    },
    {
      "epoch": 0.9572050444913898,
      "grad_norm": 0.007611459121108055,
      "learning_rate": 0.044736709028766426,
      "loss": 0.5946,
      "num_input_tokens_seen": 41114576,
      "step": 29905
    },
    {
      "epoch": 0.9573650854618783,
      "grad_norm": 0.002049668924883008,
      "learning_rate": 0.04469475225677832,
      "loss": 0.4187,
      "num_input_tokens_seen": 41121712,
      "step": 29910
    },
    {
      "epoch": 0.9575251264323666,
      "grad_norm": 0.004497111774981022,
      "learning_rate": 0.04465281172418273,
      "loss": 0.6201,
      "num_input_tokens_seen": 41128752,
      "step": 29915
    },
    {
      "epoch": 0.9576851674028551,
      "grad_norm": 0.006595288403332233,
      "learning_rate": 0.044610887437447476,
      "loss": 0.3473,
      "num_input_tokens_seen": 41135616,
      "step": 29920
    },
    {
      "epoch": 0.9578452083733435,
      "grad_norm": 0.003054210217669606,
      "learning_rate": 0.044568979403037744,
      "loss": 0.4837,
      "num_input_tokens_seen": 41142144,
      "step": 29925
    },
    {
      "epoch": 0.958005249343832,
      "grad_norm": 0.005214477423578501,
      "learning_rate": 0.04452708762741631,
      "loss": 0.6524,
      "num_input_tokens_seen": 41149456,
      "step": 29930
    },
    {
      "epoch": 0.9581652903143205,
      "grad_norm": 0.004697342403233051,
      "learning_rate": 0.044485212117043475,
      "loss": 0.4028,
      "num_input_tokens_seen": 41156352,
      "step": 29935
    },
    {
      "epoch": 0.9583253312848089,
      "grad_norm": 0.00465439772233367,
      "learning_rate": 0.04444335287837687,
      "loss": 0.456,
      "num_input_tokens_seen": 41163456,
      "step": 29940
    },
    {
      "epoch": 0.9584853722552974,
      "grad_norm": 0.0035518379881978035,
      "learning_rate": 0.04440150991787179,
      "loss": 0.3093,
      "num_input_tokens_seen": 41170624,
      "step": 29945
    },
    {
      "epoch": 0.9586454132257858,
      "grad_norm": 0.006201417651027441,
      "learning_rate": 0.04435968324198088,
      "loss": 0.5469,
      "num_input_tokens_seen": 41177648,
      "step": 29950
    },
    {
      "epoch": 0.9588054541962743,
      "grad_norm": 0.003799100173637271,
      "learning_rate": 0.04431787285715442,
      "loss": 0.611,
      "num_input_tokens_seen": 41184576,
      "step": 29955
    },
    {
      "epoch": 0.9589654951667627,
      "grad_norm": 0.009521030820906162,
      "learning_rate": 0.04427607876984004,
      "loss": 0.4403,
      "num_input_tokens_seen": 41191472,
      "step": 29960
    },
    {
      "epoch": 0.9591255361372512,
      "grad_norm": 0.004429488442838192,
      "learning_rate": 0.044234300986482886,
      "loss": 0.5978,
      "num_input_tokens_seen": 41198352,
      "step": 29965
    },
    {
      "epoch": 0.9592855771077395,
      "grad_norm": 0.003711263881996274,
      "learning_rate": 0.04419253951352566,
      "loss": 0.6305,
      "num_input_tokens_seen": 41205360,
      "step": 29970
    },
    {
      "epoch": 0.959445618078228,
      "grad_norm": 0.006884851027280092,
      "learning_rate": 0.044150794357408533,
      "loss": 0.5354,
      "num_input_tokens_seen": 41212096,
      "step": 29975
    },
    {
      "epoch": 0.9596056590487164,
      "grad_norm": 0.006715836469084024,
      "learning_rate": 0.044109065524569065,
      "loss": 0.4353,
      "num_input_tokens_seen": 41219072,
      "step": 29980
    },
    {
      "epoch": 0.9597657000192049,
      "grad_norm": 0.0037452210672199726,
      "learning_rate": 0.0440673530214424,
      "loss": 0.4684,
      "num_input_tokens_seen": 41225936,
      "step": 29985
    },
    {
      "epoch": 0.9599257409896934,
      "grad_norm": 0.0040886737406253815,
      "learning_rate": 0.04402565685446117,
      "loss": 0.4238,
      "num_input_tokens_seen": 41232512,
      "step": 29990
    },
    {
      "epoch": 0.9600857819601818,
      "grad_norm": 0.00461812037974596,
      "learning_rate": 0.04398397703005536,
      "loss": 0.4833,
      "num_input_tokens_seen": 41239392,
      "step": 29995
    },
    {
      "epoch": 0.9602458229306703,
      "grad_norm": 0.005263759754598141,
      "learning_rate": 0.043942313554652626,
      "loss": 0.5119,
      "num_input_tokens_seen": 41245904,
      "step": 30000
    },
    {
      "epoch": 0.9602458229306703,
      "eval_loss": 0.5056482553482056,
      "eval_runtime": 332.0464,
      "eval_samples_per_second": 41.819,
      "eval_steps_per_second": 20.91,
      "num_input_tokens_seen": 41245904,
      "step": 30000
    },
    {
      "epoch": 0.9604058639011587,
      "grad_norm": 0.004182521253824234,
      "learning_rate": 0.0439006664346779,
      "loss": 0.3686,
      "num_input_tokens_seen": 41253376,
      "step": 30005
    },
    {
      "epoch": 0.9605659048716472,
      "grad_norm": 0.0031444705091416836,
      "learning_rate": 0.043859035676553755,
      "loss": 0.5755,
      "num_input_tokens_seen": 41260400,
      "step": 30010
    },
    {
      "epoch": 0.9607259458421356,
      "grad_norm": 0.005837281234562397,
      "learning_rate": 0.043817421286700194,
      "loss": 0.5735,
      "num_input_tokens_seen": 41267424,
      "step": 30015
    },
    {
      "epoch": 0.960885986812624,
      "grad_norm": 0.003293517278507352,
      "learning_rate": 0.043775823271534585,
      "loss": 0.5073,
      "num_input_tokens_seen": 41274160,
      "step": 30020
    },
    {
      "epoch": 0.9610460277831124,
      "grad_norm": 0.003922365140169859,
      "learning_rate": 0.04373424163747197,
      "loss": 0.5121,
      "num_input_tokens_seen": 41280784,
      "step": 30025
    },
    {
      "epoch": 0.9612060687536009,
      "grad_norm": 0.0033154108095914125,
      "learning_rate": 0.04369267639092473,
      "loss": 0.705,
      "num_input_tokens_seen": 41287424,
      "step": 30030
    },
    {
      "epoch": 0.9613661097240893,
      "grad_norm": 0.00417615519836545,
      "learning_rate": 0.04365112753830268,
      "loss": 0.4956,
      "num_input_tokens_seen": 41294240,
      "step": 30035
    },
    {
      "epoch": 0.9615261506945778,
      "grad_norm": 0.00563512509688735,
      "learning_rate": 0.04360959508601327,
      "loss": 0.5068,
      "num_input_tokens_seen": 41301280,
      "step": 30040
    },
    {
      "epoch": 0.9616861916650663,
      "grad_norm": 0.0066095031797885895,
      "learning_rate": 0.04356807904046123,
      "loss": 0.4242,
      "num_input_tokens_seen": 41308016,
      "step": 30045
    },
    {
      "epoch": 0.9618462326355547,
      "grad_norm": 0.0032114984933286905,
      "learning_rate": 0.04352657940804892,
      "loss": 0.5373,
      "num_input_tokens_seen": 41314736,
      "step": 30050
    },
    {
      "epoch": 0.9620062736060432,
      "grad_norm": 0.004453004337847233,
      "learning_rate": 0.04348509619517613,
      "loss": 0.5761,
      "num_input_tokens_seen": 41321264,
      "step": 30055
    },
    {
      "epoch": 0.9621663145765316,
      "grad_norm": 0.005572655703872442,
      "learning_rate": 0.04344362940824002,
      "loss": 0.3633,
      "num_input_tokens_seen": 41328432,
      "step": 30060
    },
    {
      "epoch": 0.9623263555470201,
      "grad_norm": 0.011464591138064861,
      "learning_rate": 0.04340217905363533,
      "loss": 0.6197,
      "num_input_tokens_seen": 41335200,
      "step": 30065
    },
    {
      "epoch": 0.9624863965175084,
      "grad_norm": 0.004991081077605486,
      "learning_rate": 0.04336074513775425,
      "loss": 0.4345,
      "num_input_tokens_seen": 41341696,
      "step": 30070
    },
    {
      "epoch": 0.9626464374879969,
      "grad_norm": 0.0038940419908612967,
      "learning_rate": 0.04331932766698636,
      "loss": 0.5219,
      "num_input_tokens_seen": 41348688,
      "step": 30075
    },
    {
      "epoch": 0.9628064784584853,
      "grad_norm": 0.0063455840572714806,
      "learning_rate": 0.0432779266477188,
      "loss": 0.4183,
      "num_input_tokens_seen": 41355712,
      "step": 30080
    },
    {
      "epoch": 0.9629665194289738,
      "grad_norm": 0.004628051072359085,
      "learning_rate": 0.04323654208633607,
      "loss": 0.5492,
      "num_input_tokens_seen": 41362464,
      "step": 30085
    },
    {
      "epoch": 0.9631265603994623,
      "grad_norm": 0.007089382037520409,
      "learning_rate": 0.04319517398922024,
      "loss": 0.6909,
      "num_input_tokens_seen": 41369232,
      "step": 30090
    },
    {
      "epoch": 0.9632866013699507,
      "grad_norm": 0.004120769444853067,
      "learning_rate": 0.04315382236275079,
      "loss": 0.3763,
      "num_input_tokens_seen": 41376576,
      "step": 30095
    },
    {
      "epoch": 0.9634466423404392,
      "grad_norm": 0.004056065808981657,
      "learning_rate": 0.043112487213304664,
      "loss": 0.4869,
      "num_input_tokens_seen": 41383408,
      "step": 30100
    },
    {
      "epoch": 0.9636066833109276,
      "grad_norm": 0.009084658697247505,
      "learning_rate": 0.04307116854725618,
      "loss": 0.4145,
      "num_input_tokens_seen": 41390112,
      "step": 30105
    },
    {
      "epoch": 0.9637667242814161,
      "grad_norm": 0.005642089527100325,
      "learning_rate": 0.043029866370977325,
      "loss": 0.491,
      "num_input_tokens_seen": 41397584,
      "step": 30110
    },
    {
      "epoch": 0.9639267652519045,
      "grad_norm": 0.006902335211634636,
      "learning_rate": 0.04298858069083728,
      "loss": 0.6205,
      "num_input_tokens_seen": 41404240,
      "step": 30115
    },
    {
      "epoch": 0.964086806222393,
      "grad_norm": 0.005866141989827156,
      "learning_rate": 0.04294731151320295,
      "loss": 0.628,
      "num_input_tokens_seen": 41410528,
      "step": 30120
    },
    {
      "epoch": 0.9642468471928813,
      "grad_norm": 0.005988111719489098,
      "learning_rate": 0.04290605884443841,
      "loss": 0.4897,
      "num_input_tokens_seen": 41417168,
      "step": 30125
    },
    {
      "epoch": 0.9644068881633698,
      "grad_norm": 0.011159712448716164,
      "learning_rate": 0.04286482269090545,
      "loss": 0.4755,
      "num_input_tokens_seen": 41424096,
      "step": 30130
    },
    {
      "epoch": 0.9645669291338582,
      "grad_norm": 0.005171859636902809,
      "learning_rate": 0.04282360305896323,
      "loss": 0.5199,
      "num_input_tokens_seen": 41431328,
      "step": 30135
    },
    {
      "epoch": 0.9647269701043467,
      "grad_norm": 0.00738013070076704,
      "learning_rate": 0.04278239995496822,
      "loss": 0.6757,
      "num_input_tokens_seen": 41437632,
      "step": 30140
    },
    {
      "epoch": 0.9648870110748352,
      "grad_norm": 0.004529023077338934,
      "learning_rate": 0.042741213385274514,
      "loss": 0.4416,
      "num_input_tokens_seen": 41444272,
      "step": 30145
    },
    {
      "epoch": 0.9650470520453236,
      "grad_norm": 0.002443937584757805,
      "learning_rate": 0.04270004335623366,
      "loss": 0.4523,
      "num_input_tokens_seen": 41450736,
      "step": 30150
    },
    {
      "epoch": 0.9652070930158121,
      "grad_norm": 0.0056390115059912205,
      "learning_rate": 0.04265888987419448,
      "loss": 0.7975,
      "num_input_tokens_seen": 41457680,
      "step": 30155
    },
    {
      "epoch": 0.9653671339863005,
      "grad_norm": 0.00279806787148118,
      "learning_rate": 0.04261775294550346,
      "loss": 0.5685,
      "num_input_tokens_seen": 41464480,
      "step": 30160
    },
    {
      "epoch": 0.965527174956789,
      "grad_norm": 0.005159447900950909,
      "learning_rate": 0.042576632576504354,
      "loss": 0.4364,
      "num_input_tokens_seen": 41471248,
      "step": 30165
    },
    {
      "epoch": 0.9656872159272774,
      "grad_norm": 0.006091001909226179,
      "learning_rate": 0.0425355287735385,
      "loss": 0.4923,
      "num_input_tokens_seen": 41478368,
      "step": 30170
    },
    {
      "epoch": 0.9658472568977658,
      "grad_norm": 0.004740455187857151,
      "learning_rate": 0.0424944415429446,
      "loss": 0.4462,
      "num_input_tokens_seen": 41484688,
      "step": 30175
    },
    {
      "epoch": 0.9660072978682542,
      "grad_norm": 0.006315258331596851,
      "learning_rate": 0.04245337089105877,
      "loss": 0.4768,
      "num_input_tokens_seen": 41491504,
      "step": 30180
    },
    {
      "epoch": 0.9661673388387427,
      "grad_norm": 0.0030134369153529406,
      "learning_rate": 0.04241231682421467,
      "loss": 0.7093,
      "num_input_tokens_seen": 41498624,
      "step": 30185
    },
    {
      "epoch": 0.9663273798092311,
      "grad_norm": 0.002677513286471367,
      "learning_rate": 0.04237127934874337,
      "loss": 0.3901,
      "num_input_tokens_seen": 41505488,
      "step": 30190
    },
    {
      "epoch": 0.9664874207797196,
      "grad_norm": 0.00981078390032053,
      "learning_rate": 0.042330258470973305,
      "loss": 0.5975,
      "num_input_tokens_seen": 41512448,
      "step": 30195
    },
    {
      "epoch": 0.9666474617502081,
      "grad_norm": 0.007621414493769407,
      "learning_rate": 0.042289254197230515,
      "loss": 0.6609,
      "num_input_tokens_seen": 41519232,
      "step": 30200
    },
    {
      "epoch": 0.9666474617502081,
      "eval_loss": 0.5052435994148254,
      "eval_runtime": 332.2098,
      "eval_samples_per_second": 41.799,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 41519232,
      "step": 30200
    },
    {
      "epoch": 0.9668075027206965,
      "grad_norm": 0.004266004078090191,
      "learning_rate": 0.04224826653383823,
      "loss": 0.6932,
      "num_input_tokens_seen": 41525952,
      "step": 30205
    },
    {
      "epoch": 0.966967543691185,
      "grad_norm": 0.006537155248224735,
      "learning_rate": 0.04220729548711735,
      "loss": 0.5802,
      "num_input_tokens_seen": 41532752,
      "step": 30210
    },
    {
      "epoch": 0.9671275846616734,
      "grad_norm": 0.004426646512001753,
      "learning_rate": 0.04216634106338616,
      "loss": 0.3429,
      "num_input_tokens_seen": 41539168,
      "step": 30215
    },
    {
      "epoch": 0.9672876256321619,
      "grad_norm": 0.004462828394025564,
      "learning_rate": 0.04212540326896025,
      "loss": 0.6078,
      "num_input_tokens_seen": 41545920,
      "step": 30220
    },
    {
      "epoch": 0.9674476666026502,
      "grad_norm": 0.003368259174749255,
      "learning_rate": 0.0420844821101528,
      "loss": 0.3212,
      "num_input_tokens_seen": 41552704,
      "step": 30225
    },
    {
      "epoch": 0.9676077075731387,
      "grad_norm": 0.004165003541857004,
      "learning_rate": 0.04204357759327441,
      "loss": 0.3761,
      "num_input_tokens_seen": 41559728,
      "step": 30230
    },
    {
      "epoch": 0.9677677485436271,
      "grad_norm": 0.005679097957909107,
      "learning_rate": 0.042002689724632954,
      "loss": 0.5106,
      "num_input_tokens_seen": 41566496,
      "step": 30235
    },
    {
      "epoch": 0.9679277895141156,
      "grad_norm": 0.0073393359780311584,
      "learning_rate": 0.04196181851053398,
      "loss": 0.7242,
      "num_input_tokens_seen": 41573152,
      "step": 30240
    },
    {
      "epoch": 0.968087830484604,
      "grad_norm": 0.002669254317879677,
      "learning_rate": 0.041920963957280295,
      "loss": 0.537,
      "num_input_tokens_seen": 41580000,
      "step": 30245
    },
    {
      "epoch": 0.9682478714550925,
      "grad_norm": 0.009815638884902,
      "learning_rate": 0.04188012607117212,
      "loss": 0.5228,
      "num_input_tokens_seen": 41587568,
      "step": 30250
    },
    {
      "epoch": 0.968407912425581,
      "grad_norm": 0.006156508810818195,
      "learning_rate": 0.04183930485850725,
      "loss": 0.6521,
      "num_input_tokens_seen": 41594640,
      "step": 30255
    },
    {
      "epoch": 0.9685679533960694,
      "grad_norm": 0.005553813185542822,
      "learning_rate": 0.04179850032558078,
      "loss": 0.5137,
      "num_input_tokens_seen": 41601440,
      "step": 30260
    },
    {
      "epoch": 0.9687279943665579,
      "grad_norm": 0.00699286162853241,
      "learning_rate": 0.041757712478685295,
      "loss": 0.618,
      "num_input_tokens_seen": 41607920,
      "step": 30265
    },
    {
      "epoch": 0.9688880353370463,
      "grad_norm": 0.005326509475708008,
      "learning_rate": 0.04171694132411085,
      "loss": 0.5153,
      "num_input_tokens_seen": 41614768,
      "step": 30270
    },
    {
      "epoch": 0.9690480763075348,
      "grad_norm": 0.007609821856021881,
      "learning_rate": 0.04167618686814479,
      "loss": 0.6156,
      "num_input_tokens_seen": 41621648,
      "step": 30275
    },
    {
      "epoch": 0.9692081172780231,
      "grad_norm": 0.0022665841970592737,
      "learning_rate": 0.041635449117072024,
      "loss": 0.3519,
      "num_input_tokens_seen": 41628400,
      "step": 30280
    },
    {
      "epoch": 0.9693681582485116,
      "grad_norm": 0.0028270920738577843,
      "learning_rate": 0.04159472807717477,
      "loss": 0.5184,
      "num_input_tokens_seen": 41634976,
      "step": 30285
    },
    {
      "epoch": 0.969528199219,
      "grad_norm": 0.003298589028418064,
      "learning_rate": 0.041554023754732744,
      "loss": 0.5991,
      "num_input_tokens_seen": 41641920,
      "step": 30290
    },
    {
      "epoch": 0.9696882401894885,
      "grad_norm": 0.004323660396039486,
      "learning_rate": 0.04151333615602311,
      "loss": 0.6233,
      "num_input_tokens_seen": 41648928,
      "step": 30295
    },
    {
      "epoch": 0.9698482811599769,
      "grad_norm": 0.003957974724471569,
      "learning_rate": 0.04147266528732034,
      "loss": 0.5623,
      "num_input_tokens_seen": 41655664,
      "step": 30300
    },
    {
      "epoch": 0.9700083221304654,
      "grad_norm": 0.005350195337086916,
      "learning_rate": 0.0414320111548964,
      "loss": 0.4534,
      "num_input_tokens_seen": 41662736,
      "step": 30305
    },
    {
      "epoch": 0.9701683631009539,
      "grad_norm": 0.00834882166236639,
      "learning_rate": 0.04139137376502076,
      "loss": 0.6324,
      "num_input_tokens_seen": 41669664,
      "step": 30310
    },
    {
      "epoch": 0.9703284040714423,
      "grad_norm": 0.003870912827551365,
      "learning_rate": 0.04135075312396014,
      "loss": 0.4114,
      "num_input_tokens_seen": 41676064,
      "step": 30315
    },
    {
      "epoch": 0.9704884450419308,
      "grad_norm": 0.003666912205517292,
      "learning_rate": 0.04131014923797875,
      "loss": 0.512,
      "num_input_tokens_seen": 41682832,
      "step": 30320
    },
    {
      "epoch": 0.9706484860124192,
      "grad_norm": 0.004676077049225569,
      "learning_rate": 0.04126956211333819,
      "loss": 0.3988,
      "num_input_tokens_seen": 41689408,
      "step": 30325
    },
    {
      "epoch": 0.9708085269829076,
      "grad_norm": 0.005159233696758747,
      "learning_rate": 0.041228991756297545,
      "loss": 0.3858,
      "num_input_tokens_seen": 41696224,
      "step": 30330
    },
    {
      "epoch": 0.970968567953396,
      "grad_norm": 0.003974451217800379,
      "learning_rate": 0.04118843817311332,
      "loss": 0.4185,
      "num_input_tokens_seen": 41702608,
      "step": 30335
    },
    {
      "epoch": 0.9711286089238845,
      "grad_norm": 0.0069602662697434425,
      "learning_rate": 0.0411479013700393,
      "loss": 0.4654,
      "num_input_tokens_seen": 41709200,
      "step": 30340
    },
    {
      "epoch": 0.9712886498943729,
      "grad_norm": 0.0027415594086050987,
      "learning_rate": 0.0411073813533268,
      "loss": 0.4128,
      "num_input_tokens_seen": 41715968,
      "step": 30345
    },
    {
      "epoch": 0.9714486908648614,
      "grad_norm": 0.004066189751029015,
      "learning_rate": 0.04106687812922456,
      "loss": 0.3557,
      "num_input_tokens_seen": 41722608,
      "step": 30350
    },
    {
      "epoch": 0.9716087318353499,
      "grad_norm": 0.004196366295218468,
      "learning_rate": 0.041026391703978635,
      "loss": 0.3155,
      "num_input_tokens_seen": 41729424,
      "step": 30355
    },
    {
      "epoch": 0.9717687728058383,
      "grad_norm": 0.005323639139533043,
      "learning_rate": 0.04098592208383259,
      "loss": 0.5386,
      "num_input_tokens_seen": 41736288,
      "step": 30360
    },
    {
      "epoch": 0.9719288137763268,
      "grad_norm": 0.005272340029478073,
      "learning_rate": 0.040945469275027256,
      "loss": 0.5311,
      "num_input_tokens_seen": 41742592,
      "step": 30365
    },
    {
      "epoch": 0.9720888547468152,
      "grad_norm": 0.003421155270189047,
      "learning_rate": 0.04090503328380104,
      "loss": 0.5824,
      "num_input_tokens_seen": 41749808,
      "step": 30370
    },
    {
      "epoch": 0.9722488957173037,
      "grad_norm": 0.0057847644202411175,
      "learning_rate": 0.04086461411638971,
      "loss": 0.3765,
      "num_input_tokens_seen": 41756784,
      "step": 30375
    },
    {
      "epoch": 0.972408936687792,
      "grad_norm": 0.004598920699208975,
      "learning_rate": 0.04082421177902631,
      "loss": 0.3357,
      "num_input_tokens_seen": 41763664,
      "step": 30380
    },
    {
      "epoch": 0.9725689776582805,
      "grad_norm": 0.006076267454773188,
      "learning_rate": 0.04078382627794149,
      "loss": 0.4209,
      "num_input_tokens_seen": 41770288,
      "step": 30385
    },
    {
      "epoch": 0.9727290186287689,
      "grad_norm": 0.006767359562218189,
      "learning_rate": 0.04074345761936316,
      "loss": 0.6253,
      "num_input_tokens_seen": 41777328,
      "step": 30390
    },
    {
      "epoch": 0.9728890595992574,
      "grad_norm": 0.00357504072599113,
      "learning_rate": 0.04070310580951663,
      "loss": 0.4949,
      "num_input_tokens_seen": 41784272,
      "step": 30395
    },
    {
      "epoch": 0.9730491005697458,
      "grad_norm": 0.003484225133433938,
      "learning_rate": 0.040662770854624726,
      "loss": 0.5844,
      "num_input_tokens_seen": 41791520,
      "step": 30400
    },
    {
      "epoch": 0.9730491005697458,
      "eval_loss": 0.5047814846038818,
      "eval_runtime": 331.9351,
      "eval_samples_per_second": 41.833,
      "eval_steps_per_second": 20.917,
      "num_input_tokens_seen": 41791520,
      "step": 30400
    },
    {
      "epoch": 0.9732091415402343,
      "grad_norm": 0.0028732677455991507,
      "learning_rate": 0.040622452760907535,
      "loss": 0.4998,
      "num_input_tokens_seen": 41798304,
      "step": 30405
    },
    {
      "epoch": 0.9733691825107228,
      "grad_norm": 0.005665189120918512,
      "learning_rate": 0.04058215153458265,
      "loss": 0.5043,
      "num_input_tokens_seen": 41805392,
      "step": 30410
    },
    {
      "epoch": 0.9735292234812112,
      "grad_norm": 0.0038456141483038664,
      "learning_rate": 0.04054186718186507,
      "loss": 0.5423,
      "num_input_tokens_seen": 41812432,
      "step": 30415
    },
    {
      "epoch": 0.9736892644516997,
      "grad_norm": 0.00705447793006897,
      "learning_rate": 0.04050159970896708,
      "loss": 0.4498,
      "num_input_tokens_seen": 41819008,
      "step": 30420
    },
    {
      "epoch": 0.9738493054221881,
      "grad_norm": 0.009884143248200417,
      "learning_rate": 0.04046134912209843,
      "loss": 0.6537,
      "num_input_tokens_seen": 41826592,
      "step": 30425
    },
    {
      "epoch": 0.9740093463926766,
      "grad_norm": 0.005254028830677271,
      "learning_rate": 0.040421115427466354,
      "loss": 0.6034,
      "num_input_tokens_seen": 41833280,
      "step": 30430
    },
    {
      "epoch": 0.9741693873631649,
      "grad_norm": 0.0053877760656178,
      "learning_rate": 0.04038089863127529,
      "loss": 0.4931,
      "num_input_tokens_seen": 41840240,
      "step": 30435
    },
    {
      "epoch": 0.9743294283336534,
      "grad_norm": 0.0028649375308305025,
      "learning_rate": 0.04034069873972727,
      "loss": 0.4386,
      "num_input_tokens_seen": 41847440,
      "step": 30440
    },
    {
      "epoch": 0.9744894693041418,
      "grad_norm": 0.003977580461651087,
      "learning_rate": 0.040300515759021514,
      "loss": 0.5025,
      "num_input_tokens_seen": 41854192,
      "step": 30445
    },
    {
      "epoch": 0.9746495102746303,
      "grad_norm": 0.00394840631633997,
      "learning_rate": 0.04026034969535478,
      "loss": 0.4224,
      "num_input_tokens_seen": 41861136,
      "step": 30450
    },
    {
      "epoch": 0.9748095512451187,
      "grad_norm": 0.0066309645771980286,
      "learning_rate": 0.040220200554921266,
      "loss": 0.7083,
      "num_input_tokens_seen": 41867792,
      "step": 30455
    },
    {
      "epoch": 0.9749695922156072,
      "grad_norm": 0.005592024885118008,
      "learning_rate": 0.0401800683439124,
      "loss": 0.4812,
      "num_input_tokens_seen": 41874784,
      "step": 30460
    },
    {
      "epoch": 0.9751296331860957,
      "grad_norm": 0.009232495911419392,
      "learning_rate": 0.04013995306851704,
      "loss": 0.6985,
      "num_input_tokens_seen": 41881568,
      "step": 30465
    },
    {
      "epoch": 0.9752896741565841,
      "grad_norm": 0.007255019154399633,
      "learning_rate": 0.040099854734921545,
      "loss": 0.516,
      "num_input_tokens_seen": 41888224,
      "step": 30470
    },
    {
      "epoch": 0.9754497151270726,
      "grad_norm": 0.007626599166542292,
      "learning_rate": 0.0400597733493095,
      "loss": 0.4601,
      "num_input_tokens_seen": 41894960,
      "step": 30475
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.0036857191007584333,
      "learning_rate": 0.04001970891786203,
      "loss": 0.5089,
      "num_input_tokens_seen": 41901968,
      "step": 30480
    },
    {
      "epoch": 0.9757697970680494,
      "grad_norm": 0.005313207861036062,
      "learning_rate": 0.03997966144675752,
      "loss": 0.6085,
      "num_input_tokens_seen": 41908736,
      "step": 30485
    },
    {
      "epoch": 0.9759298380385378,
      "grad_norm": 0.010474101640284061,
      "learning_rate": 0.039939630942171796,
      "loss": 0.6082,
      "num_input_tokens_seen": 41915472,
      "step": 30490
    },
    {
      "epoch": 0.9760898790090263,
      "grad_norm": 0.006114896386861801,
      "learning_rate": 0.03989961741027815,
      "loss": 0.4048,
      "num_input_tokens_seen": 41922160,
      "step": 30495
    },
    {
      "epoch": 0.9762499199795147,
      "grad_norm": 0.003852691501379013,
      "learning_rate": 0.03985962085724704,
      "loss": 0.4918,
      "num_input_tokens_seen": 41928976,
      "step": 30500
    },
    {
      "epoch": 0.9764099609500032,
      "grad_norm": 0.0031516202725470066,
      "learning_rate": 0.03981964128924656,
      "loss": 0.6553,
      "num_input_tokens_seen": 41935536,
      "step": 30505
    },
    {
      "epoch": 0.9765700019204916,
      "grad_norm": 0.00689335260540247,
      "learning_rate": 0.03977967871244197,
      "loss": 0.5534,
      "num_input_tokens_seen": 41942288,
      "step": 30510
    },
    {
      "epoch": 0.9767300428909801,
      "grad_norm": 0.004145990591496229,
      "learning_rate": 0.03973973313299602,
      "loss": 0.5083,
      "num_input_tokens_seen": 41948832,
      "step": 30515
    },
    {
      "epoch": 0.9768900838614686,
      "grad_norm": 0.005371376406401396,
      "learning_rate": 0.0396998045570689,
      "loss": 0.3811,
      "num_input_tokens_seen": 41955776,
      "step": 30520
    },
    {
      "epoch": 0.977050124831957,
      "grad_norm": 0.005372428800910711,
      "learning_rate": 0.03965989299081798,
      "loss": 0.3905,
      "num_input_tokens_seen": 41962352,
      "step": 30525
    },
    {
      "epoch": 0.9772101658024455,
      "grad_norm": 0.005356648471206427,
      "learning_rate": 0.039619998440398235,
      "loss": 0.6676,
      "num_input_tokens_seen": 41969440,
      "step": 30530
    },
    {
      "epoch": 0.9773702067729338,
      "grad_norm": 0.001822542049922049,
      "learning_rate": 0.03958012091196184,
      "loss": 0.3138,
      "num_input_tokens_seen": 41976592,
      "step": 30535
    },
    {
      "epoch": 0.9775302477434223,
      "grad_norm": 0.0029457202181220055,
      "learning_rate": 0.039540260411658396,
      "loss": 0.368,
      "num_input_tokens_seen": 41983680,
      "step": 30540
    },
    {
      "epoch": 0.9776902887139107,
      "grad_norm": 0.006135148461908102,
      "learning_rate": 0.03950041694563496,
      "loss": 0.4409,
      "num_input_tokens_seen": 41990800,
      "step": 30545
    },
    {
      "epoch": 0.9778503296843992,
      "grad_norm": 0.003334303153678775,
      "learning_rate": 0.0394605905200358,
      "loss": 0.3751,
      "num_input_tokens_seen": 41997712,
      "step": 30550
    },
    {
      "epoch": 0.9780103706548876,
      "grad_norm": 0.003778248094022274,
      "learning_rate": 0.03942078114100272,
      "loss": 0.5486,
      "num_input_tokens_seen": 42004608,
      "step": 30555
    },
    {
      "epoch": 0.9781704116253761,
      "grad_norm": 0.0035832456778734922,
      "learning_rate": 0.03938098881467485,
      "loss": 0.412,
      "num_input_tokens_seen": 42012512,
      "step": 30560
    },
    {
      "epoch": 0.9783304525958645,
      "grad_norm": 0.004350563511252403,
      "learning_rate": 0.039341213547188586,
      "loss": 0.4737,
      "num_input_tokens_seen": 42019264,
      "step": 30565
    },
    {
      "epoch": 0.978490493566353,
      "grad_norm": 0.004887261427938938,
      "learning_rate": 0.03930145534467782,
      "loss": 0.5206,
      "num_input_tokens_seen": 42026112,
      "step": 30570
    },
    {
      "epoch": 0.9786505345368415,
      "grad_norm": 0.002416673582047224,
      "learning_rate": 0.0392617142132738,
      "loss": 0.4843,
      "num_input_tokens_seen": 42032640,
      "step": 30575
    },
    {
      "epoch": 0.9788105755073299,
      "grad_norm": 0.010524298064410686,
      "learning_rate": 0.03922199015910504,
      "loss": 0.4397,
      "num_input_tokens_seen": 42039376,
      "step": 30580
    },
    {
      "epoch": 0.9789706164778184,
      "grad_norm": 0.006254301406443119,
      "learning_rate": 0.039182283188297556,
      "loss": 0.5912,
      "num_input_tokens_seen": 42046128,
      "step": 30585
    },
    {
      "epoch": 0.9791306574483067,
      "grad_norm": 0.005119442008435726,
      "learning_rate": 0.039142593306974595,
      "loss": 0.4136,
      "num_input_tokens_seen": 42052816,
      "step": 30590
    },
    {
      "epoch": 0.9792906984187952,
      "grad_norm": 0.004889143630862236,
      "learning_rate": 0.039102920521256856,
      "loss": 0.509,
      "num_input_tokens_seen": 42059904,
      "step": 30595
    },
    {
      "epoch": 0.9794507393892836,
      "grad_norm": 0.007261769380420446,
      "learning_rate": 0.03906326483726243,
      "loss": 0.4285,
      "num_input_tokens_seen": 42066928,
      "step": 30600
    },
    {
      "epoch": 0.9794507393892836,
      "eval_loss": 0.503872275352478,
      "eval_runtime": 331.847,
      "eval_samples_per_second": 41.845,
      "eval_steps_per_second": 20.922,
      "num_input_tokens_seen": 42066928,
      "step": 30600
    },
    {
      "epoch": 0.9796107803597721,
      "grad_norm": 0.004351356066763401,
      "learning_rate": 0.039023626261106704,
      "loss": 0.5152,
      "num_input_tokens_seen": 42073776,
      "step": 30605
    },
    {
      "epoch": 0.9797708213302605,
      "grad_norm": 0.007582687307149172,
      "learning_rate": 0.03898400479890237,
      "loss": 0.7757,
      "num_input_tokens_seen": 42080544,
      "step": 30610
    },
    {
      "epoch": 0.979930862300749,
      "grad_norm": 0.004784265998750925,
      "learning_rate": 0.038944400456759655,
      "loss": 0.3695,
      "num_input_tokens_seen": 42087376,
      "step": 30615
    },
    {
      "epoch": 0.9800909032712375,
      "grad_norm": 0.004507001023739576,
      "learning_rate": 0.038904813240785964,
      "loss": 0.595,
      "num_input_tokens_seen": 42094672,
      "step": 30620
    },
    {
      "epoch": 0.9802509442417259,
      "grad_norm": 0.005250763613730669,
      "learning_rate": 0.03886524315708621,
      "loss": 0.5486,
      "num_input_tokens_seen": 42101664,
      "step": 30625
    },
    {
      "epoch": 0.9804109852122144,
      "grad_norm": 0.002759745577350259,
      "learning_rate": 0.03882569021176255,
      "loss": 0.4889,
      "num_input_tokens_seen": 42108656,
      "step": 30630
    },
    {
      "epoch": 0.9805710261827028,
      "grad_norm": 0.004641219973564148,
      "learning_rate": 0.038786154410914535,
      "loss": 0.4229,
      "num_input_tokens_seen": 42115136,
      "step": 30635
    },
    {
      "epoch": 0.9807310671531912,
      "grad_norm": 0.005850809160619974,
      "learning_rate": 0.03874663576063917,
      "loss": 0.5117,
      "num_input_tokens_seen": 42122080,
      "step": 30640
    },
    {
      "epoch": 0.9808911081236796,
      "grad_norm": 0.00739015219733119,
      "learning_rate": 0.038707134267030624,
      "loss": 0.4986,
      "num_input_tokens_seen": 42128704,
      "step": 30645
    },
    {
      "epoch": 0.9810511490941681,
      "grad_norm": 0.0049631111323833466,
      "learning_rate": 0.038667649936180555,
      "loss": 0.5261,
      "num_input_tokens_seen": 42135536,
      "step": 30650
    },
    {
      "epoch": 0.9812111900646565,
      "grad_norm": 0.005865690764039755,
      "learning_rate": 0.038628182774178,
      "loss": 0.5386,
      "num_input_tokens_seen": 42141984,
      "step": 30655
    },
    {
      "epoch": 0.981371231035145,
      "grad_norm": 0.005905190948396921,
      "learning_rate": 0.038588732787109226,
      "loss": 0.4747,
      "num_input_tokens_seen": 42148848,
      "step": 30660
    },
    {
      "epoch": 0.9815312720056334,
      "grad_norm": 0.008773834444582462,
      "learning_rate": 0.03854929998105795,
      "loss": 0.3718,
      "num_input_tokens_seen": 42155600,
      "step": 30665
    },
    {
      "epoch": 0.9816913129761219,
      "grad_norm": 0.004303161986172199,
      "learning_rate": 0.03850988436210518,
      "loss": 0.7295,
      "num_input_tokens_seen": 42162400,
      "step": 30670
    },
    {
      "epoch": 0.9818513539466104,
      "grad_norm": 0.004293948877602816,
      "learning_rate": 0.03847048593632933,
      "loss": 0.5562,
      "num_input_tokens_seen": 42169344,
      "step": 30675
    },
    {
      "epoch": 0.9820113949170988,
      "grad_norm": 0.004948499612510204,
      "learning_rate": 0.038431104709806096,
      "loss": 0.5939,
      "num_input_tokens_seen": 42176128,
      "step": 30680
    },
    {
      "epoch": 0.9821714358875873,
      "grad_norm": 0.003933553118258715,
      "learning_rate": 0.0383917406886086,
      "loss": 0.3848,
      "num_input_tokens_seen": 42183136,
      "step": 30685
    },
    {
      "epoch": 0.9823314768580756,
      "grad_norm": 0.006441409233957529,
      "learning_rate": 0.03835239387880722,
      "loss": 0.363,
      "num_input_tokens_seen": 42189600,
      "step": 30690
    },
    {
      "epoch": 0.9824915178285641,
      "grad_norm": 0.0033179819583892822,
      "learning_rate": 0.03831306428646979,
      "loss": 0.4803,
      "num_input_tokens_seen": 42196352,
      "step": 30695
    },
    {
      "epoch": 0.9826515587990525,
      "grad_norm": 0.0031780381686985493,
      "learning_rate": 0.03827375191766135,
      "loss": 0.5844,
      "num_input_tokens_seen": 42203024,
      "step": 30700
    },
    {
      "epoch": 0.982811599769541,
      "grad_norm": 0.005862824618816376,
      "learning_rate": 0.03823445677844446,
      "loss": 0.7369,
      "num_input_tokens_seen": 42210000,
      "step": 30705
    },
    {
      "epoch": 0.9829716407400294,
      "grad_norm": 0.0025743693113327026,
      "learning_rate": 0.03819517887487881,
      "loss": 0.433,
      "num_input_tokens_seen": 42216816,
      "step": 30710
    },
    {
      "epoch": 0.9831316817105179,
      "grad_norm": 0.005359269212931395,
      "learning_rate": 0.03815591821302161,
      "loss": 0.4613,
      "num_input_tokens_seen": 42223632,
      "step": 30715
    },
    {
      "epoch": 0.9832917226810063,
      "grad_norm": 0.007357375230640173,
      "learning_rate": 0.03811667479892739,
      "loss": 0.5833,
      "num_input_tokens_seen": 42230592,
      "step": 30720
    },
    {
      "epoch": 0.9834517636514948,
      "grad_norm": 0.0034250516910105944,
      "learning_rate": 0.03807744863864788,
      "loss": 0.3335,
      "num_input_tokens_seen": 42237216,
      "step": 30725
    },
    {
      "epoch": 0.9836118046219833,
      "grad_norm": 0.003097866429015994,
      "learning_rate": 0.03803823973823229,
      "loss": 0.4503,
      "num_input_tokens_seen": 42244112,
      "step": 30730
    },
    {
      "epoch": 0.9837718455924717,
      "grad_norm": 0.006958760786801577,
      "learning_rate": 0.03799904810372719,
      "loss": 0.4544,
      "num_input_tokens_seen": 42250944,
      "step": 30735
    },
    {
      "epoch": 0.9839318865629602,
      "grad_norm": 0.0035513355396687984,
      "learning_rate": 0.03795987374117632,
      "loss": 0.5184,
      "num_input_tokens_seen": 42257952,
      "step": 30740
    },
    {
      "epoch": 0.9840919275334485,
      "grad_norm": 0.008379076607525349,
      "learning_rate": 0.03792071665662093,
      "loss": 0.43,
      "num_input_tokens_seen": 42264592,
      "step": 30745
    },
    {
      "epoch": 0.984251968503937,
      "grad_norm": 0.003648293437436223,
      "learning_rate": 0.03788157685609952,
      "loss": 0.4049,
      "num_input_tokens_seen": 42271648,
      "step": 30750
    },
    {
      "epoch": 0.9844120094744254,
      "grad_norm": 0.00675185676664114,
      "learning_rate": 0.037842454345647876,
      "loss": 0.5147,
      "num_input_tokens_seen": 42278464,
      "step": 30755
    },
    {
      "epoch": 0.9845720504449139,
      "grad_norm": 0.005441510584205389,
      "learning_rate": 0.03780334913129929,
      "loss": 0.5909,
      "num_input_tokens_seen": 42285264,
      "step": 30760
    },
    {
      "epoch": 0.9847320914154023,
      "grad_norm": 0.008924800902605057,
      "learning_rate": 0.037764261219084175,
      "loss": 0.7674,
      "num_input_tokens_seen": 42291904,
      "step": 30765
    },
    {
      "epoch": 0.9848921323858908,
      "grad_norm": 0.0034914433490484953,
      "learning_rate": 0.037725190615030414,
      "loss": 0.4789,
      "num_input_tokens_seen": 42298608,
      "step": 30770
    },
    {
      "epoch": 0.9850521733563792,
      "grad_norm": 0.003712472738698125,
      "learning_rate": 0.037686137325163224,
      "loss": 0.3837,
      "num_input_tokens_seen": 42305184,
      "step": 30775
    },
    {
      "epoch": 0.9852122143268677,
      "grad_norm": 0.014130507595837116,
      "learning_rate": 0.037647101355505065,
      "loss": 0.6779,
      "num_input_tokens_seen": 42311968,
      "step": 30780
    },
    {
      "epoch": 0.9853722552973562,
      "grad_norm": 0.006696886382997036,
      "learning_rate": 0.03760808271207581,
      "loss": 0.3215,
      "num_input_tokens_seen": 42319008,
      "step": 30785
    },
    {
      "epoch": 0.9855322962678446,
      "grad_norm": 0.0020622864831238985,
      "learning_rate": 0.03756908140089258,
      "loss": 0.3685,
      "num_input_tokens_seen": 42325696,
      "step": 30790
    },
    {
      "epoch": 0.985692337238333,
      "grad_norm": 0.007139429450035095,
      "learning_rate": 0.03753009742796989,
      "loss": 0.5965,
      "num_input_tokens_seen": 42332496,
      "step": 30795
    },
    {
      "epoch": 0.9858523782088214,
      "grad_norm": 0.003272653790190816,
      "learning_rate": 0.037491130799319615,
      "loss": 0.56,
      "num_input_tokens_seen": 42339616,
      "step": 30800
    },
    {
      "epoch": 0.9858523782088214,
      "eval_loss": 0.5050134658813477,
      "eval_runtime": 332.3745,
      "eval_samples_per_second": 41.778,
      "eval_steps_per_second": 20.889,
      "num_input_tokens_seen": 42339616,
      "step": 30800
    },
    {
      "epoch": 0.9860124191793099,
      "grad_norm": 0.0045720478519797325,
      "learning_rate": 0.03745218152095079,
      "loss": 0.4583,
      "num_input_tokens_seen": 42346768,
      "step": 30805
    },
    {
      "epoch": 0.9861724601497983,
      "grad_norm": 0.007941075600683689,
      "learning_rate": 0.037413249598869935,
      "loss": 0.4697,
      "num_input_tokens_seen": 42353632,
      "step": 30810
    },
    {
      "epoch": 0.9863325011202868,
      "grad_norm": 0.003895925125107169,
      "learning_rate": 0.037374335039080886,
      "loss": 0.4638,
      "num_input_tokens_seen": 42360320,
      "step": 30815
    },
    {
      "epoch": 0.9864925420907752,
      "grad_norm": 0.002625021617859602,
      "learning_rate": 0.037335437847584724,
      "loss": 0.4411,
      "num_input_tokens_seen": 42366976,
      "step": 30820
    },
    {
      "epoch": 0.9866525830612637,
      "grad_norm": 0.006909089628607035,
      "learning_rate": 0.03729655803037983,
      "loss": 0.7722,
      "num_input_tokens_seen": 42373872,
      "step": 30825
    },
    {
      "epoch": 0.9868126240317522,
      "grad_norm": 0.006072274874895811,
      "learning_rate": 0.03725769559346207,
      "loss": 0.4447,
      "num_input_tokens_seen": 42380736,
      "step": 30830
    },
    {
      "epoch": 0.9869726650022406,
      "grad_norm": 0.004317422863095999,
      "learning_rate": 0.03721885054282439,
      "loss": 0.5006,
      "num_input_tokens_seen": 42387824,
      "step": 30835
    },
    {
      "epoch": 0.9871327059727291,
      "grad_norm": 0.001468182192184031,
      "learning_rate": 0.03718002288445731,
      "loss": 0.4633,
      "num_input_tokens_seen": 42394752,
      "step": 30840
    },
    {
      "epoch": 0.9872927469432174,
      "grad_norm": 0.0034142194781452417,
      "learning_rate": 0.03714121262434844,
      "loss": 0.4475,
      "num_input_tokens_seen": 42401584,
      "step": 30845
    },
    {
      "epoch": 0.9874527879137059,
      "grad_norm": 0.002945918356999755,
      "learning_rate": 0.037102419768482844,
      "loss": 0.4,
      "num_input_tokens_seen": 42408688,
      "step": 30850
    },
    {
      "epoch": 0.9876128288841943,
      "grad_norm": 0.007534039672464132,
      "learning_rate": 0.03706364432284293,
      "loss": 0.7212,
      "num_input_tokens_seen": 42415344,
      "step": 30855
    },
    {
      "epoch": 0.9877728698546828,
      "grad_norm": 0.0035782204940915108,
      "learning_rate": 0.03702488629340828,
      "loss": 0.5007,
      "num_input_tokens_seen": 42422320,
      "step": 30860
    },
    {
      "epoch": 0.9879329108251712,
      "grad_norm": 0.005990957375615835,
      "learning_rate": 0.036986145686155915,
      "loss": 0.5075,
      "num_input_tokens_seen": 42429248,
      "step": 30865
    },
    {
      "epoch": 0.9880929517956597,
      "grad_norm": 0.011117742396891117,
      "learning_rate": 0.036947422507060075,
      "loss": 0.4486,
      "num_input_tokens_seen": 42436016,
      "step": 30870
    },
    {
      "epoch": 0.9882529927661481,
      "grad_norm": 0.0022709565237164497,
      "learning_rate": 0.0369087167620924,
      "loss": 0.3778,
      "num_input_tokens_seen": 42442800,
      "step": 30875
    },
    {
      "epoch": 0.9884130337366366,
      "grad_norm": 0.005733849015086889,
      "learning_rate": 0.03687002845722183,
      "loss": 0.6049,
      "num_input_tokens_seen": 42450064,
      "step": 30880
    },
    {
      "epoch": 0.9885730747071251,
      "grad_norm": 0.004896500147879124,
      "learning_rate": 0.03683135759841451,
      "loss": 0.4074,
      "num_input_tokens_seen": 42456736,
      "step": 30885
    },
    {
      "epoch": 0.9887331156776135,
      "grad_norm": 0.004343184642493725,
      "learning_rate": 0.03679270419163406,
      "loss": 0.4369,
      "num_input_tokens_seen": 42463712,
      "step": 30890
    },
    {
      "epoch": 0.988893156648102,
      "grad_norm": 0.0069816820323467255,
      "learning_rate": 0.03675406824284127,
      "loss": 0.587,
      "num_input_tokens_seen": 42470688,
      "step": 30895
    },
    {
      "epoch": 0.9890531976185903,
      "grad_norm": 0.003735640784725547,
      "learning_rate": 0.03671544975799425,
      "loss": 0.4606,
      "num_input_tokens_seen": 42477760,
      "step": 30900
    },
    {
      "epoch": 0.9892132385890788,
      "grad_norm": 0.0064912764355540276,
      "learning_rate": 0.03667684874304854,
      "loss": 0.4895,
      "num_input_tokens_seen": 42484640,
      "step": 30905
    },
    {
      "epoch": 0.9893732795595672,
      "grad_norm": 0.0027638666797429323,
      "learning_rate": 0.03663826520395683,
      "loss": 0.5641,
      "num_input_tokens_seen": 42491504,
      "step": 30910
    },
    {
      "epoch": 0.9895333205300557,
      "grad_norm": 0.004145861603319645,
      "learning_rate": 0.03659969914666922,
      "loss": 0.4085,
      "num_input_tokens_seen": 42498464,
      "step": 30915
    },
    {
      "epoch": 0.9896933615005441,
      "grad_norm": 0.0038879045750945807,
      "learning_rate": 0.036561150577133106,
      "loss": 0.618,
      "num_input_tokens_seen": 42505280,
      "step": 30920
    },
    {
      "epoch": 0.9898534024710326,
      "grad_norm": 0.009406650438904762,
      "learning_rate": 0.036522619501293103,
      "loss": 0.5498,
      "num_input_tokens_seen": 42512016,
      "step": 30925
    },
    {
      "epoch": 0.990013443441521,
      "grad_norm": 0.0032476894557476044,
      "learning_rate": 0.03648410592509122,
      "loss": 0.6281,
      "num_input_tokens_seen": 42518928,
      "step": 30930
    },
    {
      "epoch": 0.9901734844120095,
      "grad_norm": 0.0049964310601353645,
      "learning_rate": 0.03644560985446676,
      "loss": 0.5999,
      "num_input_tokens_seen": 42526032,
      "step": 30935
    },
    {
      "epoch": 0.990333525382498,
      "grad_norm": 0.008360696956515312,
      "learning_rate": 0.036407131295356256,
      "loss": 0.471,
      "num_input_tokens_seen": 42532976,
      "step": 30940
    },
    {
      "epoch": 0.9904935663529864,
      "grad_norm": 0.005500521045178175,
      "learning_rate": 0.03636867025369362,
      "loss": 0.5598,
      "num_input_tokens_seen": 42539920,
      "step": 30945
    },
    {
      "epoch": 0.9906536073234748,
      "grad_norm": 0.004522376693785191,
      "learning_rate": 0.03633022673540999,
      "loss": 0.5289,
      "num_input_tokens_seen": 42547136,
      "step": 30950
    },
    {
      "epoch": 0.9908136482939632,
      "grad_norm": 0.007805875968188047,
      "learning_rate": 0.03629180074643385,
      "loss": 0.5364,
      "num_input_tokens_seen": 42554256,
      "step": 30955
    },
    {
      "epoch": 0.9909736892644517,
      "grad_norm": 0.004431941080838442,
      "learning_rate": 0.03625339229269102,
      "loss": 0.4718,
      "num_input_tokens_seen": 42561104,
      "step": 30960
    },
    {
      "epoch": 0.9911337302349401,
      "grad_norm": 0.007157925050705671,
      "learning_rate": 0.036215001380104535,
      "loss": 0.4334,
      "num_input_tokens_seen": 42567984,
      "step": 30965
    },
    {
      "epoch": 0.9912937712054286,
      "grad_norm": 0.0039824810810387135,
      "learning_rate": 0.03617662801459471,
      "loss": 0.5111,
      "num_input_tokens_seen": 42574944,
      "step": 30970
    },
    {
      "epoch": 0.991453812175917,
      "grad_norm": 0.003028750652447343,
      "learning_rate": 0.036138272202079276,
      "loss": 0.3662,
      "num_input_tokens_seen": 42581600,
      "step": 30975
    },
    {
      "epoch": 0.9916138531464055,
      "grad_norm": 0.003945093136280775,
      "learning_rate": 0.036099933948473106,
      "loss": 0.3908,
      "num_input_tokens_seen": 42588176,
      "step": 30980
    },
    {
      "epoch": 0.9917738941168939,
      "grad_norm": 0.007773510180413723,
      "learning_rate": 0.03606161325968851,
      "loss": 0.4335,
      "num_input_tokens_seen": 42595216,
      "step": 30985
    },
    {
      "epoch": 0.9919339350873824,
      "grad_norm": 0.003293101442977786,
      "learning_rate": 0.03602331014163496,
      "loss": 0.4047,
      "num_input_tokens_seen": 42602160,
      "step": 30990
    },
    {
      "epoch": 0.9920939760578709,
      "grad_norm": 0.006715892814099789,
      "learning_rate": 0.035985024600219295,
      "loss": 0.3847,
      "num_input_tokens_seen": 42609520,
      "step": 30995
    },
    {
      "epoch": 0.9922540170283592,
      "grad_norm": 0.005170101299881935,
      "learning_rate": 0.03594675664134569,
      "loss": 0.5263,
      "num_input_tokens_seen": 42616352,
      "step": 31000
    },
    {
      "epoch": 0.9922540170283592,
      "eval_loss": 0.502670168876648,
      "eval_runtime": 332.213,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 42616352,
      "step": 31000
    },
    {
      "epoch": 0.9924140579988477,
      "grad_norm": 0.009632695466279984,
      "learning_rate": 0.03590850627091545,
      "loss": 0.6547,
      "num_input_tokens_seen": 42623760,
      "step": 31005
    },
    {
      "epoch": 0.9925740989693361,
      "grad_norm": 0.003781449282541871,
      "learning_rate": 0.03587027349482731,
      "loss": 0.4778,
      "num_input_tokens_seen": 42631152,
      "step": 31010
    },
    {
      "epoch": 0.9927341399398246,
      "grad_norm": 0.0020188032649457455,
      "learning_rate": 0.035832058318977275,
      "loss": 0.3794,
      "num_input_tokens_seen": 42638128,
      "step": 31015
    },
    {
      "epoch": 0.992894180910313,
      "grad_norm": 0.004047616384923458,
      "learning_rate": 0.03579386074925853,
      "loss": 0.5634,
      "num_input_tokens_seen": 42645248,
      "step": 31020
    },
    {
      "epoch": 0.9930542218808015,
      "grad_norm": 0.006383772939443588,
      "learning_rate": 0.035755680791561696,
      "loss": 0.5032,
      "num_input_tokens_seen": 42651920,
      "step": 31025
    },
    {
      "epoch": 0.9932142628512899,
      "grad_norm": 0.003868738655000925,
      "learning_rate": 0.03571751845177454,
      "loss": 0.545,
      "num_input_tokens_seen": 42658720,
      "step": 31030
    },
    {
      "epoch": 0.9933743038217784,
      "grad_norm": 0.006639326456934214,
      "learning_rate": 0.03567937373578225,
      "loss": 0.6168,
      "num_input_tokens_seen": 42665600,
      "step": 31035
    },
    {
      "epoch": 0.9935343447922668,
      "grad_norm": 0.014624475501477718,
      "learning_rate": 0.03564124664946711,
      "loss": 0.6075,
      "num_input_tokens_seen": 42672608,
      "step": 31040
    },
    {
      "epoch": 0.9936943857627553,
      "grad_norm": 0.004688046872615814,
      "learning_rate": 0.035603137198708924,
      "loss": 0.3133,
      "num_input_tokens_seen": 42679680,
      "step": 31045
    },
    {
      "epoch": 0.9938544267332438,
      "grad_norm": 0.006153902038931847,
      "learning_rate": 0.035565045389384514,
      "loss": 0.4302,
      "num_input_tokens_seen": 42686848,
      "step": 31050
    },
    {
      "epoch": 0.9940144677037321,
      "grad_norm": 0.004751573316752911,
      "learning_rate": 0.03552697122736823,
      "loss": 0.5216,
      "num_input_tokens_seen": 42693472,
      "step": 31055
    },
    {
      "epoch": 0.9941745086742206,
      "grad_norm": 0.0038341344334185123,
      "learning_rate": 0.03548891471853153,
      "loss": 0.4193,
      "num_input_tokens_seen": 42700064,
      "step": 31060
    },
    {
      "epoch": 0.994334549644709,
      "grad_norm": 0.002896179910749197,
      "learning_rate": 0.03545087586874322,
      "loss": 0.4276,
      "num_input_tokens_seen": 42707376,
      "step": 31065
    },
    {
      "epoch": 0.9944945906151975,
      "grad_norm": 0.004772788845002651,
      "learning_rate": 0.03541285468386935,
      "loss": 0.4254,
      "num_input_tokens_seen": 42714400,
      "step": 31070
    },
    {
      "epoch": 0.9946546315856859,
      "grad_norm": 0.007453301455825567,
      "learning_rate": 0.03537485116977327,
      "loss": 0.5499,
      "num_input_tokens_seen": 42721088,
      "step": 31075
    },
    {
      "epoch": 0.9948146725561744,
      "grad_norm": 0.0039984253235161304,
      "learning_rate": 0.03533686533231565,
      "loss": 0.4961,
      "num_input_tokens_seen": 42727776,
      "step": 31080
    },
    {
      "epoch": 0.9949747135266628,
      "grad_norm": 0.005343568045645952,
      "learning_rate": 0.0352988971773543,
      "loss": 0.6165,
      "num_input_tokens_seen": 42734912,
      "step": 31085
    },
    {
      "epoch": 0.9951347544971513,
      "grad_norm": 0.0074682822450995445,
      "learning_rate": 0.03526094671074443,
      "loss": 0.621,
      "num_input_tokens_seen": 42741840,
      "step": 31090
    },
    {
      "epoch": 0.9952947954676398,
      "grad_norm": 0.0020774640142917633,
      "learning_rate": 0.03522301393833852,
      "loss": 0.3804,
      "num_input_tokens_seen": 42748560,
      "step": 31095
    },
    {
      "epoch": 0.9954548364381282,
      "grad_norm": 0.002663013758137822,
      "learning_rate": 0.035185098865986204,
      "loss": 0.5132,
      "num_input_tokens_seen": 42755616,
      "step": 31100
    },
    {
      "epoch": 0.9956148774086166,
      "grad_norm": 0.003491128794848919,
      "learning_rate": 0.03514720149953453,
      "loss": 0.4221,
      "num_input_tokens_seen": 42763024,
      "step": 31105
    },
    {
      "epoch": 0.995774918379105,
      "grad_norm": 0.005468434654176235,
      "learning_rate": 0.03510932184482773,
      "loss": 0.5999,
      "num_input_tokens_seen": 42769456,
      "step": 31110
    },
    {
      "epoch": 0.9959349593495935,
      "grad_norm": 0.0033295140601694584,
      "learning_rate": 0.03507145990770724,
      "loss": 0.4367,
      "num_input_tokens_seen": 42776672,
      "step": 31115
    },
    {
      "epoch": 0.9960950003200819,
      "grad_norm": 0.004082926083356142,
      "learning_rate": 0.035033615694011984,
      "loss": 0.6038,
      "num_input_tokens_seen": 42783872,
      "step": 31120
    },
    {
      "epoch": 0.9962550412905704,
      "grad_norm": 0.005119363311678171,
      "learning_rate": 0.03499578920957788,
      "loss": 0.581,
      "num_input_tokens_seen": 42790736,
      "step": 31125
    },
    {
      "epoch": 0.9964150822610588,
      "grad_norm": 0.003015493508428335,
      "learning_rate": 0.034957980460238375,
      "loss": 0.6071,
      "num_input_tokens_seen": 42797536,
      "step": 31130
    },
    {
      "epoch": 0.9965751232315473,
      "grad_norm": 0.004918608348816633,
      "learning_rate": 0.03492018945182393,
      "loss": 0.4805,
      "num_input_tokens_seen": 42804448,
      "step": 31135
    },
    {
      "epoch": 0.9967351642020357,
      "grad_norm": 0.005898555740714073,
      "learning_rate": 0.03488241619016247,
      "loss": 0.4973,
      "num_input_tokens_seen": 42811280,
      "step": 31140
    },
    {
      "epoch": 0.9968952051725242,
      "grad_norm": 0.0030683220829814672,
      "learning_rate": 0.03484466068107913,
      "loss": 0.6959,
      "num_input_tokens_seen": 42818096,
      "step": 31145
    },
    {
      "epoch": 0.9970552461430127,
      "grad_norm": 0.0030551492236554623,
      "learning_rate": 0.034806922930396195,
      "loss": 0.4207,
      "num_input_tokens_seen": 42824640,
      "step": 31150
    },
    {
      "epoch": 0.997215287113501,
      "grad_norm": 0.0051645576022565365,
      "learning_rate": 0.03476920294393337,
      "loss": 0.5028,
      "num_input_tokens_seen": 42832304,
      "step": 31155
    },
    {
      "epoch": 0.9973753280839895,
      "grad_norm": 0.005498314741998911,
      "learning_rate": 0.03473150072750755,
      "loss": 0.5112,
      "num_input_tokens_seen": 42839056,
      "step": 31160
    },
    {
      "epoch": 0.9975353690544779,
      "grad_norm": 0.004224941600114107,
      "learning_rate": 0.03469381628693284,
      "loss": 0.4212,
      "num_input_tokens_seen": 42845728,
      "step": 31165
    },
    {
      "epoch": 0.9976954100249664,
      "grad_norm": 0.006159026175737381,
      "learning_rate": 0.03465614962802072,
      "loss": 0.5653,
      "num_input_tokens_seen": 42852448,
      "step": 31170
    },
    {
      "epoch": 0.9978554509954548,
      "grad_norm": 0.00494794175028801,
      "learning_rate": 0.0346185007565798,
      "loss": 0.3917,
      "num_input_tokens_seen": 42859136,
      "step": 31175
    },
    {
      "epoch": 0.9980154919659433,
      "grad_norm": 0.0030580400489270687,
      "learning_rate": 0.03458086967841609,
      "loss": 0.4651,
      "num_input_tokens_seen": 42865808,
      "step": 31180
    },
    {
      "epoch": 0.9981755329364317,
      "grad_norm": 0.007091890089213848,
      "learning_rate": 0.03454325639933266,
      "loss": 0.5153,
      "num_input_tokens_seen": 42872816,
      "step": 31185
    },
    {
      "epoch": 0.9983355739069202,
      "grad_norm": 0.005060690455138683,
      "learning_rate": 0.03450566092513007,
      "loss": 0.4649,
      "num_input_tokens_seen": 42879584,
      "step": 31190
    },
    {
      "epoch": 0.9984956148774086,
      "grad_norm": 0.004317671991884708,
      "learning_rate": 0.034468083261605914,
      "loss": 0.483,
      "num_input_tokens_seen": 42886336,
      "step": 31195
    },
    {
      "epoch": 0.9986556558478971,
      "grad_norm": 0.006558852270245552,
      "learning_rate": 0.03443052341455522,
      "loss": 0.6078,
      "num_input_tokens_seen": 42892688,
      "step": 31200
    },
    {
      "epoch": 0.9986556558478971,
      "eval_loss": 0.503192663192749,
      "eval_runtime": 332.0792,
      "eval_samples_per_second": 41.815,
      "eval_steps_per_second": 20.908,
      "num_input_tokens_seen": 42892688,
      "step": 31200
    },
    {
      "epoch": 0.9988156968183856,
      "grad_norm": 0.005031125620007515,
      "learning_rate": 0.0343929813897701,
      "loss": 0.579,
      "num_input_tokens_seen": 42899840,
      "step": 31205
    },
    {
      "epoch": 0.9989757377888739,
      "grad_norm": 0.006498498842120171,
      "learning_rate": 0.034355457193040125,
      "loss": 0.4085,
      "num_input_tokens_seen": 42906528,
      "step": 31210
    },
    {
      "epoch": 0.9991357787593624,
      "grad_norm": 0.004423162434250116,
      "learning_rate": 0.03431795083015186,
      "loss": 0.454,
      "num_input_tokens_seen": 42913968,
      "step": 31215
    },
    {
      "epoch": 0.9992958197298508,
      "grad_norm": 0.006583699490875006,
      "learning_rate": 0.03428046230688936,
      "loss": 0.5931,
      "num_input_tokens_seen": 42921136,
      "step": 31220
    },
    {
      "epoch": 0.9994558607003393,
      "grad_norm": 0.00561115425080061,
      "learning_rate": 0.034242991629033805,
      "loss": 0.4836,
      "num_input_tokens_seen": 42928144,
      "step": 31225
    },
    {
      "epoch": 0.9996159016708277,
      "grad_norm": 0.0030267478432506323,
      "learning_rate": 0.03420553880236362,
      "loss": 0.4719,
      "num_input_tokens_seen": 42935728,
      "step": 31230
    },
    {
      "epoch": 0.9997759426413162,
      "grad_norm": 0.006453614216297865,
      "learning_rate": 0.03416810383265449,
      "loss": 0.4038,
      "num_input_tokens_seen": 42942384,
      "step": 31235
    },
    {
      "epoch": 0.9999359836118046,
      "grad_norm": 0.004878044594079256,
      "learning_rate": 0.03413068672567944,
      "loss": 0.5221,
      "num_input_tokens_seen": 42949504,
      "step": 31240
    },
    {
      "epoch": 1.000096024582293,
      "grad_norm": 0.006857697851955891,
      "learning_rate": 0.034093287487208565,
      "loss": 0.3759,
      "num_input_tokens_seen": 42956000,
      "step": 31245
    },
    {
      "epoch": 1.0002560655527815,
      "grad_norm": 0.0039894417859613895,
      "learning_rate": 0.03405590612300937,
      "loss": 0.4391,
      "num_input_tokens_seen": 42962928,
      "step": 31250
    },
    {
      "epoch": 1.00041610652327,
      "grad_norm": 0.008421187289059162,
      "learning_rate": 0.03401854263884646,
      "loss": 0.5087,
      "num_input_tokens_seen": 42969872,
      "step": 31255
    },
    {
      "epoch": 1.0005761474937584,
      "grad_norm": 0.004719887860119343,
      "learning_rate": 0.033981197040481824,
      "loss": 0.4329,
      "num_input_tokens_seen": 42976592,
      "step": 31260
    },
    {
      "epoch": 1.000736188464247,
      "grad_norm": 0.004373519215732813,
      "learning_rate": 0.03394386933367459,
      "loss": 0.555,
      "num_input_tokens_seen": 42983584,
      "step": 31265
    },
    {
      "epoch": 1.0008962294347352,
      "grad_norm": 0.004582477733492851,
      "learning_rate": 0.033906559524181104,
      "loss": 0.4437,
      "num_input_tokens_seen": 42990224,
      "step": 31270
    },
    {
      "epoch": 1.0010562704052237,
      "grad_norm": 0.006564828101545572,
      "learning_rate": 0.033869267617755085,
      "loss": 0.4065,
      "num_input_tokens_seen": 42997280,
      "step": 31275
    },
    {
      "epoch": 1.0012163113757122,
      "grad_norm": 0.003060168121010065,
      "learning_rate": 0.0338319936201474,
      "loss": 0.3827,
      "num_input_tokens_seen": 43004112,
      "step": 31280
    },
    {
      "epoch": 1.0013763523462007,
      "grad_norm": 0.003973647020757198,
      "learning_rate": 0.033794737537106136,
      "loss": 0.5959,
      "num_input_tokens_seen": 43011296,
      "step": 31285
    },
    {
      "epoch": 1.001536393316689,
      "grad_norm": 0.0033593459520488977,
      "learning_rate": 0.03375749937437671,
      "loss": 0.5007,
      "num_input_tokens_seen": 43018496,
      "step": 31290
    },
    {
      "epoch": 1.0016964342871775,
      "grad_norm": 0.0043337708339095116,
      "learning_rate": 0.033720279137701634,
      "loss": 0.2947,
      "num_input_tokens_seen": 43025104,
      "step": 31295
    },
    {
      "epoch": 1.001856475257666,
      "grad_norm": 0.003549326676875353,
      "learning_rate": 0.03368307683282078,
      "loss": 0.4594,
      "num_input_tokens_seen": 43031856,
      "step": 31300
    },
    {
      "epoch": 1.0020165162281545,
      "grad_norm": 0.005018011201173067,
      "learning_rate": 0.033645892465471235,
      "loss": 0.6017,
      "num_input_tokens_seen": 43038464,
      "step": 31305
    },
    {
      "epoch": 1.002176557198643,
      "grad_norm": 0.004560173489153385,
      "learning_rate": 0.03360872604138724,
      "loss": 0.338,
      "num_input_tokens_seen": 43045168,
      "step": 31310
    },
    {
      "epoch": 1.0023365981691312,
      "grad_norm": 0.002751851687207818,
      "learning_rate": 0.03357157756630034,
      "loss": 0.3995,
      "num_input_tokens_seen": 43052160,
      "step": 31315
    },
    {
      "epoch": 1.0024966391396197,
      "grad_norm": 0.00446275295689702,
      "learning_rate": 0.033534447045939365,
      "loss": 0.535,
      "num_input_tokens_seen": 43059056,
      "step": 31320
    },
    {
      "epoch": 1.0026566801101082,
      "grad_norm": 0.0037091001868247986,
      "learning_rate": 0.03349733448603026,
      "loss": 0.5193,
      "num_input_tokens_seen": 43065808,
      "step": 31325
    },
    {
      "epoch": 1.0028167210805967,
      "grad_norm": 0.00653747096657753,
      "learning_rate": 0.03346023989229619,
      "loss": 0.42,
      "num_input_tokens_seen": 43072304,
      "step": 31330
    },
    {
      "epoch": 1.002976762051085,
      "grad_norm": 0.0027628352399915457,
      "learning_rate": 0.03342316327045769,
      "loss": 0.3752,
      "num_input_tokens_seen": 43078976,
      "step": 31335
    },
    {
      "epoch": 1.0031368030215735,
      "grad_norm": 0.003278731368482113,
      "learning_rate": 0.033386104626232385,
      "loss": 0.4918,
      "num_input_tokens_seen": 43086032,
      "step": 31340
    },
    {
      "epoch": 1.003296843992062,
      "grad_norm": 0.0055152177810668945,
      "learning_rate": 0.03334906396533525,
      "loss": 0.5419,
      "num_input_tokens_seen": 43092496,
      "step": 31345
    },
    {
      "epoch": 1.0034568849625505,
      "grad_norm": 0.004429565276950598,
      "learning_rate": 0.033312041293478326,
      "loss": 0.4229,
      "num_input_tokens_seen": 43099088,
      "step": 31350
    },
    {
      "epoch": 1.0036169259330388,
      "grad_norm": 0.005621312651783228,
      "learning_rate": 0.03327503661637103,
      "loss": 0.5282,
      "num_input_tokens_seen": 43106128,
      "step": 31355
    },
    {
      "epoch": 1.0037769669035272,
      "grad_norm": 0.0027716541662812233,
      "learning_rate": 0.03323804993971998,
      "loss": 0.3716,
      "num_input_tokens_seen": 43113056,
      "step": 31360
    },
    {
      "epoch": 1.0039370078740157,
      "grad_norm": 0.00221290928311646,
      "learning_rate": 0.033201081269228924,
      "loss": 0.4095,
      "num_input_tokens_seen": 43120304,
      "step": 31365
    },
    {
      "epoch": 1.0040970488445042,
      "grad_norm": 0.002638467587530613,
      "learning_rate": 0.03316413061059895,
      "loss": 0.37,
      "num_input_tokens_seen": 43126928,
      "step": 31370
    },
    {
      "epoch": 1.0042570898149927,
      "grad_norm": 0.005863445345312357,
      "learning_rate": 0.03312719796952827,
      "loss": 0.6175,
      "num_input_tokens_seen": 43133408,
      "step": 31375
    },
    {
      "epoch": 1.004417130785481,
      "grad_norm": 0.0029865175019949675,
      "learning_rate": 0.03309028335171236,
      "loss": 0.4579,
      "num_input_tokens_seen": 43140176,
      "step": 31380
    },
    {
      "epoch": 1.0045771717559695,
      "grad_norm": 0.0045825522392988205,
      "learning_rate": 0.03305338676284398,
      "loss": 0.4117,
      "num_input_tokens_seen": 43146640,
      "step": 31385
    },
    {
      "epoch": 1.004737212726458,
      "grad_norm": 0.004947811830788851,
      "learning_rate": 0.03301650820861296,
      "loss": 0.3945,
      "num_input_tokens_seen": 43153184,
      "step": 31390
    },
    {
      "epoch": 1.0048972536969465,
      "grad_norm": 0.005281591787934303,
      "learning_rate": 0.03297964769470652,
      "loss": 0.4338,
      "num_input_tokens_seen": 43160096,
      "step": 31395
    },
    {
      "epoch": 1.0050572946674348,
      "grad_norm": 0.003898621303960681,
      "learning_rate": 0.032942805226808945,
      "loss": 0.4005,
      "num_input_tokens_seen": 43167792,
      "step": 31400
    },
    {
      "epoch": 1.0050572946674348,
      "eval_loss": 0.5013073086738586,
      "eval_runtime": 332.3643,
      "eval_samples_per_second": 41.779,
      "eval_steps_per_second": 20.89,
      "num_input_tokens_seen": 43167792,
      "step": 31400
    },
    {
      "epoch": 1.0052173356379233,
      "grad_norm": 0.006249357480555773,
      "learning_rate": 0.03290598081060187,
      "loss": 0.7495,
      "num_input_tokens_seen": 43175232,
      "step": 31405
    },
    {
      "epoch": 1.0053773766084118,
      "grad_norm": 0.005071481689810753,
      "learning_rate": 0.03286917445176407,
      "loss": 0.3317,
      "num_input_tokens_seen": 43181984,
      "step": 31410
    },
    {
      "epoch": 1.0055374175789002,
      "grad_norm": 0.007113703526556492,
      "learning_rate": 0.032832386155971456,
      "loss": 0.531,
      "num_input_tokens_seen": 43189552,
      "step": 31415
    },
    {
      "epoch": 1.0056974585493887,
      "grad_norm": 0.004384008701890707,
      "learning_rate": 0.032795615928897334,
      "loss": 0.4185,
      "num_input_tokens_seen": 43196864,
      "step": 31420
    },
    {
      "epoch": 1.005857499519877,
      "grad_norm": 0.004399024415761232,
      "learning_rate": 0.03275886377621215,
      "loss": 0.429,
      "num_input_tokens_seen": 43203616,
      "step": 31425
    },
    {
      "epoch": 1.0060175404903655,
      "grad_norm": 0.007805600296705961,
      "learning_rate": 0.03272212970358348,
      "loss": 0.3757,
      "num_input_tokens_seen": 43210336,
      "step": 31430
    },
    {
      "epoch": 1.006177581460854,
      "grad_norm": 0.0032252955716103315,
      "learning_rate": 0.032685413716676215,
      "loss": 0.49,
      "num_input_tokens_seen": 43217056,
      "step": 31435
    },
    {
      "epoch": 1.0063376224313425,
      "grad_norm": 0.00422258535400033,
      "learning_rate": 0.032648715821152474,
      "loss": 0.4813,
      "num_input_tokens_seen": 43223824,
      "step": 31440
    },
    {
      "epoch": 1.0064976634018308,
      "grad_norm": 0.0034140199422836304,
      "learning_rate": 0.03261203602267143,
      "loss": 0.4477,
      "num_input_tokens_seen": 43230768,
      "step": 31445
    },
    {
      "epoch": 1.0066577043723193,
      "grad_norm": 0.0026507682632654905,
      "learning_rate": 0.03257537432688966,
      "loss": 0.3746,
      "num_input_tokens_seen": 43237632,
      "step": 31450
    },
    {
      "epoch": 1.0068177453428078,
      "grad_norm": 0.009119763039052486,
      "learning_rate": 0.03253873073946077,
      "loss": 0.4513,
      "num_input_tokens_seen": 43244624,
      "step": 31455
    },
    {
      "epoch": 1.0069777863132963,
      "grad_norm": 0.004746060818433762,
      "learning_rate": 0.03250210526603572,
      "loss": 0.5226,
      "num_input_tokens_seen": 43251984,
      "step": 31460
    },
    {
      "epoch": 1.0071378272837848,
      "grad_norm": 0.004036350175738335,
      "learning_rate": 0.03246549791226266,
      "loss": 0.6145,
      "num_input_tokens_seen": 43258992,
      "step": 31465
    },
    {
      "epoch": 1.007297868254273,
      "grad_norm": 0.005580396857112646,
      "learning_rate": 0.03242890868378679,
      "loss": 0.4795,
      "num_input_tokens_seen": 43266128,
      "step": 31470
    },
    {
      "epoch": 1.0074579092247615,
      "grad_norm": 0.008087190799415112,
      "learning_rate": 0.03239233758625074,
      "loss": 0.4588,
      "num_input_tokens_seen": 43272928,
      "step": 31475
    },
    {
      "epoch": 1.00761795019525,
      "grad_norm": 0.004406728781759739,
      "learning_rate": 0.032355784625294204,
      "loss": 0.4629,
      "num_input_tokens_seen": 43280352,
      "step": 31480
    },
    {
      "epoch": 1.0077779911657385,
      "grad_norm": 0.005303187761455774,
      "learning_rate": 0.03231924980655402,
      "loss": 0.497,
      "num_input_tokens_seen": 43287392,
      "step": 31485
    },
    {
      "epoch": 1.0079380321362268,
      "grad_norm": 0.005893089808523655,
      "learning_rate": 0.032282733135664446,
      "loss": 0.4588,
      "num_input_tokens_seen": 43294144,
      "step": 31490
    },
    {
      "epoch": 1.0080980731067153,
      "grad_norm": 0.0048964484594762325,
      "learning_rate": 0.03224623461825669,
      "loss": 0.3941,
      "num_input_tokens_seen": 43300672,
      "step": 31495
    },
    {
      "epoch": 1.0082581140772038,
      "grad_norm": 0.004282733425498009,
      "learning_rate": 0.03220975425995937,
      "loss": 0.475,
      "num_input_tokens_seen": 43307296,
      "step": 31500
    },
    {
      "epoch": 1.0084181550476923,
      "grad_norm": 0.0062499819323420525,
      "learning_rate": 0.032173292066398206,
      "loss": 0.5202,
      "num_input_tokens_seen": 43314208,
      "step": 31505
    },
    {
      "epoch": 1.0085781960181806,
      "grad_norm": 0.006388874724507332,
      "learning_rate": 0.03213684804319606,
      "loss": 0.327,
      "num_input_tokens_seen": 43320848,
      "step": 31510
    },
    {
      "epoch": 1.008738236988669,
      "grad_norm": 0.0030069202184677124,
      "learning_rate": 0.03210042219597312,
      "loss": 0.3839,
      "num_input_tokens_seen": 43328208,
      "step": 31515
    },
    {
      "epoch": 1.0088982779591575,
      "grad_norm": 0.005161229055374861,
      "learning_rate": 0.03206401453034675,
      "loss": 0.564,
      "num_input_tokens_seen": 43334848,
      "step": 31520
    },
    {
      "epoch": 1.009058318929646,
      "grad_norm": 0.003937433939427137,
      "learning_rate": 0.03202762505193136,
      "loss": 0.4693,
      "num_input_tokens_seen": 43341728,
      "step": 31525
    },
    {
      "epoch": 1.0092183599001345,
      "grad_norm": 0.0037079553585499525,
      "learning_rate": 0.031991253766338754,
      "loss": 0.4604,
      "num_input_tokens_seen": 43348496,
      "step": 31530
    },
    {
      "epoch": 1.0093784008706228,
      "grad_norm": 0.005279130302369595,
      "learning_rate": 0.03195490067917778,
      "loss": 0.4995,
      "num_input_tokens_seen": 43355184,
      "step": 31535
    },
    {
      "epoch": 1.0095384418411113,
      "grad_norm": 0.003262326819822192,
      "learning_rate": 0.03191856579605461,
      "loss": 0.4231,
      "num_input_tokens_seen": 43362160,
      "step": 31540
    },
    {
      "epoch": 1.0096984828115998,
      "grad_norm": 0.008106561377644539,
      "learning_rate": 0.031882249122572454,
      "loss": 0.3389,
      "num_input_tokens_seen": 43368832,
      "step": 31545
    },
    {
      "epoch": 1.0098585237820883,
      "grad_norm": 0.003358436981216073,
      "learning_rate": 0.03184595066433188,
      "loss": 0.4404,
      "num_input_tokens_seen": 43375632,
      "step": 31550
    },
    {
      "epoch": 1.0100185647525766,
      "grad_norm": 0.004501251503825188,
      "learning_rate": 0.03180967042693049,
      "loss": 0.5663,
      "num_input_tokens_seen": 43382640,
      "step": 31555
    },
    {
      "epoch": 1.010178605723065,
      "grad_norm": 0.0027152334805577993,
      "learning_rate": 0.03177340841596323,
      "loss": 0.383,
      "num_input_tokens_seen": 43389792,
      "step": 31560
    },
    {
      "epoch": 1.0103386466935536,
      "grad_norm": 0.003825055668130517,
      "learning_rate": 0.03173716463702209,
      "loss": 0.4356,
      "num_input_tokens_seen": 43396592,
      "step": 31565
    },
    {
      "epoch": 1.010498687664042,
      "grad_norm": 0.007480144035071135,
      "learning_rate": 0.03170093909569638,
      "loss": 0.4476,
      "num_input_tokens_seen": 43403312,
      "step": 31570
    },
    {
      "epoch": 1.0106587286345305,
      "grad_norm": 0.005062256008386612,
      "learning_rate": 0.03166473179757246,
      "loss": 0.3776,
      "num_input_tokens_seen": 43409888,
      "step": 31575
    },
    {
      "epoch": 1.0108187696050188,
      "grad_norm": 0.005648891907185316,
      "learning_rate": 0.031628542748234005,
      "loss": 0.7322,
      "num_input_tokens_seen": 43417008,
      "step": 31580
    },
    {
      "epoch": 1.0109788105755073,
      "grad_norm": 0.0063338554464280605,
      "learning_rate": 0.03159237195326184,
      "loss": 0.4489,
      "num_input_tokens_seen": 43423872,
      "step": 31585
    },
    {
      "epoch": 1.0111388515459958,
      "grad_norm": 0.011873764917254448,
      "learning_rate": 0.031556219418233875,
      "loss": 0.4715,
      "num_input_tokens_seen": 43431168,
      "step": 31590
    },
    {
      "epoch": 1.0112988925164843,
      "grad_norm": 0.004673257935792208,
      "learning_rate": 0.03152008514872533,
      "loss": 0.4989,
      "num_input_tokens_seen": 43438048,
      "step": 31595
    },
    {
      "epoch": 1.0114589334869726,
      "grad_norm": 0.0037545417435467243,
      "learning_rate": 0.03148396915030862,
      "loss": 0.3178,
      "num_input_tokens_seen": 43444592,
      "step": 31600
    },
    {
      "epoch": 1.0114589334869726,
      "eval_loss": 0.5008519887924194,
      "eval_runtime": 332.1129,
      "eval_samples_per_second": 41.811,
      "eval_steps_per_second": 20.906,
      "num_input_tokens_seen": 43444592,
      "step": 31600
    },
    {
      "epoch": 1.011618974457461,
      "grad_norm": 0.006648683920502663,
      "learning_rate": 0.03144787142855318,
      "loss": 0.3275,
      "num_input_tokens_seen": 43451456,
      "step": 31605
    },
    {
      "epoch": 1.0117790154279496,
      "grad_norm": 0.004196429159492254,
      "learning_rate": 0.031411791989025835,
      "loss": 0.4914,
      "num_input_tokens_seen": 43458272,
      "step": 31610
    },
    {
      "epoch": 1.011939056398438,
      "grad_norm": 0.002869140822440386,
      "learning_rate": 0.031375730837290394,
      "loss": 0.4334,
      "num_input_tokens_seen": 43465472,
      "step": 31615
    },
    {
      "epoch": 1.0120990973689263,
      "grad_norm": 0.0096238749101758,
      "learning_rate": 0.031339687978908015,
      "loss": 0.3554,
      "num_input_tokens_seen": 43472416,
      "step": 31620
    },
    {
      "epoch": 1.0122591383394148,
      "grad_norm": 0.010065211914479733,
      "learning_rate": 0.03130366341943694,
      "loss": 0.7013,
      "num_input_tokens_seen": 43479296,
      "step": 31625
    },
    {
      "epoch": 1.0124191793099033,
      "grad_norm": 0.004222092684358358,
      "learning_rate": 0.031267657164432555,
      "loss": 0.347,
      "num_input_tokens_seen": 43486192,
      "step": 31630
    },
    {
      "epoch": 1.0125792202803918,
      "grad_norm": 0.004208588507026434,
      "learning_rate": 0.03123166921944752,
      "loss": 0.3799,
      "num_input_tokens_seen": 43492896,
      "step": 31635
    },
    {
      "epoch": 1.0127392612508803,
      "grad_norm": 0.006633377633988857,
      "learning_rate": 0.031195699590031666,
      "loss": 0.4972,
      "num_input_tokens_seen": 43499408,
      "step": 31640
    },
    {
      "epoch": 1.0128993022213686,
      "grad_norm": 0.005288217216730118,
      "learning_rate": 0.031159748281731885,
      "loss": 0.3207,
      "num_input_tokens_seen": 43506256,
      "step": 31645
    },
    {
      "epoch": 1.013059343191857,
      "grad_norm": 0.008957836776971817,
      "learning_rate": 0.031123815300092394,
      "loss": 0.6015,
      "num_input_tokens_seen": 43513312,
      "step": 31650
    },
    {
      "epoch": 1.0132193841623456,
      "grad_norm": 0.004926921799778938,
      "learning_rate": 0.031087900650654424,
      "loss": 0.6603,
      "num_input_tokens_seen": 43519968,
      "step": 31655
    },
    {
      "epoch": 1.013379425132834,
      "grad_norm": 0.006072932854294777,
      "learning_rate": 0.031052004338956534,
      "loss": 0.4403,
      "num_input_tokens_seen": 43526928,
      "step": 31660
    },
    {
      "epoch": 1.0135394661033224,
      "grad_norm": 0.004289309494197369,
      "learning_rate": 0.031016126370534407,
      "loss": 0.5432,
      "num_input_tokens_seen": 43533536,
      "step": 31665
    },
    {
      "epoch": 1.0136995070738108,
      "grad_norm": 0.004983372054994106,
      "learning_rate": 0.030980266750920804,
      "loss": 0.3912,
      "num_input_tokens_seen": 43540544,
      "step": 31670
    },
    {
      "epoch": 1.0138595480442993,
      "grad_norm": 0.0031794521491974592,
      "learning_rate": 0.030944425485645747,
      "loss": 0.4437,
      "num_input_tokens_seen": 43546992,
      "step": 31675
    },
    {
      "epoch": 1.0140195890147878,
      "grad_norm": 0.004084140062332153,
      "learning_rate": 0.03090860258023647,
      "loss": 0.4696,
      "num_input_tokens_seen": 43553696,
      "step": 31680
    },
    {
      "epoch": 1.0141796299852763,
      "grad_norm": 0.005161271430552006,
      "learning_rate": 0.030872798040217236,
      "loss": 0.5072,
      "num_input_tokens_seen": 43560432,
      "step": 31685
    },
    {
      "epoch": 1.0143396709557646,
      "grad_norm": 0.007485358510166407,
      "learning_rate": 0.03083701187110964,
      "loss": 0.472,
      "num_input_tokens_seen": 43567120,
      "step": 31690
    },
    {
      "epoch": 1.014499711926253,
      "grad_norm": 0.0020651521626859903,
      "learning_rate": 0.030801244078432294,
      "loss": 0.5172,
      "num_input_tokens_seen": 43574176,
      "step": 31695
    },
    {
      "epoch": 1.0146597528967416,
      "grad_norm": 0.006164704915136099,
      "learning_rate": 0.030765494667701024,
      "loss": 0.817,
      "num_input_tokens_seen": 43580656,
      "step": 31700
    },
    {
      "epoch": 1.01481979386723,
      "grad_norm": 0.005225050263106823,
      "learning_rate": 0.030729763644428913,
      "loss": 0.5872,
      "num_input_tokens_seen": 43587328,
      "step": 31705
    },
    {
      "epoch": 1.0149798348377184,
      "grad_norm": 0.003827726934105158,
      "learning_rate": 0.030694051014126048,
      "loss": 0.5085,
      "num_input_tokens_seen": 43594176,
      "step": 31710
    },
    {
      "epoch": 1.0151398758082069,
      "grad_norm": 0.0044885724782943726,
      "learning_rate": 0.030658356782299792,
      "loss": 0.4788,
      "num_input_tokens_seen": 43601312,
      "step": 31715
    },
    {
      "epoch": 1.0152999167786954,
      "grad_norm": 0.0022618689108639956,
      "learning_rate": 0.030622680954454726,
      "loss": 0.4145,
      "num_input_tokens_seen": 43608080,
      "step": 31720
    },
    {
      "epoch": 1.0154599577491839,
      "grad_norm": 0.0043334802612662315,
      "learning_rate": 0.030587023536092398,
      "loss": 0.4329,
      "num_input_tokens_seen": 43615280,
      "step": 31725
    },
    {
      "epoch": 1.0156199987196723,
      "grad_norm": 0.006188005208969116,
      "learning_rate": 0.03055138453271171,
      "loss": 0.5791,
      "num_input_tokens_seen": 43622384,
      "step": 31730
    },
    {
      "epoch": 1.0157800396901606,
      "grad_norm": 0.002424472477287054,
      "learning_rate": 0.03051576394980858,
      "loss": 0.3402,
      "num_input_tokens_seen": 43629280,
      "step": 31735
    },
    {
      "epoch": 1.0159400806606491,
      "grad_norm": 0.003616218687966466,
      "learning_rate": 0.030480161792876187,
      "loss": 0.4391,
      "num_input_tokens_seen": 43635792,
      "step": 31740
    },
    {
      "epoch": 1.0161001216311376,
      "grad_norm": 0.005687345284968615,
      "learning_rate": 0.030444578067404846,
      "loss": 0.3992,
      "num_input_tokens_seen": 43642816,
      "step": 31745
    },
    {
      "epoch": 1.016260162601626,
      "grad_norm": 0.0029233077075332403,
      "learning_rate": 0.030409012778881975,
      "loss": 0.3602,
      "num_input_tokens_seen": 43650048,
      "step": 31750
    },
    {
      "epoch": 1.0164202035721144,
      "grad_norm": 0.0059176161885261536,
      "learning_rate": 0.030373465932792235,
      "loss": 0.4058,
      "num_input_tokens_seen": 43656736,
      "step": 31755
    },
    {
      "epoch": 1.0165802445426029,
      "grad_norm": 0.005709989462047815,
      "learning_rate": 0.030337937534617342,
      "loss": 0.5892,
      "num_input_tokens_seen": 43663392,
      "step": 31760
    },
    {
      "epoch": 1.0167402855130914,
      "grad_norm": 0.003598557086661458,
      "learning_rate": 0.030302427589836277,
      "loss": 0.4993,
      "num_input_tokens_seen": 43670240,
      "step": 31765
    },
    {
      "epoch": 1.0169003264835799,
      "grad_norm": 0.004335499834269285,
      "learning_rate": 0.030266936103925095,
      "loss": 0.3891,
      "num_input_tokens_seen": 43677088,
      "step": 31770
    },
    {
      "epoch": 1.0170603674540681,
      "grad_norm": 0.0035840305499732494,
      "learning_rate": 0.030231463082356982,
      "loss": 0.6088,
      "num_input_tokens_seen": 43684240,
      "step": 31775
    },
    {
      "epoch": 1.0172204084245566,
      "grad_norm": 0.004715774208307266,
      "learning_rate": 0.030196008530602367,
      "loss": 0.4908,
      "num_input_tokens_seen": 43690784,
      "step": 31780
    },
    {
      "epoch": 1.0173804493950451,
      "grad_norm": 0.009396280162036419,
      "learning_rate": 0.030160572454128842,
      "loss": 0.4233,
      "num_input_tokens_seen": 43697872,
      "step": 31785
    },
    {
      "epoch": 1.0175404903655336,
      "grad_norm": 0.004617666359990835,
      "learning_rate": 0.03012515485840098,
      "loss": 0.5697,
      "num_input_tokens_seen": 43704496,
      "step": 31790
    },
    {
      "epoch": 1.0177005313360221,
      "grad_norm": 0.006753009743988514,
      "learning_rate": 0.030089755748880734,
      "loss": 0.4301,
      "num_input_tokens_seen": 43712128,
      "step": 31795
    },
    {
      "epoch": 1.0178605723065104,
      "grad_norm": 0.0047318902797997,
      "learning_rate": 0.030054375131027003,
      "loss": 0.5995,
      "num_input_tokens_seen": 43719328,
      "step": 31800
    },
    {
      "epoch": 1.0178605723065104,
      "eval_loss": 0.49977782368659973,
      "eval_runtime": 333.0305,
      "eval_samples_per_second": 41.696,
      "eval_steps_per_second": 20.848,
      "num_input_tokens_seen": 43719328,
      "step": 31800
    },
    {
      "epoch": 1.0180206132769989,
      "grad_norm": 0.004999486263841391,
      "learning_rate": 0.030019013010295942,
      "loss": 0.284,
      "num_input_tokens_seen": 43726240,
      "step": 31805
    },
    {
      "epoch": 1.0181806542474874,
      "grad_norm": 0.005801514256745577,
      "learning_rate": 0.029983669392140897,
      "loss": 0.4276,
      "num_input_tokens_seen": 43732992,
      "step": 31810
    },
    {
      "epoch": 1.0183406952179759,
      "grad_norm": 0.004635072313249111,
      "learning_rate": 0.029948344282012217,
      "loss": 0.5672,
      "num_input_tokens_seen": 43740272,
      "step": 31815
    },
    {
      "epoch": 1.0185007361884642,
      "grad_norm": 0.004680713638663292,
      "learning_rate": 0.029913037685357507,
      "loss": 0.3951,
      "num_input_tokens_seen": 43747344,
      "step": 31820
    },
    {
      "epoch": 1.0186607771589526,
      "grad_norm": 0.002749006263911724,
      "learning_rate": 0.029877749607621528,
      "loss": 0.4295,
      "num_input_tokens_seen": 43753904,
      "step": 31825
    },
    {
      "epoch": 1.0188208181294411,
      "grad_norm": 0.005992962513118982,
      "learning_rate": 0.029842480054246077,
      "loss": 0.4665,
      "num_input_tokens_seen": 43760592,
      "step": 31830
    },
    {
      "epoch": 1.0189808590999296,
      "grad_norm": 0.0021196978632360697,
      "learning_rate": 0.02980722903067022,
      "loss": 0.4755,
      "num_input_tokens_seen": 43767568,
      "step": 31835
    },
    {
      "epoch": 1.0191409000704181,
      "grad_norm": 0.0036399641539901495,
      "learning_rate": 0.029771996542330113,
      "loss": 0.3996,
      "num_input_tokens_seen": 43774400,
      "step": 31840
    },
    {
      "epoch": 1.0193009410409064,
      "grad_norm": 0.0034981591161340475,
      "learning_rate": 0.029736782594658954,
      "loss": 0.5236,
      "num_input_tokens_seen": 43781104,
      "step": 31845
    },
    {
      "epoch": 1.019460982011395,
      "grad_norm": 0.004132258705794811,
      "learning_rate": 0.029701587193087284,
      "loss": 0.5143,
      "num_input_tokens_seen": 43787744,
      "step": 31850
    },
    {
      "epoch": 1.0196210229818834,
      "grad_norm": 0.005133007653057575,
      "learning_rate": 0.0296664103430426,
      "loss": 0.4058,
      "num_input_tokens_seen": 43794256,
      "step": 31855
    },
    {
      "epoch": 1.019781063952372,
      "grad_norm": 0.0070336973294615746,
      "learning_rate": 0.029631252049949652,
      "loss": 0.7,
      "num_input_tokens_seen": 43801376,
      "step": 31860
    },
    {
      "epoch": 1.0199411049228602,
      "grad_norm": 0.004019555635750294,
      "learning_rate": 0.02959611231923031,
      "loss": 0.3382,
      "num_input_tokens_seen": 43808416,
      "step": 31865
    },
    {
      "epoch": 1.0201011458933487,
      "grad_norm": 0.002806344535201788,
      "learning_rate": 0.029560991156303507,
      "loss": 0.4107,
      "num_input_tokens_seen": 43815104,
      "step": 31870
    },
    {
      "epoch": 1.0202611868638372,
      "grad_norm": 0.004590472672134638,
      "learning_rate": 0.02952588856658544,
      "loss": 0.3703,
      "num_input_tokens_seen": 43821920,
      "step": 31875
    },
    {
      "epoch": 1.0204212278343257,
      "grad_norm": 0.0038722101598978043,
      "learning_rate": 0.029490804555489296,
      "loss": 0.6092,
      "num_input_tokens_seen": 43828688,
      "step": 31880
    },
    {
      "epoch": 1.020581268804814,
      "grad_norm": 0.003446967573836446,
      "learning_rate": 0.029455739128425484,
      "loss": 0.3062,
      "num_input_tokens_seen": 43835536,
      "step": 31885
    },
    {
      "epoch": 1.0207413097753024,
      "grad_norm": 0.004515875596553087,
      "learning_rate": 0.029420692290801607,
      "loss": 0.6137,
      "num_input_tokens_seen": 43842592,
      "step": 31890
    },
    {
      "epoch": 1.020901350745791,
      "grad_norm": 0.004395944066345692,
      "learning_rate": 0.02938566404802223,
      "loss": 0.7378,
      "num_input_tokens_seen": 43849344,
      "step": 31895
    },
    {
      "epoch": 1.0210613917162794,
      "grad_norm": 0.0055099064484238625,
      "learning_rate": 0.029350654405489195,
      "loss": 0.4959,
      "num_input_tokens_seen": 43856256,
      "step": 31900
    },
    {
      "epoch": 1.021221432686768,
      "grad_norm": 0.004733944311738014,
      "learning_rate": 0.02931566336860145,
      "loss": 0.3418,
      "num_input_tokens_seen": 43863920,
      "step": 31905
    },
    {
      "epoch": 1.0213814736572562,
      "grad_norm": 0.003711237572133541,
      "learning_rate": 0.02928069094275505,
      "loss": 0.3946,
      "num_input_tokens_seen": 43870624,
      "step": 31910
    },
    {
      "epoch": 1.0215415146277447,
      "grad_norm": 0.003835177281871438,
      "learning_rate": 0.02924573713334314,
      "loss": 0.5107,
      "num_input_tokens_seen": 43877856,
      "step": 31915
    },
    {
      "epoch": 1.0217015555982332,
      "grad_norm": 0.006076520774513483,
      "learning_rate": 0.02921080194575603,
      "loss": 0.7449,
      "num_input_tokens_seen": 43885376,
      "step": 31920
    },
    {
      "epoch": 1.0218615965687217,
      "grad_norm": 0.004815214779227972,
      "learning_rate": 0.029175885385381177,
      "loss": 0.332,
      "num_input_tokens_seen": 43891808,
      "step": 31925
    },
    {
      "epoch": 1.02202163753921,
      "grad_norm": 0.006328859832137823,
      "learning_rate": 0.029140987457603223,
      "loss": 0.5594,
      "num_input_tokens_seen": 43898848,
      "step": 31930
    },
    {
      "epoch": 1.0221816785096984,
      "grad_norm": 0.004656296689063311,
      "learning_rate": 0.029106108167803763,
      "loss": 0.47,
      "num_input_tokens_seen": 43905712,
      "step": 31935
    },
    {
      "epoch": 1.022341719480187,
      "grad_norm": 0.002508208155632019,
      "learning_rate": 0.029071247521361674,
      "loss": 0.4936,
      "num_input_tokens_seen": 43912288,
      "step": 31940
    },
    {
      "epoch": 1.0225017604506754,
      "grad_norm": 0.006341294385492802,
      "learning_rate": 0.029036405523652945,
      "loss": 0.5301,
      "num_input_tokens_seen": 43919280,
      "step": 31945
    },
    {
      "epoch": 1.022661801421164,
      "grad_norm": 0.003430477576330304,
      "learning_rate": 0.029001582180050577,
      "loss": 0.325,
      "num_input_tokens_seen": 43926032,
      "step": 31950
    },
    {
      "epoch": 1.0228218423916522,
      "grad_norm": 0.00627824617549777,
      "learning_rate": 0.02896677749592482,
      "loss": 0.3409,
      "num_input_tokens_seen": 43932368,
      "step": 31955
    },
    {
      "epoch": 1.0229818833621407,
      "grad_norm": 0.006681925151497126,
      "learning_rate": 0.028931991476642938,
      "loss": 0.3923,
      "num_input_tokens_seen": 43938912,
      "step": 31960
    },
    {
      "epoch": 1.0231419243326292,
      "grad_norm": 0.006285583600401878,
      "learning_rate": 0.028897224127569412,
      "loss": 0.4404,
      "num_input_tokens_seen": 43945968,
      "step": 31965
    },
    {
      "epoch": 1.0233019653031177,
      "grad_norm": 0.00385485147126019,
      "learning_rate": 0.028862475454065832,
      "loss": 0.441,
      "num_input_tokens_seen": 43953200,
      "step": 31970
    },
    {
      "epoch": 1.023462006273606,
      "grad_norm": 0.004775928799062967,
      "learning_rate": 0.028827745461490806,
      "loss": 0.3191,
      "num_input_tokens_seen": 43959984,
      "step": 31975
    },
    {
      "epoch": 1.0236220472440944,
      "grad_norm": 0.004070572089403868,
      "learning_rate": 0.028793034155200212,
      "loss": 0.5834,
      "num_input_tokens_seen": 43967264,
      "step": 31980
    },
    {
      "epoch": 1.023782088214583,
      "grad_norm": 0.004703065380454063,
      "learning_rate": 0.028758341540546944,
      "loss": 0.3255,
      "num_input_tokens_seen": 43973792,
      "step": 31985
    },
    {
      "epoch": 1.0239421291850714,
      "grad_norm": 0.009788008406758308,
      "learning_rate": 0.02872366762288098,
      "loss": 0.5226,
      "num_input_tokens_seen": 43980432,
      "step": 31990
    },
    {
      "epoch": 1.02410217015556,
      "grad_norm": 0.003287827130407095,
      "learning_rate": 0.028689012407549567,
      "loss": 0.3514,
      "num_input_tokens_seen": 43987312,
      "step": 31995
    },
    {
      "epoch": 1.0242622111260482,
      "grad_norm": 0.006942383944988251,
      "learning_rate": 0.028654375899896892,
      "loss": 0.4926,
      "num_input_tokens_seen": 43994064,
      "step": 32000
    },
    {
      "epoch": 1.0242622111260482,
      "eval_loss": 0.49994444847106934,
      "eval_runtime": 332.3997,
      "eval_samples_per_second": 41.775,
      "eval_steps_per_second": 20.888,
      "num_input_tokens_seen": 43994064,
      "step": 32000
    },
    {
      "epoch": 1.0244222520965367,
      "grad_norm": 0.006770405452698469,
      "learning_rate": 0.02861975810526437,
      "loss": 0.4776,
      "num_input_tokens_seen": 44000848,
      "step": 32005
    },
    {
      "epoch": 1.0245822930670252,
      "grad_norm": 0.0035388434771448374,
      "learning_rate": 0.02858515902899056,
      "loss": 0.4737,
      "num_input_tokens_seen": 44007504,
      "step": 32010
    },
    {
      "epoch": 1.0247423340375137,
      "grad_norm": 0.006658350583165884,
      "learning_rate": 0.028550578676410976,
      "loss": 0.4252,
      "num_input_tokens_seen": 44014528,
      "step": 32015
    },
    {
      "epoch": 1.024902375008002,
      "grad_norm": 0.0062209004536271095,
      "learning_rate": 0.02851601705285837,
      "loss": 0.5763,
      "num_input_tokens_seen": 44021408,
      "step": 32020
    },
    {
      "epoch": 1.0250624159784905,
      "grad_norm": 0.008203425444662571,
      "learning_rate": 0.028481474163662666,
      "loss": 0.4288,
      "num_input_tokens_seen": 44028160,
      "step": 32025
    },
    {
      "epoch": 1.025222456948979,
      "grad_norm": 0.004221934825181961,
      "learning_rate": 0.028446950014150683,
      "loss": 0.3869,
      "num_input_tokens_seen": 44034976,
      "step": 32030
    },
    {
      "epoch": 1.0253824979194675,
      "grad_norm": 0.004584951791912317,
      "learning_rate": 0.028412444609646596,
      "loss": 0.5619,
      "num_input_tokens_seen": 44041856,
      "step": 32035
    },
    {
      "epoch": 1.0255425388899557,
      "grad_norm": 0.0042715235613286495,
      "learning_rate": 0.028377957955471465,
      "loss": 0.4432,
      "num_input_tokens_seen": 44048560,
      "step": 32040
    },
    {
      "epoch": 1.0257025798604442,
      "grad_norm": 0.003785693319514394,
      "learning_rate": 0.0283434900569436,
      "loss": 0.4825,
      "num_input_tokens_seen": 44055488,
      "step": 32045
    },
    {
      "epoch": 1.0258626208309327,
      "grad_norm": 0.004762344062328339,
      "learning_rate": 0.028309040919378456,
      "loss": 0.3533,
      "num_input_tokens_seen": 44062416,
      "step": 32050
    },
    {
      "epoch": 1.0260226618014212,
      "grad_norm": 0.006989115849137306,
      "learning_rate": 0.02827461054808848,
      "loss": 0.6137,
      "num_input_tokens_seen": 44068880,
      "step": 32055
    },
    {
      "epoch": 1.0261827027719097,
      "grad_norm": 0.006619608495384455,
      "learning_rate": 0.028240198948383186,
      "loss": 0.5848,
      "num_input_tokens_seen": 44076192,
      "step": 32060
    },
    {
      "epoch": 1.026342743742398,
      "grad_norm": 0.0023639206774532795,
      "learning_rate": 0.028205806125569402,
      "loss": 0.3623,
      "num_input_tokens_seen": 44083120,
      "step": 32065
    },
    {
      "epoch": 1.0265027847128865,
      "grad_norm": 0.0043433294631540775,
      "learning_rate": 0.028171432084950834,
      "loss": 0.4405,
      "num_input_tokens_seen": 44089824,
      "step": 32070
    },
    {
      "epoch": 1.026662825683375,
      "grad_norm": 0.0033120345324277878,
      "learning_rate": 0.028137076831828478,
      "loss": 0.3591,
      "num_input_tokens_seen": 44096560,
      "step": 32075
    },
    {
      "epoch": 1.0268228666538635,
      "grad_norm": 0.004493164364248514,
      "learning_rate": 0.028102740371500238,
      "loss": 0.4572,
      "num_input_tokens_seen": 44103696,
      "step": 32080
    },
    {
      "epoch": 1.0269829076243517,
      "grad_norm": 0.0030837426893413067,
      "learning_rate": 0.0280684227092613,
      "loss": 0.2866,
      "num_input_tokens_seen": 44110432,
      "step": 32085
    },
    {
      "epoch": 1.0271429485948402,
      "grad_norm": 0.005239661782979965,
      "learning_rate": 0.02803412385040392,
      "loss": 0.4195,
      "num_input_tokens_seen": 44117104,
      "step": 32090
    },
    {
      "epoch": 1.0273029895653287,
      "grad_norm": 0.007116671651601791,
      "learning_rate": 0.027999843800217306,
      "loss": 0.4432,
      "num_input_tokens_seen": 44123680,
      "step": 32095
    },
    {
      "epoch": 1.0274630305358172,
      "grad_norm": 0.004950850270688534,
      "learning_rate": 0.027965582563987932,
      "loss": 0.4591,
      "num_input_tokens_seen": 44130496,
      "step": 32100
    },
    {
      "epoch": 1.0276230715063057,
      "grad_norm": 0.007627212908118963,
      "learning_rate": 0.027931340146999346,
      "loss": 0.4447,
      "num_input_tokens_seen": 44137568,
      "step": 32105
    },
    {
      "epoch": 1.027783112476794,
      "grad_norm": 0.006724042352288961,
      "learning_rate": 0.02789711655453208,
      "loss": 0.5263,
      "num_input_tokens_seen": 44144256,
      "step": 32110
    },
    {
      "epoch": 1.0279431534472825,
      "grad_norm": 0.0036420414689928293,
      "learning_rate": 0.02786291179186392,
      "loss": 0.518,
      "num_input_tokens_seen": 44151312,
      "step": 32115
    },
    {
      "epoch": 1.028103194417771,
      "grad_norm": 0.005884879734367132,
      "learning_rate": 0.02782872586426961,
      "loss": 0.5059,
      "num_input_tokens_seen": 44157888,
      "step": 32120
    },
    {
      "epoch": 1.0282632353882595,
      "grad_norm": 0.01384669914841652,
      "learning_rate": 0.027794558777021083,
      "loss": 0.8169,
      "num_input_tokens_seen": 44165280,
      "step": 32125
    },
    {
      "epoch": 1.0284232763587478,
      "grad_norm": 0.004975990392267704,
      "learning_rate": 0.02776041053538734,
      "loss": 0.5871,
      "num_input_tokens_seen": 44171904,
      "step": 32130
    },
    {
      "epoch": 1.0285833173292362,
      "grad_norm": 0.005807003937661648,
      "learning_rate": 0.027726281144634407,
      "loss": 0.5218,
      "num_input_tokens_seen": 44178768,
      "step": 32135
    },
    {
      "epoch": 1.0287433582997247,
      "grad_norm": 0.004243744071573019,
      "learning_rate": 0.02769217061002552,
      "loss": 0.501,
      "num_input_tokens_seen": 44186928,
      "step": 32140
    },
    {
      "epoch": 1.0289033992702132,
      "grad_norm": 0.0060686348006129265,
      "learning_rate": 0.027658078936820967,
      "loss": 0.4579,
      "num_input_tokens_seen": 44193664,
      "step": 32145
    },
    {
      "epoch": 1.0290634402407015,
      "grad_norm": 0.011510690674185753,
      "learning_rate": 0.02762400613027805,
      "loss": 0.526,
      "num_input_tokens_seen": 44200992,
      "step": 32150
    },
    {
      "epoch": 1.02922348121119,
      "grad_norm": 0.005975025240331888,
      "learning_rate": 0.027589952195651295,
      "loss": 0.3514,
      "num_input_tokens_seen": 44207824,
      "step": 32155
    },
    {
      "epoch": 1.0293835221816785,
      "grad_norm": 0.005523574072867632,
      "learning_rate": 0.027555917138192186,
      "loss": 0.3514,
      "num_input_tokens_seen": 44214496,
      "step": 32160
    },
    {
      "epoch": 1.029543563152167,
      "grad_norm": 0.0037994678132236004,
      "learning_rate": 0.027521900963149375,
      "loss": 0.3856,
      "num_input_tokens_seen": 44221360,
      "step": 32165
    },
    {
      "epoch": 1.0297036041226555,
      "grad_norm": 0.004905691836029291,
      "learning_rate": 0.027487903675768633,
      "loss": 0.4355,
      "num_input_tokens_seen": 44228416,
      "step": 32170
    },
    {
      "epoch": 1.0298636450931438,
      "grad_norm": 0.0038456881884485483,
      "learning_rate": 0.027453925281292677,
      "loss": 0.5521,
      "num_input_tokens_seen": 44235280,
      "step": 32175
    },
    {
      "epoch": 1.0300236860636323,
      "grad_norm": 0.0041946168057620525,
      "learning_rate": 0.027419965784961475,
      "loss": 0.397,
      "num_input_tokens_seen": 44242256,
      "step": 32180
    },
    {
      "epoch": 1.0301837270341208,
      "grad_norm": 0.005845197476446629,
      "learning_rate": 0.027386025192012015,
      "loss": 0.3565,
      "num_input_tokens_seen": 44249152,
      "step": 32185
    },
    {
      "epoch": 1.0303437680046093,
      "grad_norm": 0.0029871349688619375,
      "learning_rate": 0.027352103507678277,
      "loss": 0.5481,
      "num_input_tokens_seen": 44256368,
      "step": 32190
    },
    {
      "epoch": 1.0305038089750975,
      "grad_norm": 0.011519907973706722,
      "learning_rate": 0.027318200737191527,
      "loss": 0.5553,
      "num_input_tokens_seen": 44262864,
      "step": 32195
    },
    {
      "epoch": 1.030663849945586,
      "grad_norm": 0.012291817925870419,
      "learning_rate": 0.027284316885779935,
      "loss": 0.5995,
      "num_input_tokens_seen": 44269712,
      "step": 32200
    },
    {
      "epoch": 1.030663849945586,
      "eval_loss": 0.4989619851112366,
      "eval_runtime": 332.4361,
      "eval_samples_per_second": 41.77,
      "eval_steps_per_second": 20.885,
      "num_input_tokens_seen": 44269712,
      "step": 32200
    },
    {
      "epoch": 1.0308238909160745,
      "grad_norm": 0.003669235622510314,
      "learning_rate": 0.027250451958668785,
      "loss": 0.3584,
      "num_input_tokens_seen": 44276592,
      "step": 32205
    },
    {
      "epoch": 1.030983931886563,
      "grad_norm": 0.0035810228437185287,
      "learning_rate": 0.027216605961080536,
      "loss": 0.5416,
      "num_input_tokens_seen": 44283344,
      "step": 32210
    },
    {
      "epoch": 1.0311439728570515,
      "grad_norm": 0.005689981859177351,
      "learning_rate": 0.02718277889823461,
      "loss": 0.649,
      "num_input_tokens_seen": 44290064,
      "step": 32215
    },
    {
      "epoch": 1.0313040138275398,
      "grad_norm": 0.0077261063270270824,
      "learning_rate": 0.027148970775347604,
      "loss": 0.585,
      "num_input_tokens_seen": 44296864,
      "step": 32220
    },
    {
      "epoch": 1.0314640547980283,
      "grad_norm": 0.004099132958799601,
      "learning_rate": 0.027115181597633174,
      "loss": 0.5303,
      "num_input_tokens_seen": 44303264,
      "step": 32225
    },
    {
      "epoch": 1.0316240957685168,
      "grad_norm": 0.005914069712162018,
      "learning_rate": 0.027081411370301976,
      "loss": 0.5146,
      "num_input_tokens_seen": 44310544,
      "step": 32230
    },
    {
      "epoch": 1.0317841367390053,
      "grad_norm": 0.0028115855529904366,
      "learning_rate": 0.027047660098561875,
      "loss": 0.3633,
      "num_input_tokens_seen": 44317088,
      "step": 32235
    },
    {
      "epoch": 1.0319441777094935,
      "grad_norm": 0.005388657096773386,
      "learning_rate": 0.02701392778761766,
      "loss": 0.4918,
      "num_input_tokens_seen": 44324288,
      "step": 32240
    },
    {
      "epoch": 1.032104218679982,
      "grad_norm": 0.004273401107639074,
      "learning_rate": 0.02698021444267133,
      "loss": 0.5575,
      "num_input_tokens_seen": 44331920,
      "step": 32245
    },
    {
      "epoch": 1.0322642596504705,
      "grad_norm": 0.002451429609209299,
      "learning_rate": 0.026946520068921915,
      "loss": 0.4333,
      "num_input_tokens_seen": 44338800,
      "step": 32250
    },
    {
      "epoch": 1.032424300620959,
      "grad_norm": 0.004671856295317411,
      "learning_rate": 0.02691284467156547,
      "loss": 0.5022,
      "num_input_tokens_seen": 44345456,
      "step": 32255
    },
    {
      "epoch": 1.0325843415914475,
      "grad_norm": 0.004798928741365671,
      "learning_rate": 0.026879188255795182,
      "loss": 0.4086,
      "num_input_tokens_seen": 44352176,
      "step": 32260
    },
    {
      "epoch": 1.0327443825619358,
      "grad_norm": 0.005234020296484232,
      "learning_rate": 0.026845550826801328,
      "loss": 0.41,
      "num_input_tokens_seen": 44359552,
      "step": 32265
    },
    {
      "epoch": 1.0329044235324243,
      "grad_norm": 0.00680190697312355,
      "learning_rate": 0.02681193238977121,
      "loss": 0.6465,
      "num_input_tokens_seen": 44366160,
      "step": 32270
    },
    {
      "epoch": 1.0330644645029128,
      "grad_norm": 0.003701220266520977,
      "learning_rate": 0.026778332949889145,
      "loss": 0.4232,
      "num_input_tokens_seen": 44372992,
      "step": 32275
    },
    {
      "epoch": 1.0332245054734013,
      "grad_norm": 0.0029341578483581543,
      "learning_rate": 0.026744752512336673,
      "loss": 0.5078,
      "num_input_tokens_seen": 44380112,
      "step": 32280
    },
    {
      "epoch": 1.0333845464438896,
      "grad_norm": 0.011574660427868366,
      "learning_rate": 0.02671119108229225,
      "loss": 0.551,
      "num_input_tokens_seen": 44386928,
      "step": 32285
    },
    {
      "epoch": 1.033544587414378,
      "grad_norm": 0.006200016476213932,
      "learning_rate": 0.026677648664931556,
      "loss": 0.4747,
      "num_input_tokens_seen": 44393776,
      "step": 32290
    },
    {
      "epoch": 1.0337046283848665,
      "grad_norm": 0.0039056024979799986,
      "learning_rate": 0.026644125265427154,
      "loss": 0.5227,
      "num_input_tokens_seen": 44400688,
      "step": 32295
    },
    {
      "epoch": 1.033864669355355,
      "grad_norm": 0.005363674834370613,
      "learning_rate": 0.026610620888948822,
      "loss": 0.3726,
      "num_input_tokens_seen": 44407680,
      "step": 32300
    },
    {
      "epoch": 1.0340247103258433,
      "grad_norm": 0.007502575404942036,
      "learning_rate": 0.026577135540663408,
      "loss": 0.6211,
      "num_input_tokens_seen": 44414816,
      "step": 32305
    },
    {
      "epoch": 1.0341847512963318,
      "grad_norm": 0.009186947718262672,
      "learning_rate": 0.026543669225734673,
      "loss": 0.5785,
      "num_input_tokens_seen": 44422016,
      "step": 32310
    },
    {
      "epoch": 1.0343447922668203,
      "grad_norm": 0.006488041952252388,
      "learning_rate": 0.02651022194932363,
      "loss": 0.648,
      "num_input_tokens_seen": 44428816,
      "step": 32315
    },
    {
      "epoch": 1.0345048332373088,
      "grad_norm": 0.0029607191681861877,
      "learning_rate": 0.026476793716588194,
      "loss": 0.4777,
      "num_input_tokens_seen": 44435488,
      "step": 32320
    },
    {
      "epoch": 1.0346648742077973,
      "grad_norm": 0.003398042870685458,
      "learning_rate": 0.026443384532683467,
      "loss": 0.3733,
      "num_input_tokens_seen": 44442704,
      "step": 32325
    },
    {
      "epoch": 1.0348249151782856,
      "grad_norm": 0.0071847341023385525,
      "learning_rate": 0.026409994402761584,
      "loss": 0.5658,
      "num_input_tokens_seen": 44449440,
      "step": 32330
    },
    {
      "epoch": 1.034984956148774,
      "grad_norm": 0.01850356161594391,
      "learning_rate": 0.026376623331971653,
      "loss": 0.562,
      "num_input_tokens_seen": 44456688,
      "step": 32335
    },
    {
      "epoch": 1.0351449971192626,
      "grad_norm": 0.0047831665724515915,
      "learning_rate": 0.026343271325459997,
      "loss": 0.6405,
      "num_input_tokens_seen": 44463360,
      "step": 32340
    },
    {
      "epoch": 1.035305038089751,
      "grad_norm": 0.005792306736111641,
      "learning_rate": 0.02630993838836987,
      "loss": 0.3362,
      "num_input_tokens_seen": 44470336,
      "step": 32345
    },
    {
      "epoch": 1.0354650790602393,
      "grad_norm": 0.005992184393107891,
      "learning_rate": 0.026276624525841584,
      "loss": 0.4471,
      "num_input_tokens_seen": 44477408,
      "step": 32350
    },
    {
      "epoch": 1.0356251200307278,
      "grad_norm": 0.0026454103644937277,
      "learning_rate": 0.026243329743012637,
      "loss": 0.474,
      "num_input_tokens_seen": 44484304,
      "step": 32355
    },
    {
      "epoch": 1.0357851610012163,
      "grad_norm": 0.005935049150139093,
      "learning_rate": 0.026210054045017438,
      "loss": 0.3421,
      "num_input_tokens_seen": 44490784,
      "step": 32360
    },
    {
      "epoch": 1.0359452019717048,
      "grad_norm": 0.0045479414984583855,
      "learning_rate": 0.02617679743698755,
      "loss": 0.8416,
      "num_input_tokens_seen": 44497168,
      "step": 32365
    },
    {
      "epoch": 1.0361052429421933,
      "grad_norm": 0.002507840283215046,
      "learning_rate": 0.02614355992405158,
      "loss": 0.3553,
      "num_input_tokens_seen": 44503952,
      "step": 32370
    },
    {
      "epoch": 1.0362652839126816,
      "grad_norm": 0.002219023648649454,
      "learning_rate": 0.026110341511335115,
      "loss": 0.4561,
      "num_input_tokens_seen": 44510736,
      "step": 32375
    },
    {
      "epoch": 1.03642532488317,
      "grad_norm": 0.006494044791907072,
      "learning_rate": 0.02607714220396093,
      "loss": 0.4824,
      "num_input_tokens_seen": 44517760,
      "step": 32380
    },
    {
      "epoch": 1.0365853658536586,
      "grad_norm": 0.005244602914899588,
      "learning_rate": 0.02604396200704869,
      "loss": 0.5282,
      "num_input_tokens_seen": 44524544,
      "step": 32385
    },
    {
      "epoch": 1.036745406824147,
      "grad_norm": 0.002709768945351243,
      "learning_rate": 0.02601080092571523,
      "loss": 0.4687,
      "num_input_tokens_seen": 44531328,
      "step": 32390
    },
    {
      "epoch": 1.0369054477946353,
      "grad_norm": 0.0038938652724027634,
      "learning_rate": 0.025977658965074455,
      "loss": 0.4345,
      "num_input_tokens_seen": 44538880,
      "step": 32395
    },
    {
      "epoch": 1.0370654887651238,
      "grad_norm": 0.004711908753961325,
      "learning_rate": 0.02594453613023719,
      "loss": 0.5196,
      "num_input_tokens_seen": 44545408,
      "step": 32400
    },
    {
      "epoch": 1.0370654887651238,
      "eval_loss": 0.4984075725078583,
      "eval_runtime": 332.0099,
      "eval_samples_per_second": 41.824,
      "eval_steps_per_second": 20.912,
      "num_input_tokens_seen": 44545408,
      "step": 32400
    },
    {
      "epoch": 1.0372255297356123,
      "grad_norm": 0.0035178440157324076,
      "learning_rate": 0.025911432426311443,
      "loss": 0.4805,
      "num_input_tokens_seen": 44552288,
      "step": 32405
    },
    {
      "epoch": 1.0373855707061008,
      "grad_norm": 0.009475491009652615,
      "learning_rate": 0.025878347858402234,
      "loss": 0.651,
      "num_input_tokens_seen": 44559024,
      "step": 32410
    },
    {
      "epoch": 1.037545611676589,
      "grad_norm": 0.004922448191791773,
      "learning_rate": 0.025845282431611598,
      "loss": 0.509,
      "num_input_tokens_seen": 44565584,
      "step": 32415
    },
    {
      "epoch": 1.0377056526470776,
      "grad_norm": 0.003587679471820593,
      "learning_rate": 0.025812236151038608,
      "loss": 0.5097,
      "num_input_tokens_seen": 44572352,
      "step": 32420
    },
    {
      "epoch": 1.037865693617566,
      "grad_norm": 0.004238084424287081,
      "learning_rate": 0.025779209021779468,
      "loss": 0.4267,
      "num_input_tokens_seen": 44579312,
      "step": 32425
    },
    {
      "epoch": 1.0380257345880546,
      "grad_norm": 0.009633481502532959,
      "learning_rate": 0.025746201048927324,
      "loss": 0.3385,
      "num_input_tokens_seen": 44586368,
      "step": 32430
    },
    {
      "epoch": 1.038185775558543,
      "grad_norm": 0.009568444453179836,
      "learning_rate": 0.025713212237572485,
      "loss": 0.4619,
      "num_input_tokens_seen": 44593360,
      "step": 32435
    },
    {
      "epoch": 1.0383458165290314,
      "grad_norm": 0.0021972637623548508,
      "learning_rate": 0.025680242592802164,
      "loss": 0.3406,
      "num_input_tokens_seen": 44600000,
      "step": 32440
    },
    {
      "epoch": 1.0385058574995198,
      "grad_norm": 0.004054687917232513,
      "learning_rate": 0.02564729211970073,
      "loss": 0.4833,
      "num_input_tokens_seen": 44606976,
      "step": 32445
    },
    {
      "epoch": 1.0386658984700083,
      "grad_norm": 0.00868893787264824,
      "learning_rate": 0.025614360823349617,
      "loss": 0.6167,
      "num_input_tokens_seen": 44614416,
      "step": 32450
    },
    {
      "epoch": 1.0388259394404968,
      "grad_norm": 0.004024706315249205,
      "learning_rate": 0.025581448708827146,
      "loss": 0.4685,
      "num_input_tokens_seen": 44621200,
      "step": 32455
    },
    {
      "epoch": 1.038985980410985,
      "grad_norm": 0.0036677909083664417,
      "learning_rate": 0.025548555781208876,
      "loss": 0.4315,
      "num_input_tokens_seen": 44627792,
      "step": 32460
    },
    {
      "epoch": 1.0391460213814736,
      "grad_norm": 0.0036106009501963854,
      "learning_rate": 0.02551568204556721,
      "loss": 0.5295,
      "num_input_tokens_seen": 44635040,
      "step": 32465
    },
    {
      "epoch": 1.039306062351962,
      "grad_norm": 0.0029035131447017193,
      "learning_rate": 0.02548282750697173,
      "loss": 0.4302,
      "num_input_tokens_seen": 44641936,
      "step": 32470
    },
    {
      "epoch": 1.0394661033224506,
      "grad_norm": 0.0051857950165867805,
      "learning_rate": 0.02544999217048909,
      "loss": 0.4992,
      "num_input_tokens_seen": 44649072,
      "step": 32475
    },
    {
      "epoch": 1.039626144292939,
      "grad_norm": 0.0054990206845104694,
      "learning_rate": 0.025417176041182793,
      "loss": 0.5297,
      "num_input_tokens_seen": 44656016,
      "step": 32480
    },
    {
      "epoch": 1.0397861852634274,
      "grad_norm": 0.0035553397610783577,
      "learning_rate": 0.025384379124113596,
      "loss": 0.4828,
      "num_input_tokens_seen": 44662912,
      "step": 32485
    },
    {
      "epoch": 1.0399462262339159,
      "grad_norm": 0.006970374379307032,
      "learning_rate": 0.025351601424339124,
      "loss": 0.5425,
      "num_input_tokens_seen": 44669888,
      "step": 32490
    },
    {
      "epoch": 1.0401062672044044,
      "grad_norm": 0.003610244020819664,
      "learning_rate": 0.025318842946914184,
      "loss": 0.5486,
      "num_input_tokens_seen": 44676544,
      "step": 32495
    },
    {
      "epoch": 1.0402663081748929,
      "grad_norm": 0.005061222240328789,
      "learning_rate": 0.025286103696890494,
      "loss": 0.4016,
      "num_input_tokens_seen": 44683024,
      "step": 32500
    },
    {
      "epoch": 1.0404263491453811,
      "grad_norm": 0.006616665981709957,
      "learning_rate": 0.025253383679316836,
      "loss": 0.6487,
      "num_input_tokens_seen": 44690032,
      "step": 32505
    },
    {
      "epoch": 1.0405863901158696,
      "grad_norm": 0.005490491166710854,
      "learning_rate": 0.025220682899239077,
      "loss": 0.6801,
      "num_input_tokens_seen": 44696704,
      "step": 32510
    },
    {
      "epoch": 1.0407464310863581,
      "grad_norm": 0.002618449041619897,
      "learning_rate": 0.02518800136170013,
      "loss": 0.4863,
      "num_input_tokens_seen": 44703264,
      "step": 32515
    },
    {
      "epoch": 1.0409064720568466,
      "grad_norm": 0.005106160417199135,
      "learning_rate": 0.02515533907173981,
      "loss": 0.5419,
      "num_input_tokens_seen": 44709952,
      "step": 32520
    },
    {
      "epoch": 1.041066513027335,
      "grad_norm": 0.004513347987085581,
      "learning_rate": 0.025122696034395115,
      "loss": 0.4996,
      "num_input_tokens_seen": 44716848,
      "step": 32525
    },
    {
      "epoch": 1.0412265539978234,
      "grad_norm": 0.007366345264017582,
      "learning_rate": 0.025090072254700023,
      "loss": 0.5341,
      "num_input_tokens_seen": 44724256,
      "step": 32530
    },
    {
      "epoch": 1.0413865949683119,
      "grad_norm": 0.002520925598219037,
      "learning_rate": 0.025057467737685468,
      "loss": 0.5456,
      "num_input_tokens_seen": 44731152,
      "step": 32535
    },
    {
      "epoch": 1.0415466359388004,
      "grad_norm": 0.003924864809960127,
      "learning_rate": 0.025024882488379557,
      "loss": 0.5607,
      "num_input_tokens_seen": 44737680,
      "step": 32540
    },
    {
      "epoch": 1.0417066769092889,
      "grad_norm": 0.00423333840444684,
      "learning_rate": 0.02499231651180727,
      "loss": 0.375,
      "num_input_tokens_seen": 44744512,
      "step": 32545
    },
    {
      "epoch": 1.0418667178797771,
      "grad_norm": 0.0031959351617842913,
      "learning_rate": 0.024959769812990713,
      "loss": 0.354,
      "num_input_tokens_seen": 44751040,
      "step": 32550
    },
    {
      "epoch": 1.0420267588502656,
      "grad_norm": 0.007611040025949478,
      "learning_rate": 0.024927242396949045,
      "loss": 0.5363,
      "num_input_tokens_seen": 44757552,
      "step": 32555
    },
    {
      "epoch": 1.0421867998207541,
      "grad_norm": 0.0096126152202487,
      "learning_rate": 0.02489473426869836,
      "loss": 0.6909,
      "num_input_tokens_seen": 44764624,
      "step": 32560
    },
    {
      "epoch": 1.0423468407912426,
      "grad_norm": 0.0038298217114061117,
      "learning_rate": 0.024862245433251776,
      "loss": 0.5389,
      "num_input_tokens_seen": 44771392,
      "step": 32565
    },
    {
      "epoch": 1.042506881761731,
      "grad_norm": 0.004508305806666613,
      "learning_rate": 0.024829775895619577,
      "loss": 0.3835,
      "num_input_tokens_seen": 44778448,
      "step": 32570
    },
    {
      "epoch": 1.0426669227322194,
      "grad_norm": 0.002813015365973115,
      "learning_rate": 0.024797325660808882,
      "loss": 0.5623,
      "num_input_tokens_seen": 44785216,
      "step": 32575
    },
    {
      "epoch": 1.0428269637027079,
      "grad_norm": 0.0024146782234311104,
      "learning_rate": 0.02476489473382401,
      "loss": 0.4445,
      "num_input_tokens_seen": 44792096,
      "step": 32580
    },
    {
      "epoch": 1.0429870046731964,
      "grad_norm": 0.007174803875386715,
      "learning_rate": 0.024732483119666127,
      "loss": 0.4378,
      "num_input_tokens_seen": 44798560,
      "step": 32585
    },
    {
      "epoch": 1.0431470456436849,
      "grad_norm": 0.008136618882417679,
      "learning_rate": 0.024700090823333548,
      "loss": 0.5444,
      "num_input_tokens_seen": 44805648,
      "step": 32590
    },
    {
      "epoch": 1.0433070866141732,
      "grad_norm": 0.0023522446863353252,
      "learning_rate": 0.02466771784982163,
      "loss": 0.4937,
      "num_input_tokens_seen": 44812800,
      "step": 32595
    },
    {
      "epoch": 1.0434671275846616,
      "grad_norm": 0.0041697658598423,
      "learning_rate": 0.024635364204122594,
      "loss": 0.3068,
      "num_input_tokens_seen": 44819808,
      "step": 32600
    },
    {
      "epoch": 1.0434671275846616,
      "eval_loss": 0.4979678690433502,
      "eval_runtime": 331.57,
      "eval_samples_per_second": 41.88,
      "eval_steps_per_second": 20.94,
      "num_input_tokens_seen": 44819808,
      "step": 32600
    },
    {
      "epoch": 1.0436271685551501,
      "grad_norm": 0.008922497741878033,
      "learning_rate": 0.024603029891225852,
      "loss": 0.6011,
      "num_input_tokens_seen": 44826864,
      "step": 32605
    },
    {
      "epoch": 1.0437872095256386,
      "grad_norm": 0.003299591364338994,
      "learning_rate": 0.024570714916117748,
      "loss": 0.4848,
      "num_input_tokens_seen": 44833456,
      "step": 32610
    },
    {
      "epoch": 1.043947250496127,
      "grad_norm": 0.0062897903844714165,
      "learning_rate": 0.024538419283781625,
      "loss": 0.5053,
      "num_input_tokens_seen": 44840096,
      "step": 32615
    },
    {
      "epoch": 1.0441072914666154,
      "grad_norm": 0.007650772109627724,
      "learning_rate": 0.024506142999197938,
      "loss": 0.6059,
      "num_input_tokens_seen": 44846576,
      "step": 32620
    },
    {
      "epoch": 1.044267332437104,
      "grad_norm": 0.0037784064188599586,
      "learning_rate": 0.024473886067344002,
      "loss": 0.5688,
      "num_input_tokens_seen": 44853232,
      "step": 32625
    },
    {
      "epoch": 1.0444273734075924,
      "grad_norm": 0.004340303596109152,
      "learning_rate": 0.02444164849319434,
      "loss": 0.3544,
      "num_input_tokens_seen": 44859824,
      "step": 32630
    },
    {
      "epoch": 1.044587414378081,
      "grad_norm": 0.0030078929848968983,
      "learning_rate": 0.024409430281720306,
      "loss": 0.3738,
      "num_input_tokens_seen": 44866416,
      "step": 32635
    },
    {
      "epoch": 1.0447474553485692,
      "grad_norm": 0.0046110330149531364,
      "learning_rate": 0.024377231437890428,
      "loss": 0.5901,
      "num_input_tokens_seen": 44872976,
      "step": 32640
    },
    {
      "epoch": 1.0449074963190577,
      "grad_norm": 0.004951375536620617,
      "learning_rate": 0.024345051966670115,
      "loss": 0.3555,
      "num_input_tokens_seen": 44880000,
      "step": 32645
    },
    {
      "epoch": 1.0450675372895462,
      "grad_norm": 0.003626316087320447,
      "learning_rate": 0.024312891873021884,
      "loss": 0.4084,
      "num_input_tokens_seen": 44887376,
      "step": 32650
    },
    {
      "epoch": 1.0452275782600347,
      "grad_norm": 0.01239386759698391,
      "learning_rate": 0.024280751161905183,
      "loss": 0.5673,
      "num_input_tokens_seen": 44894560,
      "step": 32655
    },
    {
      "epoch": 1.045387619230523,
      "grad_norm": 0.004819459281861782,
      "learning_rate": 0.02424862983827658,
      "loss": 0.5112,
      "num_input_tokens_seen": 44901792,
      "step": 32660
    },
    {
      "epoch": 1.0455476602010114,
      "grad_norm": 0.0065665775910019875,
      "learning_rate": 0.024216527907089495,
      "loss": 0.4326,
      "num_input_tokens_seen": 44909040,
      "step": 32665
    },
    {
      "epoch": 1.0457077011715,
      "grad_norm": 0.00407314021140337,
      "learning_rate": 0.024184445373294505,
      "loss": 0.6854,
      "num_input_tokens_seen": 44915920,
      "step": 32670
    },
    {
      "epoch": 1.0458677421419884,
      "grad_norm": 0.0029248299542814493,
      "learning_rate": 0.02415238224183918,
      "loss": 0.4243,
      "num_input_tokens_seen": 44922640,
      "step": 32675
    },
    {
      "epoch": 1.0460277831124767,
      "grad_norm": 0.005708354525268078,
      "learning_rate": 0.024120338517667973,
      "loss": 0.5776,
      "num_input_tokens_seen": 44929632,
      "step": 32680
    },
    {
      "epoch": 1.0461878240829652,
      "grad_norm": 0.0048048109747469425,
      "learning_rate": 0.02408831420572247,
      "loss": 0.6127,
      "num_input_tokens_seen": 44936912,
      "step": 32685
    },
    {
      "epoch": 1.0463478650534537,
      "grad_norm": 0.006995939649641514,
      "learning_rate": 0.024056309310941264,
      "loss": 0.5003,
      "num_input_tokens_seen": 44944016,
      "step": 32690
    },
    {
      "epoch": 1.0465079060239422,
      "grad_norm": 0.003114277496933937,
      "learning_rate": 0.02402432383825982,
      "loss": 0.3774,
      "num_input_tokens_seen": 44950720,
      "step": 32695
    },
    {
      "epoch": 1.0466679469944307,
      "grad_norm": 0.005048033315688372,
      "learning_rate": 0.023992357792610792,
      "loss": 0.4732,
      "num_input_tokens_seen": 44957632,
      "step": 32700
    },
    {
      "epoch": 1.046827987964919,
      "grad_norm": 0.003422880545258522,
      "learning_rate": 0.0239604111789237,
      "loss": 0.3848,
      "num_input_tokens_seen": 44964256,
      "step": 32705
    },
    {
      "epoch": 1.0469880289354074,
      "grad_norm": 0.002802106551826,
      "learning_rate": 0.023928484002125095,
      "loss": 0.5078,
      "num_input_tokens_seen": 44971376,
      "step": 32710
    },
    {
      "epoch": 1.047148069905896,
      "grad_norm": 0.00486359465867281,
      "learning_rate": 0.023896576267138595,
      "loss": 0.4559,
      "num_input_tokens_seen": 44978416,
      "step": 32715
    },
    {
      "epoch": 1.0473081108763844,
      "grad_norm": 0.01290441956371069,
      "learning_rate": 0.02386468797888471,
      "loss": 0.5109,
      "num_input_tokens_seen": 44985696,
      "step": 32720
    },
    {
      "epoch": 1.0474681518468727,
      "grad_norm": 0.007552550174295902,
      "learning_rate": 0.023832819142281057,
      "loss": 0.5641,
      "num_input_tokens_seen": 44993184,
      "step": 32725
    },
    {
      "epoch": 1.0476281928173612,
      "grad_norm": 0.00449723144993186,
      "learning_rate": 0.02380096976224225,
      "loss": 0.4233,
      "num_input_tokens_seen": 45000256,
      "step": 32730
    },
    {
      "epoch": 1.0477882337878497,
      "grad_norm": 0.005673198029398918,
      "learning_rate": 0.023769139843679777,
      "loss": 0.5573,
      "num_input_tokens_seen": 45007104,
      "step": 32735
    },
    {
      "epoch": 1.0479482747583382,
      "grad_norm": 0.003978431690484285,
      "learning_rate": 0.023737329391502287,
      "loss": 0.5295,
      "num_input_tokens_seen": 45014368,
      "step": 32740
    },
    {
      "epoch": 1.0481083157288267,
      "grad_norm": 0.004400825593620539,
      "learning_rate": 0.023705538410615293,
      "loss": 0.4977,
      "num_input_tokens_seen": 45021872,
      "step": 32745
    },
    {
      "epoch": 1.048268356699315,
      "grad_norm": 0.004951959941536188,
      "learning_rate": 0.023673766905921396,
      "loss": 0.5605,
      "num_input_tokens_seen": 45028688,
      "step": 32750
    },
    {
      "epoch": 1.0484283976698034,
      "grad_norm": 0.0033799863886088133,
      "learning_rate": 0.0236420148823202,
      "loss": 0.3427,
      "num_input_tokens_seen": 45036032,
      "step": 32755
    },
    {
      "epoch": 1.048588438640292,
      "grad_norm": 0.004381610546261072,
      "learning_rate": 0.02361028234470816,
      "loss": 0.615,
      "num_input_tokens_seen": 45043120,
      "step": 32760
    },
    {
      "epoch": 1.0487484796107804,
      "grad_norm": 0.004634813871234655,
      "learning_rate": 0.023578569297978913,
      "loss": 0.4204,
      "num_input_tokens_seen": 45050240,
      "step": 32765
    },
    {
      "epoch": 1.0489085205812687,
      "grad_norm": 0.002790505765005946,
      "learning_rate": 0.023546875747023025,
      "loss": 0.422,
      "num_input_tokens_seen": 45057232,
      "step": 32770
    },
    {
      "epoch": 1.0490685615517572,
      "grad_norm": 0.00536394352093339,
      "learning_rate": 0.02351520169672801,
      "loss": 0.5072,
      "num_input_tokens_seen": 45064432,
      "step": 32775
    },
    {
      "epoch": 1.0492286025222457,
      "grad_norm": 0.004911845549941063,
      "learning_rate": 0.023483547151978357,
      "loss": 0.4235,
      "num_input_tokens_seen": 45071280,
      "step": 32780
    },
    {
      "epoch": 1.0493886434927342,
      "grad_norm": 0.006428249180316925,
      "learning_rate": 0.023451912117655675,
      "loss": 0.6126,
      "num_input_tokens_seen": 45078016,
      "step": 32785
    },
    {
      "epoch": 1.0495486844632227,
      "grad_norm": 0.005791246425360441,
      "learning_rate": 0.023420296598638417,
      "loss": 0.4274,
      "num_input_tokens_seen": 45084768,
      "step": 32790
    },
    {
      "epoch": 1.049708725433711,
      "grad_norm": 0.003742637811228633,
      "learning_rate": 0.023388700599802165,
      "loss": 0.4698,
      "num_input_tokens_seen": 45091152,
      "step": 32795
    },
    {
      "epoch": 1.0498687664041995,
      "grad_norm": 0.004487817641347647,
      "learning_rate": 0.023357124126019334,
      "loss": 0.4889,
      "num_input_tokens_seen": 45097904,
      "step": 32800
    },
    {
      "epoch": 1.0498687664041995,
      "eval_loss": 0.4971758723258972,
      "eval_runtime": 332.1703,
      "eval_samples_per_second": 41.804,
      "eval_steps_per_second": 20.902,
      "num_input_tokens_seen": 45097904,
      "step": 32800
    },
    {
      "epoch": 1.050028807374688,
      "grad_norm": 0.0037069853860884905,
      "learning_rate": 0.02332556718215945,
      "loss": 0.406,
      "num_input_tokens_seen": 45104928,
      "step": 32805
    },
    {
      "epoch": 1.0501888483451765,
      "grad_norm": 0.008784751407802105,
      "learning_rate": 0.023294029773089035,
      "loss": 0.413,
      "num_input_tokens_seen": 45111968,
      "step": 32810
    },
    {
      "epoch": 1.0503488893156647,
      "grad_norm": 0.004587032366544008,
      "learning_rate": 0.023262511903671484,
      "loss": 0.3708,
      "num_input_tokens_seen": 45118976,
      "step": 32815
    },
    {
      "epoch": 1.0505089302861532,
      "grad_norm": 0.004732415080070496,
      "learning_rate": 0.023231013578767324,
      "loss": 0.552,
      "num_input_tokens_seen": 45125760,
      "step": 32820
    },
    {
      "epoch": 1.0506689712566417,
      "grad_norm": 0.0036722642835229635,
      "learning_rate": 0.0231995348032339,
      "loss": 0.5094,
      "num_input_tokens_seen": 45132912,
      "step": 32825
    },
    {
      "epoch": 1.0508290122271302,
      "grad_norm": 0.00515986792743206,
      "learning_rate": 0.023168075581925685,
      "loss": 0.4263,
      "num_input_tokens_seen": 45139488,
      "step": 32830
    },
    {
      "epoch": 1.0509890531976185,
      "grad_norm": 0.005274907685816288,
      "learning_rate": 0.023136635919694126,
      "loss": 0.4778,
      "num_input_tokens_seen": 45146624,
      "step": 32835
    },
    {
      "epoch": 1.051149094168107,
      "grad_norm": 0.00964042916893959,
      "learning_rate": 0.02310521582138753,
      "loss": 0.4907,
      "num_input_tokens_seen": 45153760,
      "step": 32840
    },
    {
      "epoch": 1.0513091351385955,
      "grad_norm": 0.007004341576248407,
      "learning_rate": 0.023073815291851357,
      "loss": 0.6584,
      "num_input_tokens_seen": 45161232,
      "step": 32845
    },
    {
      "epoch": 1.051469176109084,
      "grad_norm": 0.004204645752906799,
      "learning_rate": 0.02304243433592788,
      "loss": 0.5451,
      "num_input_tokens_seen": 45168560,
      "step": 32850
    },
    {
      "epoch": 1.0516292170795725,
      "grad_norm": 0.003065898083150387,
      "learning_rate": 0.023011072958456513,
      "loss": 0.538,
      "num_input_tokens_seen": 45175360,
      "step": 32855
    },
    {
      "epoch": 1.0517892580500607,
      "grad_norm": 0.005645787809044123,
      "learning_rate": 0.022979731164273536,
      "loss": 0.4002,
      "num_input_tokens_seen": 45182320,
      "step": 32860
    },
    {
      "epoch": 1.0519492990205492,
      "grad_norm": 0.008214747533202171,
      "learning_rate": 0.022948408958212218,
      "loss": 0.5264,
      "num_input_tokens_seen": 45189312,
      "step": 32865
    },
    {
      "epoch": 1.0521093399910377,
      "grad_norm": 0.004605886060744524,
      "learning_rate": 0.022917106345102876,
      "loss": 0.5118,
      "num_input_tokens_seen": 45196304,
      "step": 32870
    },
    {
      "epoch": 1.0522693809615262,
      "grad_norm": 0.008417666889727116,
      "learning_rate": 0.022885823329772785,
      "loss": 0.547,
      "num_input_tokens_seen": 45203040,
      "step": 32875
    },
    {
      "epoch": 1.0524294219320145,
      "grad_norm": 0.007128950674086809,
      "learning_rate": 0.02285455991704612,
      "loss": 0.564,
      "num_input_tokens_seen": 45210128,
      "step": 32880
    },
    {
      "epoch": 1.052589462902503,
      "grad_norm": 0.008322256617248058,
      "learning_rate": 0.022823316111744117,
      "loss": 0.2635,
      "num_input_tokens_seen": 45217584,
      "step": 32885
    },
    {
      "epoch": 1.0527495038729915,
      "grad_norm": 0.0032905717380344868,
      "learning_rate": 0.022792091918685014,
      "loss": 0.3681,
      "num_input_tokens_seen": 45223872,
      "step": 32890
    },
    {
      "epoch": 1.05290954484348,
      "grad_norm": 0.004741070792078972,
      "learning_rate": 0.022760887342683906,
      "loss": 0.4254,
      "num_input_tokens_seen": 45230496,
      "step": 32895
    },
    {
      "epoch": 1.0530695858139685,
      "grad_norm": 0.004445821978151798,
      "learning_rate": 0.022729702388552975,
      "loss": 0.3996,
      "num_input_tokens_seen": 45237552,
      "step": 32900
    },
    {
      "epoch": 1.0532296267844568,
      "grad_norm": 0.0047763739712536335,
      "learning_rate": 0.022698537061101292,
      "loss": 0.3649,
      "num_input_tokens_seen": 45244512,
      "step": 32905
    },
    {
      "epoch": 1.0533896677549452,
      "grad_norm": 0.0034289557952433825,
      "learning_rate": 0.022667391365134962,
      "loss": 0.5235,
      "num_input_tokens_seen": 45251440,
      "step": 32910
    },
    {
      "epoch": 1.0535497087254337,
      "grad_norm": 0.008962711319327354,
      "learning_rate": 0.022636265305457065,
      "loss": 0.5782,
      "num_input_tokens_seen": 45258048,
      "step": 32915
    },
    {
      "epoch": 1.0537097496959222,
      "grad_norm": 0.00484116654843092,
      "learning_rate": 0.02260515888686764,
      "loss": 0.6658,
      "num_input_tokens_seen": 45264944,
      "step": 32920
    },
    {
      "epoch": 1.0538697906664105,
      "grad_norm": 0.0032823793590068817,
      "learning_rate": 0.022574072114163596,
      "loss": 0.5873,
      "num_input_tokens_seen": 45271760,
      "step": 32925
    },
    {
      "epoch": 1.054029831636899,
      "grad_norm": 0.004417682532221079,
      "learning_rate": 0.022543004992139005,
      "loss": 0.4205,
      "num_input_tokens_seen": 45278656,
      "step": 32930
    },
    {
      "epoch": 1.0541898726073875,
      "grad_norm": 0.004466356709599495,
      "learning_rate": 0.022511957525584745,
      "loss": 0.3852,
      "num_input_tokens_seen": 45285312,
      "step": 32935
    },
    {
      "epoch": 1.054349913577876,
      "grad_norm": 0.0022135567851364613,
      "learning_rate": 0.022480929719288778,
      "loss": 0.6433,
      "num_input_tokens_seen": 45292592,
      "step": 32940
    },
    {
      "epoch": 1.0545099545483643,
      "grad_norm": 0.003927991725504398,
      "learning_rate": 0.02244992157803592,
      "loss": 0.4598,
      "num_input_tokens_seen": 45299392,
      "step": 32945
    },
    {
      "epoch": 1.0546699955188528,
      "grad_norm": 0.0058533321134746075,
      "learning_rate": 0.022418933106608047,
      "loss": 0.4779,
      "num_input_tokens_seen": 45306640,
      "step": 32950
    },
    {
      "epoch": 1.0548300364893413,
      "grad_norm": 0.003609732259064913,
      "learning_rate": 0.022387964309784018,
      "loss": 0.6254,
      "num_input_tokens_seen": 45313888,
      "step": 32955
    },
    {
      "epoch": 1.0549900774598298,
      "grad_norm": 0.009126076474785805,
      "learning_rate": 0.022357015192339517,
      "loss": 0.4546,
      "num_input_tokens_seen": 45321136,
      "step": 32960
    },
    {
      "epoch": 1.0551501184303183,
      "grad_norm": 0.004731332417577505,
      "learning_rate": 0.02232608575904734,
      "loss": 0.531,
      "num_input_tokens_seen": 45327488,
      "step": 32965
    },
    {
      "epoch": 1.0553101594008065,
      "grad_norm": 0.004290097393095493,
      "learning_rate": 0.022295176014677225,
      "loss": 0.4324,
      "num_input_tokens_seen": 45334704,
      "step": 32970
    },
    {
      "epoch": 1.055470200371295,
      "grad_norm": 0.004569982644170523,
      "learning_rate": 0.02226428596399577,
      "loss": 0.4307,
      "num_input_tokens_seen": 45341328,
      "step": 32975
    },
    {
      "epoch": 1.0556302413417835,
      "grad_norm": 0.005070853978395462,
      "learning_rate": 0.02223341561176669,
      "loss": 0.4596,
      "num_input_tokens_seen": 45348016,
      "step": 32980
    },
    {
      "epoch": 1.055790282312272,
      "grad_norm": 0.00425802543759346,
      "learning_rate": 0.0222025649627505,
      "loss": 0.6074,
      "num_input_tokens_seen": 45355952,
      "step": 32985
    },
    {
      "epoch": 1.0559503232827603,
      "grad_norm": 0.003075108164921403,
      "learning_rate": 0.022171734021704814,
      "loss": 0.3816,
      "num_input_tokens_seen": 45362672,
      "step": 32990
    },
    {
      "epoch": 1.0561103642532488,
      "grad_norm": 0.0030916177202016115,
      "learning_rate": 0.022140922793384116,
      "loss": 0.409,
      "num_input_tokens_seen": 45369472,
      "step": 32995
    },
    {
      "epoch": 1.0562704052237373,
      "grad_norm": 0.00390783092007041,
      "learning_rate": 0.022110131282539934,
      "loss": 0.4359,
      "num_input_tokens_seen": 45376272,
      "step": 33000
    },
    {
      "epoch": 1.0562704052237373,
      "eval_loss": 0.49768003821372986,
      "eval_runtime": 332.4429,
      "eval_samples_per_second": 41.77,
      "eval_steps_per_second": 20.885,
      "num_input_tokens_seen": 45376272,
      "step": 33000
    },
    {
      "epoch": 1.0564304461942258,
      "grad_norm": 0.0034898302983492613,
      "learning_rate": 0.022079359493920675,
      "loss": 0.557,
      "num_input_tokens_seen": 45382896,
      "step": 33005
    },
    {
      "epoch": 1.0565904871647143,
      "grad_norm": 0.0023998674005270004,
      "learning_rate": 0.02204860743227169,
      "loss": 0.3931,
      "num_input_tokens_seen": 45389680,
      "step": 33010
    },
    {
      "epoch": 1.0567505281352025,
      "grad_norm": 0.007116345688700676,
      "learning_rate": 0.022017875102335365,
      "loss": 0.7375,
      "num_input_tokens_seen": 45396816,
      "step": 33015
    },
    {
      "epoch": 1.056910569105691,
      "grad_norm": 0.004805984441190958,
      "learning_rate": 0.02198716250885108,
      "loss": 0.4242,
      "num_input_tokens_seen": 45403552,
      "step": 33020
    },
    {
      "epoch": 1.0570706100761795,
      "grad_norm": 0.004746474325656891,
      "learning_rate": 0.021956469656555,
      "loss": 0.5183,
      "num_input_tokens_seen": 45410784,
      "step": 33025
    },
    {
      "epoch": 1.057230651046668,
      "grad_norm": 0.004300160799175501,
      "learning_rate": 0.0219257965501804,
      "loss": 0.5212,
      "num_input_tokens_seen": 45417264,
      "step": 33030
    },
    {
      "epoch": 1.0573906920171563,
      "grad_norm": 0.004186778794974089,
      "learning_rate": 0.021895143194457494,
      "loss": 0.3623,
      "num_input_tokens_seen": 45424000,
      "step": 33035
    },
    {
      "epoch": 1.0575507329876448,
      "grad_norm": 0.005688187666237354,
      "learning_rate": 0.021864509594113322,
      "loss": 0.5366,
      "num_input_tokens_seen": 45430752,
      "step": 33040
    },
    {
      "epoch": 1.0577107739581333,
      "grad_norm": 0.004143199883401394,
      "learning_rate": 0.02183389575387207,
      "loss": 0.5236,
      "num_input_tokens_seen": 45437296,
      "step": 33045
    },
    {
      "epoch": 1.0578708149286218,
      "grad_norm": 0.002833182457834482,
      "learning_rate": 0.021803301678454682,
      "loss": 0.486,
      "num_input_tokens_seen": 45443872,
      "step": 33050
    },
    {
      "epoch": 1.0580308558991103,
      "grad_norm": 0.0037801717408001423,
      "learning_rate": 0.021772727372579213,
      "loss": 0.431,
      "num_input_tokens_seen": 45450544,
      "step": 33055
    },
    {
      "epoch": 1.0581908968695986,
      "grad_norm": 0.0031304042786359787,
      "learning_rate": 0.02174217284096061,
      "loss": 0.2612,
      "num_input_tokens_seen": 45457504,
      "step": 33060
    },
    {
      "epoch": 1.058350937840087,
      "grad_norm": 0.00417400011792779,
      "learning_rate": 0.0217116380883107,
      "loss": 0.5713,
      "num_input_tokens_seen": 45464320,
      "step": 33065
    },
    {
      "epoch": 1.0585109788105755,
      "grad_norm": 0.005163256544619799,
      "learning_rate": 0.021681123119338425,
      "loss": 0.6851,
      "num_input_tokens_seen": 45470816,
      "step": 33070
    },
    {
      "epoch": 1.058671019781064,
      "grad_norm": 0.00485487375408411,
      "learning_rate": 0.02165062793874951,
      "loss": 0.5669,
      "num_input_tokens_seen": 45477456,
      "step": 33075
    },
    {
      "epoch": 1.0588310607515523,
      "grad_norm": 0.004850985016673803,
      "learning_rate": 0.021620152551246666,
      "loss": 0.5123,
      "num_input_tokens_seen": 45484272,
      "step": 33080
    },
    {
      "epoch": 1.0589911017220408,
      "grad_norm": 0.006177626550197601,
      "learning_rate": 0.02158969696152967,
      "loss": 0.6148,
      "num_input_tokens_seen": 45490752,
      "step": 33085
    },
    {
      "epoch": 1.0591511426925293,
      "grad_norm": 0.0032893300522118807,
      "learning_rate": 0.021559261174295057,
      "loss": 0.3723,
      "num_input_tokens_seen": 45497584,
      "step": 33090
    },
    {
      "epoch": 1.0593111836630178,
      "grad_norm": 0.0037275126669555902,
      "learning_rate": 0.02152884519423646,
      "loss": 0.4333,
      "num_input_tokens_seen": 45504416,
      "step": 33095
    },
    {
      "epoch": 1.059471224633506,
      "grad_norm": 0.002435553353279829,
      "learning_rate": 0.021498449026044447,
      "loss": 0.4582,
      "num_input_tokens_seen": 45511312,
      "step": 33100
    },
    {
      "epoch": 1.0596312656039946,
      "grad_norm": 0.004747869446873665,
      "learning_rate": 0.021468072674406414,
      "loss": 0.4805,
      "num_input_tokens_seen": 45517840,
      "step": 33105
    },
    {
      "epoch": 1.059791306574483,
      "grad_norm": 0.004522189497947693,
      "learning_rate": 0.021437716144006795,
      "loss": 0.5019,
      "num_input_tokens_seen": 45524944,
      "step": 33110
    },
    {
      "epoch": 1.0599513475449716,
      "grad_norm": 0.006973085459321737,
      "learning_rate": 0.021407379439527002,
      "loss": 0.5057,
      "num_input_tokens_seen": 45532336,
      "step": 33115
    },
    {
      "epoch": 1.06011138851546,
      "grad_norm": 0.008768550120294094,
      "learning_rate": 0.021377062565645255,
      "loss": 0.3548,
      "num_input_tokens_seen": 45539056,
      "step": 33120
    },
    {
      "epoch": 1.0602714294859483,
      "grad_norm": 0.0056605106219649315,
      "learning_rate": 0.02134676552703688,
      "loss": 0.4835,
      "num_input_tokens_seen": 45545872,
      "step": 33125
    },
    {
      "epoch": 1.0604314704564368,
      "grad_norm": 0.0029320698231458664,
      "learning_rate": 0.02131648832837398,
      "loss": 0.4398,
      "num_input_tokens_seen": 45552432,
      "step": 33130
    },
    {
      "epoch": 1.0605915114269253,
      "grad_norm": 0.003065098775550723,
      "learning_rate": 0.02128623097432574,
      "loss": 0.5497,
      "num_input_tokens_seen": 45559072,
      "step": 33135
    },
    {
      "epoch": 1.0607515523974138,
      "grad_norm": 0.005898956209421158,
      "learning_rate": 0.021255993469558192,
      "loss": 0.6699,
      "num_input_tokens_seen": 45566896,
      "step": 33140
    },
    {
      "epoch": 1.060911593367902,
      "grad_norm": 0.003052107058465481,
      "learning_rate": 0.021225775818734364,
      "loss": 0.4044,
      "num_input_tokens_seen": 45573664,
      "step": 33145
    },
    {
      "epoch": 1.0610716343383906,
      "grad_norm": 0.010270487517118454,
      "learning_rate": 0.021195578026514166,
      "loss": 0.6961,
      "num_input_tokens_seen": 45580128,
      "step": 33150
    },
    {
      "epoch": 1.061231675308879,
      "grad_norm": 0.005469068884849548,
      "learning_rate": 0.02116540009755452,
      "loss": 0.552,
      "num_input_tokens_seen": 45587136,
      "step": 33155
    },
    {
      "epoch": 1.0613917162793676,
      "grad_norm": 0.005591580644249916,
      "learning_rate": 0.021135242036509173,
      "loss": 0.46,
      "num_input_tokens_seen": 45594016,
      "step": 33160
    },
    {
      "epoch": 1.061551757249856,
      "grad_norm": 0.002982624340802431,
      "learning_rate": 0.021105103848028967,
      "loss": 0.4915,
      "num_input_tokens_seen": 45600832,
      "step": 33165
    },
    {
      "epoch": 1.0617117982203443,
      "grad_norm": 0.004548271186649799,
      "learning_rate": 0.021074985536761504,
      "loss": 0.461,
      "num_input_tokens_seen": 45607808,
      "step": 33170
    },
    {
      "epoch": 1.0618718391908328,
      "grad_norm": 0.0032681175507605076,
      "learning_rate": 0.021044887107351435,
      "loss": 0.4393,
      "num_input_tokens_seen": 45614688,
      "step": 33175
    },
    {
      "epoch": 1.0620318801613213,
      "grad_norm": 0.004026779439300299,
      "learning_rate": 0.021014808564440362,
      "loss": 0.5323,
      "num_input_tokens_seen": 45621632,
      "step": 33180
    },
    {
      "epoch": 1.0621919211318098,
      "grad_norm": 0.005754550918936729,
      "learning_rate": 0.02098474991266671,
      "loss": 0.6985,
      "num_input_tokens_seen": 45628288,
      "step": 33185
    },
    {
      "epoch": 1.062351962102298,
      "grad_norm": 0.0055718510411679745,
      "learning_rate": 0.02095471115666592,
      "loss": 0.5666,
      "num_input_tokens_seen": 45634816,
      "step": 33190
    },
    {
      "epoch": 1.0625120030727866,
      "grad_norm": 0.005915109068155289,
      "learning_rate": 0.020924692301070406,
      "loss": 0.5269,
      "num_input_tokens_seen": 45641472,
      "step": 33195
    },
    {
      "epoch": 1.062672044043275,
      "grad_norm": 0.006091408431529999,
      "learning_rate": 0.020894693350509346,
      "loss": 0.4718,
      "num_input_tokens_seen": 45647824,
      "step": 33200
    },
    {
      "epoch": 1.062672044043275,
      "eval_loss": 0.4974338412284851,
      "eval_runtime": 331.5284,
      "eval_samples_per_second": 41.885,
      "eval_steps_per_second": 20.942,
      "num_input_tokens_seen": 45647824,
      "step": 33200
    },
    {
      "epoch": 1.0628320850137636,
      "grad_norm": 0.0033159134909510612,
      "learning_rate": 0.020864714309609057,
      "loss": 0.4192,
      "num_input_tokens_seen": 45654768,
      "step": 33205
    },
    {
      "epoch": 1.0629921259842519,
      "grad_norm": 0.006033208686858416,
      "learning_rate": 0.020834755182992604,
      "loss": 0.5231,
      "num_input_tokens_seen": 45661536,
      "step": 33210
    },
    {
      "epoch": 1.0631521669547404,
      "grad_norm": 0.0014314077561721206,
      "learning_rate": 0.02080481597528011,
      "loss": 0.4025,
      "num_input_tokens_seen": 45668128,
      "step": 33215
    },
    {
      "epoch": 1.0633122079252288,
      "grad_norm": 0.0026981360279023647,
      "learning_rate": 0.020774896691088583,
      "loss": 0.3263,
      "num_input_tokens_seen": 45675040,
      "step": 33220
    },
    {
      "epoch": 1.0634722488957173,
      "grad_norm": 0.004012465476989746,
      "learning_rate": 0.020744997335031882,
      "loss": 0.4964,
      "num_input_tokens_seen": 45681856,
      "step": 33225
    },
    {
      "epoch": 1.0636322898662058,
      "grad_norm": 0.004410907160490751,
      "learning_rate": 0.02071511791172092,
      "loss": 0.3658,
      "num_input_tokens_seen": 45688704,
      "step": 33230
    },
    {
      "epoch": 1.0637923308366941,
      "grad_norm": 0.005062684416770935,
      "learning_rate": 0.02068525842576351,
      "loss": 0.5592,
      "num_input_tokens_seen": 45695408,
      "step": 33235
    },
    {
      "epoch": 1.0639523718071826,
      "grad_norm": 0.005180952604860067,
      "learning_rate": 0.020655418881764264,
      "loss": 0.5268,
      "num_input_tokens_seen": 45701984,
      "step": 33240
    },
    {
      "epoch": 1.064112412777671,
      "grad_norm": 0.0021193597931414843,
      "learning_rate": 0.020625599284324923,
      "loss": 0.3586,
      "num_input_tokens_seen": 45708944,
      "step": 33245
    },
    {
      "epoch": 1.0642724537481596,
      "grad_norm": 0.003479150589555502,
      "learning_rate": 0.02059579963804396,
      "loss": 0.3472,
      "num_input_tokens_seen": 45715824,
      "step": 33250
    },
    {
      "epoch": 1.0644324947186479,
      "grad_norm": 0.002264501992613077,
      "learning_rate": 0.02056601994751688,
      "loss": 0.4172,
      "num_input_tokens_seen": 45722544,
      "step": 33255
    },
    {
      "epoch": 1.0645925356891364,
      "grad_norm": 0.00683425972238183,
      "learning_rate": 0.02053626021733614,
      "loss": 0.4995,
      "num_input_tokens_seen": 45729792,
      "step": 33260
    },
    {
      "epoch": 1.0647525766596249,
      "grad_norm": 0.005956000182777643,
      "learning_rate": 0.02050652045209097,
      "loss": 0.4415,
      "num_input_tokens_seen": 45736608,
      "step": 33265
    },
    {
      "epoch": 1.0649126176301134,
      "grad_norm": 0.006788982544094324,
      "learning_rate": 0.020476800656367672,
      "loss": 0.3293,
      "num_input_tokens_seen": 45743600,
      "step": 33270
    },
    {
      "epoch": 1.0650726586006019,
      "grad_norm": 0.0051768627017736435,
      "learning_rate": 0.020447100834749425,
      "loss": 0.5982,
      "num_input_tokens_seen": 45750544,
      "step": 33275
    },
    {
      "epoch": 1.0652326995710901,
      "grad_norm": 0.003973050508648157,
      "learning_rate": 0.02041742099181627,
      "loss": 0.574,
      "num_input_tokens_seen": 45757664,
      "step": 33280
    },
    {
      "epoch": 1.0653927405415786,
      "grad_norm": 0.004069929476827383,
      "learning_rate": 0.02038776113214526,
      "loss": 0.4569,
      "num_input_tokens_seen": 45764576,
      "step": 33285
    },
    {
      "epoch": 1.0655527815120671,
      "grad_norm": 0.00562879117205739,
      "learning_rate": 0.0203581212603103,
      "loss": 0.5148,
      "num_input_tokens_seen": 45771408,
      "step": 33290
    },
    {
      "epoch": 1.0657128224825556,
      "grad_norm": 0.00324124563485384,
      "learning_rate": 0.02032850138088219,
      "loss": 0.4149,
      "num_input_tokens_seen": 45778608,
      "step": 33295
    },
    {
      "epoch": 1.0658728634530439,
      "grad_norm": 0.009699140675365925,
      "learning_rate": 0.020298901498428754,
      "loss": 0.5183,
      "num_input_tokens_seen": 45785376,
      "step": 33300
    },
    {
      "epoch": 1.0660329044235324,
      "grad_norm": 0.0028393473476171494,
      "learning_rate": 0.020269321617514595,
      "loss": 0.3567,
      "num_input_tokens_seen": 45791984,
      "step": 33305
    },
    {
      "epoch": 1.0661929453940209,
      "grad_norm": 0.007172558456659317,
      "learning_rate": 0.020239761742701343,
      "loss": 0.3944,
      "num_input_tokens_seen": 45798400,
      "step": 33310
    },
    {
      "epoch": 1.0663529863645094,
      "grad_norm": 0.0029326549265533686,
      "learning_rate": 0.02021022187854754,
      "loss": 0.5411,
      "num_input_tokens_seen": 45805136,
      "step": 33315
    },
    {
      "epoch": 1.0665130273349979,
      "grad_norm": 0.003911473788321018,
      "learning_rate": 0.020180702029608522,
      "loss": 0.5081,
      "num_input_tokens_seen": 45812368,
      "step": 33320
    },
    {
      "epoch": 1.0666730683054861,
      "grad_norm": 0.004182455595582724,
      "learning_rate": 0.020151202200436695,
      "loss": 0.4082,
      "num_input_tokens_seen": 45819392,
      "step": 33325
    },
    {
      "epoch": 1.0668331092759746,
      "grad_norm": 0.004299905151128769,
      "learning_rate": 0.020121722395581226,
      "loss": 0.5981,
      "num_input_tokens_seen": 45826192,
      "step": 33330
    },
    {
      "epoch": 1.0669931502464631,
      "grad_norm": 0.005276472307741642,
      "learning_rate": 0.020092262619588342,
      "loss": 0.5758,
      "num_input_tokens_seen": 45833120,
      "step": 33335
    },
    {
      "epoch": 1.0671531912169516,
      "grad_norm": 0.008574914187192917,
      "learning_rate": 0.02006282287700109,
      "loss": 0.3358,
      "num_input_tokens_seen": 45840208,
      "step": 33340
    },
    {
      "epoch": 1.06731323218744,
      "grad_norm": 0.004820812493562698,
      "learning_rate": 0.020033403172359427,
      "loss": 0.3629,
      "num_input_tokens_seen": 45846992,
      "step": 33345
    },
    {
      "epoch": 1.0674732731579284,
      "grad_norm": 0.00424268888309598,
      "learning_rate": 0.020004003510200284,
      "loss": 0.4492,
      "num_input_tokens_seen": 45853840,
      "step": 33350
    },
    {
      "epoch": 1.0676333141284169,
      "grad_norm": 0.0053993891924619675,
      "learning_rate": 0.019974623895057407,
      "loss": 0.6587,
      "num_input_tokens_seen": 45860672,
      "step": 33355
    },
    {
      "epoch": 1.0677933550989054,
      "grad_norm": 0.006989458575844765,
      "learning_rate": 0.019945264331461553,
      "loss": 0.5809,
      "num_input_tokens_seen": 45867712,
      "step": 33360
    },
    {
      "epoch": 1.0679533960693939,
      "grad_norm": 0.004114097915589809,
      "learning_rate": 0.019915924823940317,
      "loss": 0.6222,
      "num_input_tokens_seen": 45874656,
      "step": 33365
    },
    {
      "epoch": 1.0681134370398822,
      "grad_norm": 0.0031221904791891575,
      "learning_rate": 0.01988660537701816,
      "loss": 0.3863,
      "num_input_tokens_seen": 45881232,
      "step": 33370
    },
    {
      "epoch": 1.0682734780103706,
      "grad_norm": 0.00527920899912715,
      "learning_rate": 0.01985730599521659,
      "loss": 0.4545,
      "num_input_tokens_seen": 45888160,
      "step": 33375
    },
    {
      "epoch": 1.0684335189808591,
      "grad_norm": 0.0031009239610284567,
      "learning_rate": 0.019828026683053918,
      "loss": 0.3045,
      "num_input_tokens_seen": 45894960,
      "step": 33380
    },
    {
      "epoch": 1.0685935599513476,
      "grad_norm": 0.0025256844237446785,
      "learning_rate": 0.01979876744504535,
      "loss": 0.6263,
      "num_input_tokens_seen": 45901696,
      "step": 33385
    },
    {
      "epoch": 1.068753600921836,
      "grad_norm": 0.007306511979550123,
      "learning_rate": 0.019769528285703046,
      "loss": 0.5343,
      "num_input_tokens_seen": 45908176,
      "step": 33390
    },
    {
      "epoch": 1.0689136418923244,
      "grad_norm": 0.004047273192554712,
      "learning_rate": 0.019740309209536098,
      "loss": 0.5204,
      "num_input_tokens_seen": 45915136,
      "step": 33395
    },
    {
      "epoch": 1.069073682862813,
      "grad_norm": 0.006713482551276684,
      "learning_rate": 0.019711110221050387,
      "loss": 0.536,
      "num_input_tokens_seen": 45922032,
      "step": 33400
    },
    {
      "epoch": 1.069073682862813,
      "eval_loss": 0.49789419770240784,
      "eval_runtime": 331.7716,
      "eval_samples_per_second": 41.854,
      "eval_steps_per_second": 20.927,
      "num_input_tokens_seen": 45922032,
      "step": 33400
    },
    {
      "epoch": 1.0692337238333014,
      "grad_norm": 0.003820456098765135,
      "learning_rate": 0.019681931324748825,
      "loss": 0.4159,
      "num_input_tokens_seen": 45928528,
      "step": 33405
    },
    {
      "epoch": 1.0693937648037897,
      "grad_norm": 0.004163458943367004,
      "learning_rate": 0.019652772525131094,
      "loss": 0.4462,
      "num_input_tokens_seen": 45935488,
      "step": 33410
    },
    {
      "epoch": 1.0695538057742782,
      "grad_norm": 0.005181184504181147,
      "learning_rate": 0.019623633826693885,
      "loss": 0.6216,
      "num_input_tokens_seen": 45942512,
      "step": 33415
    },
    {
      "epoch": 1.0697138467447667,
      "grad_norm": 0.006020816974341869,
      "learning_rate": 0.019594515233930788,
      "loss": 0.5505,
      "num_input_tokens_seen": 45949424,
      "step": 33420
    },
    {
      "epoch": 1.0698738877152552,
      "grad_norm": 0.005609878338873386,
      "learning_rate": 0.019565416751332186,
      "loss": 0.4257,
      "num_input_tokens_seen": 45956448,
      "step": 33425
    },
    {
      "epoch": 1.0700339286857437,
      "grad_norm": 0.0034667295403778553,
      "learning_rate": 0.019536338383385497,
      "loss": 0.3778,
      "num_input_tokens_seen": 45964192,
      "step": 33430
    },
    {
      "epoch": 1.070193969656232,
      "grad_norm": 0.004945521242916584,
      "learning_rate": 0.019507280134574933,
      "loss": 0.4396,
      "num_input_tokens_seen": 45971056,
      "step": 33435
    },
    {
      "epoch": 1.0703540106267204,
      "grad_norm": 0.006562894210219383,
      "learning_rate": 0.019478242009381624,
      "loss": 0.5571,
      "num_input_tokens_seen": 45978304,
      "step": 33440
    },
    {
      "epoch": 1.070514051597209,
      "grad_norm": 0.00484421756118536,
      "learning_rate": 0.01944922401228367,
      "loss": 0.5195,
      "num_input_tokens_seen": 45985488,
      "step": 33445
    },
    {
      "epoch": 1.0706740925676974,
      "grad_norm": 0.0037500483449548483,
      "learning_rate": 0.01942022614775593,
      "loss": 0.3518,
      "num_input_tokens_seen": 45992336,
      "step": 33450
    },
    {
      "epoch": 1.0708341335381857,
      "grad_norm": 0.006436591036617756,
      "learning_rate": 0.01939124842027029,
      "loss": 0.6032,
      "num_input_tokens_seen": 45999040,
      "step": 33455
    },
    {
      "epoch": 1.0709941745086742,
      "grad_norm": 0.006869049742817879,
      "learning_rate": 0.01936229083429551,
      "loss": 0.5851,
      "num_input_tokens_seen": 46006256,
      "step": 33460
    },
    {
      "epoch": 1.0711542154791627,
      "grad_norm": 0.0034311204217374325,
      "learning_rate": 0.019333353394297148,
      "loss": 0.6348,
      "num_input_tokens_seen": 46012672,
      "step": 33465
    },
    {
      "epoch": 1.0713142564496512,
      "grad_norm": 0.004933509044349194,
      "learning_rate": 0.019304436104737754,
      "loss": 0.3541,
      "num_input_tokens_seen": 46020096,
      "step": 33470
    },
    {
      "epoch": 1.0714742974201394,
      "grad_norm": 0.005495934747159481,
      "learning_rate": 0.019275538970076778,
      "loss": 0.3636,
      "num_input_tokens_seen": 46026784,
      "step": 33475
    },
    {
      "epoch": 1.071634338390628,
      "grad_norm": 0.0038837087340652943,
      "learning_rate": 0.019246661994770434,
      "loss": 0.5056,
      "num_input_tokens_seen": 46033904,
      "step": 33480
    },
    {
      "epoch": 1.0717943793611164,
      "grad_norm": 0.003964665345847607,
      "learning_rate": 0.019217805183271985,
      "loss": 0.4563,
      "num_input_tokens_seen": 46040928,
      "step": 33485
    },
    {
      "epoch": 1.071954420331605,
      "grad_norm": 0.009059738367795944,
      "learning_rate": 0.019188968540031465,
      "loss": 0.6128,
      "num_input_tokens_seen": 46048336,
      "step": 33490
    },
    {
      "epoch": 1.0721144613020934,
      "grad_norm": 0.003866876708343625,
      "learning_rate": 0.019160152069495867,
      "loss": 0.5149,
      "num_input_tokens_seen": 46055184,
      "step": 33495
    },
    {
      "epoch": 1.0722745022725817,
      "grad_norm": 0.007202216424047947,
      "learning_rate": 0.019131355776109103,
      "loss": 0.7291,
      "num_input_tokens_seen": 46061920,
      "step": 33500
    },
    {
      "epoch": 1.0724345432430702,
      "grad_norm": 0.005483951419591904,
      "learning_rate": 0.019102579664311857,
      "loss": 0.4431,
      "num_input_tokens_seen": 46068368,
      "step": 33505
    },
    {
      "epoch": 1.0725945842135587,
      "grad_norm": 0.012482593767344952,
      "learning_rate": 0.019073823738541763,
      "loss": 0.3981,
      "num_input_tokens_seen": 46075200,
      "step": 33510
    },
    {
      "epoch": 1.0727546251840472,
      "grad_norm": 0.0026001285295933485,
      "learning_rate": 0.0190450880032334,
      "loss": 0.4358,
      "num_input_tokens_seen": 46081584,
      "step": 33515
    },
    {
      "epoch": 1.0729146661545355,
      "grad_norm": 0.0034787124022841454,
      "learning_rate": 0.019016372462818114,
      "loss": 0.4577,
      "num_input_tokens_seen": 46088288,
      "step": 33520
    },
    {
      "epoch": 1.073074707125024,
      "grad_norm": 0.008065884932875633,
      "learning_rate": 0.018987677121724278,
      "loss": 0.4469,
      "num_input_tokens_seen": 46095008,
      "step": 33525
    },
    {
      "epoch": 1.0732347480955124,
      "grad_norm": 0.006798510439693928,
      "learning_rate": 0.018959001984377,
      "loss": 0.5522,
      "num_input_tokens_seen": 46102112,
      "step": 33530
    },
    {
      "epoch": 1.073394789066001,
      "grad_norm": 0.004424516577273607,
      "learning_rate": 0.018930347055198377,
      "loss": 0.4121,
      "num_input_tokens_seen": 46108560,
      "step": 33535
    },
    {
      "epoch": 1.0735548300364894,
      "grad_norm": 0.004994677845388651,
      "learning_rate": 0.01890171233860739,
      "loss": 0.4835,
      "num_input_tokens_seen": 46115216,
      "step": 33540
    },
    {
      "epoch": 1.0737148710069777,
      "grad_norm": 0.004281857516616583,
      "learning_rate": 0.018873097839019807,
      "loss": 0.7001,
      "num_input_tokens_seen": 46122304,
      "step": 33545
    },
    {
      "epoch": 1.0738749119774662,
      "grad_norm": 0.003557079005986452,
      "learning_rate": 0.0188445035608484,
      "loss": 0.3747,
      "num_input_tokens_seen": 46129648,
      "step": 33550
    },
    {
      "epoch": 1.0740349529479547,
      "grad_norm": 0.005597129929810762,
      "learning_rate": 0.018815929508502777,
      "loss": 0.4232,
      "num_input_tokens_seen": 46136480,
      "step": 33555
    },
    {
      "epoch": 1.0741949939184432,
      "grad_norm": 0.005431739613413811,
      "learning_rate": 0.01878737568638934,
      "loss": 0.5253,
      "num_input_tokens_seen": 46143280,
      "step": 33560
    },
    {
      "epoch": 1.0743550348889315,
      "grad_norm": 0.004795936867594719,
      "learning_rate": 0.01875884209891152,
      "loss": 0.6015,
      "num_input_tokens_seen": 46150240,
      "step": 33565
    },
    {
      "epoch": 1.07451507585942,
      "grad_norm": 0.002060574246570468,
      "learning_rate": 0.018730328750469514,
      "loss": 0.5266,
      "num_input_tokens_seen": 46157280,
      "step": 33570
    },
    {
      "epoch": 1.0746751168299085,
      "grad_norm": 0.0032920334488153458,
      "learning_rate": 0.018701835645460473,
      "loss": 0.5342,
      "num_input_tokens_seen": 46163952,
      "step": 33575
    },
    {
      "epoch": 1.074835157800397,
      "grad_norm": 0.010294211097061634,
      "learning_rate": 0.01867336278827838,
      "loss": 0.4535,
      "num_input_tokens_seen": 46170656,
      "step": 33580
    },
    {
      "epoch": 1.0749951987708855,
      "grad_norm": 0.003457958111539483,
      "learning_rate": 0.018644910183314056,
      "loss": 0.4484,
      "num_input_tokens_seen": 46177248,
      "step": 33585
    },
    {
      "epoch": 1.0751552397413737,
      "grad_norm": 0.00442332960665226,
      "learning_rate": 0.01861647783495531,
      "loss": 0.3484,
      "num_input_tokens_seen": 46184176,
      "step": 33590
    },
    {
      "epoch": 1.0753152807118622,
      "grad_norm": 0.0026939052622765303,
      "learning_rate": 0.01858806574758676,
      "loss": 0.3674,
      "num_input_tokens_seen": 46190784,
      "step": 33595
    },
    {
      "epoch": 1.0754753216823507,
      "grad_norm": 0.007240551058202982,
      "learning_rate": 0.01855967392558988,
      "loss": 0.5677,
      "num_input_tokens_seen": 46197840,
      "step": 33600
    },
    {
      "epoch": 1.0754753216823507,
      "eval_loss": 0.4972416162490845,
      "eval_runtime": 332.1772,
      "eval_samples_per_second": 41.803,
      "eval_steps_per_second": 20.901,
      "num_input_tokens_seen": 46197840,
      "step": 33600
    },
    {
      "epoch": 1.0756353626528392,
      "grad_norm": 0.004185801837593317,
      "learning_rate": 0.018531302373343096,
      "loss": 0.5835,
      "num_input_tokens_seen": 46204752,
      "step": 33605
    },
    {
      "epoch": 1.0757954036233275,
      "grad_norm": 0.0036900960840284824,
      "learning_rate": 0.018502951095221588,
      "loss": 0.5176,
      "num_input_tokens_seen": 46211696,
      "step": 33610
    },
    {
      "epoch": 1.075955444593816,
      "grad_norm": 0.003173992969095707,
      "learning_rate": 0.01847462009559751,
      "loss": 0.5344,
      "num_input_tokens_seen": 46218832,
      "step": 33615
    },
    {
      "epoch": 1.0761154855643045,
      "grad_norm": 0.0019406576175242662,
      "learning_rate": 0.01844630937883992,
      "loss": 0.3456,
      "num_input_tokens_seen": 46225792,
      "step": 33620
    },
    {
      "epoch": 1.076275526534793,
      "grad_norm": 0.0060816071927547455,
      "learning_rate": 0.018418018949314573,
      "loss": 0.4044,
      "num_input_tokens_seen": 46232592,
      "step": 33625
    },
    {
      "epoch": 1.0764355675052815,
      "grad_norm": 0.006894935388118029,
      "learning_rate": 0.018389748811384315,
      "loss": 0.4767,
      "num_input_tokens_seen": 46239376,
      "step": 33630
    },
    {
      "epoch": 1.0765956084757697,
      "grad_norm": 0.004862874746322632,
      "learning_rate": 0.018361498969408658,
      "loss": 0.4038,
      "num_input_tokens_seen": 46246224,
      "step": 33635
    },
    {
      "epoch": 1.0767556494462582,
      "grad_norm": 0.003365796757861972,
      "learning_rate": 0.01833326942774415,
      "loss": 0.387,
      "num_input_tokens_seen": 46253728,
      "step": 33640
    },
    {
      "epoch": 1.0769156904167467,
      "grad_norm": 0.003583981189876795,
      "learning_rate": 0.018305060190744155,
      "loss": 0.304,
      "num_input_tokens_seen": 46260400,
      "step": 33645
    },
    {
      "epoch": 1.0770757313872352,
      "grad_norm": 0.004761639516800642,
      "learning_rate": 0.018276871262758846,
      "loss": 0.5541,
      "num_input_tokens_seen": 46267440,
      "step": 33650
    },
    {
      "epoch": 1.0772357723577235,
      "grad_norm": 0.006695516873151064,
      "learning_rate": 0.0182487026481353,
      "loss": 0.5155,
      "num_input_tokens_seen": 46274704,
      "step": 33655
    },
    {
      "epoch": 1.077395813328212,
      "grad_norm": 0.0038066250272095203,
      "learning_rate": 0.018220554351217538,
      "loss": 0.3204,
      "num_input_tokens_seen": 46281648,
      "step": 33660
    },
    {
      "epoch": 1.0775558542987005,
      "grad_norm": 0.005249334499239922,
      "learning_rate": 0.01819242637634629,
      "loss": 0.3696,
      "num_input_tokens_seen": 46289312,
      "step": 33665
    },
    {
      "epoch": 1.077715895269189,
      "grad_norm": 0.004938203375786543,
      "learning_rate": 0.01816431872785933,
      "loss": 0.3935,
      "num_input_tokens_seen": 46295984,
      "step": 33670
    },
    {
      "epoch": 1.0778759362396773,
      "grad_norm": 0.0019029354443773627,
      "learning_rate": 0.018136231410091148,
      "loss": 0.3453,
      "num_input_tokens_seen": 46302752,
      "step": 33675
    },
    {
      "epoch": 1.0780359772101658,
      "grad_norm": 0.0055668968707323074,
      "learning_rate": 0.018108164427373175,
      "loss": 0.4127,
      "num_input_tokens_seen": 46309264,
      "step": 33680
    },
    {
      "epoch": 1.0781960181806542,
      "grad_norm": 0.007208739407360554,
      "learning_rate": 0.01808011778403375,
      "loss": 0.471,
      "num_input_tokens_seen": 46316480,
      "step": 33685
    },
    {
      "epoch": 1.0783560591511427,
      "grad_norm": 0.002410548273473978,
      "learning_rate": 0.01805209148439793,
      "loss": 0.4743,
      "num_input_tokens_seen": 46323024,
      "step": 33690
    },
    {
      "epoch": 1.0785161001216312,
      "grad_norm": 0.02144007198512554,
      "learning_rate": 0.018024085532787757,
      "loss": 0.4614,
      "num_input_tokens_seen": 46329728,
      "step": 33695
    },
    {
      "epoch": 1.0786761410921195,
      "grad_norm": 0.0026814332231879234,
      "learning_rate": 0.017996099933522164,
      "loss": 0.389,
      "num_input_tokens_seen": 46337024,
      "step": 33700
    },
    {
      "epoch": 1.078836182062608,
      "grad_norm": 0.0025486184749752283,
      "learning_rate": 0.017968134690916775,
      "loss": 0.3376,
      "num_input_tokens_seen": 46343792,
      "step": 33705
    },
    {
      "epoch": 1.0789962230330965,
      "grad_norm": 0.007444345857948065,
      "learning_rate": 0.017940189809284263,
      "loss": 0.5342,
      "num_input_tokens_seen": 46350560,
      "step": 33710
    },
    {
      "epoch": 1.079156264003585,
      "grad_norm": 0.006882400251924992,
      "learning_rate": 0.017912265292934024,
      "loss": 0.6028,
      "num_input_tokens_seen": 46357664,
      "step": 33715
    },
    {
      "epoch": 1.0793163049740733,
      "grad_norm": 0.0028111287392675877,
      "learning_rate": 0.017884361146172423,
      "loss": 0.4918,
      "num_input_tokens_seen": 46364192,
      "step": 33720
    },
    {
      "epoch": 1.0794763459445618,
      "grad_norm": 0.005317118484526873,
      "learning_rate": 0.01785647737330261,
      "loss": 0.5222,
      "num_input_tokens_seen": 46370608,
      "step": 33725
    },
    {
      "epoch": 1.0796363869150503,
      "grad_norm": 0.004199802875518799,
      "learning_rate": 0.017828613978624563,
      "loss": 0.5166,
      "num_input_tokens_seen": 46377456,
      "step": 33730
    },
    {
      "epoch": 1.0797964278855388,
      "grad_norm": 0.003510815091431141,
      "learning_rate": 0.01780077096643523,
      "loss": 0.4743,
      "num_input_tokens_seen": 46384000,
      "step": 33735
    },
    {
      "epoch": 1.079956468856027,
      "grad_norm": 0.0018999662715941668,
      "learning_rate": 0.017772948341028345,
      "loss": 0.4496,
      "num_input_tokens_seen": 46391376,
      "step": 33740
    },
    {
      "epoch": 1.0801165098265155,
      "grad_norm": 0.0042709652334451675,
      "learning_rate": 0.01774514610669447,
      "loss": 0.3393,
      "num_input_tokens_seen": 46398144,
      "step": 33745
    },
    {
      "epoch": 1.080276550797004,
      "grad_norm": 0.004356301389634609,
      "learning_rate": 0.017717364267721112,
      "loss": 0.5835,
      "num_input_tokens_seen": 46404432,
      "step": 33750
    },
    {
      "epoch": 1.0804365917674925,
      "grad_norm": 0.0040536122396588326,
      "learning_rate": 0.017689602828392513,
      "loss": 0.5326,
      "num_input_tokens_seen": 46411792,
      "step": 33755
    },
    {
      "epoch": 1.080596632737981,
      "grad_norm": 0.00880168192088604,
      "learning_rate": 0.017661861792989897,
      "loss": 0.6438,
      "num_input_tokens_seen": 46418832,
      "step": 33760
    },
    {
      "epoch": 1.0807566737084693,
      "grad_norm": 0.004629158414900303,
      "learning_rate": 0.017634141165791272,
      "loss": 0.401,
      "num_input_tokens_seen": 46425712,
      "step": 33765
    },
    {
      "epoch": 1.0809167146789578,
      "grad_norm": 0.005984866060316563,
      "learning_rate": 0.017606440951071455,
      "loss": 0.4707,
      "num_input_tokens_seen": 46432672,
      "step": 33770
    },
    {
      "epoch": 1.0810767556494463,
      "grad_norm": 0.003875092137604952,
      "learning_rate": 0.017578761153102213,
      "loss": 0.4345,
      "num_input_tokens_seen": 46439520,
      "step": 33775
    },
    {
      "epoch": 1.0812367966199348,
      "grad_norm": 0.0036323554813861847,
      "learning_rate": 0.017551101776152146,
      "loss": 0.3846,
      "num_input_tokens_seen": 46446160,
      "step": 33780
    },
    {
      "epoch": 1.081396837590423,
      "grad_norm": 0.004977197851985693,
      "learning_rate": 0.017523462824486608,
      "loss": 0.468,
      "num_input_tokens_seen": 46453312,
      "step": 33785
    },
    {
      "epoch": 1.0815568785609115,
      "grad_norm": 0.0025732952635735273,
      "learning_rate": 0.01749584430236794,
      "loss": 0.4185,
      "num_input_tokens_seen": 46460400,
      "step": 33790
    },
    {
      "epoch": 1.0817169195314,
      "grad_norm": 0.006091427523642778,
      "learning_rate": 0.01746824621405524,
      "loss": 0.5288,
      "num_input_tokens_seen": 46467808,
      "step": 33795
    },
    {
      "epoch": 1.0818769605018885,
      "grad_norm": 0.010724578984081745,
      "learning_rate": 0.017440668563804412,
      "loss": 0.4757,
      "num_input_tokens_seen": 46474848,
      "step": 33800
    },
    {
      "epoch": 1.0818769605018885,
      "eval_loss": 0.4973933696746826,
      "eval_runtime": 332.8331,
      "eval_samples_per_second": 41.721,
      "eval_steps_per_second": 20.86,
      "num_input_tokens_seen": 46474848,
      "step": 33800
    },
    {
      "epoch": 1.082037001472377,
      "grad_norm": 0.0026118203531950712,
      "learning_rate": 0.017413111355868392,
      "loss": 0.494,
      "num_input_tokens_seen": 46481344,
      "step": 33805
    },
    {
      "epoch": 1.0821970424428653,
      "grad_norm": 0.0044539631344377995,
      "learning_rate": 0.017385574594496748,
      "loss": 0.5277,
      "num_input_tokens_seen": 46488464,
      "step": 33810
    },
    {
      "epoch": 1.0823570834133538,
      "grad_norm": 0.006038842722773552,
      "learning_rate": 0.01735805828393605,
      "loss": 0.4529,
      "num_input_tokens_seen": 46495664,
      "step": 33815
    },
    {
      "epoch": 1.0825171243838423,
      "grad_norm": 0.0037114887963980436,
      "learning_rate": 0.017330562428429667,
      "loss": 0.4028,
      "num_input_tokens_seen": 46502448,
      "step": 33820
    },
    {
      "epoch": 1.0826771653543308,
      "grad_norm": 0.004454685375094414,
      "learning_rate": 0.01730308703221776,
      "loss": 0.5178,
      "num_input_tokens_seen": 46509376,
      "step": 33825
    },
    {
      "epoch": 1.082837206324819,
      "grad_norm": 0.002815501531586051,
      "learning_rate": 0.01727563209953744,
      "loss": 0.3177,
      "num_input_tokens_seen": 46516144,
      "step": 33830
    },
    {
      "epoch": 1.0829972472953076,
      "grad_norm": 0.0030841969419270754,
      "learning_rate": 0.017248197634622535,
      "loss": 0.4716,
      "num_input_tokens_seen": 46523344,
      "step": 33835
    },
    {
      "epoch": 1.083157288265796,
      "grad_norm": 0.004793687257915735,
      "learning_rate": 0.01722078364170383,
      "loss": 0.4647,
      "num_input_tokens_seen": 46530144,
      "step": 33840
    },
    {
      "epoch": 1.0833173292362845,
      "grad_norm": 0.004467438440769911,
      "learning_rate": 0.017193390125008905,
      "loss": 0.4241,
      "num_input_tokens_seen": 46537040,
      "step": 33845
    },
    {
      "epoch": 1.083477370206773,
      "grad_norm": 0.005678160581737757,
      "learning_rate": 0.017166017088762153,
      "loss": 0.5065,
      "num_input_tokens_seen": 46543824,
      "step": 33850
    },
    {
      "epoch": 1.0836374111772613,
      "grad_norm": 0.003246394218876958,
      "learning_rate": 0.017138664537184878,
      "loss": 0.54,
      "num_input_tokens_seen": 46550864,
      "step": 33855
    },
    {
      "epoch": 1.0837974521477498,
      "grad_norm": 0.008824368938803673,
      "learning_rate": 0.017111332474495172,
      "loss": 0.3091,
      "num_input_tokens_seen": 46557536,
      "step": 33860
    },
    {
      "epoch": 1.0839574931182383,
      "grad_norm": 0.0038983433041721582,
      "learning_rate": 0.017084020904907998,
      "loss": 0.5905,
      "num_input_tokens_seen": 46564480,
      "step": 33865
    },
    {
      "epoch": 1.0841175340887268,
      "grad_norm": 0.005078893154859543,
      "learning_rate": 0.017056729832635103,
      "loss": 0.5771,
      "num_input_tokens_seen": 46571680,
      "step": 33870
    },
    {
      "epoch": 1.084277575059215,
      "grad_norm": 0.00471279164776206,
      "learning_rate": 0.017029459261885153,
      "loss": 0.4136,
      "num_input_tokens_seen": 46578608,
      "step": 33875
    },
    {
      "epoch": 1.0844376160297036,
      "grad_norm": 0.005074393004179001,
      "learning_rate": 0.01700220919686359,
      "loss": 0.4686,
      "num_input_tokens_seen": 46585200,
      "step": 33880
    },
    {
      "epoch": 1.084597657000192,
      "grad_norm": 0.006886638700962067,
      "learning_rate": 0.016974979641772723,
      "loss": 0.3949,
      "num_input_tokens_seen": 46592096,
      "step": 33885
    },
    {
      "epoch": 1.0847576979706806,
      "grad_norm": 0.006898552179336548,
      "learning_rate": 0.01694777060081169,
      "loss": 0.2655,
      "num_input_tokens_seen": 46599280,
      "step": 33890
    },
    {
      "epoch": 1.084917738941169,
      "grad_norm": 0.007567889057099819,
      "learning_rate": 0.016920582078176444,
      "loss": 0.6148,
      "num_input_tokens_seen": 46606560,
      "step": 33895
    },
    {
      "epoch": 1.0850777799116573,
      "grad_norm": 0.005608276929706335,
      "learning_rate": 0.016893414078059863,
      "loss": 0.6356,
      "num_input_tokens_seen": 46613392,
      "step": 33900
    },
    {
      "epoch": 1.0852378208821458,
      "grad_norm": 0.00812701415270567,
      "learning_rate": 0.016866266604651535,
      "loss": 0.4482,
      "num_input_tokens_seen": 46620144,
      "step": 33905
    },
    {
      "epoch": 1.0853978618526343,
      "grad_norm": 0.001093674567528069,
      "learning_rate": 0.016839139662137976,
      "loss": 0.3688,
      "num_input_tokens_seen": 46626800,
      "step": 33910
    },
    {
      "epoch": 1.0855579028231228,
      "grad_norm": 0.0038771010003983974,
      "learning_rate": 0.01681203325470245,
      "loss": 0.5185,
      "num_input_tokens_seen": 46633504,
      "step": 33915
    },
    {
      "epoch": 1.085717943793611,
      "grad_norm": 0.004595627076923847,
      "learning_rate": 0.016784947386525157,
      "loss": 0.4643,
      "num_input_tokens_seen": 46640128,
      "step": 33920
    },
    {
      "epoch": 1.0858779847640996,
      "grad_norm": 0.0027545730117708445,
      "learning_rate": 0.01675788206178308,
      "loss": 0.4017,
      "num_input_tokens_seen": 46646832,
      "step": 33925
    },
    {
      "epoch": 1.086038025734588,
      "grad_norm": 0.0058073909021914005,
      "learning_rate": 0.016730837284649986,
      "loss": 0.4038,
      "num_input_tokens_seen": 46653648,
      "step": 33930
    },
    {
      "epoch": 1.0861980667050766,
      "grad_norm": 0.0048252069391310215,
      "learning_rate": 0.016703813059296583,
      "loss": 0.4604,
      "num_input_tokens_seen": 46661040,
      "step": 33935
    },
    {
      "epoch": 1.0863581076755648,
      "grad_norm": 0.003427425865083933,
      "learning_rate": 0.016676809389890294,
      "loss": 0.4972,
      "num_input_tokens_seen": 46667840,
      "step": 33940
    },
    {
      "epoch": 1.0865181486460533,
      "grad_norm": 0.006038637366145849,
      "learning_rate": 0.016649826280595435,
      "loss": 0.4486,
      "num_input_tokens_seen": 46674960,
      "step": 33945
    },
    {
      "epoch": 1.0866781896165418,
      "grad_norm": 0.004207894206047058,
      "learning_rate": 0.016622863735573163,
      "loss": 0.5453,
      "num_input_tokens_seen": 46681600,
      "step": 33950
    },
    {
      "epoch": 1.0868382305870303,
      "grad_norm": 0.005749468691647053,
      "learning_rate": 0.016595921758981395,
      "loss": 0.6079,
      "num_input_tokens_seen": 46688864,
      "step": 33955
    },
    {
      "epoch": 1.0869982715575188,
      "grad_norm": 0.005117243155837059,
      "learning_rate": 0.01656900035497495,
      "loss": 0.4517,
      "num_input_tokens_seen": 46695584,
      "step": 33960
    },
    {
      "epoch": 1.087158312528007,
      "grad_norm": 0.003817556658759713,
      "learning_rate": 0.016542099527705485,
      "loss": 0.3435,
      "num_input_tokens_seen": 46702352,
      "step": 33965
    },
    {
      "epoch": 1.0873183534984956,
      "grad_norm": 0.00522800674661994,
      "learning_rate": 0.01651521928132138,
      "loss": 0.5019,
      "num_input_tokens_seen": 46709184,
      "step": 33970
    },
    {
      "epoch": 1.087478394468984,
      "grad_norm": 0.004615446552634239,
      "learning_rate": 0.01648835961996794,
      "loss": 0.395,
      "num_input_tokens_seen": 46715856,
      "step": 33975
    },
    {
      "epoch": 1.0876384354394726,
      "grad_norm": 0.004133419133722782,
      "learning_rate": 0.016461520547787285,
      "loss": 0.5897,
      "num_input_tokens_seen": 46723008,
      "step": 33980
    },
    {
      "epoch": 1.0877984764099609,
      "grad_norm": 0.0047527700662612915,
      "learning_rate": 0.016434702068918266,
      "loss": 0.4503,
      "num_input_tokens_seen": 46729856,
      "step": 33985
    },
    {
      "epoch": 1.0879585173804494,
      "grad_norm": 0.002684782724827528,
      "learning_rate": 0.01640790418749673,
      "loss": 0.5112,
      "num_input_tokens_seen": 46736960,
      "step": 33990
    },
    {
      "epoch": 1.0881185583509378,
      "grad_norm": 0.005792920012027025,
      "learning_rate": 0.016381126907655134,
      "loss": 0.4959,
      "num_input_tokens_seen": 46743616,
      "step": 33995
    },
    {
      "epoch": 1.0882785993214263,
      "grad_norm": 0.0062135374173521996,
      "learning_rate": 0.016354370233522948,
      "loss": 0.5645,
      "num_input_tokens_seen": 46749824,
      "step": 34000
    },
    {
      "epoch": 1.0882785993214263,
      "eval_loss": 0.4970117211341858,
      "eval_runtime": 332.0187,
      "eval_samples_per_second": 41.823,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 46749824,
      "step": 34000
    },
    {
      "epoch": 1.0884386402919146,
      "grad_norm": 0.0029537237714976072,
      "learning_rate": 0.016327634169226394,
      "loss": 0.5307,
      "num_input_tokens_seen": 46756768,
      "step": 34005
    },
    {
      "epoch": 1.0885986812624031,
      "grad_norm": 0.0035628117620944977,
      "learning_rate": 0.016300918718888485,
      "loss": 0.4221,
      "num_input_tokens_seen": 46763584,
      "step": 34010
    },
    {
      "epoch": 1.0887587222328916,
      "grad_norm": 0.003221015678718686,
      "learning_rate": 0.016274223886629052,
      "loss": 0.4663,
      "num_input_tokens_seen": 46770144,
      "step": 34015
    },
    {
      "epoch": 1.08891876320338,
      "grad_norm": 0.0030806383583694696,
      "learning_rate": 0.01624754967656482,
      "loss": 0.4312,
      "num_input_tokens_seen": 46776688,
      "step": 34020
    },
    {
      "epoch": 1.0890788041738686,
      "grad_norm": 0.005625811405479908,
      "learning_rate": 0.016220896092809235,
      "loss": 0.4829,
      "num_input_tokens_seen": 46783728,
      "step": 34025
    },
    {
      "epoch": 1.0892388451443569,
      "grad_norm": 0.005357911344617605,
      "learning_rate": 0.01619426313947267,
      "loss": 0.5088,
      "num_input_tokens_seen": 46790704,
      "step": 34030
    },
    {
      "epoch": 1.0893988861148454,
      "grad_norm": 0.00422472832724452,
      "learning_rate": 0.016167650820662228,
      "loss": 0.4904,
      "num_input_tokens_seen": 46797616,
      "step": 34035
    },
    {
      "epoch": 1.0895589270853339,
      "grad_norm": 0.0035242680460214615,
      "learning_rate": 0.016141059140481855,
      "loss": 0.3215,
      "num_input_tokens_seen": 46805200,
      "step": 34040
    },
    {
      "epoch": 1.0897189680558224,
      "grad_norm": 0.008718174882233143,
      "learning_rate": 0.016114488103032374,
      "loss": 0.4433,
      "num_input_tokens_seen": 46811680,
      "step": 34045
    },
    {
      "epoch": 1.0898790090263106,
      "grad_norm": 0.008846844546496868,
      "learning_rate": 0.016087937712411293,
      "loss": 0.4343,
      "num_input_tokens_seen": 46818864,
      "step": 34050
    },
    {
      "epoch": 1.0900390499967991,
      "grad_norm": 0.007690564729273319,
      "learning_rate": 0.01606140797271308,
      "loss": 0.4848,
      "num_input_tokens_seen": 46825376,
      "step": 34055
    },
    {
      "epoch": 1.0901990909672876,
      "grad_norm": 0.0032156382221728563,
      "learning_rate": 0.01603489888802897,
      "loss": 0.4518,
      "num_input_tokens_seen": 46832384,
      "step": 34060
    },
    {
      "epoch": 1.0903591319377761,
      "grad_norm": 0.006973397918045521,
      "learning_rate": 0.016008410462446918,
      "loss": 0.4506,
      "num_input_tokens_seen": 46839312,
      "step": 34065
    },
    {
      "epoch": 1.0905191729082646,
      "grad_norm": 0.007051127962768078,
      "learning_rate": 0.01598194270005185,
      "loss": 0.6185,
      "num_input_tokens_seen": 46846288,
      "step": 34070
    },
    {
      "epoch": 1.0906792138787529,
      "grad_norm": 0.0045518758706748486,
      "learning_rate": 0.015955495604925356,
      "loss": 0.4187,
      "num_input_tokens_seen": 46852928,
      "step": 34075
    },
    {
      "epoch": 1.0908392548492414,
      "grad_norm": 0.006195869296789169,
      "learning_rate": 0.01592906918114598,
      "loss": 0.4287,
      "num_input_tokens_seen": 46859872,
      "step": 34080
    },
    {
      "epoch": 1.0909992958197299,
      "grad_norm": 0.007799689192324877,
      "learning_rate": 0.015902663432788965,
      "loss": 0.6461,
      "num_input_tokens_seen": 46866704,
      "step": 34085
    },
    {
      "epoch": 1.0911593367902184,
      "grad_norm": 0.0060691130347549915,
      "learning_rate": 0.01587627836392643,
      "loss": 0.434,
      "num_input_tokens_seen": 46873120,
      "step": 34090
    },
    {
      "epoch": 1.0913193777607066,
      "grad_norm": 0.005861303303390741,
      "learning_rate": 0.01584991397862726,
      "loss": 0.429,
      "num_input_tokens_seen": 46879840,
      "step": 34095
    },
    {
      "epoch": 1.0914794187311951,
      "grad_norm": 0.002163552911952138,
      "learning_rate": 0.015823570280957214,
      "loss": 0.3514,
      "num_input_tokens_seen": 46886304,
      "step": 34100
    },
    {
      "epoch": 1.0916394597016836,
      "grad_norm": 0.004648755304515362,
      "learning_rate": 0.015797247274978766,
      "loss": 0.5257,
      "num_input_tokens_seen": 46893104,
      "step": 34105
    },
    {
      "epoch": 1.0917995006721721,
      "grad_norm": 0.00511144520714879,
      "learning_rate": 0.015770944964751326,
      "loss": 0.4658,
      "num_input_tokens_seen": 46899920,
      "step": 34110
    },
    {
      "epoch": 1.0919595416426606,
      "grad_norm": 0.0043505411595106125,
      "learning_rate": 0.015744663354330956,
      "loss": 0.45,
      "num_input_tokens_seen": 46906768,
      "step": 34115
    },
    {
      "epoch": 1.092119582613149,
      "grad_norm": 0.0034392494708299637,
      "learning_rate": 0.015718402447770664,
      "loss": 0.3699,
      "num_input_tokens_seen": 46913712,
      "step": 34120
    },
    {
      "epoch": 1.0922796235836374,
      "grad_norm": 0.0027504432946443558,
      "learning_rate": 0.015692162249120224,
      "loss": 0.3896,
      "num_input_tokens_seen": 46920448,
      "step": 34125
    },
    {
      "epoch": 1.0924396645541259,
      "grad_norm": 0.0047865561209619045,
      "learning_rate": 0.01566594276242615,
      "loss": 0.4486,
      "num_input_tokens_seen": 46927392,
      "step": 34130
    },
    {
      "epoch": 1.0925997055246144,
      "grad_norm": 0.003015264868736267,
      "learning_rate": 0.015639743991731857,
      "loss": 0.4435,
      "num_input_tokens_seen": 46933920,
      "step": 34135
    },
    {
      "epoch": 1.0927597464951027,
      "grad_norm": 0.002604391425848007,
      "learning_rate": 0.01561356594107755,
      "loss": 0.5014,
      "num_input_tokens_seen": 46941104,
      "step": 34140
    },
    {
      "epoch": 1.0929197874655912,
      "grad_norm": 0.00510431407019496,
      "learning_rate": 0.015587408614500147,
      "loss": 0.4244,
      "num_input_tokens_seen": 46947760,
      "step": 34145
    },
    {
      "epoch": 1.0930798284360796,
      "grad_norm": 0.0033818059600889683,
      "learning_rate": 0.015561272016033505,
      "loss": 0.4183,
      "num_input_tokens_seen": 46954688,
      "step": 34150
    },
    {
      "epoch": 1.0932398694065681,
      "grad_norm": 0.0036775749176740646,
      "learning_rate": 0.015535156149708167,
      "loss": 0.5922,
      "num_input_tokens_seen": 46961024,
      "step": 34155
    },
    {
      "epoch": 1.0933999103770566,
      "grad_norm": 0.005269724875688553,
      "learning_rate": 0.015509061019551528,
      "loss": 0.4539,
      "num_input_tokens_seen": 46968336,
      "step": 34160
    },
    {
      "epoch": 1.093559951347545,
      "grad_norm": 0.004728043917566538,
      "learning_rate": 0.015482986629587818,
      "loss": 0.352,
      "num_input_tokens_seen": 46974912,
      "step": 34165
    },
    {
      "epoch": 1.0937199923180334,
      "grad_norm": 0.002409764099866152,
      "learning_rate": 0.01545693298383799,
      "loss": 0.4375,
      "num_input_tokens_seen": 46982464,
      "step": 34170
    },
    {
      "epoch": 1.093880033288522,
      "grad_norm": 0.003894820809364319,
      "learning_rate": 0.015430900086319858,
      "loss": 0.4564,
      "num_input_tokens_seen": 46989808,
      "step": 34175
    },
    {
      "epoch": 1.0940400742590104,
      "grad_norm": 0.0038975130300968885,
      "learning_rate": 0.015404887941048084,
      "loss": 0.4346,
      "num_input_tokens_seen": 46996560,
      "step": 34180
    },
    {
      "epoch": 1.0942001152294987,
      "grad_norm": 0.004261436872184277,
      "learning_rate": 0.01537889655203397,
      "loss": 0.5346,
      "num_input_tokens_seen": 47003184,
      "step": 34185
    },
    {
      "epoch": 1.0943601561999872,
      "grad_norm": 0.010551093146204948,
      "learning_rate": 0.015352925923285798,
      "loss": 0.8109,
      "num_input_tokens_seen": 47010528,
      "step": 34190
    },
    {
      "epoch": 1.0945201971704757,
      "grad_norm": 0.004058667924255133,
      "learning_rate": 0.015326976058808511,
      "loss": 0.4917,
      "num_input_tokens_seen": 47017168,
      "step": 34195
    },
    {
      "epoch": 1.0946802381409642,
      "grad_norm": 0.003274489426985383,
      "learning_rate": 0.015301046962603908,
      "loss": 0.5164,
      "num_input_tokens_seen": 47023856,
      "step": 34200
    },
    {
      "epoch": 1.0946802381409642,
      "eval_loss": 0.4961601793766022,
      "eval_runtime": 332.3195,
      "eval_samples_per_second": 41.785,
      "eval_steps_per_second": 20.893,
      "num_input_tokens_seen": 47023856,
      "step": 34200
    },
    {
      "epoch": 1.0948402791114524,
      "grad_norm": 0.006860094144940376,
      "learning_rate": 0.015275138638670626,
      "loss": 0.5435,
      "num_input_tokens_seen": 47030720,
      "step": 34205
    },
    {
      "epoch": 1.095000320081941,
      "grad_norm": 0.005170451942831278,
      "learning_rate": 0.015249251091004001,
      "loss": 0.4498,
      "num_input_tokens_seen": 47037536,
      "step": 34210
    },
    {
      "epoch": 1.0951603610524294,
      "grad_norm": 0.003160931868478656,
      "learning_rate": 0.01522338432359624,
      "loss": 0.379,
      "num_input_tokens_seen": 47044656,
      "step": 34215
    },
    {
      "epoch": 1.095320402022918,
      "grad_norm": 0.004995007999241352,
      "learning_rate": 0.01519753834043635,
      "loss": 0.4898,
      "num_input_tokens_seen": 47051744,
      "step": 34220
    },
    {
      "epoch": 1.0954804429934064,
      "grad_norm": 0.003614323679357767,
      "learning_rate": 0.015171713145510095,
      "loss": 0.4561,
      "num_input_tokens_seen": 47058720,
      "step": 34225
    },
    {
      "epoch": 1.0956404839638947,
      "grad_norm": 0.005900608375668526,
      "learning_rate": 0.01514590874279999,
      "loss": 0.562,
      "num_input_tokens_seen": 47065760,
      "step": 34230
    },
    {
      "epoch": 1.0958005249343832,
      "grad_norm": 0.0036051305942237377,
      "learning_rate": 0.015120125136285467,
      "loss": 0.4555,
      "num_input_tokens_seen": 47072912,
      "step": 34235
    },
    {
      "epoch": 1.0959605659048717,
      "grad_norm": 0.004243907518684864,
      "learning_rate": 0.015094362329942629,
      "loss": 0.4626,
      "num_input_tokens_seen": 47079824,
      "step": 34240
    },
    {
      "epoch": 1.0961206068753602,
      "grad_norm": 0.0028542843647301197,
      "learning_rate": 0.01506862032774448,
      "loss": 0.3617,
      "num_input_tokens_seen": 47087328,
      "step": 34245
    },
    {
      "epoch": 1.0962806478458484,
      "grad_norm": 0.023545680567622185,
      "learning_rate": 0.015042899133660697,
      "loss": 0.5883,
      "num_input_tokens_seen": 47094064,
      "step": 34250
    },
    {
      "epoch": 1.096440688816337,
      "grad_norm": 0.003729801159352064,
      "learning_rate": 0.01501719875165789,
      "loss": 0.4248,
      "num_input_tokens_seen": 47100800,
      "step": 34255
    },
    {
      "epoch": 1.0966007297868254,
      "grad_norm": 0.0058340211398899555,
      "learning_rate": 0.014991519185699286,
      "loss": 0.5814,
      "num_input_tokens_seen": 47107680,
      "step": 34260
    },
    {
      "epoch": 1.096760770757314,
      "grad_norm": 0.005712676793336868,
      "learning_rate": 0.014965860439745054,
      "loss": 0.5967,
      "num_input_tokens_seen": 47114592,
      "step": 34265
    },
    {
      "epoch": 1.0969208117278022,
      "grad_norm": 0.0018451310461387038,
      "learning_rate": 0.01494022251775211,
      "loss": 0.3798,
      "num_input_tokens_seen": 47121840,
      "step": 34270
    },
    {
      "epoch": 1.0970808526982907,
      "grad_norm": 0.004457391332834959,
      "learning_rate": 0.014914605423674109,
      "loss": 0.4064,
      "num_input_tokens_seen": 47128768,
      "step": 34275
    },
    {
      "epoch": 1.0972408936687792,
      "grad_norm": 0.006662935949862003,
      "learning_rate": 0.014889009161461525,
      "loss": 0.421,
      "num_input_tokens_seen": 47135744,
      "step": 34280
    },
    {
      "epoch": 1.0974009346392677,
      "grad_norm": 0.004116032272577286,
      "learning_rate": 0.014863433735061665,
      "loss": 0.3821,
      "num_input_tokens_seen": 47142784,
      "step": 34285
    },
    {
      "epoch": 1.0975609756097562,
      "grad_norm": 0.004145745187997818,
      "learning_rate": 0.014837879148418541,
      "loss": 0.4377,
      "num_input_tokens_seen": 47149536,
      "step": 34290
    },
    {
      "epoch": 1.0977210165802445,
      "grad_norm": 0.0054071638733148575,
      "learning_rate": 0.01481234540547302,
      "loss": 0.56,
      "num_input_tokens_seen": 47156560,
      "step": 34295
    },
    {
      "epoch": 1.097881057550733,
      "grad_norm": 0.005286841187626123,
      "learning_rate": 0.014786832510162717,
      "loss": 0.4148,
      "num_input_tokens_seen": 47163424,
      "step": 34300
    },
    {
      "epoch": 1.0980410985212214,
      "grad_norm": 0.005147294607013464,
      "learning_rate": 0.014761340466422017,
      "loss": 0.3876,
      "num_input_tokens_seen": 47170192,
      "step": 34305
    },
    {
      "epoch": 1.09820113949171,
      "grad_norm": 0.004133713897317648,
      "learning_rate": 0.014735869278182144,
      "loss": 0.3803,
      "num_input_tokens_seen": 47177360,
      "step": 34310
    },
    {
      "epoch": 1.0983611804621982,
      "grad_norm": 0.005607676226645708,
      "learning_rate": 0.014710418949371057,
      "loss": 0.5351,
      "num_input_tokens_seen": 47184272,
      "step": 34315
    },
    {
      "epoch": 1.0985212214326867,
      "grad_norm": 0.005900274496525526,
      "learning_rate": 0.014684989483913495,
      "loss": 0.4602,
      "num_input_tokens_seen": 47190848,
      "step": 34320
    },
    {
      "epoch": 1.0986812624031752,
      "grad_norm": 0.004068249370902777,
      "learning_rate": 0.014659580885731077,
      "loss": 0.5077,
      "num_input_tokens_seen": 47197680,
      "step": 34325
    },
    {
      "epoch": 1.0988413033736637,
      "grad_norm": 0.0037337092217057943,
      "learning_rate": 0.014634193158742047,
      "loss": 0.4334,
      "num_input_tokens_seen": 47204704,
      "step": 34330
    },
    {
      "epoch": 1.0990013443441522,
      "grad_norm": 0.005801777821034193,
      "learning_rate": 0.014608826306861576,
      "loss": 0.487,
      "num_input_tokens_seen": 47211808,
      "step": 34335
    },
    {
      "epoch": 1.0991613853146405,
      "grad_norm": 0.00587339187040925,
      "learning_rate": 0.014583480334001486,
      "loss": 0.3362,
      "num_input_tokens_seen": 47218672,
      "step": 34340
    },
    {
      "epoch": 1.099321426285129,
      "grad_norm": 0.007643110118806362,
      "learning_rate": 0.014558155244070496,
      "loss": 0.4407,
      "num_input_tokens_seen": 47225664,
      "step": 34345
    },
    {
      "epoch": 1.0994814672556175,
      "grad_norm": 0.00304663204587996,
      "learning_rate": 0.014532851040974036,
      "loss": 0.4234,
      "num_input_tokens_seen": 47233200,
      "step": 34350
    },
    {
      "epoch": 1.099641508226106,
      "grad_norm": 0.005155749619007111,
      "learning_rate": 0.014507567728614335,
      "loss": 0.4642,
      "num_input_tokens_seen": 47240064,
      "step": 34355
    },
    {
      "epoch": 1.0998015491965942,
      "grad_norm": 0.006004240829497576,
      "learning_rate": 0.01448230531089037,
      "loss": 0.3537,
      "num_input_tokens_seen": 47246800,
      "step": 34360
    },
    {
      "epoch": 1.0999615901670827,
      "grad_norm": 0.0021962698083370924,
      "learning_rate": 0.014457063791697993,
      "loss": 0.2799,
      "num_input_tokens_seen": 47253312,
      "step": 34365
    },
    {
      "epoch": 1.1001216311375712,
      "grad_norm": 0.006833389867097139,
      "learning_rate": 0.01443184317492971,
      "loss": 0.3842,
      "num_input_tokens_seen": 47260064,
      "step": 34370
    },
    {
      "epoch": 1.1002816721080597,
      "grad_norm": 0.0024265472311526537,
      "learning_rate": 0.014406643464474822,
      "loss": 0.569,
      "num_input_tokens_seen": 47266976,
      "step": 34375
    },
    {
      "epoch": 1.1004417130785482,
      "grad_norm": 0.006044729147106409,
      "learning_rate": 0.014381464664219539,
      "loss": 0.4717,
      "num_input_tokens_seen": 47273856,
      "step": 34380
    },
    {
      "epoch": 1.1006017540490365,
      "grad_norm": 0.0034876754507422447,
      "learning_rate": 0.014356306778046656,
      "loss": 0.7818,
      "num_input_tokens_seen": 47280752,
      "step": 34385
    },
    {
      "epoch": 1.100761795019525,
      "grad_norm": 0.003437182167544961,
      "learning_rate": 0.014331169809835885,
      "loss": 0.6478,
      "num_input_tokens_seen": 47288032,
      "step": 34390
    },
    {
      "epoch": 1.1009218359900135,
      "grad_norm": 0.004317691083997488,
      "learning_rate": 0.014306053763463644,
      "loss": 0.5038,
      "num_input_tokens_seen": 47294832,
      "step": 34395
    },
    {
      "epoch": 1.101081876960502,
      "grad_norm": 0.00439574895426631,
      "learning_rate": 0.014280958642803147,
      "loss": 0.4217,
      "num_input_tokens_seen": 47301520,
      "step": 34400
    },
    {
      "epoch": 1.101081876960502,
      "eval_loss": 0.4973459541797638,
      "eval_runtime": 332.144,
      "eval_samples_per_second": 41.807,
      "eval_steps_per_second": 20.904,
      "num_input_tokens_seen": 47301520,
      "step": 34400
    },
    {
      "epoch": 1.1012419179309902,
      "grad_norm": 0.005847639869898558,
      "learning_rate": 0.014255884451724404,
      "loss": 0.4002,
      "num_input_tokens_seen": 47308416,
      "step": 34405
    },
    {
      "epoch": 1.1014019589014787,
      "grad_norm": 0.008742746897041798,
      "learning_rate": 0.014230831194094101,
      "loss": 0.5451,
      "num_input_tokens_seen": 47315040,
      "step": 34410
    },
    {
      "epoch": 1.1015619998719672,
      "grad_norm": 0.0039924876764416695,
      "learning_rate": 0.014205798873775865,
      "loss": 0.4951,
      "num_input_tokens_seen": 47321776,
      "step": 34415
    },
    {
      "epoch": 1.1017220408424557,
      "grad_norm": 0.00728240329772234,
      "learning_rate": 0.014180787494629893,
      "loss": 0.5943,
      "num_input_tokens_seen": 47328512,
      "step": 34420
    },
    {
      "epoch": 1.1018820818129442,
      "grad_norm": 0.0077343168668448925,
      "learning_rate": 0.014155797060513314,
      "loss": 0.5668,
      "num_input_tokens_seen": 47335488,
      "step": 34425
    },
    {
      "epoch": 1.1020421227834325,
      "grad_norm": 0.005767846945673227,
      "learning_rate": 0.014130827575279963,
      "loss": 0.4097,
      "num_input_tokens_seen": 47342128,
      "step": 34430
    },
    {
      "epoch": 1.102202163753921,
      "grad_norm": 0.004521414637565613,
      "learning_rate": 0.014105879042780427,
      "loss": 0.3966,
      "num_input_tokens_seen": 47348736,
      "step": 34435
    },
    {
      "epoch": 1.1023622047244095,
      "grad_norm": 0.005898920353502035,
      "learning_rate": 0.014080951466862113,
      "loss": 0.4826,
      "num_input_tokens_seen": 47355488,
      "step": 34440
    },
    {
      "epoch": 1.102522245694898,
      "grad_norm": 0.005965698976069689,
      "learning_rate": 0.014056044851369126,
      "loss": 0.4293,
      "num_input_tokens_seen": 47362624,
      "step": 34445
    },
    {
      "epoch": 1.1026822866653863,
      "grad_norm": 0.00495936768129468,
      "learning_rate": 0.014031159200142428,
      "loss": 0.4986,
      "num_input_tokens_seen": 47369312,
      "step": 34450
    },
    {
      "epoch": 1.1028423276358748,
      "grad_norm": 0.002392883412539959,
      "learning_rate": 0.014006294517019667,
      "loss": 0.2474,
      "num_input_tokens_seen": 47376096,
      "step": 34455
    },
    {
      "epoch": 1.1030023686063632,
      "grad_norm": 0.005422604735940695,
      "learning_rate": 0.013981450805835276,
      "loss": 0.477,
      "num_input_tokens_seen": 47383136,
      "step": 34460
    },
    {
      "epoch": 1.1031624095768517,
      "grad_norm": 0.001863356214016676,
      "learning_rate": 0.01395662807042049,
      "loss": 0.499,
      "num_input_tokens_seen": 47390416,
      "step": 34465
    },
    {
      "epoch": 1.10332245054734,
      "grad_norm": 0.003452565986663103,
      "learning_rate": 0.013931826314603296,
      "loss": 0.4639,
      "num_input_tokens_seen": 47397488,
      "step": 34470
    },
    {
      "epoch": 1.1034824915178285,
      "grad_norm": 0.00467110238969326,
      "learning_rate": 0.013907045542208401,
      "loss": 0.6368,
      "num_input_tokens_seen": 47404112,
      "step": 34475
    },
    {
      "epoch": 1.103642532488317,
      "grad_norm": 0.004067651927471161,
      "learning_rate": 0.013882285757057333,
      "loss": 0.6415,
      "num_input_tokens_seen": 47411184,
      "step": 34480
    },
    {
      "epoch": 1.1038025734588055,
      "grad_norm": 0.0038568072486668825,
      "learning_rate": 0.013857546962968403,
      "loss": 0.5708,
      "num_input_tokens_seen": 47417600,
      "step": 34485
    },
    {
      "epoch": 1.103962614429294,
      "grad_norm": 0.0035952518228441477,
      "learning_rate": 0.013832829163756577,
      "loss": 0.4451,
      "num_input_tokens_seen": 47424800,
      "step": 34490
    },
    {
      "epoch": 1.1041226553997823,
      "grad_norm": 0.004210071638226509,
      "learning_rate": 0.013808132363233689,
      "loss": 0.4613,
      "num_input_tokens_seen": 47431248,
      "step": 34495
    },
    {
      "epoch": 1.1042826963702708,
      "grad_norm": 0.005411107558757067,
      "learning_rate": 0.013783456565208256,
      "loss": 0.4458,
      "num_input_tokens_seen": 47438352,
      "step": 34500
    },
    {
      "epoch": 1.1044427373407593,
      "grad_norm": 0.005864849779754877,
      "learning_rate": 0.01375880177348564,
      "loss": 0.4441,
      "num_input_tokens_seen": 47445072,
      "step": 34505
    },
    {
      "epoch": 1.1046027783112478,
      "grad_norm": 0.008215149864554405,
      "learning_rate": 0.013734167991867928,
      "loss": 0.5287,
      "num_input_tokens_seen": 47452048,
      "step": 34510
    },
    {
      "epoch": 1.104762819281736,
      "grad_norm": 0.0033775311894714832,
      "learning_rate": 0.013709555224153935,
      "loss": 0.4837,
      "num_input_tokens_seen": 47459296,
      "step": 34515
    },
    {
      "epoch": 1.1049228602522245,
      "grad_norm": 0.0021294187754392624,
      "learning_rate": 0.013684963474139222,
      "loss": 0.4513,
      "num_input_tokens_seen": 47466384,
      "step": 34520
    },
    {
      "epoch": 1.105082901222713,
      "grad_norm": 0.004168936517089605,
      "learning_rate": 0.013660392745616224,
      "loss": 0.4692,
      "num_input_tokens_seen": 47472816,
      "step": 34525
    },
    {
      "epoch": 1.1052429421932015,
      "grad_norm": 0.002648143796250224,
      "learning_rate": 0.013635843042373974,
      "loss": 0.4081,
      "num_input_tokens_seen": 47479920,
      "step": 34530
    },
    {
      "epoch": 1.1054029831636898,
      "grad_norm": 0.0075241029262542725,
      "learning_rate": 0.01361131436819843,
      "loss": 0.4745,
      "num_input_tokens_seen": 47486704,
      "step": 34535
    },
    {
      "epoch": 1.1055630241341783,
      "grad_norm": 0.005116224754601717,
      "learning_rate": 0.013586806726872147,
      "loss": 0.6904,
      "num_input_tokens_seen": 47493696,
      "step": 34540
    },
    {
      "epoch": 1.1057230651046668,
      "grad_norm": 0.003794788382947445,
      "learning_rate": 0.013562320122174537,
      "loss": 0.6044,
      "num_input_tokens_seen": 47500320,
      "step": 34545
    },
    {
      "epoch": 1.1058831060751553,
      "grad_norm": 0.004426529165357351,
      "learning_rate": 0.013537854557881762,
      "loss": 0.5133,
      "num_input_tokens_seen": 47507152,
      "step": 34550
    },
    {
      "epoch": 1.1060431470456438,
      "grad_norm": 0.004873438738286495,
      "learning_rate": 0.013513410037766687,
      "loss": 0.5889,
      "num_input_tokens_seen": 47513424,
      "step": 34555
    },
    {
      "epoch": 1.106203188016132,
      "grad_norm": 0.0050610629841685295,
      "learning_rate": 0.013488986565598998,
      "loss": 0.4631,
      "num_input_tokens_seen": 47520720,
      "step": 34560
    },
    {
      "epoch": 1.1063632289866205,
      "grad_norm": 0.0033341855742037296,
      "learning_rate": 0.013464584145145097,
      "loss": 0.4754,
      "num_input_tokens_seen": 47527456,
      "step": 34565
    },
    {
      "epoch": 1.106523269957109,
      "grad_norm": 0.005665253847837448,
      "learning_rate": 0.013440202780168109,
      "loss": 0.3896,
      "num_input_tokens_seen": 47533968,
      "step": 34570
    },
    {
      "epoch": 1.1066833109275975,
      "grad_norm": 0.005559862591326237,
      "learning_rate": 0.01341584247442799,
      "loss": 0.5866,
      "num_input_tokens_seen": 47540736,
      "step": 34575
    },
    {
      "epoch": 1.1068433518980858,
      "grad_norm": 0.003181230276823044,
      "learning_rate": 0.013391503231681355,
      "loss": 0.4132,
      "num_input_tokens_seen": 47547456,
      "step": 34580
    },
    {
      "epoch": 1.1070033928685743,
      "grad_norm": 0.005067066289484501,
      "learning_rate": 0.013367185055681685,
      "loss": 0.6001,
      "num_input_tokens_seen": 47554384,
      "step": 34585
    },
    {
      "epoch": 1.1071634338390628,
      "grad_norm": 0.010008140467107296,
      "learning_rate": 0.013342887950179095,
      "loss": 0.4178,
      "num_input_tokens_seen": 47561248,
      "step": 34590
    },
    {
      "epoch": 1.1073234748095513,
      "grad_norm": 0.004184138495475054,
      "learning_rate": 0.013318611918920554,
      "loss": 0.4893,
      "num_input_tokens_seen": 47568112,
      "step": 34595
    },
    {
      "epoch": 1.1074835157800398,
      "grad_norm": 0.0033792173489928246,
      "learning_rate": 0.01329435696564965,
      "loss": 0.5314,
      "num_input_tokens_seen": 47574864,
      "step": 34600
    },
    {
      "epoch": 1.1074835157800398,
      "eval_loss": 0.49658483266830444,
      "eval_runtime": 332.4488,
      "eval_samples_per_second": 41.769,
      "eval_steps_per_second": 20.884,
      "num_input_tokens_seen": 47574864,
      "step": 34600
    },
    {
      "epoch": 1.107643556750528,
      "grad_norm": 0.004267814569175243,
      "learning_rate": 0.013270123094106894,
      "loss": 0.3593,
      "num_input_tokens_seen": 47581456,
      "step": 34605
    },
    {
      "epoch": 1.1078035977210166,
      "grad_norm": 0.006555644795298576,
      "learning_rate": 0.013245910308029395,
      "loss": 0.456,
      "num_input_tokens_seen": 47588656,
      "step": 34610
    },
    {
      "epoch": 1.107963638691505,
      "grad_norm": 0.004877245519310236,
      "learning_rate": 0.0132217186111511,
      "loss": 0.3809,
      "num_input_tokens_seen": 47595312,
      "step": 34615
    },
    {
      "epoch": 1.1081236796619935,
      "grad_norm": 0.0034337970428168774,
      "learning_rate": 0.013197548007202626,
      "loss": 0.3906,
      "num_input_tokens_seen": 47602064,
      "step": 34620
    },
    {
      "epoch": 1.1082837206324818,
      "grad_norm": 0.007278581149876118,
      "learning_rate": 0.01317339849991142,
      "loss": 0.4925,
      "num_input_tokens_seen": 47609072,
      "step": 34625
    },
    {
      "epoch": 1.1084437616029703,
      "grad_norm": 0.0047384933568537235,
      "learning_rate": 0.013149270093001675,
      "loss": 0.4106,
      "num_input_tokens_seen": 47615856,
      "step": 34630
    },
    {
      "epoch": 1.1086038025734588,
      "grad_norm": 0.002383928745985031,
      "learning_rate": 0.013125162790194227,
      "loss": 0.4735,
      "num_input_tokens_seen": 47622784,
      "step": 34635
    },
    {
      "epoch": 1.1087638435439473,
      "grad_norm": 0.0025475916918367147,
      "learning_rate": 0.01310107659520674,
      "loss": 0.3981,
      "num_input_tokens_seen": 47629952,
      "step": 34640
    },
    {
      "epoch": 1.1089238845144358,
      "grad_norm": 0.004494798835366964,
      "learning_rate": 0.013077011511753655,
      "loss": 0.5177,
      "num_input_tokens_seen": 47636896,
      "step": 34645
    },
    {
      "epoch": 1.109083925484924,
      "grad_norm": 0.005967758595943451,
      "learning_rate": 0.013052967543546056,
      "loss": 0.5766,
      "num_input_tokens_seen": 47644112,
      "step": 34650
    },
    {
      "epoch": 1.1092439664554126,
      "grad_norm": 0.002871428383514285,
      "learning_rate": 0.01302894469429186,
      "loss": 0.3682,
      "num_input_tokens_seen": 47651440,
      "step": 34655
    },
    {
      "epoch": 1.109404007425901,
      "grad_norm": 0.006339244544506073,
      "learning_rate": 0.013004942967695653,
      "loss": 0.6204,
      "num_input_tokens_seen": 47658704,
      "step": 34660
    },
    {
      "epoch": 1.1095640483963896,
      "grad_norm": 0.006165169645100832,
      "learning_rate": 0.012980962367458859,
      "loss": 0.3485,
      "num_input_tokens_seen": 47665792,
      "step": 34665
    },
    {
      "epoch": 1.1097240893668778,
      "grad_norm": 0.007102289702743292,
      "learning_rate": 0.012957002897279567,
      "loss": 0.5237,
      "num_input_tokens_seen": 47673264,
      "step": 34670
    },
    {
      "epoch": 1.1098841303373663,
      "grad_norm": 0.006835692562162876,
      "learning_rate": 0.012933064560852576,
      "loss": 0.5448,
      "num_input_tokens_seen": 47680336,
      "step": 34675
    },
    {
      "epoch": 1.1100441713078548,
      "grad_norm": 0.007236690726131201,
      "learning_rate": 0.012909147361869527,
      "loss": 0.6804,
      "num_input_tokens_seen": 47687536,
      "step": 34680
    },
    {
      "epoch": 1.1102042122783433,
      "grad_norm": 0.00699644023552537,
      "learning_rate": 0.012885251304018774,
      "loss": 0.4599,
      "num_input_tokens_seen": 47694624,
      "step": 34685
    },
    {
      "epoch": 1.1103642532488318,
      "grad_norm": 0.0046707140281796455,
      "learning_rate": 0.012861376390985335,
      "loss": 0.4763,
      "num_input_tokens_seen": 47701696,
      "step": 34690
    },
    {
      "epoch": 1.11052429421932,
      "grad_norm": 0.005084825213998556,
      "learning_rate": 0.012837522626451063,
      "loss": 0.537,
      "num_input_tokens_seen": 47708736,
      "step": 34695
    },
    {
      "epoch": 1.1106843351898086,
      "grad_norm": 0.007291075307875872,
      "learning_rate": 0.01281369001409447,
      "loss": 0.3845,
      "num_input_tokens_seen": 47715920,
      "step": 34700
    },
    {
      "epoch": 1.110844376160297,
      "grad_norm": 0.006376693956553936,
      "learning_rate": 0.012789878557590877,
      "loss": 0.4017,
      "num_input_tokens_seen": 47722672,
      "step": 34705
    },
    {
      "epoch": 1.1110044171307856,
      "grad_norm": 0.006528443656861782,
      "learning_rate": 0.012766088260612334,
      "loss": 0.5844,
      "num_input_tokens_seen": 47729248,
      "step": 34710
    },
    {
      "epoch": 1.1111644581012738,
      "grad_norm": 0.0031107694376260042,
      "learning_rate": 0.012742319126827523,
      "loss": 0.506,
      "num_input_tokens_seen": 47735888,
      "step": 34715
    },
    {
      "epoch": 1.1113244990717623,
      "grad_norm": 0.0036205993965268135,
      "learning_rate": 0.012718571159902008,
      "loss": 0.4964,
      "num_input_tokens_seen": 47742768,
      "step": 34720
    },
    {
      "epoch": 1.1114845400422508,
      "grad_norm": 0.005526447668671608,
      "learning_rate": 0.01269484436349803,
      "loss": 0.6656,
      "num_input_tokens_seen": 47749888,
      "step": 34725
    },
    {
      "epoch": 1.1116445810127393,
      "grad_norm": 0.004012514371424913,
      "learning_rate": 0.012671138741274528,
      "loss": 0.389,
      "num_input_tokens_seen": 47756608,
      "step": 34730
    },
    {
      "epoch": 1.1118046219832276,
      "grad_norm": 0.006667267996817827,
      "learning_rate": 0.012647454296887194,
      "loss": 0.3875,
      "num_input_tokens_seen": 47763712,
      "step": 34735
    },
    {
      "epoch": 1.111964662953716,
      "grad_norm": 0.008901041932404041,
      "learning_rate": 0.012623791033988507,
      "loss": 0.4908,
      "num_input_tokens_seen": 47770624,
      "step": 34740
    },
    {
      "epoch": 1.1121247039242046,
      "grad_norm": 0.006587791256606579,
      "learning_rate": 0.012600148956227597,
      "loss": 0.3702,
      "num_input_tokens_seen": 47777312,
      "step": 34745
    },
    {
      "epoch": 1.112284744894693,
      "grad_norm": 0.005488347262144089,
      "learning_rate": 0.012576528067250414,
      "loss": 0.4326,
      "num_input_tokens_seen": 47784176,
      "step": 34750
    },
    {
      "epoch": 1.1124447858651816,
      "grad_norm": 0.005210684612393379,
      "learning_rate": 0.012552928370699561,
      "loss": 0.3649,
      "num_input_tokens_seen": 47790944,
      "step": 34755
    },
    {
      "epoch": 1.1126048268356699,
      "grad_norm": 0.0016225893050432205,
      "learning_rate": 0.012529349870214411,
      "loss": 0.3667,
      "num_input_tokens_seen": 47798192,
      "step": 34760
    },
    {
      "epoch": 1.1127648678061584,
      "grad_norm": 0.0030015353113412857,
      "learning_rate": 0.012505792569431106,
      "loss": 0.3453,
      "num_input_tokens_seen": 47805168,
      "step": 34765
    },
    {
      "epoch": 1.1129249087766468,
      "grad_norm": 0.009078576229512691,
      "learning_rate": 0.012482256471982422,
      "loss": 0.6375,
      "num_input_tokens_seen": 47812080,
      "step": 34770
    },
    {
      "epoch": 1.1130849497471353,
      "grad_norm": 0.0035730344243347645,
      "learning_rate": 0.012458741581497956,
      "loss": 0.3594,
      "num_input_tokens_seen": 47818752,
      "step": 34775
    },
    {
      "epoch": 1.1132449907176236,
      "grad_norm": 0.004656337667256594,
      "learning_rate": 0.012435247901603974,
      "loss": 0.5067,
      "num_input_tokens_seen": 47825808,
      "step": 34780
    },
    {
      "epoch": 1.1134050316881121,
      "grad_norm": 0.0022171619348227978,
      "learning_rate": 0.012411775435923528,
      "loss": 0.4267,
      "num_input_tokens_seen": 47832736,
      "step": 34785
    },
    {
      "epoch": 1.1135650726586006,
      "grad_norm": 0.00617660116404295,
      "learning_rate": 0.012388324188076354,
      "loss": 0.6077,
      "num_input_tokens_seen": 47839840,
      "step": 34790
    },
    {
      "epoch": 1.113725113629089,
      "grad_norm": 0.003971413243561983,
      "learning_rate": 0.012364894161678913,
      "loss": 0.4764,
      "num_input_tokens_seen": 47846784,
      "step": 34795
    },
    {
      "epoch": 1.1138851545995774,
      "grad_norm": 0.006209602579474449,
      "learning_rate": 0.012341485360344445,
      "loss": 0.7575,
      "num_input_tokens_seen": 47853888,
      "step": 34800
    },
    {
      "epoch": 1.1138851545995774,
      "eval_loss": 0.4967350363731384,
      "eval_runtime": 332.5743,
      "eval_samples_per_second": 41.753,
      "eval_steps_per_second": 20.877,
      "num_input_tokens_seen": 47853888,
      "step": 34800
    },
    {
      "epoch": 1.1140451955700659,
      "grad_norm": 0.006388406734913588,
      "learning_rate": 0.01231809778768283,
      "loss": 0.4938,
      "num_input_tokens_seen": 47860848,
      "step": 34805
    },
    {
      "epoch": 1.1142052365405544,
      "grad_norm": 0.002810681238770485,
      "learning_rate": 0.012294731447300799,
      "loss": 0.3942,
      "num_input_tokens_seen": 47867968,
      "step": 34810
    },
    {
      "epoch": 1.1143652775110429,
      "grad_norm": 0.003456980222836137,
      "learning_rate": 0.012271386342801671,
      "loss": 0.6023,
      "num_input_tokens_seen": 47875008,
      "step": 34815
    },
    {
      "epoch": 1.1145253184815314,
      "grad_norm": 0.0031651409808546305,
      "learning_rate": 0.012248062477785565,
      "loss": 0.6054,
      "num_input_tokens_seen": 47882528,
      "step": 34820
    },
    {
      "epoch": 1.1146853594520196,
      "grad_norm": 0.0027446888852864504,
      "learning_rate": 0.012224759855849305,
      "loss": 0.4067,
      "num_input_tokens_seen": 47889088,
      "step": 34825
    },
    {
      "epoch": 1.1148454004225081,
      "grad_norm": 0.0023706061765551567,
      "learning_rate": 0.012201478480586513,
      "loss": 0.3978,
      "num_input_tokens_seen": 47896144,
      "step": 34830
    },
    {
      "epoch": 1.1150054413929966,
      "grad_norm": 0.009126980789005756,
      "learning_rate": 0.012178218355587389,
      "loss": 0.5899,
      "num_input_tokens_seen": 47903280,
      "step": 34835
    },
    {
      "epoch": 1.1151654823634851,
      "grad_norm": 0.0025365150067955256,
      "learning_rate": 0.01215497948443896,
      "loss": 0.4048,
      "num_input_tokens_seen": 47909952,
      "step": 34840
    },
    {
      "epoch": 1.1153255233339734,
      "grad_norm": 0.004058655817061663,
      "learning_rate": 0.012131761870724993,
      "loss": 0.6433,
      "num_input_tokens_seen": 47916816,
      "step": 34845
    },
    {
      "epoch": 1.1154855643044619,
      "grad_norm": 0.006296806503087282,
      "learning_rate": 0.012108565518025893,
      "loss": 0.5209,
      "num_input_tokens_seen": 47924192,
      "step": 34850
    },
    {
      "epoch": 1.1156456052749504,
      "grad_norm": 0.0056610568426549435,
      "learning_rate": 0.012085390429918862,
      "loss": 0.6429,
      "num_input_tokens_seen": 47931088,
      "step": 34855
    },
    {
      "epoch": 1.1158056462454389,
      "grad_norm": 0.0041609290055930614,
      "learning_rate": 0.012062236609977744,
      "loss": 0.4981,
      "num_input_tokens_seen": 47938416,
      "step": 34860
    },
    {
      "epoch": 1.1159656872159274,
      "grad_norm": 0.007099317852407694,
      "learning_rate": 0.01203910406177318,
      "loss": 0.5209,
      "num_input_tokens_seen": 47945168,
      "step": 34865
    },
    {
      "epoch": 1.1161257281864156,
      "grad_norm": 0.0060038380324840546,
      "learning_rate": 0.01201599278887252,
      "loss": 0.4394,
      "num_input_tokens_seen": 47952192,
      "step": 34870
    },
    {
      "epoch": 1.1162857691569041,
      "grad_norm": 0.0037921417970210314,
      "learning_rate": 0.011992902794839744,
      "loss": 0.3383,
      "num_input_tokens_seen": 47959296,
      "step": 34875
    },
    {
      "epoch": 1.1164458101273926,
      "grad_norm": 0.004088313784450293,
      "learning_rate": 0.011969834083235703,
      "loss": 0.4847,
      "num_input_tokens_seen": 47965936,
      "step": 34880
    },
    {
      "epoch": 1.1166058510978811,
      "grad_norm": 0.010645509697496891,
      "learning_rate": 0.011946786657617836,
      "loss": 0.4659,
      "num_input_tokens_seen": 47973248,
      "step": 34885
    },
    {
      "epoch": 1.1167658920683694,
      "grad_norm": 0.006852872669696808,
      "learning_rate": 0.011923760521540332,
      "loss": 0.649,
      "num_input_tokens_seen": 47979968,
      "step": 34890
    },
    {
      "epoch": 1.116925933038858,
      "grad_norm": 0.009064104408025742,
      "learning_rate": 0.011900755678554153,
      "loss": 0.5471,
      "num_input_tokens_seen": 47986672,
      "step": 34895
    },
    {
      "epoch": 1.1170859740093464,
      "grad_norm": 0.005368274636566639,
      "learning_rate": 0.011877772132206893,
      "loss": 0.5125,
      "num_input_tokens_seen": 47993840,
      "step": 34900
    },
    {
      "epoch": 1.117246014979835,
      "grad_norm": 0.0038959637749940157,
      "learning_rate": 0.011854809886042915,
      "loss": 0.3984,
      "num_input_tokens_seen": 48000256,
      "step": 34905
    },
    {
      "epoch": 1.1174060559503234,
      "grad_norm": 0.004808226600289345,
      "learning_rate": 0.011831868943603325,
      "loss": 0.6418,
      "num_input_tokens_seen": 48006944,
      "step": 34910
    },
    {
      "epoch": 1.1175660969208117,
      "grad_norm": 0.0045028068125247955,
      "learning_rate": 0.011808949308425836,
      "loss": 0.4779,
      "num_input_tokens_seen": 48014064,
      "step": 34915
    },
    {
      "epoch": 1.1177261378913002,
      "grad_norm": 0.005580935161560774,
      "learning_rate": 0.01178605098404501,
      "loss": 0.5394,
      "num_input_tokens_seen": 48020704,
      "step": 34920
    },
    {
      "epoch": 1.1178861788617886,
      "grad_norm": 0.0035062748938798904,
      "learning_rate": 0.011763173973992002,
      "loss": 0.5657,
      "num_input_tokens_seen": 48028112,
      "step": 34925
    },
    {
      "epoch": 1.1180462198322771,
      "grad_norm": 0.00302502210251987,
      "learning_rate": 0.011740318281794776,
      "loss": 0.3034,
      "num_input_tokens_seen": 48034912,
      "step": 34930
    },
    {
      "epoch": 1.1182062608027654,
      "grad_norm": 0.004934550262987614,
      "learning_rate": 0.01171748391097796,
      "loss": 0.4839,
      "num_input_tokens_seen": 48041648,
      "step": 34935
    },
    {
      "epoch": 1.118366301773254,
      "grad_norm": 0.004142760299146175,
      "learning_rate": 0.011694670865062873,
      "loss": 0.4891,
      "num_input_tokens_seen": 48048016,
      "step": 34940
    },
    {
      "epoch": 1.1185263427437424,
      "grad_norm": 0.0027160390745848417,
      "learning_rate": 0.011671879147567616,
      "loss": 0.421,
      "num_input_tokens_seen": 48054864,
      "step": 34945
    },
    {
      "epoch": 1.118686383714231,
      "grad_norm": 0.003927722107619047,
      "learning_rate": 0.011649108762006893,
      "loss": 0.4609,
      "num_input_tokens_seen": 48061392,
      "step": 34950
    },
    {
      "epoch": 1.1188464246847194,
      "grad_norm": 0.008076130412518978,
      "learning_rate": 0.011626359711892265,
      "loss": 0.4645,
      "num_input_tokens_seen": 48068240,
      "step": 34955
    },
    {
      "epoch": 1.1190064656552077,
      "grad_norm": 0.004010592587292194,
      "learning_rate": 0.01160363200073189,
      "loss": 0.3282,
      "num_input_tokens_seen": 48075200,
      "step": 34960
    },
    {
      "epoch": 1.1191665066256962,
      "grad_norm": 0.004792303778231144,
      "learning_rate": 0.011580925632030614,
      "loss": 0.4947,
      "num_input_tokens_seen": 48081776,
      "step": 34965
    },
    {
      "epoch": 1.1193265475961847,
      "grad_norm": 0.0048374212346971035,
      "learning_rate": 0.011558240609290104,
      "loss": 0.5961,
      "num_input_tokens_seen": 48088544,
      "step": 34970
    },
    {
      "epoch": 1.1194865885666732,
      "grad_norm": 0.003087596734985709,
      "learning_rate": 0.011535576936008679,
      "loss": 0.4258,
      "num_input_tokens_seen": 48095680,
      "step": 34975
    },
    {
      "epoch": 1.1196466295371614,
      "grad_norm": 0.003389539662748575,
      "learning_rate": 0.011512934615681309,
      "loss": 0.3843,
      "num_input_tokens_seen": 48102272,
      "step": 34980
    },
    {
      "epoch": 1.11980667050765,
      "grad_norm": 0.004725723993033171,
      "learning_rate": 0.011490313651799765,
      "loss": 0.4124,
      "num_input_tokens_seen": 48109200,
      "step": 34985
    },
    {
      "epoch": 1.1199667114781384,
      "grad_norm": 0.005685134790837765,
      "learning_rate": 0.011467714047852512,
      "loss": 0.4429,
      "num_input_tokens_seen": 48116528,
      "step": 34990
    },
    {
      "epoch": 1.120126752448627,
      "grad_norm": 0.0030685134697705507,
      "learning_rate": 0.011445135807324624,
      "loss": 0.4074,
      "num_input_tokens_seen": 48123296,
      "step": 34995
    },
    {
      "epoch": 1.1202867934191152,
      "grad_norm": 0.006290927529335022,
      "learning_rate": 0.011422578933698002,
      "loss": 0.3661,
      "num_input_tokens_seen": 48129792,
      "step": 35000
    },
    {
      "epoch": 1.1202867934191152,
      "eval_loss": 0.49555814266204834,
      "eval_runtime": 331.9591,
      "eval_samples_per_second": 41.83,
      "eval_steps_per_second": 20.915,
      "num_input_tokens_seen": 48129792,
      "step": 35000
    },
    {
      "epoch": 1.1204468343896037,
      "grad_norm": 0.0065249730832874775,
      "learning_rate": 0.011400043430451161,
      "loss": 0.4854,
      "num_input_tokens_seen": 48137376,
      "step": 35005
    },
    {
      "epoch": 1.1206068753600922,
      "grad_norm": 0.0030266724061220884,
      "learning_rate": 0.011377529301059392,
      "loss": 0.6083,
      "num_input_tokens_seen": 48144016,
      "step": 35010
    },
    {
      "epoch": 1.1207669163305807,
      "grad_norm": 0.008696246892213821,
      "learning_rate": 0.011355036548994646,
      "loss": 0.5077,
      "num_input_tokens_seen": 48150864,
      "step": 35015
    },
    {
      "epoch": 1.1209269573010692,
      "grad_norm": 0.003933447413146496,
      "learning_rate": 0.011332565177725584,
      "loss": 0.4292,
      "num_input_tokens_seen": 48157520,
      "step": 35020
    },
    {
      "epoch": 1.1210869982715574,
      "grad_norm": 0.005215534940361977,
      "learning_rate": 0.011310115190717585,
      "loss": 0.3599,
      "num_input_tokens_seen": 48164416,
      "step": 35025
    },
    {
      "epoch": 1.121247039242046,
      "grad_norm": 0.005581690464168787,
      "learning_rate": 0.01128768659143271,
      "loss": 0.5082,
      "num_input_tokens_seen": 48171680,
      "step": 35030
    },
    {
      "epoch": 1.1214070802125344,
      "grad_norm": 0.00650504045188427,
      "learning_rate": 0.011265279383329713,
      "loss": 0.4567,
      "num_input_tokens_seen": 48178400,
      "step": 35035
    },
    {
      "epoch": 1.121567121183023,
      "grad_norm": 0.006137509364634752,
      "learning_rate": 0.01124289356986411,
      "loss": 0.6211,
      "num_input_tokens_seen": 48185280,
      "step": 35040
    },
    {
      "epoch": 1.1217271621535112,
      "grad_norm": 0.005041657481342554,
      "learning_rate": 0.011220529154488023,
      "loss": 0.5536,
      "num_input_tokens_seen": 48192256,
      "step": 35045
    },
    {
      "epoch": 1.1218872031239997,
      "grad_norm": 0.0037345015443861485,
      "learning_rate": 0.011198186140650346,
      "loss": 0.6067,
      "num_input_tokens_seen": 48199248,
      "step": 35050
    },
    {
      "epoch": 1.1220472440944882,
      "grad_norm": 0.0034465952776372433,
      "learning_rate": 0.011175864531796685,
      "loss": 0.432,
      "num_input_tokens_seen": 48206160,
      "step": 35055
    },
    {
      "epoch": 1.1222072850649767,
      "grad_norm": 0.003733027260750532,
      "learning_rate": 0.011153564331369258,
      "loss": 0.4899,
      "num_input_tokens_seen": 48213472,
      "step": 35060
    },
    {
      "epoch": 1.122367326035465,
      "grad_norm": 0.006488953251391649,
      "learning_rate": 0.011131285542807078,
      "loss": 0.3333,
      "num_input_tokens_seen": 48219872,
      "step": 35065
    },
    {
      "epoch": 1.1225273670059535,
      "grad_norm": 0.00872192345559597,
      "learning_rate": 0.011109028169545815,
      "loss": 0.4274,
      "num_input_tokens_seen": 48226608,
      "step": 35070
    },
    {
      "epoch": 1.122687407976442,
      "grad_norm": 0.005541889928281307,
      "learning_rate": 0.011086792215017804,
      "loss": 0.3835,
      "num_input_tokens_seen": 48233248,
      "step": 35075
    },
    {
      "epoch": 1.1228474489469304,
      "grad_norm": 0.002086354186758399,
      "learning_rate": 0.011064577682652137,
      "loss": 0.4007,
      "num_input_tokens_seen": 48240544,
      "step": 35080
    },
    {
      "epoch": 1.123007489917419,
      "grad_norm": 0.005407347809523344,
      "learning_rate": 0.011042384575874559,
      "loss": 0.4754,
      "num_input_tokens_seen": 48247920,
      "step": 35085
    },
    {
      "epoch": 1.1231675308879072,
      "grad_norm": 0.003877029288560152,
      "learning_rate": 0.011020212898107512,
      "loss": 0.5712,
      "num_input_tokens_seen": 48254800,
      "step": 35090
    },
    {
      "epoch": 1.1233275718583957,
      "grad_norm": 0.006165441125631332,
      "learning_rate": 0.010998062652770197,
      "loss": 0.4057,
      "num_input_tokens_seen": 48261888,
      "step": 35095
    },
    {
      "epoch": 1.1234876128288842,
      "grad_norm": 0.007499439641833305,
      "learning_rate": 0.010975933843278428,
      "loss": 0.4345,
      "num_input_tokens_seen": 48268688,
      "step": 35100
    },
    {
      "epoch": 1.1236476537993727,
      "grad_norm": 0.006409372203052044,
      "learning_rate": 0.010953826473044714,
      "loss": 0.425,
      "num_input_tokens_seen": 48275376,
      "step": 35105
    },
    {
      "epoch": 1.123807694769861,
      "grad_norm": 0.00399366021156311,
      "learning_rate": 0.010931740545478357,
      "loss": 0.3908,
      "num_input_tokens_seen": 48281936,
      "step": 35110
    },
    {
      "epoch": 1.1239677357403495,
      "grad_norm": 0.004824942909181118,
      "learning_rate": 0.010909676063985218,
      "loss": 0.5453,
      "num_input_tokens_seen": 48288976,
      "step": 35115
    },
    {
      "epoch": 1.124127776710838,
      "grad_norm": 0.004033960402011871,
      "learning_rate": 0.010887633031967974,
      "loss": 0.4571,
      "num_input_tokens_seen": 48295648,
      "step": 35120
    },
    {
      "epoch": 1.1242878176813265,
      "grad_norm": 0.003779635764658451,
      "learning_rate": 0.01086561145282589,
      "loss": 0.4422,
      "num_input_tokens_seen": 48302016,
      "step": 35125
    },
    {
      "epoch": 1.124447858651815,
      "grad_norm": 0.013538713566958904,
      "learning_rate": 0.010843611329954983,
      "loss": 0.5979,
      "num_input_tokens_seen": 48308976,
      "step": 35130
    },
    {
      "epoch": 1.1246078996223032,
      "grad_norm": 0.0021594264544546604,
      "learning_rate": 0.010821632666747988,
      "loss": 0.4944,
      "num_input_tokens_seen": 48315856,
      "step": 35135
    },
    {
      "epoch": 1.1247679405927917,
      "grad_norm": 0.005431813653558493,
      "learning_rate": 0.010799675466594244,
      "loss": 0.5035,
      "num_input_tokens_seen": 48322832,
      "step": 35140
    },
    {
      "epoch": 1.1249279815632802,
      "grad_norm": 0.00533763924613595,
      "learning_rate": 0.010777739732879826,
      "loss": 0.6125,
      "num_input_tokens_seen": 48329360,
      "step": 35145
    },
    {
      "epoch": 1.1250880225337687,
      "grad_norm": 0.002612403128296137,
      "learning_rate": 0.010755825468987562,
      "loss": 0.4625,
      "num_input_tokens_seen": 48336368,
      "step": 35150
    },
    {
      "epoch": 1.125248063504257,
      "grad_norm": 0.005487783811986446,
      "learning_rate": 0.010733932678296814,
      "loss": 0.6623,
      "num_input_tokens_seen": 48343136,
      "step": 35155
    },
    {
      "epoch": 1.1254081044747455,
      "grad_norm": 0.004527638200670481,
      "learning_rate": 0.010712061364183817,
      "loss": 0.5515,
      "num_input_tokens_seen": 48349872,
      "step": 35160
    },
    {
      "epoch": 1.125568145445234,
      "grad_norm": 0.004126188810914755,
      "learning_rate": 0.010690211530021337,
      "loss": 0.3491,
      "num_input_tokens_seen": 48356928,
      "step": 35165
    },
    {
      "epoch": 1.1257281864157225,
      "grad_norm": 0.005329851992428303,
      "learning_rate": 0.01066838317917893,
      "loss": 0.6471,
      "num_input_tokens_seen": 48363728,
      "step": 35170
    },
    {
      "epoch": 1.125888227386211,
      "grad_norm": 0.005638018250465393,
      "learning_rate": 0.010646576315022787,
      "loss": 0.4535,
      "num_input_tokens_seen": 48370496,
      "step": 35175
    },
    {
      "epoch": 1.1260482683566992,
      "grad_norm": 0.004161605145782232,
      "learning_rate": 0.010624790940915785,
      "loss": 0.4618,
      "num_input_tokens_seen": 48377136,
      "step": 35180
    },
    {
      "epoch": 1.1262083093271877,
      "grad_norm": 0.002996673109009862,
      "learning_rate": 0.0106030270602175,
      "loss": 0.3926,
      "num_input_tokens_seen": 48384176,
      "step": 35185
    },
    {
      "epoch": 1.1263683502976762,
      "grad_norm": 0.00442452123388648,
      "learning_rate": 0.010581284676284252,
      "loss": 0.4925,
      "num_input_tokens_seen": 48391072,
      "step": 35190
    },
    {
      "epoch": 1.1265283912681647,
      "grad_norm": 0.00481450418010354,
      "learning_rate": 0.010559563792468923,
      "loss": 0.4806,
      "num_input_tokens_seen": 48398064,
      "step": 35195
    },
    {
      "epoch": 1.126688432238653,
      "grad_norm": 0.004983024671673775,
      "learning_rate": 0.010537864412121217,
      "loss": 0.5163,
      "num_input_tokens_seen": 48405024,
      "step": 35200
    },
    {
      "epoch": 1.126688432238653,
      "eval_loss": 0.49560001492500305,
      "eval_runtime": 332.3811,
      "eval_samples_per_second": 41.777,
      "eval_steps_per_second": 20.889,
      "num_input_tokens_seen": 48405024,
      "step": 35200
    },
    {
      "epoch": 1.1268484732091415,
      "grad_norm": 0.00558898551389575,
      "learning_rate": 0.010516186538587357,
      "loss": 0.6054,
      "num_input_tokens_seen": 48412112,
      "step": 35205
    },
    {
      "epoch": 1.12700851417963,
      "grad_norm": 0.0030861308332532644,
      "learning_rate": 0.01049453017521042,
      "loss": 0.3462,
      "num_input_tokens_seen": 48418704,
      "step": 35210
    },
    {
      "epoch": 1.1271685551501185,
      "grad_norm": 0.0053512235172092915,
      "learning_rate": 0.010472895325330083,
      "loss": 0.4603,
      "num_input_tokens_seen": 48424992,
      "step": 35215
    },
    {
      "epoch": 1.127328596120607,
      "grad_norm": 0.004028015304356813,
      "learning_rate": 0.010451281992282662,
      "loss": 0.4728,
      "num_input_tokens_seen": 48431552,
      "step": 35220
    },
    {
      "epoch": 1.1274886370910953,
      "grad_norm": 0.002501525217667222,
      "learning_rate": 0.01042969017940124,
      "loss": 0.4025,
      "num_input_tokens_seen": 48438096,
      "step": 35225
    },
    {
      "epoch": 1.1276486780615838,
      "grad_norm": 0.0031953982543200254,
      "learning_rate": 0.01040811989001557,
      "loss": 0.4237,
      "num_input_tokens_seen": 48444976,
      "step": 35230
    },
    {
      "epoch": 1.1278087190320722,
      "grad_norm": 0.0032332446426153183,
      "learning_rate": 0.010386571127451992,
      "loss": 0.4949,
      "num_input_tokens_seen": 48451568,
      "step": 35235
    },
    {
      "epoch": 1.1279687600025607,
      "grad_norm": 0.00264010694809258,
      "learning_rate": 0.010365043895033682,
      "loss": 0.3639,
      "num_input_tokens_seen": 48458224,
      "step": 35240
    },
    {
      "epoch": 1.128128800973049,
      "grad_norm": 0.002913682721555233,
      "learning_rate": 0.010343538196080365,
      "loss": 0.4758,
      "num_input_tokens_seen": 48464800,
      "step": 35245
    },
    {
      "epoch": 1.1282888419435375,
      "grad_norm": 0.003127164440229535,
      "learning_rate": 0.010322054033908457,
      "loss": 0.5548,
      "num_input_tokens_seen": 48471248,
      "step": 35250
    },
    {
      "epoch": 1.128448882914026,
      "grad_norm": 0.006904010660946369,
      "learning_rate": 0.010300591411831156,
      "loss": 0.561,
      "num_input_tokens_seen": 48478560,
      "step": 35255
    },
    {
      "epoch": 1.1286089238845145,
      "grad_norm": 0.003179936669766903,
      "learning_rate": 0.010279150333158198,
      "loss": 0.4862,
      "num_input_tokens_seen": 48485264,
      "step": 35260
    },
    {
      "epoch": 1.128768964855003,
      "grad_norm": 0.005084563512355089,
      "learning_rate": 0.010257730801196107,
      "loss": 0.4487,
      "num_input_tokens_seen": 48492208,
      "step": 35265
    },
    {
      "epoch": 1.1289290058254913,
      "grad_norm": 0.003928446210920811,
      "learning_rate": 0.010236332819248056,
      "loss": 0.5162,
      "num_input_tokens_seen": 48499456,
      "step": 35270
    },
    {
      "epoch": 1.1290890467959798,
      "grad_norm": 0.006152627989649773,
      "learning_rate": 0.010214956390613854,
      "loss": 0.706,
      "num_input_tokens_seen": 48506464,
      "step": 35275
    },
    {
      "epoch": 1.1292490877664683,
      "grad_norm": 0.007377176079899073,
      "learning_rate": 0.010193601518590034,
      "loss": 0.6233,
      "num_input_tokens_seen": 48513040,
      "step": 35280
    },
    {
      "epoch": 1.1294091287369568,
      "grad_norm": 0.004592000972479582,
      "learning_rate": 0.010172268206469758,
      "loss": 0.5163,
      "num_input_tokens_seen": 48520032,
      "step": 35285
    },
    {
      "epoch": 1.129569169707445,
      "grad_norm": 0.012983410619199276,
      "learning_rate": 0.010150956457542897,
      "loss": 0.5424,
      "num_input_tokens_seen": 48526832,
      "step": 35290
    },
    {
      "epoch": 1.1297292106779335,
      "grad_norm": 0.005496793892234564,
      "learning_rate": 0.010129666275096054,
      "loss": 0.4456,
      "num_input_tokens_seen": 48533824,
      "step": 35295
    },
    {
      "epoch": 1.129889251648422,
      "grad_norm": 0.004772995598614216,
      "learning_rate": 0.010108397662412338,
      "loss": 0.537,
      "num_input_tokens_seen": 48541200,
      "step": 35300
    },
    {
      "epoch": 1.1300492926189105,
      "grad_norm": 0.0029099106322973967,
      "learning_rate": 0.010087150622771707,
      "loss": 0.4295,
      "num_input_tokens_seen": 48547744,
      "step": 35305
    },
    {
      "epoch": 1.1302093335893988,
      "grad_norm": 0.003640561830252409,
      "learning_rate": 0.010065925159450739,
      "loss": 0.287,
      "num_input_tokens_seen": 48554416,
      "step": 35310
    },
    {
      "epoch": 1.1303693745598873,
      "grad_norm": 0.00403902493417263,
      "learning_rate": 0.010044721275722618,
      "loss": 0.5058,
      "num_input_tokens_seen": 48561168,
      "step": 35315
    },
    {
      "epoch": 1.1305294155303758,
      "grad_norm": 0.00622000964358449,
      "learning_rate": 0.01002353897485726,
      "loss": 0.6401,
      "num_input_tokens_seen": 48568160,
      "step": 35320
    },
    {
      "epoch": 1.1306894565008643,
      "grad_norm": 0.003462266642600298,
      "learning_rate": 0.010002378260121236,
      "loss": 0.4489,
      "num_input_tokens_seen": 48575264,
      "step": 35325
    },
    {
      "epoch": 1.1308494974713525,
      "grad_norm": 0.005886807106435299,
      "learning_rate": 0.009981239134777786,
      "loss": 0.4584,
      "num_input_tokens_seen": 48582128,
      "step": 35330
    },
    {
      "epoch": 1.131009538441841,
      "grad_norm": 0.004950274247676134,
      "learning_rate": 0.009960121602086884,
      "loss": 0.4701,
      "num_input_tokens_seen": 48589296,
      "step": 35335
    },
    {
      "epoch": 1.1311695794123295,
      "grad_norm": 0.003445202950388193,
      "learning_rate": 0.009939025665305062,
      "loss": 0.5801,
      "num_input_tokens_seen": 48595856,
      "step": 35340
    },
    {
      "epoch": 1.131329620382818,
      "grad_norm": 0.006278247572481632,
      "learning_rate": 0.009917951327685597,
      "loss": 0.608,
      "num_input_tokens_seen": 48603088,
      "step": 35345
    },
    {
      "epoch": 1.1314896613533065,
      "grad_norm": 0.007731104269623756,
      "learning_rate": 0.009896898592478425,
      "loss": 0.7546,
      "num_input_tokens_seen": 48610512,
      "step": 35350
    },
    {
      "epoch": 1.1316497023237948,
      "grad_norm": 0.003196925623342395,
      "learning_rate": 0.009875867462930132,
      "loss": 0.4416,
      "num_input_tokens_seen": 48617280,
      "step": 35355
    },
    {
      "epoch": 1.1318097432942833,
      "grad_norm": 0.004949409980326891,
      "learning_rate": 0.009854857942284006,
      "loss": 0.4784,
      "num_input_tokens_seen": 48624384,
      "step": 35360
    },
    {
      "epoch": 1.1319697842647718,
      "grad_norm": 0.0034026955254375935,
      "learning_rate": 0.009833870033779923,
      "loss": 0.5125,
      "num_input_tokens_seen": 48631456,
      "step": 35365
    },
    {
      "epoch": 1.1321298252352603,
      "grad_norm": 0.007087767589837313,
      "learning_rate": 0.009812903740654527,
      "loss": 0.4751,
      "num_input_tokens_seen": 48638432,
      "step": 35370
    },
    {
      "epoch": 1.1322898662057486,
      "grad_norm": 0.0029207677580416203,
      "learning_rate": 0.009791959066141097,
      "loss": 0.2821,
      "num_input_tokens_seen": 48645056,
      "step": 35375
    },
    {
      "epoch": 1.132449907176237,
      "grad_norm": 0.004646902438253164,
      "learning_rate": 0.009771036013469537,
      "loss": 0.3927,
      "num_input_tokens_seen": 48651728,
      "step": 35380
    },
    {
      "epoch": 1.1326099481467256,
      "grad_norm": 0.00688741123303771,
      "learning_rate": 0.00975013458586646,
      "loss": 0.6035,
      "num_input_tokens_seen": 48658064,
      "step": 35385
    },
    {
      "epoch": 1.132769989117214,
      "grad_norm": 0.005057658534497023,
      "learning_rate": 0.009729254786555107,
      "loss": 0.7348,
      "num_input_tokens_seen": 48664752,
      "step": 35390
    },
    {
      "epoch": 1.1329300300877025,
      "grad_norm": 0.007386889774352312,
      "learning_rate": 0.009708396618755421,
      "loss": 0.5889,
      "num_input_tokens_seen": 48671824,
      "step": 35395
    },
    {
      "epoch": 1.1330900710581908,
      "grad_norm": 0.004695496056228876,
      "learning_rate": 0.009687560085683994,
      "loss": 0.3735,
      "num_input_tokens_seen": 48678592,
      "step": 35400
    },
    {
      "epoch": 1.1330900710581908,
      "eval_loss": 0.4951114058494568,
      "eval_runtime": 332.2789,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 48678592,
      "step": 35400
    },
    {
      "epoch": 1.1332501120286793,
      "grad_norm": 0.003835204755887389,
      "learning_rate": 0.009666745190554054,
      "loss": 0.3439,
      "num_input_tokens_seen": 48685184,
      "step": 35405
    },
    {
      "epoch": 1.1334101529991678,
      "grad_norm": 0.00365918199531734,
      "learning_rate": 0.009645951936575553,
      "loss": 0.496,
      "num_input_tokens_seen": 48691888,
      "step": 35410
    },
    {
      "epoch": 1.1335701939696563,
      "grad_norm": 0.004145956598222256,
      "learning_rate": 0.00962518032695509,
      "loss": 0.4455,
      "num_input_tokens_seen": 48698208,
      "step": 35415
    },
    {
      "epoch": 1.1337302349401446,
      "grad_norm": 0.003021609503775835,
      "learning_rate": 0.009604430364895855,
      "loss": 0.4988,
      "num_input_tokens_seen": 48705440,
      "step": 35420
    },
    {
      "epoch": 1.133890275910633,
      "grad_norm": 0.0063857585191726685,
      "learning_rate": 0.00958370205359777,
      "loss": 0.5825,
      "num_input_tokens_seen": 48712544,
      "step": 35425
    },
    {
      "epoch": 1.1340503168811216,
      "grad_norm": 0.004139254335314035,
      "learning_rate": 0.009562995396257445,
      "loss": 0.6075,
      "num_input_tokens_seen": 48719520,
      "step": 35430
    },
    {
      "epoch": 1.13421035785161,
      "grad_norm": 0.0047048842534422874,
      "learning_rate": 0.009542310396068026,
      "loss": 0.532,
      "num_input_tokens_seen": 48726144,
      "step": 35435
    },
    {
      "epoch": 1.1343703988220986,
      "grad_norm": 0.005951245781034231,
      "learning_rate": 0.009521647056219495,
      "loss": 0.6302,
      "num_input_tokens_seen": 48733152,
      "step": 35440
    },
    {
      "epoch": 1.1345304397925868,
      "grad_norm": 0.010308322496712208,
      "learning_rate": 0.00950100537989832,
      "loss": 0.6289,
      "num_input_tokens_seen": 48739920,
      "step": 35445
    },
    {
      "epoch": 1.1346904807630753,
      "grad_norm": 0.004138196352869272,
      "learning_rate": 0.00948038537028772,
      "loss": 0.4085,
      "num_input_tokens_seen": 48746624,
      "step": 35450
    },
    {
      "epoch": 1.1348505217335638,
      "grad_norm": 0.00492518674582243,
      "learning_rate": 0.009459787030567617,
      "loss": 0.7543,
      "num_input_tokens_seen": 48753536,
      "step": 35455
    },
    {
      "epoch": 1.1350105627040523,
      "grad_norm": 0.0031509327236562967,
      "learning_rate": 0.00943921036391449,
      "loss": 0.5804,
      "num_input_tokens_seen": 48760240,
      "step": 35460
    },
    {
      "epoch": 1.1351706036745406,
      "grad_norm": 0.0035189404152333736,
      "learning_rate": 0.009418655373501483,
      "loss": 0.5004,
      "num_input_tokens_seen": 48767024,
      "step": 35465
    },
    {
      "epoch": 1.135330644645029,
      "grad_norm": 0.005343197844922543,
      "learning_rate": 0.00939812206249851,
      "loss": 0.7749,
      "num_input_tokens_seen": 48774368,
      "step": 35470
    },
    {
      "epoch": 1.1354906856155176,
      "grad_norm": 0.004991778172552586,
      "learning_rate": 0.009377610434072004,
      "loss": 0.4622,
      "num_input_tokens_seen": 48781824,
      "step": 35475
    },
    {
      "epoch": 1.135650726586006,
      "grad_norm": 0.004775755573064089,
      "learning_rate": 0.009357120491385167,
      "loss": 0.5426,
      "num_input_tokens_seen": 48788544,
      "step": 35480
    },
    {
      "epoch": 1.1358107675564946,
      "grad_norm": 0.0052804783917963505,
      "learning_rate": 0.009336652237597743,
      "loss": 0.4814,
      "num_input_tokens_seen": 48795328,
      "step": 35485
    },
    {
      "epoch": 1.1359708085269828,
      "grad_norm": 0.0031406376510858536,
      "learning_rate": 0.009316205675866251,
      "loss": 0.3589,
      "num_input_tokens_seen": 48801840,
      "step": 35490
    },
    {
      "epoch": 1.1361308494974713,
      "grad_norm": 0.008515434339642525,
      "learning_rate": 0.00929578080934379,
      "loss": 0.5625,
      "num_input_tokens_seen": 48808752,
      "step": 35495
    },
    {
      "epoch": 1.1362908904679598,
      "grad_norm": 0.0033396112266927958,
      "learning_rate": 0.00927537764118012,
      "loss": 0.3845,
      "num_input_tokens_seen": 48815760,
      "step": 35500
    },
    {
      "epoch": 1.1364509314384483,
      "grad_norm": 0.005546616390347481,
      "learning_rate": 0.009254996174521678,
      "loss": 0.7584,
      "num_input_tokens_seen": 48822880,
      "step": 35505
    },
    {
      "epoch": 1.1366109724089366,
      "grad_norm": 0.006399017293006182,
      "learning_rate": 0.009234636412511531,
      "loss": 0.3536,
      "num_input_tokens_seen": 48829536,
      "step": 35510
    },
    {
      "epoch": 1.136771013379425,
      "grad_norm": 0.0030711547005921602,
      "learning_rate": 0.009214298358289418,
      "loss": 0.5268,
      "num_input_tokens_seen": 48836464,
      "step": 35515
    },
    {
      "epoch": 1.1369310543499136,
      "grad_norm": 0.003802296705543995,
      "learning_rate": 0.00919398201499173,
      "loss": 0.4604,
      "num_input_tokens_seen": 48843808,
      "step": 35520
    },
    {
      "epoch": 1.137091095320402,
      "grad_norm": 0.005507424473762512,
      "learning_rate": 0.009173687385751495,
      "loss": 0.3965,
      "num_input_tokens_seen": 48850320,
      "step": 35525
    },
    {
      "epoch": 1.1372511362908906,
      "grad_norm": 0.0034667784348130226,
      "learning_rate": 0.009153414473698407,
      "loss": 0.4488,
      "num_input_tokens_seen": 48856992,
      "step": 35530
    },
    {
      "epoch": 1.1374111772613789,
      "grad_norm": 0.004225895740091801,
      "learning_rate": 0.009133163281958784,
      "loss": 0.4379,
      "num_input_tokens_seen": 48863552,
      "step": 35535
    },
    {
      "epoch": 1.1375712182318674,
      "grad_norm": 0.0029377867467701435,
      "learning_rate": 0.009112933813655627,
      "loss": 0.3095,
      "num_input_tokens_seen": 48870368,
      "step": 35540
    },
    {
      "epoch": 1.1377312592023558,
      "grad_norm": 0.007109504658728838,
      "learning_rate": 0.009092726071908573,
      "loss": 0.395,
      "num_input_tokens_seen": 48877168,
      "step": 35545
    },
    {
      "epoch": 1.1378913001728443,
      "grad_norm": 0.005049775820225477,
      "learning_rate": 0.0090725400598339,
      "loss": 0.5353,
      "num_input_tokens_seen": 48883904,
      "step": 35550
    },
    {
      "epoch": 1.1380513411433326,
      "grad_norm": 0.003530661342665553,
      "learning_rate": 0.009052375780544563,
      "loss": 0.4284,
      "num_input_tokens_seen": 48890832,
      "step": 35555
    },
    {
      "epoch": 1.1382113821138211,
      "grad_norm": 0.006311892997473478,
      "learning_rate": 0.009032233237150144,
      "loss": 0.4267,
      "num_input_tokens_seen": 48897296,
      "step": 35560
    },
    {
      "epoch": 1.1383714230843096,
      "grad_norm": 0.0029975499492138624,
      "learning_rate": 0.009012112432756875,
      "loss": 0.5344,
      "num_input_tokens_seen": 48904096,
      "step": 35565
    },
    {
      "epoch": 1.138531464054798,
      "grad_norm": 0.005149828735738993,
      "learning_rate": 0.008992013370467605,
      "loss": 0.6004,
      "num_input_tokens_seen": 48911120,
      "step": 35570
    },
    {
      "epoch": 1.1386915050252864,
      "grad_norm": 0.008045009337365627,
      "learning_rate": 0.008971936053381924,
      "loss": 0.5003,
      "num_input_tokens_seen": 48918384,
      "step": 35575
    },
    {
      "epoch": 1.1388515459957749,
      "grad_norm": 0.007399325724691153,
      "learning_rate": 0.008951880484595953,
      "loss": 0.4257,
      "num_input_tokens_seen": 48926000,
      "step": 35580
    },
    {
      "epoch": 1.1390115869662634,
      "grad_norm": 0.003955399617552757,
      "learning_rate": 0.008931846667202552,
      "loss": 0.5971,
      "num_input_tokens_seen": 48933168,
      "step": 35585
    },
    {
      "epoch": 1.1391716279367519,
      "grad_norm": 0.003982385620474815,
      "learning_rate": 0.008911834604291152,
      "loss": 0.2967,
      "num_input_tokens_seen": 48939856,
      "step": 35590
    },
    {
      "epoch": 1.1393316689072401,
      "grad_norm": 0.003334322478622198,
      "learning_rate": 0.008891844298947882,
      "loss": 0.4478,
      "num_input_tokens_seen": 48946656,
      "step": 35595
    },
    {
      "epoch": 1.1394917098777286,
      "grad_norm": 0.0049562896601855755,
      "learning_rate": 0.008871875754255508,
      "loss": 0.3249,
      "num_input_tokens_seen": 48954048,
      "step": 35600
    },
    {
      "epoch": 1.1394917098777286,
      "eval_loss": 0.49483349919319153,
      "eval_runtime": 332.1469,
      "eval_samples_per_second": 41.807,
      "eval_steps_per_second": 20.903,
      "num_input_tokens_seen": 48954048,
      "step": 35600
    },
    {
      "epoch": 1.1396517508482171,
      "grad_norm": 0.004365826956927776,
      "learning_rate": 0.008851928973293422,
      "loss": 0.6106,
      "num_input_tokens_seen": 48960800,
      "step": 35605
    },
    {
      "epoch": 1.1398117918187056,
      "grad_norm": 0.0026680631563067436,
      "learning_rate": 0.00883200395913764,
      "loss": 0.3307,
      "num_input_tokens_seen": 48967968,
      "step": 35610
    },
    {
      "epoch": 1.1399718327891941,
      "grad_norm": 0.0028037249576300383,
      "learning_rate": 0.00881210071486091,
      "loss": 0.419,
      "num_input_tokens_seen": 48974576,
      "step": 35615
    },
    {
      "epoch": 1.1401318737596824,
      "grad_norm": 0.00466673681512475,
      "learning_rate": 0.008792219243532505,
      "loss": 0.5525,
      "num_input_tokens_seen": 48981424,
      "step": 35620
    },
    {
      "epoch": 1.1402919147301709,
      "grad_norm": 0.0028763804584741592,
      "learning_rate": 0.008772359548218428,
      "loss": 0.3373,
      "num_input_tokens_seen": 48988320,
      "step": 35625
    },
    {
      "epoch": 1.1404519557006594,
      "grad_norm": 0.005273839458823204,
      "learning_rate": 0.008752521631981274,
      "loss": 0.5771,
      "num_input_tokens_seen": 48995376,
      "step": 35630
    },
    {
      "epoch": 1.1406119966711479,
      "grad_norm": 0.007281958591192961,
      "learning_rate": 0.008732705497880315,
      "loss": 0.5927,
      "num_input_tokens_seen": 49002288,
      "step": 35635
    },
    {
      "epoch": 1.1407720376416361,
      "grad_norm": 0.007062505930662155,
      "learning_rate": 0.008712911148971459,
      "loss": 0.4532,
      "num_input_tokens_seen": 49009696,
      "step": 35640
    },
    {
      "epoch": 1.1409320786121246,
      "grad_norm": 0.004252405371516943,
      "learning_rate": 0.008693138588307208,
      "loss": 0.3876,
      "num_input_tokens_seen": 49016736,
      "step": 35645
    },
    {
      "epoch": 1.1410921195826131,
      "grad_norm": 0.00579685578122735,
      "learning_rate": 0.008673387818936762,
      "loss": 0.5449,
      "num_input_tokens_seen": 49024384,
      "step": 35650
    },
    {
      "epoch": 1.1412521605531016,
      "grad_norm": 0.003453951794654131,
      "learning_rate": 0.008653658843905948,
      "loss": 0.388,
      "num_input_tokens_seen": 49031552,
      "step": 35655
    },
    {
      "epoch": 1.1414122015235901,
      "grad_norm": 0.004426838830113411,
      "learning_rate": 0.0086339516662572,
      "loss": 0.4552,
      "num_input_tokens_seen": 49038528,
      "step": 35660
    },
    {
      "epoch": 1.1415722424940784,
      "grad_norm": 0.003580277319997549,
      "learning_rate": 0.008614266289029638,
      "loss": 0.3398,
      "num_input_tokens_seen": 49045904,
      "step": 35665
    },
    {
      "epoch": 1.141732283464567,
      "grad_norm": 0.005821858067065477,
      "learning_rate": 0.008594602715258965,
      "loss": 0.388,
      "num_input_tokens_seen": 49052800,
      "step": 35670
    },
    {
      "epoch": 1.1418923244350554,
      "grad_norm": 0.00487197982147336,
      "learning_rate": 0.008574960947977573,
      "loss": 0.2983,
      "num_input_tokens_seen": 49059712,
      "step": 35675
    },
    {
      "epoch": 1.142052365405544,
      "grad_norm": 0.0042257835157215595,
      "learning_rate": 0.008555340990214438,
      "loss": 0.4482,
      "num_input_tokens_seen": 49066656,
      "step": 35680
    },
    {
      "epoch": 1.1422124063760322,
      "grad_norm": 0.006164203863590956,
      "learning_rate": 0.008535742844995258,
      "loss": 0.6007,
      "num_input_tokens_seen": 49073600,
      "step": 35685
    },
    {
      "epoch": 1.1423724473465207,
      "grad_norm": 0.003127355594187975,
      "learning_rate": 0.008516166515342266,
      "loss": 0.4541,
      "num_input_tokens_seen": 49080496,
      "step": 35690
    },
    {
      "epoch": 1.1425324883170092,
      "grad_norm": 0.005853383336216211,
      "learning_rate": 0.008496612004274411,
      "loss": 0.5494,
      "num_input_tokens_seen": 49087408,
      "step": 35695
    },
    {
      "epoch": 1.1426925292874976,
      "grad_norm": 0.004047082737088203,
      "learning_rate": 0.008477079314807201,
      "loss": 0.6598,
      "num_input_tokens_seen": 49094448,
      "step": 35700
    },
    {
      "epoch": 1.1428525702579861,
      "grad_norm": 0.004687349312007427,
      "learning_rate": 0.008457568449952874,
      "loss": 0.5227,
      "num_input_tokens_seen": 49100832,
      "step": 35705
    },
    {
      "epoch": 1.1430126112284744,
      "grad_norm": 0.008439511992037296,
      "learning_rate": 0.008438079412720189,
      "loss": 0.4452,
      "num_input_tokens_seen": 49108160,
      "step": 35710
    },
    {
      "epoch": 1.143172652198963,
      "grad_norm": 0.003785319160670042,
      "learning_rate": 0.00841861220611466,
      "loss": 0.4982,
      "num_input_tokens_seen": 49114864,
      "step": 35715
    },
    {
      "epoch": 1.1433326931694514,
      "grad_norm": 0.005455794744193554,
      "learning_rate": 0.008399166833138355,
      "loss": 0.4518,
      "num_input_tokens_seen": 49121808,
      "step": 35720
    },
    {
      "epoch": 1.14349273413994,
      "grad_norm": 0.004483853466808796,
      "learning_rate": 0.008379743296789987,
      "loss": 0.4004,
      "num_input_tokens_seen": 49128944,
      "step": 35725
    },
    {
      "epoch": 1.1436527751104282,
      "grad_norm": 0.0031429564114660025,
      "learning_rate": 0.008360341600064896,
      "loss": 0.4461,
      "num_input_tokens_seen": 49135616,
      "step": 35730
    },
    {
      "epoch": 1.1438128160809167,
      "grad_norm": 0.008185423910617828,
      "learning_rate": 0.008340961745955121,
      "loss": 0.4531,
      "num_input_tokens_seen": 49142320,
      "step": 35735
    },
    {
      "epoch": 1.1439728570514052,
      "grad_norm": 0.004328944720327854,
      "learning_rate": 0.008321603737449224,
      "loss": 0.5407,
      "num_input_tokens_seen": 49149072,
      "step": 35740
    },
    {
      "epoch": 1.1441328980218937,
      "grad_norm": 0.003812620881944895,
      "learning_rate": 0.008302267577532479,
      "loss": 0.5192,
      "num_input_tokens_seen": 49156080,
      "step": 35745
    },
    {
      "epoch": 1.1442929389923822,
      "grad_norm": 0.0058805495500564575,
      "learning_rate": 0.008282953269186771,
      "loss": 0.5608,
      "num_input_tokens_seen": 49162768,
      "step": 35750
    },
    {
      "epoch": 1.1444529799628704,
      "grad_norm": 0.006566069554537535,
      "learning_rate": 0.008263660815390567,
      "loss": 0.4156,
      "num_input_tokens_seen": 49170256,
      "step": 35755
    },
    {
      "epoch": 1.144613020933359,
      "grad_norm": 0.004773540422320366,
      "learning_rate": 0.008244390219119069,
      "loss": 0.5179,
      "num_input_tokens_seen": 49176992,
      "step": 35760
    },
    {
      "epoch": 1.1447730619038474,
      "grad_norm": 0.0036939894780516624,
      "learning_rate": 0.008225141483343967,
      "loss": 0.4728,
      "num_input_tokens_seen": 49183728,
      "step": 35765
    },
    {
      "epoch": 1.144933102874336,
      "grad_norm": 0.004562904126942158,
      "learning_rate": 0.00820591461103372,
      "loss": 0.4462,
      "num_input_tokens_seen": 49190448,
      "step": 35770
    },
    {
      "epoch": 1.1450931438448242,
      "grad_norm": 0.00679488480091095,
      "learning_rate": 0.008186709605153358,
      "loss": 0.435,
      "num_input_tokens_seen": 49197536,
      "step": 35775
    },
    {
      "epoch": 1.1452531848153127,
      "grad_norm": 0.006547866854816675,
      "learning_rate": 0.008167526468664492,
      "loss": 0.4369,
      "num_input_tokens_seen": 49204720,
      "step": 35780
    },
    {
      "epoch": 1.1454132257858012,
      "grad_norm": 0.0023674475960433483,
      "learning_rate": 0.008148365204525443,
      "loss": 0.4437,
      "num_input_tokens_seen": 49211680,
      "step": 35785
    },
    {
      "epoch": 1.1455732667562897,
      "grad_norm": 0.0032252015080302954,
      "learning_rate": 0.00812922581569106,
      "loss": 0.465,
      "num_input_tokens_seen": 49218976,
      "step": 35790
    },
    {
      "epoch": 1.1457333077267782,
      "grad_norm": 0.005085574463009834,
      "learning_rate": 0.008110108305112934,
      "loss": 0.4394,
      "num_input_tokens_seen": 49225984,
      "step": 35795
    },
    {
      "epoch": 1.1458933486972664,
      "grad_norm": 0.005919540300965309,
      "learning_rate": 0.008091012675739223,
      "loss": 0.3284,
      "num_input_tokens_seen": 49232480,
      "step": 35800
    },
    {
      "epoch": 1.1458933486972664,
      "eval_loss": 0.49492040276527405,
      "eval_runtime": 332.0296,
      "eval_samples_per_second": 41.822,
      "eval_steps_per_second": 20.911,
      "num_input_tokens_seen": 49232480,
      "step": 35800
    },
    {
      "epoch": 1.146053389667755,
      "grad_norm": 0.00530350673943758,
      "learning_rate": 0.008071938930514671,
      "loss": 0.3578,
      "num_input_tokens_seen": 49239376,
      "step": 35805
    },
    {
      "epoch": 1.1462134306382434,
      "grad_norm": 0.006298502907156944,
      "learning_rate": 0.008052887072380726,
      "loss": 0.7346,
      "num_input_tokens_seen": 49246352,
      "step": 35810
    },
    {
      "epoch": 1.146373471608732,
      "grad_norm": 0.0032224273309111595,
      "learning_rate": 0.008033857104275437,
      "loss": 0.343,
      "num_input_tokens_seen": 49252768,
      "step": 35815
    },
    {
      "epoch": 1.1465335125792202,
      "grad_norm": 0.005254375748336315,
      "learning_rate": 0.008014849029133424,
      "loss": 0.5381,
      "num_input_tokens_seen": 49259472,
      "step": 35820
    },
    {
      "epoch": 1.1466935535497087,
      "grad_norm": 0.005399032030254602,
      "learning_rate": 0.007995862849885975,
      "loss": 0.5531,
      "num_input_tokens_seen": 49266176,
      "step": 35825
    },
    {
      "epoch": 1.1468535945201972,
      "grad_norm": 0.004455006215721369,
      "learning_rate": 0.007976898569461032,
      "loss": 0.4824,
      "num_input_tokens_seen": 49272864,
      "step": 35830
    },
    {
      "epoch": 1.1470136354906857,
      "grad_norm": 0.002742842771112919,
      "learning_rate": 0.007957956190783088,
      "loss": 0.3657,
      "num_input_tokens_seen": 49279680,
      "step": 35835
    },
    {
      "epoch": 1.1471736764611742,
      "grad_norm": 0.001974700950086117,
      "learning_rate": 0.007939035716773324,
      "loss": 0.4359,
      "num_input_tokens_seen": 49286576,
      "step": 35840
    },
    {
      "epoch": 1.1473337174316625,
      "grad_norm": 0.005404572933912277,
      "learning_rate": 0.007920137150349487,
      "loss": 0.4976,
      "num_input_tokens_seen": 49293376,
      "step": 35845
    },
    {
      "epoch": 1.147493758402151,
      "grad_norm": 0.006981116719543934,
      "learning_rate": 0.007901260494425981,
      "loss": 0.6546,
      "num_input_tokens_seen": 49300032,
      "step": 35850
    },
    {
      "epoch": 1.1476537993726394,
      "grad_norm": 0.0032912814058363438,
      "learning_rate": 0.007882405751913861,
      "loss": 0.4343,
      "num_input_tokens_seen": 49306848,
      "step": 35855
    },
    {
      "epoch": 1.1478138403431277,
      "grad_norm": 0.0038581627886742353,
      "learning_rate": 0.007863572925720702,
      "loss": 0.4917,
      "num_input_tokens_seen": 49313760,
      "step": 35860
    },
    {
      "epoch": 1.1479738813136162,
      "grad_norm": 0.0041815368458628654,
      "learning_rate": 0.007844762018750827,
      "loss": 0.4522,
      "num_input_tokens_seen": 49320960,
      "step": 35865
    },
    {
      "epoch": 1.1481339222841047,
      "grad_norm": 0.004106204956769943,
      "learning_rate": 0.007825973033905054,
      "loss": 0.4714,
      "num_input_tokens_seen": 49327664,
      "step": 35870
    },
    {
      "epoch": 1.1482939632545932,
      "grad_norm": 0.0063057453371584415,
      "learning_rate": 0.007807205974080927,
      "loss": 0.5116,
      "num_input_tokens_seen": 49334576,
      "step": 35875
    },
    {
      "epoch": 1.1484540042250817,
      "grad_norm": 0.003270576475188136,
      "learning_rate": 0.007788460842172551,
      "loss": 0.599,
      "num_input_tokens_seen": 49341552,
      "step": 35880
    },
    {
      "epoch": 1.14861404519557,
      "grad_norm": 0.0026485836133360863,
      "learning_rate": 0.0077697376410706285,
      "loss": 0.3366,
      "num_input_tokens_seen": 49348176,
      "step": 35885
    },
    {
      "epoch": 1.1487740861660585,
      "grad_norm": 0.005677375011146069,
      "learning_rate": 0.007751036373662567,
      "loss": 0.3684,
      "num_input_tokens_seen": 49354944,
      "step": 35890
    },
    {
      "epoch": 1.148934127136547,
      "grad_norm": 0.00496818358078599,
      "learning_rate": 0.00773235704283231,
      "loss": 0.4106,
      "num_input_tokens_seen": 49361744,
      "step": 35895
    },
    {
      "epoch": 1.1490941681070355,
      "grad_norm": 0.005227888468652964,
      "learning_rate": 0.007713699651460437,
      "loss": 0.4786,
      "num_input_tokens_seen": 49368768,
      "step": 35900
    },
    {
      "epoch": 1.1492542090775237,
      "grad_norm": 0.003126192605122924,
      "learning_rate": 0.007695064202424162,
      "loss": 0.4362,
      "num_input_tokens_seen": 49375856,
      "step": 35905
    },
    {
      "epoch": 1.1494142500480122,
      "grad_norm": 0.0055552092380821705,
      "learning_rate": 0.007676450698597286,
      "loss": 0.3975,
      "num_input_tokens_seen": 49382816,
      "step": 35910
    },
    {
      "epoch": 1.1495742910185007,
      "grad_norm": 0.0023901076056063175,
      "learning_rate": 0.007657859142850265,
      "loss": 0.3326,
      "num_input_tokens_seen": 49389808,
      "step": 35915
    },
    {
      "epoch": 1.1497343319889892,
      "grad_norm": 0.00520074600353837,
      "learning_rate": 0.0076392895380501535,
      "loss": 0.524,
      "num_input_tokens_seen": 49396496,
      "step": 35920
    },
    {
      "epoch": 1.1498943729594777,
      "grad_norm": 0.0050339773297309875,
      "learning_rate": 0.007620741887060611,
      "loss": 0.4504,
      "num_input_tokens_seen": 49403152,
      "step": 35925
    },
    {
      "epoch": 1.150054413929966,
      "grad_norm": 0.007176522631198168,
      "learning_rate": 0.007602216192741901,
      "loss": 0.4432,
      "num_input_tokens_seen": 49409728,
      "step": 35930
    },
    {
      "epoch": 1.1502144549004545,
      "grad_norm": 0.0030684347730129957,
      "learning_rate": 0.007583712457950969,
      "loss": 0.3442,
      "num_input_tokens_seen": 49416704,
      "step": 35935
    },
    {
      "epoch": 1.150374495870943,
      "grad_norm": 0.003803878789767623,
      "learning_rate": 0.007565230685541269,
      "loss": 0.2641,
      "num_input_tokens_seen": 49423344,
      "step": 35940
    },
    {
      "epoch": 1.1505345368414315,
      "grad_norm": 0.0030132909305393696,
      "learning_rate": 0.007546770878362968,
      "loss": 0.5275,
      "num_input_tokens_seen": 49429920,
      "step": 35945
    },
    {
      "epoch": 1.1506945778119197,
      "grad_norm": 0.0036366942804306746,
      "learning_rate": 0.0075283330392627405,
      "loss": 0.5021,
      "num_input_tokens_seen": 49436512,
      "step": 35950
    },
    {
      "epoch": 1.1508546187824082,
      "grad_norm": 0.005811620038002729,
      "learning_rate": 0.007509917171083979,
      "loss": 0.4568,
      "num_input_tokens_seen": 49443376,
      "step": 35955
    },
    {
      "epoch": 1.1510146597528967,
      "grad_norm": 0.004804416559636593,
      "learning_rate": 0.007491523276666662,
      "loss": 0.3485,
      "num_input_tokens_seen": 49450160,
      "step": 35960
    },
    {
      "epoch": 1.1511747007233852,
      "grad_norm": 0.0033156173303723335,
      "learning_rate": 0.007473151358847318,
      "loss": 0.3974,
      "num_input_tokens_seen": 49457088,
      "step": 35965
    },
    {
      "epoch": 1.1513347416938737,
      "grad_norm": 0.005051989108324051,
      "learning_rate": 0.007454801420459117,
      "loss": 0.365,
      "num_input_tokens_seen": 49463856,
      "step": 35970
    },
    {
      "epoch": 1.151494782664362,
      "grad_norm": 0.006222884636372328,
      "learning_rate": 0.0074364734643319105,
      "loss": 0.4879,
      "num_input_tokens_seen": 49470544,
      "step": 35975
    },
    {
      "epoch": 1.1516548236348505,
      "grad_norm": 0.0061817350797355175,
      "learning_rate": 0.007418167493292022,
      "loss": 0.4719,
      "num_input_tokens_seen": 49477392,
      "step": 35980
    },
    {
      "epoch": 1.151814864605339,
      "grad_norm": 0.006528633646667004,
      "learning_rate": 0.0073998835101625245,
      "loss": 0.3234,
      "num_input_tokens_seen": 49484240,
      "step": 35985
    },
    {
      "epoch": 1.1519749055758275,
      "grad_norm": 0.0033503160811960697,
      "learning_rate": 0.007381621517762998,
      "loss": 0.6021,
      "num_input_tokens_seen": 49491360,
      "step": 35990
    },
    {
      "epoch": 1.1521349465463158,
      "grad_norm": 0.005164242349565029,
      "learning_rate": 0.007363381518909689,
      "loss": 0.3407,
      "num_input_tokens_seen": 49498432,
      "step": 35995
    },
    {
      "epoch": 1.1522949875168043,
      "grad_norm": 0.0038019560743123293,
      "learning_rate": 0.007345163516415448,
      "loss": 0.6086,
      "num_input_tokens_seen": 49505040,
      "step": 36000
    },
    {
      "epoch": 1.1522949875168043,
      "eval_loss": 0.49462342262268066,
      "eval_runtime": 331.7497,
      "eval_samples_per_second": 41.857,
      "eval_steps_per_second": 20.928,
      "num_input_tokens_seen": 49505040,
      "step": 36000
    },
    {
      "epoch": 1.1524550284872928,
      "grad_norm": 0.006080867722630501,
      "learning_rate": 0.007326967513089693,
      "loss": 0.4018,
      "num_input_tokens_seen": 49511744,
      "step": 36005
    },
    {
      "epoch": 1.1526150694577812,
      "grad_norm": 0.004734555259346962,
      "learning_rate": 0.0073087935117384815,
      "loss": 0.6019,
      "num_input_tokens_seen": 49518832,
      "step": 36010
    },
    {
      "epoch": 1.1527751104282697,
      "grad_norm": 0.005272861570119858,
      "learning_rate": 0.007290641515164503,
      "loss": 0.3557,
      "num_input_tokens_seen": 49526416,
      "step": 36015
    },
    {
      "epoch": 1.152935151398758,
      "grad_norm": 0.006960590369999409,
      "learning_rate": 0.007272511526166986,
      "loss": 0.4328,
      "num_input_tokens_seen": 49533152,
      "step": 36020
    },
    {
      "epoch": 1.1530951923692465,
      "grad_norm": 0.0038200884591788054,
      "learning_rate": 0.0072544035475418265,
      "loss": 0.5333,
      "num_input_tokens_seen": 49539808,
      "step": 36025
    },
    {
      "epoch": 1.153255233339735,
      "grad_norm": 0.007526843808591366,
      "learning_rate": 0.007236317582081475,
      "loss": 0.5096,
      "num_input_tokens_seen": 49546496,
      "step": 36030
    },
    {
      "epoch": 1.1534152743102235,
      "grad_norm": 0.004504538141191006,
      "learning_rate": 0.007218253632575066,
      "loss": 0.4376,
      "num_input_tokens_seen": 49553440,
      "step": 36035
    },
    {
      "epoch": 1.1535753152807118,
      "grad_norm": 0.0032910152804106474,
      "learning_rate": 0.007200211701808223,
      "loss": 0.5453,
      "num_input_tokens_seen": 49560416,
      "step": 36040
    },
    {
      "epoch": 1.1537353562512003,
      "grad_norm": 0.01030594389885664,
      "learning_rate": 0.007182191792563286,
      "loss": 0.5864,
      "num_input_tokens_seen": 49566992,
      "step": 36045
    },
    {
      "epoch": 1.1538953972216888,
      "grad_norm": 0.0036132773384451866,
      "learning_rate": 0.0071641939076191145,
      "loss": 0.4442,
      "num_input_tokens_seen": 49574448,
      "step": 36050
    },
    {
      "epoch": 1.1540554381921773,
      "grad_norm": 0.005694159772247076,
      "learning_rate": 0.007146218049751257,
      "loss": 0.4895,
      "num_input_tokens_seen": 49581040,
      "step": 36055
    },
    {
      "epoch": 1.1542154791626658,
      "grad_norm": 0.004303858615458012,
      "learning_rate": 0.0071282642217317775,
      "loss": 0.4236,
      "num_input_tokens_seen": 49587904,
      "step": 36060
    },
    {
      "epoch": 1.154375520133154,
      "grad_norm": 0.0031334301456809044,
      "learning_rate": 0.007110332426329396,
      "loss": 0.6002,
      "num_input_tokens_seen": 49594912,
      "step": 36065
    },
    {
      "epoch": 1.1545355611036425,
      "grad_norm": 0.005132773891091347,
      "learning_rate": 0.007092422666309417,
      "loss": 0.5705,
      "num_input_tokens_seen": 49602096,
      "step": 36070
    },
    {
      "epoch": 1.154695602074131,
      "grad_norm": 0.009426699951291084,
      "learning_rate": 0.0070745349444337295,
      "loss": 0.7114,
      "num_input_tokens_seen": 49608624,
      "step": 36075
    },
    {
      "epoch": 1.1548556430446195,
      "grad_norm": 0.00474542798474431,
      "learning_rate": 0.007056669263460913,
      "loss": 0.5152,
      "num_input_tokens_seen": 49615104,
      "step": 36080
    },
    {
      "epoch": 1.1550156840151078,
      "grad_norm": 0.004278380889445543,
      "learning_rate": 0.007038825626145995,
      "loss": 0.5258,
      "num_input_tokens_seen": 49621664,
      "step": 36085
    },
    {
      "epoch": 1.1551757249855963,
      "grad_norm": 0.005653567146509886,
      "learning_rate": 0.007021004035240724,
      "loss": 0.429,
      "num_input_tokens_seen": 49628544,
      "step": 36090
    },
    {
      "epoch": 1.1553357659560848,
      "grad_norm": 0.009274338372051716,
      "learning_rate": 0.007003204493493453,
      "loss": 0.5967,
      "num_input_tokens_seen": 49635296,
      "step": 36095
    },
    {
      "epoch": 1.1554958069265733,
      "grad_norm": 0.0031700057443231344,
      "learning_rate": 0.006985427003649036,
      "loss": 0.4796,
      "num_input_tokens_seen": 49642048,
      "step": 36100
    },
    {
      "epoch": 1.1556558478970618,
      "grad_norm": 0.006145069375634193,
      "learning_rate": 0.006967671568449013,
      "loss": 0.4598,
      "num_input_tokens_seen": 49648688,
      "step": 36105
    },
    {
      "epoch": 1.15581588886755,
      "grad_norm": 0.003386633237823844,
      "learning_rate": 0.006949938190631511,
      "loss": 0.4929,
      "num_input_tokens_seen": 49655440,
      "step": 36110
    },
    {
      "epoch": 1.1559759298380385,
      "grad_norm": 0.004743883851915598,
      "learning_rate": 0.0069322268729311905,
      "loss": 0.4194,
      "num_input_tokens_seen": 49662336,
      "step": 36115
    },
    {
      "epoch": 1.156135970808527,
      "grad_norm": 0.005414207465946674,
      "learning_rate": 0.006914537618079403,
      "loss": 0.4485,
      "num_input_tokens_seen": 49668976,
      "step": 36120
    },
    {
      "epoch": 1.1562960117790153,
      "grad_norm": 0.0027444621082395315,
      "learning_rate": 0.006896870428804031,
      "loss": 0.4539,
      "num_input_tokens_seen": 49675616,
      "step": 36125
    },
    {
      "epoch": 1.1564560527495038,
      "grad_norm": 0.0030506306793540716,
      "learning_rate": 0.006879225307829595,
      "loss": 0.4983,
      "num_input_tokens_seen": 49682400,
      "step": 36130
    },
    {
      "epoch": 1.1566160937199923,
      "grad_norm": 0.005013727582991123,
      "learning_rate": 0.00686160225787717,
      "loss": 0.5554,
      "num_input_tokens_seen": 49689104,
      "step": 36135
    },
    {
      "epoch": 1.1567761346904808,
      "grad_norm": 0.007647054269909859,
      "learning_rate": 0.006844001281664463,
      "loss": 0.48,
      "num_input_tokens_seen": 49695936,
      "step": 36140
    },
    {
      "epoch": 1.1569361756609693,
      "grad_norm": 0.009494922123849392,
      "learning_rate": 0.006826422381905789,
      "loss": 0.6024,
      "num_input_tokens_seen": 49703232,
      "step": 36145
    },
    {
      "epoch": 1.1570962166314576,
      "grad_norm": 0.00546389352530241,
      "learning_rate": 0.006808865561311994,
      "loss": 0.4088,
      "num_input_tokens_seen": 49710304,
      "step": 36150
    },
    {
      "epoch": 1.157256257601946,
      "grad_norm": 0.005439416505396366,
      "learning_rate": 0.00679133082259058,
      "loss": 0.5086,
      "num_input_tokens_seen": 49717120,
      "step": 36155
    },
    {
      "epoch": 1.1574162985724346,
      "grad_norm": 0.004277387168258429,
      "learning_rate": 0.00677381816844565,
      "loss": 0.5641,
      "num_input_tokens_seen": 49724016,
      "step": 36160
    },
    {
      "epoch": 1.157576339542923,
      "grad_norm": 0.004766235128045082,
      "learning_rate": 0.0067563276015778434,
      "loss": 0.4672,
      "num_input_tokens_seen": 49730976,
      "step": 36165
    },
    {
      "epoch": 1.1577363805134113,
      "grad_norm": 0.0044263191521167755,
      "learning_rate": 0.006738859124684437,
      "loss": 0.4824,
      "num_input_tokens_seen": 49737568,
      "step": 36170
    },
    {
      "epoch": 1.1578964214838998,
      "grad_norm": 0.004518838599324226,
      "learning_rate": 0.006721412740459259,
      "loss": 0.5647,
      "num_input_tokens_seen": 49744464,
      "step": 36175
    },
    {
      "epoch": 1.1580564624543883,
      "grad_norm": 0.005081568378955126,
      "learning_rate": 0.006703988451592824,
      "loss": 0.3856,
      "num_input_tokens_seen": 49751056,
      "step": 36180
    },
    {
      "epoch": 1.1582165034248768,
      "grad_norm": 0.0029997967649251223,
      "learning_rate": 0.006686586260772114,
      "loss": 0.346,
      "num_input_tokens_seen": 49757920,
      "step": 36185
    },
    {
      "epoch": 1.1583765443953653,
      "grad_norm": 0.006208151113241911,
      "learning_rate": 0.006669206170680819,
      "loss": 0.36,
      "num_input_tokens_seen": 49765056,
      "step": 36190
    },
    {
      "epoch": 1.1585365853658536,
      "grad_norm": 0.0033217500895261765,
      "learning_rate": 0.0066518481839991095,
      "loss": 0.5119,
      "num_input_tokens_seen": 49771888,
      "step": 36195
    },
    {
      "epoch": 1.158696626336342,
      "grad_norm": 0.0024224440567195415,
      "learning_rate": 0.006634512303403861,
      "loss": 0.4185,
      "num_input_tokens_seen": 49778864,
      "step": 36200
    },
    {
      "epoch": 1.158696626336342,
      "eval_loss": 0.4948018193244934,
      "eval_runtime": 331.8406,
      "eval_samples_per_second": 41.845,
      "eval_steps_per_second": 20.923,
      "num_input_tokens_seen": 49778864,
      "step": 36200
    },
    {
      "epoch": 1.1588566673068306,
      "grad_norm": 0.0033167044166475534,
      "learning_rate": 0.0066171985315684355,
      "loss": 0.6615,
      "num_input_tokens_seen": 49786304,
      "step": 36205
    },
    {
      "epoch": 1.159016708277319,
      "grad_norm": 0.003979257307946682,
      "learning_rate": 0.0065999068711628806,
      "loss": 0.5652,
      "num_input_tokens_seen": 49793056,
      "step": 36210
    },
    {
      "epoch": 1.1591767492478073,
      "grad_norm": 0.004520156886428595,
      "learning_rate": 0.0065826373248537295,
      "loss": 0.4927,
      "num_input_tokens_seen": 49799504,
      "step": 36215
    },
    {
      "epoch": 1.1593367902182958,
      "grad_norm": 0.006080855615437031,
      "learning_rate": 0.006565389895304218,
      "loss": 0.4886,
      "num_input_tokens_seen": 49806320,
      "step": 36220
    },
    {
      "epoch": 1.1594968311887843,
      "grad_norm": 0.003885553451254964,
      "learning_rate": 0.006548164585174104,
      "loss": 0.4447,
      "num_input_tokens_seen": 49813328,
      "step": 36225
    },
    {
      "epoch": 1.1596568721592728,
      "grad_norm": 0.005184997338801622,
      "learning_rate": 0.006530961397119728,
      "loss": 0.5316,
      "num_input_tokens_seen": 49819888,
      "step": 36230
    },
    {
      "epoch": 1.1598169131297613,
      "grad_norm": 0.0025026288349181414,
      "learning_rate": 0.00651378033379405,
      "loss": 0.3896,
      "num_input_tokens_seen": 49826352,
      "step": 36235
    },
    {
      "epoch": 1.1599769541002496,
      "grad_norm": 0.005065357778221369,
      "learning_rate": 0.006496621397846619,
      "loss": 0.5149,
      "num_input_tokens_seen": 49833168,
      "step": 36240
    },
    {
      "epoch": 1.160136995070738,
      "grad_norm": 0.00215232465416193,
      "learning_rate": 0.006479484591923518,
      "loss": 0.3936,
      "num_input_tokens_seen": 49840176,
      "step": 36245
    },
    {
      "epoch": 1.1602970360412266,
      "grad_norm": 0.0035383133217692375,
      "learning_rate": 0.006462369918667515,
      "loss": 0.4409,
      "num_input_tokens_seen": 49846816,
      "step": 36250
    },
    {
      "epoch": 1.160457077011715,
      "grad_norm": 0.007924159057438374,
      "learning_rate": 0.006445277380717851,
      "loss": 0.5055,
      "num_input_tokens_seen": 49853872,
      "step": 36255
    },
    {
      "epoch": 1.1606171179822034,
      "grad_norm": 0.003608043771237135,
      "learning_rate": 0.006428206980710466,
      "loss": 0.3601,
      "num_input_tokens_seen": 49861264,
      "step": 36260
    },
    {
      "epoch": 1.1607771589526918,
      "grad_norm": 0.002232238184660673,
      "learning_rate": 0.006411158721277788,
      "loss": 0.2772,
      "num_input_tokens_seen": 49868032,
      "step": 36265
    },
    {
      "epoch": 1.1609371999231803,
      "grad_norm": 0.00701266061514616,
      "learning_rate": 0.00639413260504888,
      "loss": 0.3969,
      "num_input_tokens_seen": 49875296,
      "step": 36270
    },
    {
      "epoch": 1.1610972408936688,
      "grad_norm": 0.004802758805453777,
      "learning_rate": 0.006377128634649376,
      "loss": 0.4645,
      "num_input_tokens_seen": 49881936,
      "step": 36275
    },
    {
      "epoch": 1.1612572818641573,
      "grad_norm": 0.007270507514476776,
      "learning_rate": 0.006360146812701528,
      "loss": 0.4237,
      "num_input_tokens_seen": 49888704,
      "step": 36280
    },
    {
      "epoch": 1.1614173228346456,
      "grad_norm": 0.004024856258183718,
      "learning_rate": 0.006343187141824125,
      "loss": 0.3884,
      "num_input_tokens_seen": 49895984,
      "step": 36285
    },
    {
      "epoch": 1.161577363805134,
      "grad_norm": 0.00425652414560318,
      "learning_rate": 0.00632624962463259,
      "loss": 0.3774,
      "num_input_tokens_seen": 49902784,
      "step": 36290
    },
    {
      "epoch": 1.1617374047756226,
      "grad_norm": 0.00622596638277173,
      "learning_rate": 0.006309334263738853,
      "loss": 0.5502,
      "num_input_tokens_seen": 49909936,
      "step": 36295
    },
    {
      "epoch": 1.161897445746111,
      "grad_norm": 0.00818493403494358,
      "learning_rate": 0.006292441061751508,
      "loss": 0.5759,
      "num_input_tokens_seen": 49917024,
      "step": 36300
    },
    {
      "epoch": 1.1620574867165994,
      "grad_norm": 0.004469107836484909,
      "learning_rate": 0.0062755700212757054,
      "loss": 0.4747,
      "num_input_tokens_seen": 49923664,
      "step": 36305
    },
    {
      "epoch": 1.1622175276870879,
      "grad_norm": 0.0032313030678778887,
      "learning_rate": 0.006258721144913148,
      "loss": 0.4888,
      "num_input_tokens_seen": 49930528,
      "step": 36310
    },
    {
      "epoch": 1.1623775686575764,
      "grad_norm": 0.004326009191572666,
      "learning_rate": 0.0062418944352621575,
      "loss": 0.5002,
      "num_input_tokens_seen": 49937376,
      "step": 36315
    },
    {
      "epoch": 1.1625376096280648,
      "grad_norm": 0.006141102407127619,
      "learning_rate": 0.0062250898949176405,
      "loss": 0.661,
      "num_input_tokens_seen": 49944416,
      "step": 36320
    },
    {
      "epoch": 1.1626976505985533,
      "grad_norm": 0.0065824412740767,
      "learning_rate": 0.006208307526471041,
      "loss": 0.5202,
      "num_input_tokens_seen": 49951392,
      "step": 36325
    },
    {
      "epoch": 1.1628576915690416,
      "grad_norm": 0.0021605079527944326,
      "learning_rate": 0.006191547332510405,
      "loss": 0.4252,
      "num_input_tokens_seen": 49958080,
      "step": 36330
    },
    {
      "epoch": 1.1630177325395301,
      "grad_norm": 0.005201001651585102,
      "learning_rate": 0.006174809315620416,
      "loss": 0.5305,
      "num_input_tokens_seen": 49964640,
      "step": 36335
    },
    {
      "epoch": 1.1631777735100186,
      "grad_norm": 0.008339080959558487,
      "learning_rate": 0.00615809347838221,
      "loss": 0.584,
      "num_input_tokens_seen": 49971376,
      "step": 36340
    },
    {
      "epoch": 1.163337814480507,
      "grad_norm": 0.006151407025754452,
      "learning_rate": 0.006141399823373655,
      "loss": 0.6121,
      "num_input_tokens_seen": 49978096,
      "step": 36345
    },
    {
      "epoch": 1.1634978554509954,
      "grad_norm": 0.0052436101250350475,
      "learning_rate": 0.0061247283531690455,
      "loss": 0.4406,
      "num_input_tokens_seen": 49984944,
      "step": 36350
    },
    {
      "epoch": 1.1636578964214839,
      "grad_norm": 0.004715960472822189,
      "learning_rate": 0.0061080790703393895,
      "loss": 0.569,
      "num_input_tokens_seen": 49991856,
      "step": 36355
    },
    {
      "epoch": 1.1638179373919724,
      "grad_norm": 0.005726311821490526,
      "learning_rate": 0.006091451977452217,
      "loss": 0.5395,
      "num_input_tokens_seen": 49998368,
      "step": 36360
    },
    {
      "epoch": 1.1639779783624609,
      "grad_norm": 0.004672985058277845,
      "learning_rate": 0.00607484707707161,
      "loss": 0.5575,
      "num_input_tokens_seen": 50005024,
      "step": 36365
    },
    {
      "epoch": 1.1641380193329494,
      "grad_norm": 0.006925644818693399,
      "learning_rate": 0.006058264371758254,
      "loss": 0.4905,
      "num_input_tokens_seen": 50011232,
      "step": 36370
    },
    {
      "epoch": 1.1642980603034376,
      "grad_norm": 0.004702966194599867,
      "learning_rate": 0.00604170386406942,
      "loss": 0.4701,
      "num_input_tokens_seen": 50017872,
      "step": 36375
    },
    {
      "epoch": 1.1644581012739261,
      "grad_norm": 0.008965685032308102,
      "learning_rate": 0.006025165556558931,
      "loss": 0.4409,
      "num_input_tokens_seen": 50024832,
      "step": 36380
    },
    {
      "epoch": 1.1646181422444146,
      "grad_norm": 0.006790578365325928,
      "learning_rate": 0.006008649451777248,
      "loss": 0.5106,
      "num_input_tokens_seen": 50031760,
      "step": 36385
    },
    {
      "epoch": 1.164778183214903,
      "grad_norm": 0.004151122644543648,
      "learning_rate": 0.005992155552271283,
      "loss": 0.3967,
      "num_input_tokens_seen": 50038608,
      "step": 36390
    },
    {
      "epoch": 1.1649382241853914,
      "grad_norm": 0.013651443645358086,
      "learning_rate": 0.005975683860584685,
      "loss": 0.5682,
      "num_input_tokens_seen": 50045072,
      "step": 36395
    },
    {
      "epoch": 1.1650982651558799,
      "grad_norm": 0.004722527228295803,
      "learning_rate": 0.0059592343792575385,
      "loss": 0.4373,
      "num_input_tokens_seen": 50051632,
      "step": 36400
    },
    {
      "epoch": 1.1650982651558799,
      "eval_loss": 0.4945104122161865,
      "eval_runtime": 332.0357,
      "eval_samples_per_second": 41.821,
      "eval_steps_per_second": 20.91,
      "num_input_tokens_seen": 50051632,
      "step": 36400
    },
    {
      "epoch": 1.1652583061263684,
      "grad_norm": 0.005065003409981728,
      "learning_rate": 0.0059428071108265975,
      "loss": 0.4271,
      "num_input_tokens_seen": 50058496,
      "step": 36405
    },
    {
      "epoch": 1.1654183470968569,
      "grad_norm": 0.0037255515344440937,
      "learning_rate": 0.005926402057825136,
      "loss": 0.4751,
      "num_input_tokens_seen": 50065280,
      "step": 36410
    },
    {
      "epoch": 1.1655783880673452,
      "grad_norm": 0.006223740056157112,
      "learning_rate": 0.005910019222782997,
      "loss": 0.3272,
      "num_input_tokens_seen": 50072144,
      "step": 36415
    },
    {
      "epoch": 1.1657384290378336,
      "grad_norm": 0.010068373754620552,
      "learning_rate": 0.005893658608226643,
      "loss": 0.5183,
      "num_input_tokens_seen": 50078880,
      "step": 36420
    },
    {
      "epoch": 1.1658984700083221,
      "grad_norm": 0.003425035858526826,
      "learning_rate": 0.0058773202166791045,
      "loss": 0.4176,
      "num_input_tokens_seen": 50086080,
      "step": 36425
    },
    {
      "epoch": 1.1660585109788106,
      "grad_norm": 0.00663079135119915,
      "learning_rate": 0.005861004050659918,
      "loss": 0.5752,
      "num_input_tokens_seen": 50092736,
      "step": 36430
    },
    {
      "epoch": 1.166218551949299,
      "grad_norm": 0.004620142746716738,
      "learning_rate": 0.005844710112685286,
      "loss": 0.4696,
      "num_input_tokens_seen": 50099536,
      "step": 36435
    },
    {
      "epoch": 1.1663785929197874,
      "grad_norm": 0.005238753743469715,
      "learning_rate": 0.005828438405267933,
      "loss": 0.4597,
      "num_input_tokens_seen": 50106752,
      "step": 36440
    },
    {
      "epoch": 1.166538633890276,
      "grad_norm": 0.004398778546601534,
      "learning_rate": 0.00581218893091715,
      "loss": 0.433,
      "num_input_tokens_seen": 50113536,
      "step": 36445
    },
    {
      "epoch": 1.1666986748607644,
      "grad_norm": 0.0033706529065966606,
      "learning_rate": 0.005795961692138801,
      "loss": 0.4228,
      "num_input_tokens_seen": 50120576,
      "step": 36450
    },
    {
      "epoch": 1.166858715831253,
      "grad_norm": 0.005578034557402134,
      "learning_rate": 0.00577975669143535,
      "loss": 0.5659,
      "num_input_tokens_seen": 50127520,
      "step": 36455
    },
    {
      "epoch": 1.1670187568017412,
      "grad_norm": 0.0023819864727556705,
      "learning_rate": 0.005763573931305782,
      "loss": 0.4746,
      "num_input_tokens_seen": 50134160,
      "step": 36460
    },
    {
      "epoch": 1.1671787977722297,
      "grad_norm": 0.00422499468550086,
      "learning_rate": 0.005747413414245733,
      "loss": 0.4544,
      "num_input_tokens_seen": 50140864,
      "step": 36465
    },
    {
      "epoch": 1.1673388387427182,
      "grad_norm": 0.004116878379136324,
      "learning_rate": 0.005731275142747294,
      "loss": 0.564,
      "num_input_tokens_seen": 50147808,
      "step": 36470
    },
    {
      "epoch": 1.1674988797132066,
      "grad_norm": 0.004922518972307444,
      "learning_rate": 0.005715159119299256,
      "loss": 0.5019,
      "num_input_tokens_seen": 50154832,
      "step": 36475
    },
    {
      "epoch": 1.167658920683695,
      "grad_norm": 0.003586389822885394,
      "learning_rate": 0.005699065346386867,
      "loss": 0.431,
      "num_input_tokens_seen": 50161712,
      "step": 36480
    },
    {
      "epoch": 1.1678189616541834,
      "grad_norm": 0.0029092621989548206,
      "learning_rate": 0.0056829938264919885,
      "loss": 0.4028,
      "num_input_tokens_seen": 50168176,
      "step": 36485
    },
    {
      "epoch": 1.167979002624672,
      "grad_norm": 0.003058107802644372,
      "learning_rate": 0.005666944562093074,
      "loss": 0.4907,
      "num_input_tokens_seen": 50174992,
      "step": 36490
    },
    {
      "epoch": 1.1681390435951604,
      "grad_norm": 0.0029946924187242985,
      "learning_rate": 0.005650917555665108,
      "loss": 0.3595,
      "num_input_tokens_seen": 50181520,
      "step": 36495
    },
    {
      "epoch": 1.168299084565649,
      "grad_norm": 0.004749383311718702,
      "learning_rate": 0.005634912809679632,
      "loss": 0.5433,
      "num_input_tokens_seen": 50188608,
      "step": 36500
    },
    {
      "epoch": 1.1684591255361372,
      "grad_norm": 0.004009828437119722,
      "learning_rate": 0.005618930326604854,
      "loss": 0.4196,
      "num_input_tokens_seen": 50195696,
      "step": 36505
    },
    {
      "epoch": 1.1686191665066257,
      "grad_norm": 0.0063039762899279594,
      "learning_rate": 0.005602970108905386,
      "loss": 0.3851,
      "num_input_tokens_seen": 50202816,
      "step": 36510
    },
    {
      "epoch": 1.1687792074771142,
      "grad_norm": 0.004721956793218851,
      "learning_rate": 0.005587032159042543,
      "loss": 0.3712,
      "num_input_tokens_seen": 50209856,
      "step": 36515
    },
    {
      "epoch": 1.1689392484476027,
      "grad_norm": 0.0058103702031075954,
      "learning_rate": 0.005571116479474158,
      "loss": 0.3761,
      "num_input_tokens_seen": 50216496,
      "step": 36520
    },
    {
      "epoch": 1.169099289418091,
      "grad_norm": 0.006209950894117355,
      "learning_rate": 0.005555223072654619,
      "loss": 0.5875,
      "num_input_tokens_seen": 50223088,
      "step": 36525
    },
    {
      "epoch": 1.1692593303885794,
      "grad_norm": 0.004858403466641903,
      "learning_rate": 0.005539351941034881,
      "loss": 0.5129,
      "num_input_tokens_seen": 50229632,
      "step": 36530
    },
    {
      "epoch": 1.169419371359068,
      "grad_norm": 0.0039984057657420635,
      "learning_rate": 0.0055235030870624865,
      "loss": 0.57,
      "num_input_tokens_seen": 50236832,
      "step": 36535
    },
    {
      "epoch": 1.1695794123295564,
      "grad_norm": 0.0037336405366659164,
      "learning_rate": 0.005507676513181514,
      "loss": 0.4413,
      "num_input_tokens_seen": 50243552,
      "step": 36540
    },
    {
      "epoch": 1.169739453300045,
      "grad_norm": 0.002701146760955453,
      "learning_rate": 0.005491872221832628,
      "loss": 0.4811,
      "num_input_tokens_seen": 50250496,
      "step": 36545
    },
    {
      "epoch": 1.1698994942705332,
      "grad_norm": 0.006100587081164122,
      "learning_rate": 0.005476090215453061,
      "loss": 0.4586,
      "num_input_tokens_seen": 50256976,
      "step": 36550
    },
    {
      "epoch": 1.1700595352410217,
      "grad_norm": 0.0038848428521305323,
      "learning_rate": 0.0054603304964765675,
      "loss": 0.4402,
      "num_input_tokens_seen": 50263680,
      "step": 36555
    },
    {
      "epoch": 1.1702195762115102,
      "grad_norm": 0.0038948666770011187,
      "learning_rate": 0.005444593067333519,
      "loss": 0.5389,
      "num_input_tokens_seen": 50270656,
      "step": 36560
    },
    {
      "epoch": 1.1703796171819987,
      "grad_norm": 0.007499884348362684,
      "learning_rate": 0.00542887793045081,
      "loss": 0.362,
      "num_input_tokens_seen": 50277520,
      "step": 36565
    },
    {
      "epoch": 1.170539658152487,
      "grad_norm": 0.004700482357293367,
      "learning_rate": 0.005413185088251932,
      "loss": 0.3458,
      "num_input_tokens_seen": 50284496,
      "step": 36570
    },
    {
      "epoch": 1.1706996991229754,
      "grad_norm": 0.004670621827244759,
      "learning_rate": 0.005397514543156884,
      "loss": 0.4278,
      "num_input_tokens_seen": 50291568,
      "step": 36575
    },
    {
      "epoch": 1.170859740093464,
      "grad_norm": 0.0043118540197610855,
      "learning_rate": 0.0053818662975822825,
      "loss": 0.5591,
      "num_input_tokens_seen": 50298176,
      "step": 36580
    },
    {
      "epoch": 1.1710197810639524,
      "grad_norm": 0.00433729961514473,
      "learning_rate": 0.005366240353941315,
      "loss": 0.3932,
      "num_input_tokens_seen": 50305328,
      "step": 36585
    },
    {
      "epoch": 1.171179822034441,
      "grad_norm": 0.005615759640932083,
      "learning_rate": 0.005350636714643636,
      "loss": 0.3995,
      "num_input_tokens_seen": 50312048,
      "step": 36590
    },
    {
      "epoch": 1.1713398630049292,
      "grad_norm": 0.007564937695860863,
      "learning_rate": 0.005335055382095555,
      "loss": 0.3707,
      "num_input_tokens_seen": 50318816,
      "step": 36595
    },
    {
      "epoch": 1.1714999039754177,
      "grad_norm": 0.002965350402519107,
      "learning_rate": 0.005319496358699915,
      "loss": 0.3587,
      "num_input_tokens_seen": 50325888,
      "step": 36600
    },
    {
      "epoch": 1.1714999039754177,
      "eval_loss": 0.49457159638404846,
      "eval_runtime": 332.3926,
      "eval_samples_per_second": 41.776,
      "eval_steps_per_second": 20.888,
      "num_input_tokens_seen": 50325888,
      "step": 36600
    },
    {
      "epoch": 1.1716599449459062,
      "grad_norm": 0.0036918334662914276,
      "learning_rate": 0.005303959646856099,
      "loss": 0.3934,
      "num_input_tokens_seen": 50332880,
      "step": 36605
    },
    {
      "epoch": 1.1718199859163947,
      "grad_norm": 0.0035038013011217117,
      "learning_rate": 0.005288445248960089,
      "loss": 0.5224,
      "num_input_tokens_seen": 50339744,
      "step": 36610
    },
    {
      "epoch": 1.171980026886883,
      "grad_norm": 0.003950017970055342,
      "learning_rate": 0.005272953167404354,
      "loss": 0.3482,
      "num_input_tokens_seen": 50346592,
      "step": 36615
    },
    {
      "epoch": 1.1721400678573715,
      "grad_norm": 0.003387419506907463,
      "learning_rate": 0.005257483404578017,
      "loss": 0.4307,
      "num_input_tokens_seen": 50353280,
      "step": 36620
    },
    {
      "epoch": 1.17230010882786,
      "grad_norm": 0.004829625133424997,
      "learning_rate": 0.0052420359628666865,
      "loss": 0.4528,
      "num_input_tokens_seen": 50359792,
      "step": 36625
    },
    {
      "epoch": 1.1724601497983484,
      "grad_norm": 0.006984615698456764,
      "learning_rate": 0.00522661084465254,
      "loss": 0.618,
      "num_input_tokens_seen": 50366512,
      "step": 36630
    },
    {
      "epoch": 1.172620190768837,
      "grad_norm": 0.006019179709255695,
      "learning_rate": 0.005211208052314326,
      "loss": 0.5368,
      "num_input_tokens_seen": 50373456,
      "step": 36635
    },
    {
      "epoch": 1.1727802317393252,
      "grad_norm": 0.004453564994037151,
      "learning_rate": 0.005195827588227391,
      "loss": 0.3506,
      "num_input_tokens_seen": 50380304,
      "step": 36640
    },
    {
      "epoch": 1.1729402727098137,
      "grad_norm": 0.0057545010931789875,
      "learning_rate": 0.0051804694547635255,
      "loss": 0.5176,
      "num_input_tokens_seen": 50387200,
      "step": 36645
    },
    {
      "epoch": 1.1731003136803022,
      "grad_norm": 0.009066884405910969,
      "learning_rate": 0.005165133654291232,
      "loss": 0.628,
      "num_input_tokens_seen": 50393984,
      "step": 36650
    },
    {
      "epoch": 1.1732603546507905,
      "grad_norm": 0.0025219323579221964,
      "learning_rate": 0.005149820189175402,
      "loss": 0.528,
      "num_input_tokens_seen": 50401024,
      "step": 36655
    },
    {
      "epoch": 1.173420395621279,
      "grad_norm": 0.006848813965916634,
      "learning_rate": 0.005134529061777598,
      "loss": 0.6989,
      "num_input_tokens_seen": 50407600,
      "step": 36660
    },
    {
      "epoch": 1.1735804365917675,
      "grad_norm": 0.004874207079410553,
      "learning_rate": 0.005119260274455933,
      "loss": 0.556,
      "num_input_tokens_seen": 50414592,
      "step": 36665
    },
    {
      "epoch": 1.173740477562256,
      "grad_norm": 0.0033866448793560266,
      "learning_rate": 0.005104013829565007,
      "loss": 0.4001,
      "num_input_tokens_seen": 50421376,
      "step": 36670
    },
    {
      "epoch": 1.1739005185327445,
      "grad_norm": 0.004143770318478346,
      "learning_rate": 0.005088789729456006,
      "loss": 0.427,
      "num_input_tokens_seen": 50428000,
      "step": 36675
    },
    {
      "epoch": 1.1740605595032327,
      "grad_norm": 0.0036504289600998163,
      "learning_rate": 0.005073587976476735,
      "loss": 0.5054,
      "num_input_tokens_seen": 50434784,
      "step": 36680
    },
    {
      "epoch": 1.1742206004737212,
      "grad_norm": 0.004122561775147915,
      "learning_rate": 0.005058408572971418,
      "loss": 0.465,
      "num_input_tokens_seen": 50441408,
      "step": 36685
    },
    {
      "epoch": 1.1743806414442097,
      "grad_norm": 0.006175803951919079,
      "learning_rate": 0.005043251521280983,
      "loss": 0.3224,
      "num_input_tokens_seen": 50448608,
      "step": 36690
    },
    {
      "epoch": 1.1745406824146982,
      "grad_norm": 0.008248276077210903,
      "learning_rate": 0.005028116823742795,
      "loss": 0.4573,
      "num_input_tokens_seen": 50455280,
      "step": 36695
    },
    {
      "epoch": 1.1747007233851865,
      "grad_norm": 0.006195095833390951,
      "learning_rate": 0.005013004482690819,
      "loss": 0.5527,
      "num_input_tokens_seen": 50461920,
      "step": 36700
    },
    {
      "epoch": 1.174860764355675,
      "grad_norm": 0.006930625066161156,
      "learning_rate": 0.0049979145004555746,
      "loss": 0.4197,
      "num_input_tokens_seen": 50468352,
      "step": 36705
    },
    {
      "epoch": 1.1750208053261635,
      "grad_norm": 0.004082698840647936,
      "learning_rate": 0.004982846879364116,
      "loss": 0.3796,
      "num_input_tokens_seen": 50475584,
      "step": 36710
    },
    {
      "epoch": 1.175180846296652,
      "grad_norm": 0.0035015938337892294,
      "learning_rate": 0.0049678016217400535,
      "loss": 0.3953,
      "num_input_tokens_seen": 50482320,
      "step": 36715
    },
    {
      "epoch": 1.1753408872671405,
      "grad_norm": 0.005933046340942383,
      "learning_rate": 0.004952778729903595,
      "loss": 0.5793,
      "num_input_tokens_seen": 50489568,
      "step": 36720
    },
    {
      "epoch": 1.1755009282376288,
      "grad_norm": 0.0044236741960048676,
      "learning_rate": 0.004937778206171422,
      "loss": 0.4889,
      "num_input_tokens_seen": 50496368,
      "step": 36725
    },
    {
      "epoch": 1.1756609692081172,
      "grad_norm": 0.004398435819894075,
      "learning_rate": 0.004922800052856835,
      "loss": 0.3682,
      "num_input_tokens_seen": 50503632,
      "step": 36730
    },
    {
      "epoch": 1.1758210101786057,
      "grad_norm": 0.010893863625824451,
      "learning_rate": 0.004907844272269602,
      "loss": 0.5445,
      "num_input_tokens_seen": 50510656,
      "step": 36735
    },
    {
      "epoch": 1.1759810511490942,
      "grad_norm": 0.005226455628871918,
      "learning_rate": 0.004892910866716144,
      "loss": 0.3593,
      "num_input_tokens_seen": 50517568,
      "step": 36740
    },
    {
      "epoch": 1.1761410921195825,
      "grad_norm": 0.003786679357290268,
      "learning_rate": 0.004877999838499369,
      "loss": 0.3558,
      "num_input_tokens_seen": 50525008,
      "step": 36745
    },
    {
      "epoch": 1.176301133090071,
      "grad_norm": 0.007125907111912966,
      "learning_rate": 0.0048631111899187065,
      "loss": 0.5586,
      "num_input_tokens_seen": 50531952,
      "step": 36750
    },
    {
      "epoch": 1.1764611740605595,
      "grad_norm": 0.004767056088894606,
      "learning_rate": 0.0048482449232702335,
      "loss": 0.4599,
      "num_input_tokens_seen": 50538832,
      "step": 36755
    },
    {
      "epoch": 1.176621215031048,
      "grad_norm": 0.004042964894324541,
      "learning_rate": 0.004833401040846469,
      "loss": 0.5742,
      "num_input_tokens_seen": 50545648,
      "step": 36760
    },
    {
      "epoch": 1.1767812560015365,
      "grad_norm": 0.005094586405903101,
      "learning_rate": 0.004818579544936546,
      "loss": 0.3242,
      "num_input_tokens_seen": 50552800,
      "step": 36765
    },
    {
      "epoch": 1.1769412969720248,
      "grad_norm": 0.005646170116961002,
      "learning_rate": 0.004803780437826121,
      "loss": 0.4706,
      "num_input_tokens_seen": 50560000,
      "step": 36770
    },
    {
      "epoch": 1.1771013379425133,
      "grad_norm": 0.004985054023563862,
      "learning_rate": 0.004789003721797402,
      "loss": 0.4528,
      "num_input_tokens_seen": 50566992,
      "step": 36775
    },
    {
      "epoch": 1.1772613789130018,
      "grad_norm": 0.003569540102034807,
      "learning_rate": 0.004774249399129132,
      "loss": 0.439,
      "num_input_tokens_seen": 50573904,
      "step": 36780
    },
    {
      "epoch": 1.1774214198834903,
      "grad_norm": 0.005541276652365923,
      "learning_rate": 0.004759517472096642,
      "loss": 0.4632,
      "num_input_tokens_seen": 50580528,
      "step": 36785
    },
    {
      "epoch": 1.1775814608539785,
      "grad_norm": 0.005850478541105986,
      "learning_rate": 0.004744807942971746,
      "loss": 0.5811,
      "num_input_tokens_seen": 50587136,
      "step": 36790
    },
    {
      "epoch": 1.177741501824467,
      "grad_norm": 0.008787091821432114,
      "learning_rate": 0.004730120814022881,
      "loss": 0.675,
      "num_input_tokens_seen": 50594160,
      "step": 36795
    },
    {
      "epoch": 1.1779015427949555,
      "grad_norm": 0.009428287856280804,
      "learning_rate": 0.004715456087514935,
      "loss": 0.456,
      "num_input_tokens_seen": 50601136,
      "step": 36800
    },
    {
      "epoch": 1.1779015427949555,
      "eval_loss": 0.49454107880592346,
      "eval_runtime": 332.0665,
      "eval_samples_per_second": 41.817,
      "eval_steps_per_second": 20.908,
      "num_input_tokens_seen": 50601136,
      "step": 36800
    },
    {
      "epoch": 1.178061583765444,
      "grad_norm": 0.0061097294092178345,
      "learning_rate": 0.004700813765709432,
      "loss": 0.6338,
      "num_input_tokens_seen": 50607632,
      "step": 36805
    },
    {
      "epoch": 1.1782216247359325,
      "grad_norm": 0.00563511997461319,
      "learning_rate": 0.004686193850864401,
      "loss": 0.4394,
      "num_input_tokens_seen": 50614480,
      "step": 36810
    },
    {
      "epoch": 1.1783816657064208,
      "grad_norm": 0.00845090951770544,
      "learning_rate": 0.004671596345234385,
      "loss": 0.4789,
      "num_input_tokens_seen": 50621504,
      "step": 36815
    },
    {
      "epoch": 1.1785417066769093,
      "grad_norm": 0.004003356210887432,
      "learning_rate": 0.00465702125107052,
      "loss": 0.3265,
      "num_input_tokens_seen": 50628896,
      "step": 36820
    },
    {
      "epoch": 1.1787017476473978,
      "grad_norm": 0.010941797867417336,
      "learning_rate": 0.004642468570620506,
      "loss": 0.5308,
      "num_input_tokens_seen": 50635168,
      "step": 36825
    },
    {
      "epoch": 1.1788617886178863,
      "grad_norm": 0.005209183320403099,
      "learning_rate": 0.004627938306128482,
      "loss": 0.436,
      "num_input_tokens_seen": 50641888,
      "step": 36830
    },
    {
      "epoch": 1.1790218295883745,
      "grad_norm": 0.006434236653149128,
      "learning_rate": 0.004613430459835255,
      "loss": 0.3462,
      "num_input_tokens_seen": 50648544,
      "step": 36835
    },
    {
      "epoch": 1.179181870558863,
      "grad_norm": 0.0051168096251785755,
      "learning_rate": 0.004598945033978085,
      "loss": 0.4719,
      "num_input_tokens_seen": 50655120,
      "step": 36840
    },
    {
      "epoch": 1.1793419115293515,
      "grad_norm": 0.005274592898786068,
      "learning_rate": 0.004584482030790804,
      "loss": 0.5401,
      "num_input_tokens_seen": 50662576,
      "step": 36845
    },
    {
      "epoch": 1.17950195249984,
      "grad_norm": 0.007091801147907972,
      "learning_rate": 0.004570041452503826,
      "loss": 0.5388,
      "num_input_tokens_seen": 50669904,
      "step": 36850
    },
    {
      "epoch": 1.1796619934703285,
      "grad_norm": 0.005469180177897215,
      "learning_rate": 0.004555623301344003,
      "loss": 0.4273,
      "num_input_tokens_seen": 50676896,
      "step": 36855
    },
    {
      "epoch": 1.1798220344408168,
      "grad_norm": 0.004714481998234987,
      "learning_rate": 0.004541227579534857,
      "loss": 0.4093,
      "num_input_tokens_seen": 50683984,
      "step": 36860
    },
    {
      "epoch": 1.1799820754113053,
      "grad_norm": 0.002727366518229246,
      "learning_rate": 0.004526854289296378,
      "loss": 0.4285,
      "num_input_tokens_seen": 50690896,
      "step": 36865
    },
    {
      "epoch": 1.1801421163817938,
      "grad_norm": 0.004717142786830664,
      "learning_rate": 0.004512503432845078,
      "loss": 0.4638,
      "num_input_tokens_seen": 50697744,
      "step": 36870
    },
    {
      "epoch": 1.1803021573522823,
      "grad_norm": 0.00345995114184916,
      "learning_rate": 0.004498175012394068,
      "loss": 0.4996,
      "num_input_tokens_seen": 50704432,
      "step": 36875
    },
    {
      "epoch": 1.1804621983227706,
      "grad_norm": 0.006167851388454437,
      "learning_rate": 0.004483869030152965,
      "loss": 0.4505,
      "num_input_tokens_seen": 50711296,
      "step": 36880
    },
    {
      "epoch": 1.180622239293259,
      "grad_norm": 0.00579822389408946,
      "learning_rate": 0.004469585488327904,
      "loss": 0.5208,
      "num_input_tokens_seen": 50718208,
      "step": 36885
    },
    {
      "epoch": 1.1807822802637475,
      "grad_norm": 0.0023832221049815416,
      "learning_rate": 0.0044553243891216395,
      "loss": 0.4552,
      "num_input_tokens_seen": 50724944,
      "step": 36890
    },
    {
      "epoch": 1.180942321234236,
      "grad_norm": 0.00562801631167531,
      "learning_rate": 0.004441085734733363,
      "loss": 0.3455,
      "num_input_tokens_seen": 50731680,
      "step": 36895
    },
    {
      "epoch": 1.1811023622047245,
      "grad_norm": 0.0064949276857078075,
      "learning_rate": 0.004426869527358884,
      "loss": 0.5215,
      "num_input_tokens_seen": 50738432,
      "step": 36900
    },
    {
      "epoch": 1.1812624031752128,
      "grad_norm": 0.009330031462013721,
      "learning_rate": 0.0044126757691905156,
      "loss": 0.4395,
      "num_input_tokens_seen": 50744800,
      "step": 36905
    },
    {
      "epoch": 1.1814224441457013,
      "grad_norm": 0.00558133190497756,
      "learning_rate": 0.004398504462417107,
      "loss": 0.447,
      "num_input_tokens_seen": 50751856,
      "step": 36910
    },
    {
      "epoch": 1.1815824851161898,
      "grad_norm": 0.010124285705387592,
      "learning_rate": 0.0043843556092240605,
      "loss": 0.5145,
      "num_input_tokens_seen": 50759168,
      "step": 36915
    },
    {
      "epoch": 1.181742526086678,
      "grad_norm": 0.004358256701380014,
      "learning_rate": 0.004370229211793281,
      "loss": 0.537,
      "num_input_tokens_seen": 50766816,
      "step": 36920
    },
    {
      "epoch": 1.1819025670571666,
      "grad_norm": 0.005930743645876646,
      "learning_rate": 0.0043561252723032405,
      "loss": 0.5926,
      "num_input_tokens_seen": 50773664,
      "step": 36925
    },
    {
      "epoch": 1.182062608027655,
      "grad_norm": 0.004348406568169594,
      "learning_rate": 0.004342043792929001,
      "loss": 0.5325,
      "num_input_tokens_seen": 50780752,
      "step": 36930
    },
    {
      "epoch": 1.1822226489981436,
      "grad_norm": 0.003992403857409954,
      "learning_rate": 0.004327984775842025,
      "loss": 0.5215,
      "num_input_tokens_seen": 50787216,
      "step": 36935
    },
    {
      "epoch": 1.182382689968632,
      "grad_norm": 0.003406292526051402,
      "learning_rate": 0.004313948223210428,
      "loss": 0.3945,
      "num_input_tokens_seen": 50794048,
      "step": 36940
    },
    {
      "epoch": 1.1825427309391203,
      "grad_norm": 0.0035708036739379168,
      "learning_rate": 0.004299934137198846,
      "loss": 0.5447,
      "num_input_tokens_seen": 50801120,
      "step": 36945
    },
    {
      "epoch": 1.1827027719096088,
      "grad_norm": 0.010761725716292858,
      "learning_rate": 0.004285942519968383,
      "loss": 0.5203,
      "num_input_tokens_seen": 50807936,
      "step": 36950
    },
    {
      "epoch": 1.1828628128800973,
      "grad_norm": 0.0031546212267130613,
      "learning_rate": 0.004271973373676746,
      "loss": 0.4241,
      "num_input_tokens_seen": 50814528,
      "step": 36955
    },
    {
      "epoch": 1.1830228538505858,
      "grad_norm": 0.002430264139547944,
      "learning_rate": 0.004258026700478146,
      "loss": 0.3486,
      "num_input_tokens_seen": 50821472,
      "step": 36960
    },
    {
      "epoch": 1.183182894821074,
      "grad_norm": 0.005086023360490799,
      "learning_rate": 0.004244102502523328,
      "loss": 0.434,
      "num_input_tokens_seen": 50828560,
      "step": 36965
    },
    {
      "epoch": 1.1833429357915626,
      "grad_norm": 0.0035805513616651297,
      "learning_rate": 0.004230200781959592,
      "loss": 0.568,
      "num_input_tokens_seen": 50835232,
      "step": 36970
    },
    {
      "epoch": 1.183502976762051,
      "grad_norm": 0.004182775970548391,
      "learning_rate": 0.004216321540930756,
      "loss": 0.367,
      "num_input_tokens_seen": 50842112,
      "step": 36975
    },
    {
      "epoch": 1.1836630177325396,
      "grad_norm": 0.012934768572449684,
      "learning_rate": 0.004202464781577175,
      "loss": 0.4457,
      "num_input_tokens_seen": 50848960,
      "step": 36980
    },
    {
      "epoch": 1.183823058703028,
      "grad_norm": 0.0015873006777837873,
      "learning_rate": 0.00418863050603574,
      "loss": 0.3147,
      "num_input_tokens_seen": 50856176,
      "step": 36985
    },
    {
      "epoch": 1.1839830996735163,
      "grad_norm": 0.00847261305898428,
      "learning_rate": 0.004174818716439843,
      "loss": 0.5428,
      "num_input_tokens_seen": 50863424,
      "step": 36990
    },
    {
      "epoch": 1.1841431406440048,
      "grad_norm": 0.0055985478684306145,
      "learning_rate": 0.004161029414919464,
      "loss": 0.4449,
      "num_input_tokens_seen": 50870144,
      "step": 36995
    },
    {
      "epoch": 1.1843031816144933,
      "grad_norm": 0.006205114535987377,
      "learning_rate": 0.004147262603601071,
      "loss": 0.4731,
      "num_input_tokens_seen": 50876992,
      "step": 37000
    },
    {
      "epoch": 1.1843031816144933,
      "eval_loss": 0.49473559856414795,
      "eval_runtime": 331.6146,
      "eval_samples_per_second": 41.874,
      "eval_steps_per_second": 20.937,
      "num_input_tokens_seen": 50876992,
      "step": 37000
    },
    {
      "epoch": 1.1844632225849818,
      "grad_norm": 0.0035024622920900583,
      "learning_rate": 0.004133518284607679,
      "loss": 0.7116,
      "num_input_tokens_seen": 50884400,
      "step": 37005
    },
    {
      "epoch": 1.18462326355547,
      "grad_norm": 0.0070081427693367004,
      "learning_rate": 0.004119796460058861,
      "loss": 0.646,
      "num_input_tokens_seen": 50891680,
      "step": 37010
    },
    {
      "epoch": 1.1847833045259586,
      "grad_norm": 0.005415678024291992,
      "learning_rate": 0.00410609713207064,
      "loss": 0.4292,
      "num_input_tokens_seen": 50898592,
      "step": 37015
    },
    {
      "epoch": 1.184943345496447,
      "grad_norm": 0.006048806942999363,
      "learning_rate": 0.004092420302755678,
      "loss": 0.514,
      "num_input_tokens_seen": 50905376,
      "step": 37020
    },
    {
      "epoch": 1.1851033864669356,
      "grad_norm": 0.005015702452510595,
      "learning_rate": 0.004078765974223103,
      "loss": 0.3633,
      "num_input_tokens_seen": 50912064,
      "step": 37025
    },
    {
      "epoch": 1.185263427437424,
      "grad_norm": 0.0058105215430259705,
      "learning_rate": 0.004065134148578564,
      "loss": 0.6136,
      "num_input_tokens_seen": 50919072,
      "step": 37030
    },
    {
      "epoch": 1.1854234684079124,
      "grad_norm": 0.004524353425949812,
      "learning_rate": 0.004051524827924279,
      "loss": 0.4807,
      "num_input_tokens_seen": 50926064,
      "step": 37035
    },
    {
      "epoch": 1.1855835093784008,
      "grad_norm": 0.006399575620889664,
      "learning_rate": 0.004037938014358955,
      "loss": 0.3596,
      "num_input_tokens_seen": 50932832,
      "step": 37040
    },
    {
      "epoch": 1.1857435503488893,
      "grad_norm": 0.007814478129148483,
      "learning_rate": 0.004024373709977863,
      "loss": 0.4873,
      "num_input_tokens_seen": 50939488,
      "step": 37045
    },
    {
      "epoch": 1.1859035913193778,
      "grad_norm": 0.0038421114441007376,
      "learning_rate": 0.004010831916872814,
      "loss": 0.3821,
      "num_input_tokens_seen": 50946144,
      "step": 37050
    },
    {
      "epoch": 1.186063632289866,
      "grad_norm": 0.005414944142103195,
      "learning_rate": 0.003997312637132089,
      "loss": 0.5429,
      "num_input_tokens_seen": 50952864,
      "step": 37055
    },
    {
      "epoch": 1.1862236732603546,
      "grad_norm": 0.0029421336948871613,
      "learning_rate": 0.003983815872840535,
      "loss": 0.6438,
      "num_input_tokens_seen": 50959488,
      "step": 37060
    },
    {
      "epoch": 1.186383714230843,
      "grad_norm": 0.0025893552228808403,
      "learning_rate": 0.003970341626079521,
      "loss": 0.3395,
      "num_input_tokens_seen": 50966336,
      "step": 37065
    },
    {
      "epoch": 1.1865437552013316,
      "grad_norm": 0.003167859511449933,
      "learning_rate": 0.003956889898926952,
      "loss": 0.5652,
      "num_input_tokens_seen": 50972960,
      "step": 37070
    },
    {
      "epoch": 1.18670379617182,
      "grad_norm": 0.00364072248339653,
      "learning_rate": 0.0039434606934572675,
      "loss": 0.5676,
      "num_input_tokens_seen": 50979696,
      "step": 37075
    },
    {
      "epoch": 1.1868638371423084,
      "grad_norm": 0.004356313496828079,
      "learning_rate": 0.003930054011741396,
      "loss": 0.3929,
      "num_input_tokens_seen": 50986704,
      "step": 37080
    },
    {
      "epoch": 1.1870238781127969,
      "grad_norm": 0.00578012689948082,
      "learning_rate": 0.0039166698558468155,
      "loss": 0.473,
      "num_input_tokens_seen": 50994384,
      "step": 37085
    },
    {
      "epoch": 1.1871839190832854,
      "grad_norm": 0.0027666911482810974,
      "learning_rate": 0.0039033082278375594,
      "loss": 0.3565,
      "num_input_tokens_seen": 51001232,
      "step": 37090
    },
    {
      "epoch": 1.1873439600537739,
      "grad_norm": 0.0033735681790858507,
      "learning_rate": 0.003889969129774112,
      "loss": 0.4245,
      "num_input_tokens_seen": 51007920,
      "step": 37095
    },
    {
      "epoch": 1.1875040010242621,
      "grad_norm": 0.005647812969982624,
      "learning_rate": 0.0038766525637135784,
      "loss": 0.4284,
      "num_input_tokens_seen": 51015216,
      "step": 37100
    },
    {
      "epoch": 1.1876640419947506,
      "grad_norm": 0.0054421694949269295,
      "learning_rate": 0.0038633585317095318,
      "loss": 0.477,
      "num_input_tokens_seen": 51022224,
      "step": 37105
    },
    {
      "epoch": 1.1878240829652391,
      "grad_norm": 0.0039529805071651936,
      "learning_rate": 0.00385008703581205,
      "loss": 0.3215,
      "num_input_tokens_seen": 51028768,
      "step": 37110
    },
    {
      "epoch": 1.1879841239357276,
      "grad_norm": 0.004323923960328102,
      "learning_rate": 0.0038368380780677944,
      "loss": 0.4727,
      "num_input_tokens_seen": 51035680,
      "step": 37115
    },
    {
      "epoch": 1.188144164906216,
      "grad_norm": 0.004896258469671011,
      "learning_rate": 0.003823611660519882,
      "loss": 0.5166,
      "num_input_tokens_seen": 51042368,
      "step": 37120
    },
    {
      "epoch": 1.1883042058767044,
      "grad_norm": 0.004491232335567474,
      "learning_rate": 0.0038104077852080475,
      "loss": 0.3922,
      "num_input_tokens_seen": 51049072,
      "step": 37125
    },
    {
      "epoch": 1.1884642468471929,
      "grad_norm": 0.0049053519032895565,
      "learning_rate": 0.003797226454168462,
      "loss": 0.4798,
      "num_input_tokens_seen": 51055920,
      "step": 37130
    },
    {
      "epoch": 1.1886242878176814,
      "grad_norm": 0.008300881832838058,
      "learning_rate": 0.003784067669433849,
      "loss": 0.5636,
      "num_input_tokens_seen": 51063376,
      "step": 37135
    },
    {
      "epoch": 1.1887843287881699,
      "grad_norm": 0.0032347622327506542,
      "learning_rate": 0.0037709314330334528,
      "loss": 0.4173,
      "num_input_tokens_seen": 51070304,
      "step": 37140
    },
    {
      "epoch": 1.1889443697586581,
      "grad_norm": 0.0023680638987571,
      "learning_rate": 0.003757817746993086,
      "loss": 0.2776,
      "num_input_tokens_seen": 51076992,
      "step": 37145
    },
    {
      "epoch": 1.1891044107291466,
      "grad_norm": 0.006647027563303709,
      "learning_rate": 0.0037447266133349977,
      "loss": 0.4789,
      "num_input_tokens_seen": 51083616,
      "step": 37150
    },
    {
      "epoch": 1.1892644516996351,
      "grad_norm": 0.004438698757439852,
      "learning_rate": 0.003731658034078039,
      "loss": 0.3741,
      "num_input_tokens_seen": 51090080,
      "step": 37155
    },
    {
      "epoch": 1.1894244926701236,
      "grad_norm": 0.0059242891147732735,
      "learning_rate": 0.0037186120112375153,
      "loss": 0.4438,
      "num_input_tokens_seen": 51097184,
      "step": 37160
    },
    {
      "epoch": 1.1895845336406121,
      "grad_norm": 0.004264279268682003,
      "learning_rate": 0.003705588546825317,
      "loss": 0.3661,
      "num_input_tokens_seen": 51104064,
      "step": 37165
    },
    {
      "epoch": 1.1897445746111004,
      "grad_norm": 0.0044828178361058235,
      "learning_rate": 0.0036925876428498205,
      "loss": 0.3928,
      "num_input_tokens_seen": 51110896,
      "step": 37170
    },
    {
      "epoch": 1.1899046155815889,
      "grad_norm": 0.0048158904537558556,
      "learning_rate": 0.0036796093013159057,
      "loss": 0.4014,
      "num_input_tokens_seen": 51117936,
      "step": 37175
    },
    {
      "epoch": 1.1900646565520774,
      "grad_norm": 0.0055523342452943325,
      "learning_rate": 0.0036666535242250217,
      "loss": 0.4357,
      "num_input_tokens_seen": 51124848,
      "step": 37180
    },
    {
      "epoch": 1.1902246975225657,
      "grad_norm": 0.007943877018988132,
      "learning_rate": 0.003653720313575104,
      "loss": 0.6933,
      "num_input_tokens_seen": 51132064,
      "step": 37185
    },
    {
      "epoch": 1.1903847384930542,
      "grad_norm": 0.004285513423383236,
      "learning_rate": 0.003640809671360623,
      "loss": 0.6558,
      "num_input_tokens_seen": 51139168,
      "step": 37190
    },
    {
      "epoch": 1.1905447794635426,
      "grad_norm": 0.0016548354178667068,
      "learning_rate": 0.003627921599572553,
      "loss": 0.534,
      "num_input_tokens_seen": 51146240,
      "step": 37195
    },
    {
      "epoch": 1.1907048204340311,
      "grad_norm": 0.0032153252977877855,
      "learning_rate": 0.003615056100198405,
      "loss": 0.5226,
      "num_input_tokens_seen": 51153296,
      "step": 37200
    },
    {
      "epoch": 1.1907048204340311,
      "eval_loss": 0.49478664994239807,
      "eval_runtime": 332.201,
      "eval_samples_per_second": 41.8,
      "eval_steps_per_second": 20.9,
      "num_input_tokens_seen": 51153296,
      "step": 37200
    },
    {
      "epoch": 1.1908648614045196,
      "grad_norm": 0.00698826601728797,
      "learning_rate": 0.003602213175222174,
      "loss": 0.4391,
      "num_input_tokens_seen": 51160400,
      "step": 37205
    },
    {
      "epoch": 1.191024902375008,
      "grad_norm": 0.0034265269059687853,
      "learning_rate": 0.0035893928266244432,
      "loss": 0.2921,
      "num_input_tokens_seen": 51166800,
      "step": 37210
    },
    {
      "epoch": 1.1911849433454964,
      "grad_norm": 0.004622171632945538,
      "learning_rate": 0.003576595056382248,
      "loss": 0.4306,
      "num_input_tokens_seen": 51173648,
      "step": 37215
    },
    {
      "epoch": 1.191344984315985,
      "grad_norm": 0.003535840194672346,
      "learning_rate": 0.0035638198664691423,
      "loss": 0.6976,
      "num_input_tokens_seen": 51180288,
      "step": 37220
    },
    {
      "epoch": 1.1915050252864734,
      "grad_norm": 0.0041128601878881454,
      "learning_rate": 0.003551067258855267,
      "loss": 0.4578,
      "num_input_tokens_seen": 51187440,
      "step": 37225
    },
    {
      "epoch": 1.1916650662569617,
      "grad_norm": 0.005069581791758537,
      "learning_rate": 0.0035383372355071996,
      "loss": 0.5181,
      "num_input_tokens_seen": 51194000,
      "step": 37230
    },
    {
      "epoch": 1.1918251072274502,
      "grad_norm": 0.0056878263130784035,
      "learning_rate": 0.0035256297983881023,
      "loss": 0.347,
      "num_input_tokens_seen": 51200880,
      "step": 37235
    },
    {
      "epoch": 1.1919851481979387,
      "grad_norm": 0.002794434316456318,
      "learning_rate": 0.0035129449494575747,
      "loss": 0.5085,
      "num_input_tokens_seen": 51207440,
      "step": 37240
    },
    {
      "epoch": 1.1921451891684272,
      "grad_norm": 0.003606947371736169,
      "learning_rate": 0.0035002826906718187,
      "loss": 0.5107,
      "num_input_tokens_seen": 51214432,
      "step": 37245
    },
    {
      "epoch": 1.1923052301389157,
      "grad_norm": 0.007273622788488865,
      "learning_rate": 0.003487643023983522,
      "loss": 0.5703,
      "num_input_tokens_seen": 51221360,
      "step": 37250
    },
    {
      "epoch": 1.192465271109404,
      "grad_norm": 0.003216383047401905,
      "learning_rate": 0.003475025951341842,
      "loss": 0.3114,
      "num_input_tokens_seen": 51228384,
      "step": 37255
    },
    {
      "epoch": 1.1926253120798924,
      "grad_norm": 0.0017874775221571326,
      "learning_rate": 0.00346243147469249,
      "loss": 0.4918,
      "num_input_tokens_seen": 51235120,
      "step": 37260
    },
    {
      "epoch": 1.192785353050381,
      "grad_norm": 0.006030111573636532,
      "learning_rate": 0.0034498595959777446,
      "loss": 0.2973,
      "num_input_tokens_seen": 51241776,
      "step": 37265
    },
    {
      "epoch": 1.1929453940208694,
      "grad_norm": 0.00451083667576313,
      "learning_rate": 0.003437310317136305,
      "loss": 0.5419,
      "num_input_tokens_seen": 51248352,
      "step": 37270
    },
    {
      "epoch": 1.1931054349913577,
      "grad_norm": 0.005437041167169809,
      "learning_rate": 0.0034247836401034236,
      "loss": 0.428,
      "num_input_tokens_seen": 51254976,
      "step": 37275
    },
    {
      "epoch": 1.1932654759618462,
      "grad_norm": 0.004491240251809359,
      "learning_rate": 0.003412279566810905,
      "loss": 0.5486,
      "num_input_tokens_seen": 51262144,
      "step": 37280
    },
    {
      "epoch": 1.1934255169323347,
      "grad_norm": 0.005862137768417597,
      "learning_rate": 0.00339979809918699,
      "loss": 0.5623,
      "num_input_tokens_seen": 51268800,
      "step": 37285
    },
    {
      "epoch": 1.1935855579028232,
      "grad_norm": 0.0034641604870557785,
      "learning_rate": 0.0033873392391565228,
      "loss": 0.4963,
      "num_input_tokens_seen": 51275776,
      "step": 37290
    },
    {
      "epoch": 1.1937455988733117,
      "grad_norm": 0.003106712829321623,
      "learning_rate": 0.003374902988640782,
      "loss": 0.4109,
      "num_input_tokens_seen": 51282672,
      "step": 37295
    },
    {
      "epoch": 1.1939056398438,
      "grad_norm": 0.0047300816513597965,
      "learning_rate": 0.0033624893495576014,
      "loss": 0.7242,
      "num_input_tokens_seen": 51290032,
      "step": 37300
    },
    {
      "epoch": 1.1940656808142884,
      "grad_norm": 0.009214648976922035,
      "learning_rate": 0.0033500983238213323,
      "loss": 0.5466,
      "num_input_tokens_seen": 51296704,
      "step": 37305
    },
    {
      "epoch": 1.194225721784777,
      "grad_norm": 0.004711587447673082,
      "learning_rate": 0.0033377299133428126,
      "loss": 0.42,
      "num_input_tokens_seen": 51303520,
      "step": 37310
    },
    {
      "epoch": 1.1943857627552654,
      "grad_norm": 0.0031406311318278313,
      "learning_rate": 0.003325384120029434,
      "loss": 0.5193,
      "num_input_tokens_seen": 51310320,
      "step": 37315
    },
    {
      "epoch": 1.1945458037257537,
      "grad_norm": 0.004502577241510153,
      "learning_rate": 0.0033130609457850233,
      "loss": 0.4585,
      "num_input_tokens_seen": 51317488,
      "step": 37320
    },
    {
      "epoch": 1.1947058446962422,
      "grad_norm": 0.00774729810655117,
      "learning_rate": 0.0033007603925100104,
      "loss": 0.6162,
      "num_input_tokens_seen": 51324704,
      "step": 37325
    },
    {
      "epoch": 1.1948658856667307,
      "grad_norm": 0.004502689465880394,
      "learning_rate": 0.003288482462101294,
      "loss": 0.404,
      "num_input_tokens_seen": 51331472,
      "step": 37330
    },
    {
      "epoch": 1.1950259266372192,
      "grad_norm": 0.005096551030874252,
      "learning_rate": 0.0032762271564522605,
      "loss": 0.5208,
      "num_input_tokens_seen": 51338384,
      "step": 37335
    },
    {
      "epoch": 1.1951859676077077,
      "grad_norm": 0.005679071880877018,
      "learning_rate": 0.003263994477452864,
      "loss": 0.4537,
      "num_input_tokens_seen": 51345296,
      "step": 37340
    },
    {
      "epoch": 1.195346008578196,
      "grad_norm": 0.002985923318192363,
      "learning_rate": 0.0032517844269895125,
      "loss": 0.5476,
      "num_input_tokens_seen": 51352416,
      "step": 37345
    },
    {
      "epoch": 1.1955060495486844,
      "grad_norm": 0.002346570137888193,
      "learning_rate": 0.0032395970069451496,
      "loss": 0.3492,
      "num_input_tokens_seen": 51359504,
      "step": 37350
    },
    {
      "epoch": 1.195666090519173,
      "grad_norm": 0.004166434518992901,
      "learning_rate": 0.0032274322191992388,
      "loss": 0.545,
      "num_input_tokens_seen": 51366144,
      "step": 37355
    },
    {
      "epoch": 1.1958261314896614,
      "grad_norm": 0.003997460473328829,
      "learning_rate": 0.0032152900656277294,
      "loss": 0.6567,
      "num_input_tokens_seen": 51372752,
      "step": 37360
    },
    {
      "epoch": 1.1959861724601497,
      "grad_norm": 0.0030404478311538696,
      "learning_rate": 0.0032031705481030902,
      "loss": 0.5782,
      "num_input_tokens_seen": 51380032,
      "step": 37365
    },
    {
      "epoch": 1.1961462134306382,
      "grad_norm": 0.005390360951423645,
      "learning_rate": 0.0031910736684943428,
      "loss": 0.4855,
      "num_input_tokens_seen": 51387104,
      "step": 37370
    },
    {
      "epoch": 1.1963062544011267,
      "grad_norm": 0.0028043242637068033,
      "learning_rate": 0.0031789994286669453,
      "loss": 0.4279,
      "num_input_tokens_seen": 51393872,
      "step": 37375
    },
    {
      "epoch": 1.1964662953716152,
      "grad_norm": 0.0023232344537973404,
      "learning_rate": 0.003166947830482908,
      "loss": 0.6431,
      "num_input_tokens_seen": 51400848,
      "step": 37380
    },
    {
      "epoch": 1.1966263363421037,
      "grad_norm": 0.006884138099849224,
      "learning_rate": 0.003154918875800727,
      "loss": 0.4705,
      "num_input_tokens_seen": 51407632,
      "step": 37385
    },
    {
      "epoch": 1.196786377312592,
      "grad_norm": 0.0034040044993162155,
      "learning_rate": 0.00314291256647542,
      "loss": 0.3862,
      "num_input_tokens_seen": 51414352,
      "step": 37390
    },
    {
      "epoch": 1.1969464182830805,
      "grad_norm": 0.00753494817763567,
      "learning_rate": 0.0031309289043585375,
      "loss": 0.421,
      "num_input_tokens_seen": 51421136,
      "step": 37395
    },
    {
      "epoch": 1.197106459253569,
      "grad_norm": 0.004166200757026672,
      "learning_rate": 0.003118967891298069,
      "loss": 0.5659,
      "num_input_tokens_seen": 51427552,
      "step": 37400
    },
    {
      "epoch": 1.197106459253569,
      "eval_loss": 0.49485260248184204,
      "eval_runtime": 332.067,
      "eval_samples_per_second": 41.817,
      "eval_steps_per_second": 20.908,
      "num_input_tokens_seen": 51427552,
      "step": 37400
    },
    {
      "epoch": 1.1972665002240575,
      "grad_norm": 0.0041641429997980595,
      "learning_rate": 0.003107029529138572,
      "loss": 0.3761,
      "num_input_tokens_seen": 51434368,
      "step": 37405
    },
    {
      "epoch": 1.1974265411945457,
      "grad_norm": 0.0038380632176995277,
      "learning_rate": 0.0030951138197211235,
      "loss": 0.4212,
      "num_input_tokens_seen": 51441008,
      "step": 37410
    },
    {
      "epoch": 1.1975865821650342,
      "grad_norm": 0.004098996054381132,
      "learning_rate": 0.0030832207648832377,
      "loss": 0.374,
      "num_input_tokens_seen": 51447856,
      "step": 37415
    },
    {
      "epoch": 1.1977466231355227,
      "grad_norm": 0.003653896739706397,
      "learning_rate": 0.0030713503664589635,
      "loss": 0.4525,
      "num_input_tokens_seen": 51454752,
      "step": 37420
    },
    {
      "epoch": 1.1979066641060112,
      "grad_norm": 0.004703962709754705,
      "learning_rate": 0.0030595026262788872,
      "loss": 0.4086,
      "num_input_tokens_seen": 51461392,
      "step": 37425
    },
    {
      "epoch": 1.1980667050764997,
      "grad_norm": 0.004945757798850536,
      "learning_rate": 0.00304767754617008,
      "loss": 0.6059,
      "num_input_tokens_seen": 51468896,
      "step": 37430
    },
    {
      "epoch": 1.198226746046988,
      "grad_norm": 0.0057912953197956085,
      "learning_rate": 0.003035875127956117,
      "loss": 0.4457,
      "num_input_tokens_seen": 51475328,
      "step": 37435
    },
    {
      "epoch": 1.1983867870174765,
      "grad_norm": 0.004716579336673021,
      "learning_rate": 0.0030240953734570752,
      "loss": 0.3249,
      "num_input_tokens_seen": 51481888,
      "step": 37440
    },
    {
      "epoch": 1.198546827987965,
      "grad_norm": 0.002302904613316059,
      "learning_rate": 0.003012338284489535,
      "loss": 0.4658,
      "num_input_tokens_seen": 51489088,
      "step": 37445
    },
    {
      "epoch": 1.1987068689584532,
      "grad_norm": 0.004434864968061447,
      "learning_rate": 0.0030006038628665964,
      "loss": 0.5201,
      "num_input_tokens_seen": 51495712,
      "step": 37450
    },
    {
      "epoch": 1.1988669099289417,
      "grad_norm": 0.006176404654979706,
      "learning_rate": 0.002988892110397845,
      "loss": 0.505,
      "num_input_tokens_seen": 51502624,
      "step": 37455
    },
    {
      "epoch": 1.1990269508994302,
      "grad_norm": 0.0015338921220973134,
      "learning_rate": 0.0029772030288894025,
      "loss": 0.3923,
      "num_input_tokens_seen": 51509520,
      "step": 37460
    },
    {
      "epoch": 1.1991869918699187,
      "grad_norm": 0.00706163514405489,
      "learning_rate": 0.0029655366201438438,
      "loss": 0.534,
      "num_input_tokens_seen": 51516320,
      "step": 37465
    },
    {
      "epoch": 1.1993470328404072,
      "grad_norm": 0.003672623308375478,
      "learning_rate": 0.0029538928859602965,
      "loss": 0.4683,
      "num_input_tokens_seen": 51523456,
      "step": 37470
    },
    {
      "epoch": 1.1995070738108955,
      "grad_norm": 0.00635412335395813,
      "learning_rate": 0.002942271828134374,
      "loss": 0.5349,
      "num_input_tokens_seen": 51530624,
      "step": 37475
    },
    {
      "epoch": 1.199667114781384,
      "grad_norm": 0.013976378366351128,
      "learning_rate": 0.00293067344845816,
      "loss": 0.6264,
      "num_input_tokens_seen": 51537856,
      "step": 37480
    },
    {
      "epoch": 1.1998271557518725,
      "grad_norm": 0.0031386043410748243,
      "learning_rate": 0.0029190977487202896,
      "loss": 0.4769,
      "num_input_tokens_seen": 51544448,
      "step": 37485
    },
    {
      "epoch": 1.199987196722361,
      "grad_norm": 0.005817289464175701,
      "learning_rate": 0.0029075447307058853,
      "loss": 0.4973,
      "num_input_tokens_seen": 51551584,
      "step": 37490
    },
    {
      "epoch": 1.2001472376928493,
      "grad_norm": 0.004557703621685505,
      "learning_rate": 0.0028960143961965722,
      "loss": 0.3873,
      "num_input_tokens_seen": 51558096,
      "step": 37495
    },
    {
      "epoch": 1.2003072786633378,
      "grad_norm": 0.0026946936268359423,
      "learning_rate": 0.002884506746970461,
      "loss": 0.4046,
      "num_input_tokens_seen": 51564944,
      "step": 37500
    },
    {
      "epoch": 1.2004673196338262,
      "grad_norm": 0.002609725808724761,
      "learning_rate": 0.0028730217848021654,
      "loss": 0.2946,
      "num_input_tokens_seen": 51572144,
      "step": 37505
    },
    {
      "epoch": 1.2006273606043147,
      "grad_norm": 0.006738999392837286,
      "learning_rate": 0.0028615595114628188,
      "loss": 0.6318,
      "num_input_tokens_seen": 51579056,
      "step": 37510
    },
    {
      "epoch": 1.2007874015748032,
      "grad_norm": 0.002239349763840437,
      "learning_rate": 0.002850119928720074,
      "loss": 0.5031,
      "num_input_tokens_seen": 51585840,
      "step": 37515
    },
    {
      "epoch": 1.2009474425452915,
      "grad_norm": 0.003337301779538393,
      "learning_rate": 0.0028387030383380195,
      "loss": 0.5057,
      "num_input_tokens_seen": 51592464,
      "step": 37520
    },
    {
      "epoch": 1.20110748351578,
      "grad_norm": 0.005562020465731621,
      "learning_rate": 0.0028273088420772974,
      "loss": 0.5529,
      "num_input_tokens_seen": 51599568,
      "step": 37525
    },
    {
      "epoch": 1.2012675244862685,
      "grad_norm": 0.002198809524998069,
      "learning_rate": 0.002815937341695068,
      "loss": 0.4444,
      "num_input_tokens_seen": 51606128,
      "step": 37530
    },
    {
      "epoch": 1.201427565456757,
      "grad_norm": 0.01072604488581419,
      "learning_rate": 0.0028045885389448963,
      "loss": 0.4015,
      "num_input_tokens_seen": 51612880,
      "step": 37535
    },
    {
      "epoch": 1.2015876064272453,
      "grad_norm": 0.006756288930773735,
      "learning_rate": 0.002793262435576965,
      "loss": 0.5342,
      "num_input_tokens_seen": 51619552,
      "step": 37540
    },
    {
      "epoch": 1.2017476473977338,
      "grad_norm": 0.0063164737075567245,
      "learning_rate": 0.0027819590333378772,
      "loss": 0.491,
      "num_input_tokens_seen": 51626416,
      "step": 37545
    },
    {
      "epoch": 1.2019076883682223,
      "grad_norm": 0.0045535326935350895,
      "learning_rate": 0.002770678333970755,
      "loss": 0.4083,
      "num_input_tokens_seen": 51633184,
      "step": 37550
    },
    {
      "epoch": 1.2020677293387108,
      "grad_norm": 0.0043660718947649,
      "learning_rate": 0.0027594203392152573,
      "loss": 0.509,
      "num_input_tokens_seen": 51640704,
      "step": 37555
    },
    {
      "epoch": 1.2022277703091993,
      "grad_norm": 0.0031886848155409098,
      "learning_rate": 0.002748185050807478,
      "loss": 0.3967,
      "num_input_tokens_seen": 51647856,
      "step": 37560
    },
    {
      "epoch": 1.2023878112796875,
      "grad_norm": 0.0023159754928201437,
      "learning_rate": 0.002736972470480031,
      "loss": 0.4207,
      "num_input_tokens_seen": 51654720,
      "step": 37565
    },
    {
      "epoch": 1.202547852250176,
      "grad_norm": 0.008807273581624031,
      "learning_rate": 0.002725782599962068,
      "loss": 0.4309,
      "num_input_tokens_seen": 51661520,
      "step": 37570
    },
    {
      "epoch": 1.2027078932206645,
      "grad_norm": 0.0034483091440051794,
      "learning_rate": 0.0027146154409791734,
      "loss": 0.3741,
      "num_input_tokens_seen": 51668464,
      "step": 37575
    },
    {
      "epoch": 1.202867934191153,
      "grad_norm": 0.0044174399226903915,
      "learning_rate": 0.002703470995253504,
      "loss": 0.4649,
      "num_input_tokens_seen": 51675584,
      "step": 37580
    },
    {
      "epoch": 1.2030279751616413,
      "grad_norm": 0.003903629258275032,
      "learning_rate": 0.0026923492645036184,
      "loss": 0.6037,
      "num_input_tokens_seen": 51682608,
      "step": 37585
    },
    {
      "epoch": 1.2031880161321298,
      "grad_norm": 0.00546781113371253,
      "learning_rate": 0.0026812502504446776,
      "loss": 0.9068,
      "num_input_tokens_seen": 51693216,
      "step": 37590
    },
    {
      "epoch": 1.2033480571026183,
      "grad_norm": 0.003576785558834672,
      "learning_rate": 0.0026701739547882798,
      "loss": 0.4642,
      "num_input_tokens_seen": 51700160,
      "step": 37595
    },
    {
      "epoch": 1.2035080980731068,
      "grad_norm": 0.005472043063491583,
      "learning_rate": 0.0026591203792425077,
      "loss": 0.5782,
      "num_input_tokens_seen": 51707088,
      "step": 37600
    },
    {
      "epoch": 1.2035080980731068,
      "eval_loss": 0.49448031187057495,
      "eval_runtime": 331.5728,
      "eval_samples_per_second": 41.879,
      "eval_steps_per_second": 20.94,
      "num_input_tokens_seen": 51707088,
      "step": 37600
    },
    {
      "epoch": 1.2036681390435953,
      "grad_norm": 0.00345657323487103,
      "learning_rate": 0.0026480895255119818,
      "loss": 0.4673,
      "num_input_tokens_seen": 51713504,
      "step": 37605
    },
    {
      "epoch": 1.2038281800140835,
      "grad_norm": 0.004961661994457245,
      "learning_rate": 0.002637081395297791,
      "loss": 0.5855,
      "num_input_tokens_seen": 51720144,
      "step": 37610
    },
    {
      "epoch": 1.203988220984572,
      "grad_norm": 0.003531404072418809,
      "learning_rate": 0.0026260959902975113,
      "loss": 0.5447,
      "num_input_tokens_seen": 51726992,
      "step": 37615
    },
    {
      "epoch": 1.2041482619550605,
      "grad_norm": 0.00401865690946579,
      "learning_rate": 0.00261513331220527,
      "loss": 0.4375,
      "num_input_tokens_seen": 51733552,
      "step": 37620
    },
    {
      "epoch": 1.204308302925549,
      "grad_norm": 0.00800019409507513,
      "learning_rate": 0.0026041933627116154,
      "loss": 0.5474,
      "num_input_tokens_seen": 51740096,
      "step": 37625
    },
    {
      "epoch": 1.2044683438960373,
      "grad_norm": 0.004533146042376757,
      "learning_rate": 0.0025932761435036476,
      "loss": 0.492,
      "num_input_tokens_seen": 51747488,
      "step": 37630
    },
    {
      "epoch": 1.2046283848665258,
      "grad_norm": 0.003319478826597333,
      "learning_rate": 0.002582381656264904,
      "loss": 0.3928,
      "num_input_tokens_seen": 51754912,
      "step": 37635
    },
    {
      "epoch": 1.2047884258370143,
      "grad_norm": 0.006012297701090574,
      "learning_rate": 0.0025715099026754895,
      "loss": 0.6827,
      "num_input_tokens_seen": 51761760,
      "step": 37640
    },
    {
      "epoch": 1.2049484668075028,
      "grad_norm": 0.002202601172029972,
      "learning_rate": 0.002560660884411947,
      "loss": 0.2867,
      "num_input_tokens_seen": 51768624,
      "step": 37645
    },
    {
      "epoch": 1.2051085077779913,
      "grad_norm": 0.004860536195337772,
      "learning_rate": 0.0025498346031473385,
      "loss": 0.265,
      "num_input_tokens_seen": 51776032,
      "step": 37650
    },
    {
      "epoch": 1.2052685487484796,
      "grad_norm": 0.003069827100262046,
      "learning_rate": 0.0025390310605511945,
      "loss": 0.4473,
      "num_input_tokens_seen": 51782800,
      "step": 37655
    },
    {
      "epoch": 1.205428589718968,
      "grad_norm": 0.008505474776029587,
      "learning_rate": 0.0025282502582895995,
      "loss": 0.5008,
      "num_input_tokens_seen": 51789568,
      "step": 37660
    },
    {
      "epoch": 1.2055886306894565,
      "grad_norm": 0.006629033945500851,
      "learning_rate": 0.002517492198025023,
      "loss": 0.5283,
      "num_input_tokens_seen": 51796272,
      "step": 37665
    },
    {
      "epoch": 1.205748671659945,
      "grad_norm": 0.0036985757760703564,
      "learning_rate": 0.0025067568814165554,
      "loss": 0.4398,
      "num_input_tokens_seen": 51803088,
      "step": 37670
    },
    {
      "epoch": 1.2059087126304333,
      "grad_norm": 0.005220736376941204,
      "learning_rate": 0.0024960443101196884,
      "loss": 0.5933,
      "num_input_tokens_seen": 51809968,
      "step": 37675
    },
    {
      "epoch": 1.2060687536009218,
      "grad_norm": 0.005303553305566311,
      "learning_rate": 0.002485354485786434,
      "loss": 0.4644,
      "num_input_tokens_seen": 51817664,
      "step": 37680
    },
    {
      "epoch": 1.2062287945714103,
      "grad_norm": 0.0039671617560088634,
      "learning_rate": 0.002474687410065307,
      "loss": 0.4541,
      "num_input_tokens_seen": 51824464,
      "step": 37685
    },
    {
      "epoch": 1.2063888355418988,
      "grad_norm": 0.004118709824979305,
      "learning_rate": 0.002464043084601308,
      "loss": 0.4769,
      "num_input_tokens_seen": 51831360,
      "step": 37690
    },
    {
      "epoch": 1.2065488765123873,
      "grad_norm": 0.0028447227086871862,
      "learning_rate": 0.0024534215110358915,
      "loss": 0.3701,
      "num_input_tokens_seen": 51838528,
      "step": 37695
    },
    {
      "epoch": 1.2067089174828756,
      "grad_norm": 0.0062091476283967495,
      "learning_rate": 0.002442822691007096,
      "loss": 0.4877,
      "num_input_tokens_seen": 51845344,
      "step": 37700
    },
    {
      "epoch": 1.206868958453364,
      "grad_norm": 0.003849530592560768,
      "learning_rate": 0.002432246626149348,
      "loss": 0.6433,
      "num_input_tokens_seen": 51852096,
      "step": 37705
    },
    {
      "epoch": 1.2070289994238526,
      "grad_norm": 0.006120924372226,
      "learning_rate": 0.002421693318093626,
      "loss": 0.5117,
      "num_input_tokens_seen": 51858976,
      "step": 37710
    },
    {
      "epoch": 1.2071890403943408,
      "grad_norm": 0.005292080342769623,
      "learning_rate": 0.0024111627684673784,
      "loss": 0.5391,
      "num_input_tokens_seen": 51865360,
      "step": 37715
    },
    {
      "epoch": 1.2073490813648293,
      "grad_norm": 0.005032422952353954,
      "learning_rate": 0.0024006549788945395,
      "loss": 0.3384,
      "num_input_tokens_seen": 51872688,
      "step": 37720
    },
    {
      "epoch": 1.2075091223353178,
      "grad_norm": 0.0034779792185872793,
      "learning_rate": 0.0023901699509955463,
      "loss": 0.4699,
      "num_input_tokens_seen": 51879552,
      "step": 37725
    },
    {
      "epoch": 1.2076691633058063,
      "grad_norm": 0.004116583149880171,
      "learning_rate": 0.0023797076863873554,
      "loss": 0.4118,
      "num_input_tokens_seen": 51886208,
      "step": 37730
    },
    {
      "epoch": 1.2078292042762948,
      "grad_norm": 0.006547530181705952,
      "learning_rate": 0.0023692681866833262,
      "loss": 0.5768,
      "num_input_tokens_seen": 51893232,
      "step": 37735
    },
    {
      "epoch": 1.207989245246783,
      "grad_norm": 0.005495474208146334,
      "learning_rate": 0.0023588514534934046,
      "loss": 0.4759,
      "num_input_tokens_seen": 51900016,
      "step": 37740
    },
    {
      "epoch": 1.2081492862172716,
      "grad_norm": 0.0030898882541805506,
      "learning_rate": 0.002348457488423955,
      "loss": 0.4169,
      "num_input_tokens_seen": 51906800,
      "step": 37745
    },
    {
      "epoch": 1.20830932718776,
      "grad_norm": 0.0038420758210122585,
      "learning_rate": 0.0023380862930778624,
      "loss": 0.4068,
      "num_input_tokens_seen": 51913296,
      "step": 37750
    },
    {
      "epoch": 1.2084693681582486,
      "grad_norm": 0.0054997969418764114,
      "learning_rate": 0.0023277378690545135,
      "loss": 0.4922,
      "num_input_tokens_seen": 51920224,
      "step": 37755
    },
    {
      "epoch": 1.2086294091287368,
      "grad_norm": 0.005701628979295492,
      "learning_rate": 0.0023174122179497325,
      "loss": 0.4614,
      "num_input_tokens_seen": 51927472,
      "step": 37760
    },
    {
      "epoch": 1.2087894500992253,
      "grad_norm": 0.005697545595467091,
      "learning_rate": 0.0023071093413558784,
      "loss": 0.5023,
      "num_input_tokens_seen": 51934048,
      "step": 37765
    },
    {
      "epoch": 1.2089494910697138,
      "grad_norm": 0.006868821568787098,
      "learning_rate": 0.002296829240861814,
      "loss": 0.4466,
      "num_input_tokens_seen": 51940560,
      "step": 37770
    },
    {
      "epoch": 1.2091095320402023,
      "grad_norm": 0.006649292539805174,
      "learning_rate": 0.002286571918052821,
      "loss": 0.414,
      "num_input_tokens_seen": 51947392,
      "step": 37775
    },
    {
      "epoch": 1.2092695730106908,
      "grad_norm": 0.004237966611981392,
      "learning_rate": 0.0022763373745107174,
      "loss": 0.6077,
      "num_input_tokens_seen": 51954144,
      "step": 37780
    },
    {
      "epoch": 1.209429613981179,
      "grad_norm": 0.004989630077034235,
      "learning_rate": 0.0022661256118138074,
      "loss": 0.4074,
      "num_input_tokens_seen": 51960656,
      "step": 37785
    },
    {
      "epoch": 1.2095896549516676,
      "grad_norm": 0.005081071052700281,
      "learning_rate": 0.0022559366315368645,
      "loss": 0.6488,
      "num_input_tokens_seen": 51967616,
      "step": 37790
    },
    {
      "epoch": 1.209749695922156,
      "grad_norm": 0.0041352505795657635,
      "learning_rate": 0.002245770435251182,
      "loss": 0.2866,
      "num_input_tokens_seen": 51974656,
      "step": 37795
    },
    {
      "epoch": 1.2099097368926446,
      "grad_norm": 0.003549562068656087,
      "learning_rate": 0.002235627024524456,
      "loss": 0.4875,
      "num_input_tokens_seen": 51981712,
      "step": 37800
    },
    {
      "epoch": 1.2099097368926446,
      "eval_loss": 0.4944447875022888,
      "eval_runtime": 332.2779,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 51981712,
      "step": 37800
    },
    {
      "epoch": 1.2100697778631329,
      "grad_norm": 0.002502589952200651,
      "learning_rate": 0.0022255064009209847,
      "loss": 0.5194,
      "num_input_tokens_seen": 51988448,
      "step": 37805
    },
    {
      "epoch": 1.2102298188336214,
      "grad_norm": 0.004654050804674625,
      "learning_rate": 0.0022154085660014864,
      "loss": 0.5509,
      "num_input_tokens_seen": 51994976,
      "step": 37810
    },
    {
      "epoch": 1.2103898598041098,
      "grad_norm": 0.005585929378867149,
      "learning_rate": 0.0022053335213231494,
      "loss": 0.4261,
      "num_input_tokens_seen": 52001776,
      "step": 37815
    },
    {
      "epoch": 1.2105499007745983,
      "grad_norm": 0.006900034844875336,
      "learning_rate": 0.002195281268439697,
      "loss": 0.6202,
      "num_input_tokens_seen": 52008688,
      "step": 37820
    },
    {
      "epoch": 1.2107099417450868,
      "grad_norm": 0.0041821799241006374,
      "learning_rate": 0.002185251808901306,
      "loss": 0.2849,
      "num_input_tokens_seen": 52015312,
      "step": 37825
    },
    {
      "epoch": 1.210869982715575,
      "grad_norm": 0.006498284637928009,
      "learning_rate": 0.0021752451442546227,
      "loss": 0.4063,
      "num_input_tokens_seen": 52021792,
      "step": 37830
    },
    {
      "epoch": 1.2110300236860636,
      "grad_norm": 0.005033135414123535,
      "learning_rate": 0.0021652612760428456,
      "loss": 0.5458,
      "num_input_tokens_seen": 52029088,
      "step": 37835
    },
    {
      "epoch": 1.211190064656552,
      "grad_norm": 0.004876869730651379,
      "learning_rate": 0.0021553002058055603,
      "loss": 0.5603,
      "num_input_tokens_seen": 52035792,
      "step": 37840
    },
    {
      "epoch": 1.2113501056270406,
      "grad_norm": 0.002826679963618517,
      "learning_rate": 0.0021453619350789376,
      "loss": 0.3767,
      "num_input_tokens_seen": 52042864,
      "step": 37845
    },
    {
      "epoch": 1.2115101465975289,
      "grad_norm": 0.004688182380050421,
      "learning_rate": 0.0021354464653955516,
      "loss": 0.4854,
      "num_input_tokens_seen": 52049376,
      "step": 37850
    },
    {
      "epoch": 1.2116701875680174,
      "grad_norm": 0.004747004713863134,
      "learning_rate": 0.002125553798284513,
      "loss": 0.4698,
      "num_input_tokens_seen": 52055936,
      "step": 37855
    },
    {
      "epoch": 1.2118302285385059,
      "grad_norm": 0.004254607949405909,
      "learning_rate": 0.002115683935271384,
      "loss": 0.3603,
      "num_input_tokens_seen": 52062560,
      "step": 37860
    },
    {
      "epoch": 1.2119902695089944,
      "grad_norm": 0.004081706050783396,
      "learning_rate": 0.0021058368778782144,
      "loss": 0.4753,
      "num_input_tokens_seen": 52069120,
      "step": 37865
    },
    {
      "epoch": 1.2121503104794829,
      "grad_norm": 0.006633030250668526,
      "learning_rate": 0.002096012627623539,
      "loss": 0.5133,
      "num_input_tokens_seen": 52075680,
      "step": 37870
    },
    {
      "epoch": 1.2123103514499711,
      "grad_norm": 0.0065363808535039425,
      "learning_rate": 0.00208621118602243,
      "loss": 0.3186,
      "num_input_tokens_seen": 52082400,
      "step": 37875
    },
    {
      "epoch": 1.2124703924204596,
      "grad_norm": 0.0033019694965332747,
      "learning_rate": 0.002076432554586327,
      "loss": 0.5374,
      "num_input_tokens_seen": 52089488,
      "step": 37880
    },
    {
      "epoch": 1.2126304333909481,
      "grad_norm": 0.002624279586598277,
      "learning_rate": 0.002066676734823258,
      "loss": 0.3532,
      "num_input_tokens_seen": 52096288,
      "step": 37885
    },
    {
      "epoch": 1.2127904743614366,
      "grad_norm": 0.0027215206064283848,
      "learning_rate": 0.0020569437282376866,
      "loss": 0.4061,
      "num_input_tokens_seen": 52103216,
      "step": 37890
    },
    {
      "epoch": 1.2129505153319249,
      "grad_norm": 0.00249391864053905,
      "learning_rate": 0.002047233536330545,
      "loss": 0.5874,
      "num_input_tokens_seen": 52110160,
      "step": 37895
    },
    {
      "epoch": 1.2131105563024134,
      "grad_norm": 0.003597497707232833,
      "learning_rate": 0.0020375461605993015,
      "loss": 0.3967,
      "num_input_tokens_seen": 52117040,
      "step": 37900
    },
    {
      "epoch": 1.2132705972729019,
      "grad_norm": 0.00502021424472332,
      "learning_rate": 0.002027881602537845,
      "loss": 0.4844,
      "num_input_tokens_seen": 52123568,
      "step": 37905
    },
    {
      "epoch": 1.2134306382433904,
      "grad_norm": 0.005049047525972128,
      "learning_rate": 0.002018239863636567,
      "loss": 0.4077,
      "num_input_tokens_seen": 52130512,
      "step": 37910
    },
    {
      "epoch": 1.2135906792138789,
      "grad_norm": 0.006411998998373747,
      "learning_rate": 0.002008620945382378,
      "loss": 0.5228,
      "num_input_tokens_seen": 52137728,
      "step": 37915
    },
    {
      "epoch": 1.2137507201843671,
      "grad_norm": 0.004356993827968836,
      "learning_rate": 0.001999024849258607,
      "loss": 0.5118,
      "num_input_tokens_seen": 52144336,
      "step": 37920
    },
    {
      "epoch": 1.2139107611548556,
      "grad_norm": 0.014700163155794144,
      "learning_rate": 0.001989451576745105,
      "loss": 0.6387,
      "num_input_tokens_seen": 52150864,
      "step": 37925
    },
    {
      "epoch": 1.2140708021253441,
      "grad_norm": 0.004069983959197998,
      "learning_rate": 0.00197990112931819,
      "loss": 0.5159,
      "num_input_tokens_seen": 52157408,
      "step": 37930
    },
    {
      "epoch": 1.2142308430958326,
      "grad_norm": 0.006164034362882376,
      "learning_rate": 0.0019703735084506345,
      "loss": 0.5174,
      "num_input_tokens_seen": 52164224,
      "step": 37935
    },
    {
      "epoch": 1.214390884066321,
      "grad_norm": 0.006282851565629244,
      "learning_rate": 0.001960868715611763,
      "loss": 0.3536,
      "num_input_tokens_seen": 52171088,
      "step": 37940
    },
    {
      "epoch": 1.2145509250368094,
      "grad_norm": 0.004910497460514307,
      "learning_rate": 0.0019513867522673034,
      "loss": 0.3817,
      "num_input_tokens_seen": 52178368,
      "step": 37945
    },
    {
      "epoch": 1.2147109660072979,
      "grad_norm": 0.006411328911781311,
      "learning_rate": 0.001941927619879502,
      "loss": 0.4616,
      "num_input_tokens_seen": 52185232,
      "step": 37950
    },
    {
      "epoch": 1.2148710069777864,
      "grad_norm": 0.006674763280898333,
      "learning_rate": 0.0019324913199070758,
      "loss": 0.5261,
      "num_input_tokens_seen": 52192544,
      "step": 37955
    },
    {
      "epoch": 1.2150310479482749,
      "grad_norm": 0.003513848874717951,
      "learning_rate": 0.0019230778538052106,
      "loss": 0.4378,
      "num_input_tokens_seen": 52199520,
      "step": 37960
    },
    {
      "epoch": 1.2151910889187632,
      "grad_norm": 0.004401682410389185,
      "learning_rate": 0.0019136872230255952,
      "loss": 0.3863,
      "num_input_tokens_seen": 52206368,
      "step": 37965
    },
    {
      "epoch": 1.2153511298892516,
      "grad_norm": 0.0045182895846664906,
      "learning_rate": 0.0019043194290164045,
      "loss": 0.623,
      "num_input_tokens_seen": 52213072,
      "step": 37970
    },
    {
      "epoch": 1.2155111708597401,
      "grad_norm": 0.0034658710937947035,
      "learning_rate": 0.0018949744732222162,
      "loss": 0.4157,
      "num_input_tokens_seen": 52219712,
      "step": 37975
    },
    {
      "epoch": 1.2156712118302284,
      "grad_norm": 0.00583378691226244,
      "learning_rate": 0.0018856523570841776,
      "loss": 0.3688,
      "num_input_tokens_seen": 52226432,
      "step": 37980
    },
    {
      "epoch": 1.215831252800717,
      "grad_norm": 0.0028320602141320705,
      "learning_rate": 0.0018763530820398555,
      "loss": 0.4661,
      "num_input_tokens_seen": 52232960,
      "step": 37985
    },
    {
      "epoch": 1.2159912937712054,
      "grad_norm": 0.007274656090885401,
      "learning_rate": 0.0018670766495233525,
      "loss": 0.4222,
      "num_input_tokens_seen": 52240192,
      "step": 37990
    },
    {
      "epoch": 1.216151334741694,
      "grad_norm": 0.006830787751823664,
      "learning_rate": 0.001857823060965158,
      "loss": 0.5689,
      "num_input_tokens_seen": 52247584,
      "step": 37995
    },
    {
      "epoch": 1.2163113757121824,
      "grad_norm": 0.0065620592795312405,
      "learning_rate": 0.0018485923177923467,
      "loss": 0.3913,
      "num_input_tokens_seen": 52254352,
      "step": 38000
    },
    {
      "epoch": 1.2163113757121824,
      "eval_loss": 0.4942423105239868,
      "eval_runtime": 332.3098,
      "eval_samples_per_second": 41.786,
      "eval_steps_per_second": 20.893,
      "num_input_tokens_seen": 52254352,
      "step": 38000
    },
    {
      "epoch": 1.2164714166826707,
      "grad_norm": 0.0036922285798937082,
      "learning_rate": 0.001839384421428364,
      "loss": 0.4646,
      "num_input_tokens_seen": 52261024,
      "step": 38005
    },
    {
      "epoch": 1.2166314576531592,
      "grad_norm": 0.0048538558185100555,
      "learning_rate": 0.0018301993732932065,
      "loss": 0.4477,
      "num_input_tokens_seen": 52267696,
      "step": 38010
    },
    {
      "epoch": 1.2167914986236477,
      "grad_norm": 0.004865501541644335,
      "learning_rate": 0.0018210371748033248,
      "loss": 0.6419,
      "num_input_tokens_seen": 52274736,
      "step": 38015
    },
    {
      "epoch": 1.2169515395941362,
      "grad_norm": 0.006305660121142864,
      "learning_rate": 0.0018118978273716556,
      "loss": 0.5015,
      "num_input_tokens_seen": 52281616,
      "step": 38020
    },
    {
      "epoch": 1.2171115805646244,
      "grad_norm": 0.006767162121832371,
      "learning_rate": 0.001802781332407588,
      "loss": 0.4999,
      "num_input_tokens_seen": 52288064,
      "step": 38025
    },
    {
      "epoch": 1.217271621535113,
      "grad_norm": 0.004686116706579924,
      "learning_rate": 0.0017936876913169806,
      "loss": 0.4877,
      "num_input_tokens_seen": 52295056,
      "step": 38030
    },
    {
      "epoch": 1.2174316625056014,
      "grad_norm": 0.006330892443656921,
      "learning_rate": 0.0017846169055022287,
      "loss": 0.4319,
      "num_input_tokens_seen": 52302032,
      "step": 38035
    },
    {
      "epoch": 1.21759170347609,
      "grad_norm": 0.006969547364860773,
      "learning_rate": 0.0017755689763621295,
      "loss": 0.458,
      "num_input_tokens_seen": 52309024,
      "step": 38040
    },
    {
      "epoch": 1.2177517444465784,
      "grad_norm": 0.007313392125070095,
      "learning_rate": 0.0017665439052920173,
      "loss": 0.51,
      "num_input_tokens_seen": 52315616,
      "step": 38045
    },
    {
      "epoch": 1.2179117854170667,
      "grad_norm": 0.007591276429593563,
      "learning_rate": 0.0017575416936836286,
      "loss": 0.5139,
      "num_input_tokens_seen": 52322512,
      "step": 38050
    },
    {
      "epoch": 1.2180718263875552,
      "grad_norm": 0.020196879282593727,
      "learning_rate": 0.0017485623429252528,
      "loss": 0.7124,
      "num_input_tokens_seen": 52329392,
      "step": 38055
    },
    {
      "epoch": 1.2182318673580437,
      "grad_norm": 0.01161432359367609,
      "learning_rate": 0.0017396058544016156,
      "loss": 0.4615,
      "num_input_tokens_seen": 52336096,
      "step": 38060
    },
    {
      "epoch": 1.2183919083285322,
      "grad_norm": 0.006821849383413792,
      "learning_rate": 0.0017306722294938958,
      "loss": 0.4952,
      "num_input_tokens_seen": 52343008,
      "step": 38065
    },
    {
      "epoch": 1.2185519492990204,
      "grad_norm": 0.004588835407048464,
      "learning_rate": 0.0017217614695798078,
      "loss": 0.5958,
      "num_input_tokens_seen": 52349712,
      "step": 38070
    },
    {
      "epoch": 1.218711990269509,
      "grad_norm": 0.004674999974668026,
      "learning_rate": 0.001712873576033469,
      "loss": 0.6228,
      "num_input_tokens_seen": 52356608,
      "step": 38075
    },
    {
      "epoch": 1.2188720312399974,
      "grad_norm": 0.0034351374488323927,
      "learning_rate": 0.0017040085502255163,
      "loss": 0.4032,
      "num_input_tokens_seen": 52363440,
      "step": 38080
    },
    {
      "epoch": 1.219032072210486,
      "grad_norm": 0.005266483873128891,
      "learning_rate": 0.0016951663935230565,
      "loss": 0.538,
      "num_input_tokens_seen": 52370224,
      "step": 38085
    },
    {
      "epoch": 1.2191921131809744,
      "grad_norm": 0.005302841309458017,
      "learning_rate": 0.0016863471072896485,
      "loss": 0.5236,
      "num_input_tokens_seen": 52377200,
      "step": 38090
    },
    {
      "epoch": 1.2193521541514627,
      "grad_norm": 0.005473876837641001,
      "learning_rate": 0.0016775506928853377,
      "loss": 0.4813,
      "num_input_tokens_seen": 52384016,
      "step": 38095
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.003788235830143094,
      "learning_rate": 0.001668777151666656,
      "loss": 0.5696,
      "num_input_tokens_seen": 52391184,
      "step": 38100
    },
    {
      "epoch": 1.2196722360924397,
      "grad_norm": 0.0032365359365940094,
      "learning_rate": 0.0016600264849865709,
      "loss": 0.3891,
      "num_input_tokens_seen": 52398416,
      "step": 38105
    },
    {
      "epoch": 1.2198322770629282,
      "grad_norm": 0.0058924248442053795,
      "learning_rate": 0.0016512986941945695,
      "loss": 0.3849,
      "num_input_tokens_seen": 52405024,
      "step": 38110
    },
    {
      "epoch": 1.2199923180334165,
      "grad_norm": 0.004244147799909115,
      "learning_rate": 0.0016425937806365753,
      "loss": 0.3971,
      "num_input_tokens_seen": 52411856,
      "step": 38115
    },
    {
      "epoch": 1.220152359003905,
      "grad_norm": 0.007573212031275034,
      "learning_rate": 0.0016339117456549979,
      "loss": 0.4869,
      "num_input_tokens_seen": 52418608,
      "step": 38120
    },
    {
      "epoch": 1.2203123999743934,
      "grad_norm": 0.006094940472394228,
      "learning_rate": 0.0016252525905886995,
      "loss": 0.4421,
      "num_input_tokens_seen": 52425920,
      "step": 38125
    },
    {
      "epoch": 1.220472440944882,
      "grad_norm": 0.003213937859982252,
      "learning_rate": 0.0016166163167730617,
      "loss": 0.4643,
      "num_input_tokens_seen": 52432816,
      "step": 38130
    },
    {
      "epoch": 1.2206324819153704,
      "grad_norm": 0.0037356261163949966,
      "learning_rate": 0.0016080029255398864,
      "loss": 0.3627,
      "num_input_tokens_seen": 52440000,
      "step": 38135
    },
    {
      "epoch": 1.2207925228858587,
      "grad_norm": 0.004666863474994898,
      "learning_rate": 0.0015994124182174606,
      "loss": 0.4388,
      "num_input_tokens_seen": 52446672,
      "step": 38140
    },
    {
      "epoch": 1.2209525638563472,
      "grad_norm": 0.003979919943958521,
      "learning_rate": 0.001590844796130575,
      "loss": 0.528,
      "num_input_tokens_seen": 52454704,
      "step": 38145
    },
    {
      "epoch": 1.2211126048268357,
      "grad_norm": 0.0034371220972388983,
      "learning_rate": 0.001582300060600439,
      "loss": 0.4431,
      "num_input_tokens_seen": 52461840,
      "step": 38150
    },
    {
      "epoch": 1.2212726457973242,
      "grad_norm": 0.0031891532707959414,
      "learning_rate": 0.0015737782129447652,
      "loss": 0.3778,
      "num_input_tokens_seen": 52468624,
      "step": 38155
    },
    {
      "epoch": 1.2214326867678125,
      "grad_norm": 0.0052624717354774475,
      "learning_rate": 0.0015652792544777361,
      "loss": 0.4683,
      "num_input_tokens_seen": 52475584,
      "step": 38160
    },
    {
      "epoch": 1.221592727738301,
      "grad_norm": 0.0049736094661056995,
      "learning_rate": 0.0015568031865099863,
      "loss": 0.4785,
      "num_input_tokens_seen": 52482160,
      "step": 38165
    },
    {
      "epoch": 1.2217527687087895,
      "grad_norm": 0.003539264900609851,
      "learning_rate": 0.0015483500103486369,
      "loss": 0.4628,
      "num_input_tokens_seen": 52489072,
      "step": 38170
    },
    {
      "epoch": 1.221912809679278,
      "grad_norm": 0.007754683960229158,
      "learning_rate": 0.0015399197272972787,
      "loss": 0.4871,
      "num_input_tokens_seen": 52495808,
      "step": 38175
    },
    {
      "epoch": 1.2220728506497665,
      "grad_norm": 0.0034586223773658276,
      "learning_rate": 0.0015315123386559714,
      "loss": 0.4151,
      "num_input_tokens_seen": 52502560,
      "step": 38180
    },
    {
      "epoch": 1.2222328916202547,
      "grad_norm": 0.004801864270120859,
      "learning_rate": 0.0015231278457212283,
      "loss": 0.4238,
      "num_input_tokens_seen": 52509184,
      "step": 38185
    },
    {
      "epoch": 1.2223929325907432,
      "grad_norm": 0.0024493092205375433,
      "learning_rate": 0.001514766249786048,
      "loss": 0.3155,
      "num_input_tokens_seen": 52515632,
      "step": 38190
    },
    {
      "epoch": 1.2225529735612317,
      "grad_norm": 0.0035395040176808834,
      "learning_rate": 0.0015064275521398994,
      "loss": 0.4165,
      "num_input_tokens_seen": 52522272,
      "step": 38195
    },
    {
      "epoch": 1.2227130145317202,
      "grad_norm": 0.0059364717453718185,
      "learning_rate": 0.0014981117540686872,
      "loss": 0.4959,
      "num_input_tokens_seen": 52529584,
      "step": 38200
    },
    {
      "epoch": 1.2227130145317202,
      "eval_loss": 0.49443164467811584,
      "eval_runtime": 332.2812,
      "eval_samples_per_second": 41.79,
      "eval_steps_per_second": 20.895,
      "num_input_tokens_seen": 52529584,
      "step": 38200
    },
    {
      "epoch": 1.2228730555022085,
      "grad_norm": 0.006221086718142033,
      "learning_rate": 0.0014898188568548687,
      "loss": 0.6326,
      "num_input_tokens_seen": 52536192,
      "step": 38205
    },
    {
      "epoch": 1.223033096472697,
      "grad_norm": 0.004443804733455181,
      "learning_rate": 0.0014815488617772542,
      "loss": 0.292,
      "num_input_tokens_seen": 52543824,
      "step": 38210
    },
    {
      "epoch": 1.2231931374431855,
      "grad_norm": 0.010401077568531036,
      "learning_rate": 0.0014733017701112072,
      "loss": 0.5598,
      "num_input_tokens_seen": 52551280,
      "step": 38215
    },
    {
      "epoch": 1.223353178413674,
      "grad_norm": 0.007343114819377661,
      "learning_rate": 0.0014650775831285435,
      "loss": 0.427,
      "num_input_tokens_seen": 52557824,
      "step": 38220
    },
    {
      "epoch": 1.2235132193841625,
      "grad_norm": 0.0036359275691211224,
      "learning_rate": 0.001456876302097515,
      "loss": 0.6019,
      "num_input_tokens_seen": 52565184,
      "step": 38225
    },
    {
      "epoch": 1.2236732603546507,
      "grad_norm": 0.0038748530205339193,
      "learning_rate": 0.0014486979282828604,
      "loss": 0.705,
      "num_input_tokens_seen": 52572112,
      "step": 38230
    },
    {
      "epoch": 1.2238333013251392,
      "grad_norm": 0.007222411688417196,
      "learning_rate": 0.001440542462945804,
      "loss": 0.4877,
      "num_input_tokens_seen": 52579024,
      "step": 38235
    },
    {
      "epoch": 1.2239933422956277,
      "grad_norm": 0.008396290242671967,
      "learning_rate": 0.0014324099073440232,
      "loss": 0.4928,
      "num_input_tokens_seen": 52585904,
      "step": 38240
    },
    {
      "epoch": 1.224153383266116,
      "grad_norm": 0.004604568239301443,
      "learning_rate": 0.0014243002627316482,
      "loss": 0.4994,
      "num_input_tokens_seen": 52592768,
      "step": 38245
    },
    {
      "epoch": 1.2243134242366045,
      "grad_norm": 0.0065934439189732075,
      "learning_rate": 0.0014162135303592781,
      "loss": 0.4072,
      "num_input_tokens_seen": 52599856,
      "step": 38250
    },
    {
      "epoch": 1.224473465207093,
      "grad_norm": 0.004499227274209261,
      "learning_rate": 0.001408149711474016,
      "loss": 0.4761,
      "num_input_tokens_seen": 52606672,
      "step": 38255
    },
    {
      "epoch": 1.2246335061775815,
      "grad_norm": 0.005030625034123659,
      "learning_rate": 0.0014001088073193834,
      "loss": 0.4949,
      "num_input_tokens_seen": 52613216,
      "step": 38260
    },
    {
      "epoch": 1.22479354714807,
      "grad_norm": 0.0025866797659546137,
      "learning_rate": 0.0013920908191354052,
      "loss": 0.5919,
      "num_input_tokens_seen": 52619792,
      "step": 38265
    },
    {
      "epoch": 1.2249535881185583,
      "grad_norm": 0.004137560725212097,
      "learning_rate": 0.001384095748158526,
      "loss": 0.3788,
      "num_input_tokens_seen": 52626880,
      "step": 38270
    },
    {
      "epoch": 1.2251136290890468,
      "grad_norm": 0.0034598412457853556,
      "learning_rate": 0.0013761235956217255,
      "loss": 0.4078,
      "num_input_tokens_seen": 52633392,
      "step": 38275
    },
    {
      "epoch": 1.2252736700595352,
      "grad_norm": 0.008274504914879799,
      "learning_rate": 0.0013681743627543873,
      "loss": 0.4629,
      "num_input_tokens_seen": 52640528,
      "step": 38280
    },
    {
      "epoch": 1.2254337110300237,
      "grad_norm": 0.00707358680665493,
      "learning_rate": 0.001360248050782381,
      "loss": 0.5669,
      "num_input_tokens_seen": 52647248,
      "step": 38285
    },
    {
      "epoch": 1.225593752000512,
      "grad_norm": 0.003910686355084181,
      "learning_rate": 0.001352344660928062,
      "loss": 0.4845,
      "num_input_tokens_seen": 52653840,
      "step": 38290
    },
    {
      "epoch": 1.2257537929710005,
      "grad_norm": 0.006633446551859379,
      "learning_rate": 0.0013444641944102052,
      "loss": 0.3936,
      "num_input_tokens_seen": 52661376,
      "step": 38295
    },
    {
      "epoch": 1.225913833941489,
      "grad_norm": 0.0035084241535514593,
      "learning_rate": 0.0013366066524441056,
      "loss": 0.4275,
      "num_input_tokens_seen": 52668064,
      "step": 38300
    },
    {
      "epoch": 1.2260738749119775,
      "grad_norm": 0.0051333112642169,
      "learning_rate": 0.0013287720362414768,
      "loss": 0.5425,
      "num_input_tokens_seen": 52674656,
      "step": 38305
    },
    {
      "epoch": 1.226233915882466,
      "grad_norm": 0.0032335123978555202,
      "learning_rate": 0.0013209603470105025,
      "loss": 0.3221,
      "num_input_tokens_seen": 52681296,
      "step": 38310
    },
    {
      "epoch": 1.2263939568529543,
      "grad_norm": 0.0057871416211128235,
      "learning_rate": 0.0013131715859558857,
      "loss": 0.5096,
      "num_input_tokens_seen": 52688176,
      "step": 38315
    },
    {
      "epoch": 1.2265539978234428,
      "grad_norm": 0.002932984381914139,
      "learning_rate": 0.001305405754278699,
      "loss": 0.4271,
      "num_input_tokens_seen": 52695424,
      "step": 38320
    },
    {
      "epoch": 1.2267140387939313,
      "grad_norm": 0.004664380569010973,
      "learning_rate": 0.0012976628531765843,
      "loss": 0.6504,
      "num_input_tokens_seen": 52702320,
      "step": 38325
    },
    {
      "epoch": 1.2268740797644198,
      "grad_norm": 0.003021653275936842,
      "learning_rate": 0.0012899428838435533,
      "loss": 0.4046,
      "num_input_tokens_seen": 52709360,
      "step": 38330
    },
    {
      "epoch": 1.227034120734908,
      "grad_norm": 0.005719138775020838,
      "learning_rate": 0.001282245847470137,
      "loss": 0.4143,
      "num_input_tokens_seen": 52716336,
      "step": 38335
    },
    {
      "epoch": 1.2271941617053965,
      "grad_norm": 0.005214115604758263,
      "learning_rate": 0.001274571745243319,
      "loss": 0.3706,
      "num_input_tokens_seen": 52723024,
      "step": 38340
    },
    {
      "epoch": 1.227354202675885,
      "grad_norm": 0.004730105400085449,
      "learning_rate": 0.0012669205783465364,
      "loss": 0.5117,
      "num_input_tokens_seen": 52729696,
      "step": 38345
    },
    {
      "epoch": 1.2275142436463735,
      "grad_norm": 0.004977860953658819,
      "learning_rate": 0.001259292347959695,
      "loss": 0.4474,
      "num_input_tokens_seen": 52736464,
      "step": 38350
    },
    {
      "epoch": 1.227674284616862,
      "grad_norm": 0.008242937736213207,
      "learning_rate": 0.0012516870552591707,
      "loss": 0.4881,
      "num_input_tokens_seen": 52743648,
      "step": 38355
    },
    {
      "epoch": 1.2278343255873503,
      "grad_norm": 0.005095736123621464,
      "learning_rate": 0.001244104701417792,
      "loss": 0.4457,
      "num_input_tokens_seen": 52750560,
      "step": 38360
    },
    {
      "epoch": 1.2279943665578388,
      "grad_norm": 0.004387345165014267,
      "learning_rate": 0.0012365452876048565,
      "loss": 0.4011,
      "num_input_tokens_seen": 52757328,
      "step": 38365
    },
    {
      "epoch": 1.2281544075283273,
      "grad_norm": 0.002512105507776141,
      "learning_rate": 0.001229008814986099,
      "loss": 0.3716,
      "num_input_tokens_seen": 52764176,
      "step": 38370
    },
    {
      "epoch": 1.2283144484988158,
      "grad_norm": 0.007819879800081253,
      "learning_rate": 0.0012214952847237725,
      "loss": 0.7637,
      "num_input_tokens_seen": 52771248,
      "step": 38375
    },
    {
      "epoch": 1.228474489469304,
      "grad_norm": 0.00337930372916162,
      "learning_rate": 0.0012140046979765339,
      "loss": 0.3653,
      "num_input_tokens_seen": 52777664,
      "step": 38380
    },
    {
      "epoch": 1.2286345304397925,
      "grad_norm": 0.006432283669710159,
      "learning_rate": 0.0012065370558995258,
      "loss": 0.3685,
      "num_input_tokens_seen": 52784096,
      "step": 38385
    },
    {
      "epoch": 1.228794571410281,
      "grad_norm": 0.002266120631247759,
      "learning_rate": 0.0011990923596443602,
      "loss": 0.4494,
      "num_input_tokens_seen": 52790416,
      "step": 38390
    },
    {
      "epoch": 1.2289546123807695,
      "grad_norm": 0.0035012250300496817,
      "learning_rate": 0.001191670610359119,
      "loss": 0.4572,
      "num_input_tokens_seen": 52796944,
      "step": 38395
    },
    {
      "epoch": 1.229114653351258,
      "grad_norm": 0.0032000350765883923,
      "learning_rate": 0.0011842718091882865,
      "loss": 0.7156,
      "num_input_tokens_seen": 52803776,
      "step": 38400
    },
    {
      "epoch": 1.229114653351258,
      "eval_loss": 0.49427172541618347,
      "eval_runtime": 331.7599,
      "eval_samples_per_second": 41.856,
      "eval_steps_per_second": 20.928,
      "num_input_tokens_seen": 52803776,
      "step": 38400
    },
    {
      "epoch": 1.2292746943217463,
      "grad_norm": 0.005836928728967905,
      "learning_rate": 0.0011768959572729,
      "loss": 0.4229,
      "num_input_tokens_seen": 52810624,
      "step": 38405
    },
    {
      "epoch": 1.2294347352922348,
      "grad_norm": 0.005299444776028395,
      "learning_rate": 0.001169543055750366,
      "loss": 0.5016,
      "num_input_tokens_seen": 52817376,
      "step": 38410
    },
    {
      "epoch": 1.2295947762627233,
      "grad_norm": 0.005750004667788744,
      "learning_rate": 0.0011622131057546115,
      "loss": 0.4077,
      "num_input_tokens_seen": 52824080,
      "step": 38415
    },
    {
      "epoch": 1.2297548172332118,
      "grad_norm": 0.00697875814512372,
      "learning_rate": 0.0011549061084160316,
      "loss": 0.5814,
      "num_input_tokens_seen": 52830960,
      "step": 38420
    },
    {
      "epoch": 1.2299148582037,
      "grad_norm": 0.002885367488488555,
      "learning_rate": 0.0011476220648614088,
      "loss": 0.3686,
      "num_input_tokens_seen": 52837728,
      "step": 38425
    },
    {
      "epoch": 1.2300748991741886,
      "grad_norm": 0.004807812161743641,
      "learning_rate": 0.0011403609762140777,
      "loss": 0.4383,
      "num_input_tokens_seen": 52844272,
      "step": 38430
    },
    {
      "epoch": 1.230234940144677,
      "grad_norm": 0.003970978315919638,
      "learning_rate": 0.0011331228435937756,
      "loss": 0.4849,
      "num_input_tokens_seen": 52851152,
      "step": 38435
    },
    {
      "epoch": 1.2303949811151655,
      "grad_norm": 0.0039862883277237415,
      "learning_rate": 0.0011259076681166935,
      "loss": 0.3475,
      "num_input_tokens_seen": 52858096,
      "step": 38440
    },
    {
      "epoch": 1.230555022085654,
      "grad_norm": 0.004667701665312052,
      "learning_rate": 0.0011187154508955244,
      "loss": 0.4232,
      "num_input_tokens_seen": 52864816,
      "step": 38445
    },
    {
      "epoch": 1.2307150630561423,
      "grad_norm": 0.0045471484772861,
      "learning_rate": 0.001111546193039381,
      "loss": 0.5463,
      "num_input_tokens_seen": 52871872,
      "step": 38450
    },
    {
      "epoch": 1.2308751040266308,
      "grad_norm": 0.006277645472437143,
      "learning_rate": 0.0011043998956538792,
      "loss": 0.5571,
      "num_input_tokens_seen": 52878768,
      "step": 38455
    },
    {
      "epoch": 1.2310351449971193,
      "grad_norm": 0.00587460957467556,
      "learning_rate": 0.0010972765598410538,
      "loss": 0.5151,
      "num_input_tokens_seen": 52885568,
      "step": 38460
    },
    {
      "epoch": 1.2311951859676078,
      "grad_norm": 0.006565357558429241,
      "learning_rate": 0.0010901761866993931,
      "loss": 0.6458,
      "num_input_tokens_seen": 52892512,
      "step": 38465
    },
    {
      "epoch": 1.231355226938096,
      "grad_norm": 0.003329262137413025,
      "learning_rate": 0.0010830987773238876,
      "loss": 0.4196,
      "num_input_tokens_seen": 52900048,
      "step": 38470
    },
    {
      "epoch": 1.2315152679085846,
      "grad_norm": 0.005394903942942619,
      "learning_rate": 0.0010760443328059644,
      "loss": 0.704,
      "num_input_tokens_seen": 52907104,
      "step": 38475
    },
    {
      "epoch": 1.231675308879073,
      "grad_norm": 0.004950906150043011,
      "learning_rate": 0.001069012854233503,
      "loss": 0.4415,
      "num_input_tokens_seen": 52914048,
      "step": 38480
    },
    {
      "epoch": 1.2318353498495616,
      "grad_norm": 0.005165800452232361,
      "learning_rate": 0.0010620043426908365,
      "loss": 0.3068,
      "num_input_tokens_seen": 52920896,
      "step": 38485
    },
    {
      "epoch": 1.23199539082005,
      "grad_norm": 0.0036529176868498325,
      "learning_rate": 0.0010550187992587833,
      "loss": 0.4075,
      "num_input_tokens_seen": 52927408,
      "step": 38490
    },
    {
      "epoch": 1.2321554317905383,
      "grad_norm": 0.004302415996789932,
      "learning_rate": 0.0010480562250145653,
      "loss": 0.3417,
      "num_input_tokens_seen": 52934128,
      "step": 38495
    },
    {
      "epoch": 1.2323154727610268,
      "grad_norm": 0.005094338208436966,
      "learning_rate": 0.0010411166210319567,
      "loss": 0.3321,
      "num_input_tokens_seen": 52940416,
      "step": 38500
    },
    {
      "epoch": 1.2324755137315153,
      "grad_norm": 0.004951664712280035,
      "learning_rate": 0.0010341999883810848,
      "loss": 0.49,
      "num_input_tokens_seen": 52947040,
      "step": 38505
    },
    {
      "epoch": 1.2326355547020036,
      "grad_norm": 0.0028256187215447426,
      "learning_rate": 0.0010273063281285965,
      "loss": 0.4484,
      "num_input_tokens_seen": 52953872,
      "step": 38510
    },
    {
      "epoch": 1.232795595672492,
      "grad_norm": 0.005660233553498983,
      "learning_rate": 0.0010204356413375747,
      "loss": 0.5058,
      "num_input_tokens_seen": 52960720,
      "step": 38515
    },
    {
      "epoch": 1.2329556366429806,
      "grad_norm": 0.003135640872642398,
      "learning_rate": 0.001013587929067572,
      "loss": 0.3039,
      "num_input_tokens_seen": 52967440,
      "step": 38520
    },
    {
      "epoch": 1.233115677613469,
      "grad_norm": 0.001499876962043345,
      "learning_rate": 0.00100676319237461,
      "loss": 0.5376,
      "num_input_tokens_seen": 52974704,
      "step": 38525
    },
    {
      "epoch": 1.2332757185839576,
      "grad_norm": 0.005511278752237558,
      "learning_rate": 0.0009999614323110972,
      "loss": 0.3099,
      "num_input_tokens_seen": 52981408,
      "step": 38530
    },
    {
      "epoch": 1.2334357595544458,
      "grad_norm": 0.004028907511383295,
      "learning_rate": 0.000993182649926011,
      "loss": 0.5114,
      "num_input_tokens_seen": 52988368,
      "step": 38535
    },
    {
      "epoch": 1.2335958005249343,
      "grad_norm": 0.0037813305389136076,
      "learning_rate": 0.000986426846264682,
      "loss": 0.3526,
      "num_input_tokens_seen": 52995152,
      "step": 38540
    },
    {
      "epoch": 1.2337558414954228,
      "grad_norm": 0.0036934176459908485,
      "learning_rate": 0.00097969402236896,
      "loss": 0.5794,
      "num_input_tokens_seen": 53002160,
      "step": 38545
    },
    {
      "epoch": 1.2339158824659113,
      "grad_norm": 0.009250643663108349,
      "learning_rate": 0.0009729841792771143,
      "loss": 0.6582,
      "num_input_tokens_seen": 53008496,
      "step": 38550
    },
    {
      "epoch": 1.2340759234363996,
      "grad_norm": 0.003495891345664859,
      "learning_rate": 0.0009662973180239176,
      "loss": 0.3058,
      "num_input_tokens_seen": 53015296,
      "step": 38555
    },
    {
      "epoch": 1.234235964406888,
      "grad_norm": 0.0042145587503910065,
      "learning_rate": 0.0009596334396405448,
      "loss": 0.569,
      "num_input_tokens_seen": 53022272,
      "step": 38560
    },
    {
      "epoch": 1.2343960053773766,
      "grad_norm": 0.003986872732639313,
      "learning_rate": 0.0009529925451546406,
      "loss": 0.2802,
      "num_input_tokens_seen": 53029616,
      "step": 38565
    },
    {
      "epoch": 1.234556046347865,
      "grad_norm": 0.0048362137749791145,
      "learning_rate": 0.0009463746355903357,
      "loss": 0.5946,
      "num_input_tokens_seen": 53036816,
      "step": 38570
    },
    {
      "epoch": 1.2347160873183536,
      "grad_norm": 0.006794227287173271,
      "learning_rate": 0.0009397797119681971,
      "loss": 0.5145,
      "num_input_tokens_seen": 53043968,
      "step": 38575
    },
    {
      "epoch": 1.2348761282888419,
      "grad_norm": 0.0057630701921880245,
      "learning_rate": 0.0009332077753052281,
      "loss": 0.5853,
      "num_input_tokens_seen": 53051184,
      "step": 38580
    },
    {
      "epoch": 1.2350361692593304,
      "grad_norm": 0.00254358840174973,
      "learning_rate": 0.0009266588266149011,
      "loss": 0.5072,
      "num_input_tokens_seen": 53058016,
      "step": 38585
    },
    {
      "epoch": 1.2351962102298188,
      "grad_norm": 0.0035435694735497236,
      "learning_rate": 0.0009201328669071584,
      "loss": 0.366,
      "num_input_tokens_seen": 53064896,
      "step": 38590
    },
    {
      "epoch": 1.2353562512003073,
      "grad_norm": 0.00579440500587225,
      "learning_rate": 0.0009136298971883949,
      "loss": 0.4778,
      "num_input_tokens_seen": 53071824,
      "step": 38595
    },
    {
      "epoch": 1.2355162921707956,
      "grad_norm": 0.0047518168576061726,
      "learning_rate": 0.0009071499184614251,
      "loss": 0.4096,
      "num_input_tokens_seen": 53078736,
      "step": 38600
    },
    {
      "epoch": 1.2355162921707956,
      "eval_loss": 0.4942466914653778,
      "eval_runtime": 332.7481,
      "eval_samples_per_second": 41.731,
      "eval_steps_per_second": 20.866,
      "num_input_tokens_seen": 53078736,
      "step": 38600
    },
    {
      "epoch": 1.2356763331412841,
      "grad_norm": 0.005554442293941975,
      "learning_rate": 0.0009006929317255663,
      "loss": 0.5513,
      "num_input_tokens_seen": 53085808,
      "step": 38605
    },
    {
      "epoch": 1.2358363741117726,
      "grad_norm": 0.005096230190247297,
      "learning_rate": 0.0008942589379765387,
      "loss": 0.3798,
      "num_input_tokens_seen": 53092848,
      "step": 38610
    },
    {
      "epoch": 1.235996415082261,
      "grad_norm": 0.0031896084547042847,
      "learning_rate": 0.0008878479382065817,
      "loss": 0.5173,
      "num_input_tokens_seen": 53100256,
      "step": 38615
    },
    {
      "epoch": 1.2361564560527496,
      "grad_norm": 0.005086616147309542,
      "learning_rate": 0.0008814599334043215,
      "loss": 0.4875,
      "num_input_tokens_seen": 53106560,
      "step": 38620
    },
    {
      "epoch": 1.2363164970232379,
      "grad_norm": 0.00459028547629714,
      "learning_rate": 0.0008750949245548866,
      "loss": 0.3277,
      "num_input_tokens_seen": 53113312,
      "step": 38625
    },
    {
      "epoch": 1.2364765379937264,
      "grad_norm": 0.005172635428607464,
      "learning_rate": 0.0008687529126398252,
      "loss": 0.6403,
      "num_input_tokens_seen": 53120480,
      "step": 38630
    },
    {
      "epoch": 1.2366365789642149,
      "grad_norm": 0.0070792813785374165,
      "learning_rate": 0.0008624338986371715,
      "loss": 0.4773,
      "num_input_tokens_seen": 53127312,
      "step": 38635
    },
    {
      "epoch": 1.2367966199347034,
      "grad_norm": 0.005865468177944422,
      "learning_rate": 0.0008561378835213962,
      "loss": 0.5831,
      "num_input_tokens_seen": 53133952,
      "step": 38640
    },
    {
      "epoch": 1.2369566609051916,
      "grad_norm": 0.0023000899236649275,
      "learning_rate": 0.0008498648682634058,
      "loss": 0.4374,
      "num_input_tokens_seen": 53141136,
      "step": 38645
    },
    {
      "epoch": 1.2371167018756801,
      "grad_norm": 0.005375709384679794,
      "learning_rate": 0.0008436148538306099,
      "loss": 0.4984,
      "num_input_tokens_seen": 53148032,
      "step": 38650
    },
    {
      "epoch": 1.2372767428461686,
      "grad_norm": 0.003543257713317871,
      "learning_rate": 0.0008373878411868041,
      "loss": 0.46,
      "num_input_tokens_seen": 53154784,
      "step": 38655
    },
    {
      "epoch": 1.2374367838166571,
      "grad_norm": 0.0036615405697375536,
      "learning_rate": 0.000831183831292287,
      "loss": 0.5927,
      "num_input_tokens_seen": 53161664,
      "step": 38660
    },
    {
      "epoch": 1.2375968247871456,
      "grad_norm": 0.008875002153217793,
      "learning_rate": 0.0008250028251037933,
      "loss": 0.5127,
      "num_input_tokens_seen": 53168400,
      "step": 38665
    },
    {
      "epoch": 1.2377568657576339,
      "grad_norm": 0.005824389401823282,
      "learning_rate": 0.0008188448235745271,
      "loss": 0.6372,
      "num_input_tokens_seen": 53175120,
      "step": 38670
    },
    {
      "epoch": 1.2379169067281224,
      "grad_norm": 0.005908405873924494,
      "learning_rate": 0.0008127098276541122,
      "loss": 0.4745,
      "num_input_tokens_seen": 53181968,
      "step": 38675
    },
    {
      "epoch": 1.2380769476986109,
      "grad_norm": 0.005894474685192108,
      "learning_rate": 0.0008065978382886418,
      "loss": 0.4693,
      "num_input_tokens_seen": 53188528,
      "step": 38680
    },
    {
      "epoch": 1.2382369886690994,
      "grad_norm": 0.006712683476507664,
      "learning_rate": 0.0008005088564206785,
      "loss": 0.5316,
      "num_input_tokens_seen": 53195088,
      "step": 38685
    },
    {
      "epoch": 1.2383970296395876,
      "grad_norm": 0.008207999169826508,
      "learning_rate": 0.0007944428829891881,
      "loss": 0.566,
      "num_input_tokens_seen": 53202016,
      "step": 38690
    },
    {
      "epoch": 1.2385570706100761,
      "grad_norm": 0.005718984641134739,
      "learning_rate": 0.0007883999189296386,
      "loss": 0.5171,
      "num_input_tokens_seen": 53208928,
      "step": 38695
    },
    {
      "epoch": 1.2387171115805646,
      "grad_norm": 0.004030950367450714,
      "learning_rate": 0.0007823799651739515,
      "loss": 0.415,
      "num_input_tokens_seen": 53215632,
      "step": 38700
    },
    {
      "epoch": 1.2388771525510531,
      "grad_norm": 0.005254174116998911,
      "learning_rate": 0.0007763830226504509,
      "loss": 0.4096,
      "num_input_tokens_seen": 53222608,
      "step": 38705
    },
    {
      "epoch": 1.2390371935215416,
      "grad_norm": 0.0043302271515131,
      "learning_rate": 0.0007704090922839468,
      "loss": 0.4123,
      "num_input_tokens_seen": 53229296,
      "step": 38710
    },
    {
      "epoch": 1.23919723449203,
      "grad_norm": 0.0024688243865966797,
      "learning_rate": 0.0007644581749957025,
      "loss": 0.4384,
      "num_input_tokens_seen": 53236656,
      "step": 38715
    },
    {
      "epoch": 1.2393572754625184,
      "grad_norm": 0.005219480954110622,
      "learning_rate": 0.000758530271703417,
      "loss": 0.6074,
      "num_input_tokens_seen": 53243696,
      "step": 38720
    },
    {
      "epoch": 1.2395173164330069,
      "grad_norm": 0.0063684904016554356,
      "learning_rate": 0.0007526253833212426,
      "loss": 0.5059,
      "num_input_tokens_seen": 53250608,
      "step": 38725
    },
    {
      "epoch": 1.2396773574034954,
      "grad_norm": 0.00792943686246872,
      "learning_rate": 0.0007467435107598008,
      "loss": 0.4106,
      "num_input_tokens_seen": 53257488,
      "step": 38730
    },
    {
      "epoch": 1.2398373983739837,
      "grad_norm": 0.006986995227634907,
      "learning_rate": 0.0007408846549261328,
      "loss": 0.4838,
      "num_input_tokens_seen": 53264032,
      "step": 38735
    },
    {
      "epoch": 1.2399974393444722,
      "grad_norm": 0.003962451126426458,
      "learning_rate": 0.0007350488167237656,
      "loss": 0.5654,
      "num_input_tokens_seen": 53270800,
      "step": 38740
    },
    {
      "epoch": 1.2401574803149606,
      "grad_norm": 0.005302033852785826,
      "learning_rate": 0.0007292359970526629,
      "loss": 0.5188,
      "num_input_tokens_seen": 53277248,
      "step": 38745
    },
    {
      "epoch": 1.2403175212854491,
      "grad_norm": 0.004309420473873615,
      "learning_rate": 0.0007234461968092076,
      "loss": 0.3981,
      "num_input_tokens_seen": 53284144,
      "step": 38750
    },
    {
      "epoch": 1.2404775622559376,
      "grad_norm": 0.0036135998088866472,
      "learning_rate": 0.0007176794168862854,
      "loss": 0.3653,
      "num_input_tokens_seen": 53291008,
      "step": 38755
    },
    {
      "epoch": 1.240637603226426,
      "grad_norm": 0.0037934775464236736,
      "learning_rate": 0.000711935658173185,
      "loss": 0.3214,
      "num_input_tokens_seen": 53297824,
      "step": 38760
    },
    {
      "epoch": 1.2407976441969144,
      "grad_norm": 0.005449882708489895,
      "learning_rate": 0.0007062149215556812,
      "loss": 0.6034,
      "num_input_tokens_seen": 53304624,
      "step": 38765
    },
    {
      "epoch": 1.240957685167403,
      "grad_norm": 0.0048272316344082355,
      "learning_rate": 0.0007005172079159849,
      "loss": 0.3694,
      "num_input_tokens_seen": 53311632,
      "step": 38770
    },
    {
      "epoch": 1.2411177261378914,
      "grad_norm": 0.00360177387483418,
      "learning_rate": 0.0006948425181327267,
      "loss": 0.4172,
      "num_input_tokens_seen": 53318816,
      "step": 38775
    },
    {
      "epoch": 1.2412777671083797,
      "grad_norm": 0.0054395473562181,
      "learning_rate": 0.000689190853081073,
      "loss": 0.277,
      "num_input_tokens_seen": 53325504,
      "step": 38780
    },
    {
      "epoch": 1.2414378080788682,
      "grad_norm": 0.004170718602836132,
      "learning_rate": 0.000683562213632527,
      "loss": 0.3356,
      "num_input_tokens_seen": 53332416,
      "step": 38785
    },
    {
      "epoch": 1.2415978490493567,
      "grad_norm": 0.00293007236905396,
      "learning_rate": 0.0006779566006551108,
      "loss": 0.4679,
      "num_input_tokens_seen": 53338928,
      "step": 38790
    },
    {
      "epoch": 1.2417578900198452,
      "grad_norm": 0.009629223495721817,
      "learning_rate": 0.0006723740150132995,
      "loss": 0.5528,
      "num_input_tokens_seen": 53345792,
      "step": 38795
    },
    {
      "epoch": 1.2419179309903334,
      "grad_norm": 0.0077569554559886456,
      "learning_rate": 0.0006668144575679713,
      "loss": 0.5082,
      "num_input_tokens_seen": 53352672,
      "step": 38800
    },
    {
      "epoch": 1.2419179309903334,
      "eval_loss": 0.49414047598838806,
      "eval_runtime": 332.1348,
      "eval_samples_per_second": 41.808,
      "eval_steps_per_second": 20.904,
      "num_input_tokens_seen": 53352672,
      "step": 38800
    },
    {
      "epoch": 1.242077971960822,
      "grad_norm": 0.00482446514070034,
      "learning_rate": 0.0006612779291765069,
      "loss": 0.4349,
      "num_input_tokens_seen": 53359648,
      "step": 38805
    },
    {
      "epoch": 1.2422380129313104,
      "grad_norm": 0.004347620997577906,
      "learning_rate": 0.0006557644306926736,
      "loss": 0.52,
      "num_input_tokens_seen": 53366240,
      "step": 38810
    },
    {
      "epoch": 1.242398053901799,
      "grad_norm": 0.003358032088726759,
      "learning_rate": 0.0006502739629667575,
      "loss": 0.5631,
      "num_input_tokens_seen": 53372832,
      "step": 38815
    },
    {
      "epoch": 1.2425580948722872,
      "grad_norm": 0.0027748080901801586,
      "learning_rate": 0.0006448065268454317,
      "loss": 0.3369,
      "num_input_tokens_seen": 53379680,
      "step": 38820
    },
    {
      "epoch": 1.2427181358427757,
      "grad_norm": 0.008385967463254929,
      "learning_rate": 0.0006393621231718549,
      "loss": 0.5123,
      "num_input_tokens_seen": 53386624,
      "step": 38825
    },
    {
      "epoch": 1.2428781768132642,
      "grad_norm": 0.003858807496726513,
      "learning_rate": 0.0006339407527856389,
      "loss": 0.5135,
      "num_input_tokens_seen": 53393488,
      "step": 38830
    },
    {
      "epoch": 1.2430382177837527,
      "grad_norm": 0.006302970461547375,
      "learning_rate": 0.0006285424165227982,
      "loss": 0.5079,
      "num_input_tokens_seen": 53400032,
      "step": 38835
    },
    {
      "epoch": 1.2431982587542412,
      "grad_norm": 0.0016115088947117329,
      "learning_rate": 0.0006231671152158169,
      "loss": 0.3202,
      "num_input_tokens_seen": 53406864,
      "step": 38840
    },
    {
      "epoch": 1.2433582997247294,
      "grad_norm": 0.007327782455831766,
      "learning_rate": 0.0006178148496936819,
      "loss": 0.4415,
      "num_input_tokens_seen": 53413456,
      "step": 38845
    },
    {
      "epoch": 1.243518340695218,
      "grad_norm": 0.009474254213273525,
      "learning_rate": 0.000612485620781733,
      "loss": 0.5081,
      "num_input_tokens_seen": 53420208,
      "step": 38850
    },
    {
      "epoch": 1.2436783816657064,
      "grad_norm": 0.006685761269181967,
      "learning_rate": 0.0006071794293018296,
      "loss": 0.536,
      "num_input_tokens_seen": 53426768,
      "step": 38855
    },
    {
      "epoch": 1.243838422636195,
      "grad_norm": 0.0021923126187175512,
      "learning_rate": 0.0006018962760722501,
      "loss": 0.3645,
      "num_input_tokens_seen": 53433568,
      "step": 38860
    },
    {
      "epoch": 1.2439984636066832,
      "grad_norm": 0.004739042837172747,
      "learning_rate": 0.0005966361619077098,
      "loss": 0.565,
      "num_input_tokens_seen": 53440512,
      "step": 38865
    },
    {
      "epoch": 1.2441585045771717,
      "grad_norm": 0.004555658437311649,
      "learning_rate": 0.000591399087619393,
      "loss": 0.4296,
      "num_input_tokens_seen": 53447472,
      "step": 38870
    },
    {
      "epoch": 1.2443185455476602,
      "grad_norm": 0.011196939274668694,
      "learning_rate": 0.0005861850540149371,
      "loss": 0.4837,
      "num_input_tokens_seen": 53454416,
      "step": 38875
    },
    {
      "epoch": 1.2444785865181487,
      "grad_norm": 0.006841036956757307,
      "learning_rate": 0.0005809940618983822,
      "loss": 0.5126,
      "num_input_tokens_seen": 53461488,
      "step": 38880
    },
    {
      "epoch": 1.2446386274886372,
      "grad_norm": 0.005910431034862995,
      "learning_rate": 0.0005758261120702712,
      "loss": 0.3723,
      "num_input_tokens_seen": 53468128,
      "step": 38885
    },
    {
      "epoch": 1.2447986684591255,
      "grad_norm": 0.005321643780916929,
      "learning_rate": 0.0005706812053275501,
      "loss": 0.303,
      "num_input_tokens_seen": 53475440,
      "step": 38890
    },
    {
      "epoch": 1.244958709429614,
      "grad_norm": 0.0032623179722577333,
      "learning_rate": 0.0005655593424636173,
      "loss": 0.4936,
      "num_input_tokens_seen": 53482736,
      "step": 38895
    },
    {
      "epoch": 1.2451187504001024,
      "grad_norm": 0.004520474001765251,
      "learning_rate": 0.0005604605242683746,
      "loss": 0.446,
      "num_input_tokens_seen": 53489536,
      "step": 38900
    },
    {
      "epoch": 1.245278791370591,
      "grad_norm": 0.0026818227488547564,
      "learning_rate": 0.0005553847515280596,
      "loss": 0.481,
      "num_input_tokens_seen": 53496352,
      "step": 38905
    },
    {
      "epoch": 1.2454388323410792,
      "grad_norm": 0.004349122755229473,
      "learning_rate": 0.0005503320250254795,
      "loss": 0.4734,
      "num_input_tokens_seen": 53503664,
      "step": 38910
    },
    {
      "epoch": 1.2455988733115677,
      "grad_norm": 0.002866022288799286,
      "learning_rate": 0.0005453023455397943,
      "loss": 0.3158,
      "num_input_tokens_seen": 53510352,
      "step": 38915
    },
    {
      "epoch": 1.2457589142820562,
      "grad_norm": 0.007979368790984154,
      "learning_rate": 0.0005402957138466502,
      "loss": 0.5149,
      "num_input_tokens_seen": 53517104,
      "step": 38920
    },
    {
      "epoch": 1.2459189552525447,
      "grad_norm": 0.00698988139629364,
      "learning_rate": 0.0005353121307181463,
      "loss": 0.6064,
      "num_input_tokens_seen": 53523952,
      "step": 38925
    },
    {
      "epoch": 1.2460789962230332,
      "grad_norm": 0.005124262534081936,
      "learning_rate": 0.0005303515969227845,
      "loss": 0.4046,
      "num_input_tokens_seen": 53531168,
      "step": 38930
    },
    {
      "epoch": 1.2462390371935215,
      "grad_norm": 0.005605926737189293,
      "learning_rate": 0.0005254141132255862,
      "loss": 0.6232,
      "num_input_tokens_seen": 53538144,
      "step": 38935
    },
    {
      "epoch": 1.24639907816401,
      "grad_norm": 0.003926759120076895,
      "learning_rate": 0.0005204996803879258,
      "loss": 0.5239,
      "num_input_tokens_seen": 53544848,
      "step": 38940
    },
    {
      "epoch": 1.2465591191344985,
      "grad_norm": 0.0054542506113648415,
      "learning_rate": 0.0005156082991676969,
      "loss": 0.5271,
      "num_input_tokens_seen": 53551728,
      "step": 38945
    },
    {
      "epoch": 1.246719160104987,
      "grad_norm": 0.002914400538429618,
      "learning_rate": 0.0005107399703192127,
      "loss": 0.5451,
      "num_input_tokens_seen": 53558960,
      "step": 38950
    },
    {
      "epoch": 1.2468792010754752,
      "grad_norm": 0.004053548909723759,
      "learning_rate": 0.0005058946945932063,
      "loss": 0.4497,
      "num_input_tokens_seen": 53565616,
      "step": 38955
    },
    {
      "epoch": 1.2470392420459637,
      "grad_norm": 0.0038266200572252274,
      "learning_rate": 0.0005010724727369131,
      "loss": 0.3026,
      "num_input_tokens_seen": 53572736,
      "step": 38960
    },
    {
      "epoch": 1.2471992830164522,
      "grad_norm": 0.0034394545946270227,
      "learning_rate": 0.000496273305493955,
      "loss": 0.3012,
      "num_input_tokens_seen": 53579856,
      "step": 38965
    },
    {
      "epoch": 1.2473593239869407,
      "grad_norm": 0.006511283107101917,
      "learning_rate": 0.0004914971936044399,
      "loss": 0.5809,
      "num_input_tokens_seen": 53586896,
      "step": 38970
    },
    {
      "epoch": 1.2475193649574292,
      "grad_norm": 0.003153369762003422,
      "learning_rate": 0.00048674413780491196,
      "loss": 0.4153,
      "num_input_tokens_seen": 53593872,
      "step": 38975
    },
    {
      "epoch": 1.2476794059279175,
      "grad_norm": 0.0038882463704794645,
      "learning_rate": 0.0004820141388283183,
      "loss": 0.3752,
      "num_input_tokens_seen": 53600992,
      "step": 38980
    },
    {
      "epoch": 1.247839446898406,
      "grad_norm": 0.0026238569989800453,
      "learning_rate": 0.00047730719740410874,
      "loss": 0.354,
      "num_input_tokens_seen": 53607888,
      "step": 38985
    },
    {
      "epoch": 1.2479994878688945,
      "grad_norm": 0.0062745073810219765,
      "learning_rate": 0.00047262331425816927,
      "loss": 0.5048,
      "num_input_tokens_seen": 53615088,
      "step": 38990
    },
    {
      "epoch": 1.248159528839383,
      "grad_norm": 0.005943346302956343,
      "learning_rate": 0.00046796249011277213,
      "loss": 0.3583,
      "num_input_tokens_seen": 53621680,
      "step": 38995
    },
    {
      "epoch": 1.2483195698098712,
      "grad_norm": 0.002950184280052781,
      "learning_rate": 0.00046332472568669236,
      "loss": 0.3781,
      "num_input_tokens_seen": 53628768,
      "step": 39000
    },
    {
      "epoch": 1.2483195698098712,
      "eval_loss": 0.4942808449268341,
      "eval_runtime": 332.2134,
      "eval_samples_per_second": 41.798,
      "eval_steps_per_second": 20.899,
      "num_input_tokens_seen": 53628768,
      "step": 39000
    },
    {
      "epoch": 1.2484796107803597,
      "grad_norm": 0.005643114447593689,
      "learning_rate": 0.0004587100216951578,
      "loss": 0.4964,
      "num_input_tokens_seen": 53636016,
      "step": 39005
    },
    {
      "epoch": 1.2486396517508482,
      "grad_norm": 0.008469357155263424,
      "learning_rate": 0.00045411837884978265,
      "loss": 0.4254,
      "num_input_tokens_seen": 53642896,
      "step": 39010
    },
    {
      "epoch": 1.2487996927213367,
      "grad_norm": 0.005195247940719128,
      "learning_rate": 0.00044954979785865045,
      "loss": 0.5744,
      "num_input_tokens_seen": 53650000,
      "step": 39015
    },
    {
      "epoch": 1.2489597336918252,
      "grad_norm": 0.0031536349561065435,
      "learning_rate": 0.00044500427942631426,
      "loss": 0.4934,
      "num_input_tokens_seen": 53657008,
      "step": 39020
    },
    {
      "epoch": 1.2491197746623135,
      "grad_norm": 0.002244573784992099,
      "learning_rate": 0.0004404818242537467,
      "loss": 0.4433,
      "num_input_tokens_seen": 53664352,
      "step": 39025
    },
    {
      "epoch": 1.249279815632802,
      "grad_norm": 0.005669764243066311,
      "learning_rate": 0.00043598243303837324,
      "loss": 0.4962,
      "num_input_tokens_seen": 53671536,
      "step": 39030
    },
    {
      "epoch": 1.2494398566032905,
      "grad_norm": 0.0036997543647885323,
      "learning_rate": 0.00043150610647403885,
      "loss": 0.5341,
      "num_input_tokens_seen": 53678784,
      "step": 39035
    },
    {
      "epoch": 1.249599897573779,
      "grad_norm": 0.002956325188279152,
      "learning_rate": 0.00042705284525104134,
      "loss": 0.3857,
      "num_input_tokens_seen": 53685536,
      "step": 39040
    },
    {
      "epoch": 1.2497599385442673,
      "grad_norm": 0.004811753518879414,
      "learning_rate": 0.0004226226500561647,
      "loss": 0.6308,
      "num_input_tokens_seen": 53692208,
      "step": 39045
    },
    {
      "epoch": 1.2499199795147558,
      "grad_norm": 0.008247412741184235,
      "learning_rate": 0.0004182155215725791,
      "loss": 0.8387,
      "num_input_tokens_seen": 53699328,
      "step": 39050
    },
    {
      "epoch": 1.2500800204852442,
      "grad_norm": 0.0025834462139755487,
      "learning_rate": 0.00041383146047992424,
      "loss": 0.4569,
      "num_input_tokens_seen": 53706112,
      "step": 39055
    },
    {
      "epoch": 1.2502400614557327,
      "grad_norm": 0.0035139238461852074,
      "learning_rate": 0.00040947046745427597,
      "loss": 0.5698,
      "num_input_tokens_seen": 53713312,
      "step": 39060
    },
    {
      "epoch": 1.2504001024262212,
      "grad_norm": 0.0053372071124613285,
      "learning_rate": 0.00040513254316814625,
      "loss": 0.5139,
      "num_input_tokens_seen": 53720288,
      "step": 39065
    },
    {
      "epoch": 1.2505601433967095,
      "grad_norm": 0.004988524597138166,
      "learning_rate": 0.0004008176882905168,
      "loss": 0.3627,
      "num_input_tokens_seen": 53727584,
      "step": 39070
    },
    {
      "epoch": 1.250720184367198,
      "grad_norm": 0.0034148471895605326,
      "learning_rate": 0.00039652590348677184,
      "loss": 0.5869,
      "num_input_tokens_seen": 53734224,
      "step": 39075
    },
    {
      "epoch": 1.2508802253376865,
      "grad_norm": 0.006672468967735767,
      "learning_rate": 0.00039225718941878206,
      "loss": 0.5469,
      "num_input_tokens_seen": 53741440,
      "step": 39080
    },
    {
      "epoch": 1.2510402663081748,
      "grad_norm": 0.004523755982518196,
      "learning_rate": 0.00038801154674480417,
      "loss": 0.3808,
      "num_input_tokens_seen": 53748464,
      "step": 39085
    },
    {
      "epoch": 1.2512003072786633,
      "grad_norm": 0.004328970331698656,
      "learning_rate": 0.00038378897611959784,
      "loss": 0.596,
      "num_input_tokens_seen": 53754960,
      "step": 39090
    },
    {
      "epoch": 1.2513603482491518,
      "grad_norm": 0.0030844248831272125,
      "learning_rate": 0.00037958947819430875,
      "loss": 0.3969,
      "num_input_tokens_seen": 53761984,
      "step": 39095
    },
    {
      "epoch": 1.2515203892196403,
      "grad_norm": 0.00515583623200655,
      "learning_rate": 0.0003754130536165856,
      "loss": 0.4496,
      "num_input_tokens_seen": 53768512,
      "step": 39100
    },
    {
      "epoch": 1.2516804301901288,
      "grad_norm": 0.004791393410414457,
      "learning_rate": 0.0003712597030304632,
      "loss": 0.3729,
      "num_input_tokens_seen": 53775424,
      "step": 39105
    },
    {
      "epoch": 1.2518404711606173,
      "grad_norm": 0.005838549230247736,
      "learning_rate": 0.00036712942707646247,
      "loss": 0.4548,
      "num_input_tokens_seen": 53782320,
      "step": 39110
    },
    {
      "epoch": 1.2520005121311055,
      "grad_norm": 0.0070443348959088326,
      "learning_rate": 0.00036302222639149063,
      "loss": 0.4781,
      "num_input_tokens_seen": 53789408,
      "step": 39115
    },
    {
      "epoch": 1.252160553101594,
      "grad_norm": 0.005138637498021126,
      "learning_rate": 0.000358938101608941,
      "loss": 0.6729,
      "num_input_tokens_seen": 53796192,
      "step": 39120
    },
    {
      "epoch": 1.2523205940720825,
      "grad_norm": 0.003858147654682398,
      "learning_rate": 0.0003548770533586598,
      "loss": 0.5007,
      "num_input_tokens_seen": 53802992,
      "step": 39125
    },
    {
      "epoch": 1.2524806350425708,
      "grad_norm": 0.0030665674712508917,
      "learning_rate": 0.0003508390822668961,
      "loss": 0.6105,
      "num_input_tokens_seen": 53809808,
      "step": 39130
    },
    {
      "epoch": 1.2526406760130593,
      "grad_norm": 0.0062347715720534325,
      "learning_rate": 0.00034682418895633503,
      "loss": 0.5688,
      "num_input_tokens_seen": 53816656,
      "step": 39135
    },
    {
      "epoch": 1.2528007169835478,
      "grad_norm": 0.004710826091468334,
      "learning_rate": 0.0003428323740461647,
      "loss": 0.3471,
      "num_input_tokens_seen": 53823712,
      "step": 39140
    },
    {
      "epoch": 1.2529607579540363,
      "grad_norm": 0.0031796349212527275,
      "learning_rate": 0.00033886363815194276,
      "loss": 0.5581,
      "num_input_tokens_seen": 53830832,
      "step": 39145
    },
    {
      "epoch": 1.2531207989245248,
      "grad_norm": 0.003713067388162017,
      "learning_rate": 0.0003349179818857129,
      "loss": 0.4595,
      "num_input_tokens_seen": 53837760,
      "step": 39150
    },
    {
      "epoch": 1.253280839895013,
      "grad_norm": 0.0035050990991294384,
      "learning_rate": 0.0003309954058559383,
      "loss": 0.467,
      "num_input_tokens_seen": 53844976,
      "step": 39155
    },
    {
      "epoch": 1.2534408808655015,
      "grad_norm": 0.0020010757725685835,
      "learning_rate": 0.0003270959106675186,
      "loss": 0.3217,
      "num_input_tokens_seen": 53851792,
      "step": 39160
    },
    {
      "epoch": 1.25360092183599,
      "grad_norm": 0.0064308373257517815,
      "learning_rate": 0.0003232194969218227,
      "loss": 0.5081,
      "num_input_tokens_seen": 53858656,
      "step": 39165
    },
    {
      "epoch": 1.2537609628064785,
      "grad_norm": 0.007823617197573185,
      "learning_rate": 0.00031936616521663905,
      "loss": 0.4925,
      "num_input_tokens_seen": 53865712,
      "step": 39170
    },
    {
      "epoch": 1.2539210037769668,
      "grad_norm": 0.0042335218749940395,
      "learning_rate": 0.00031553591614619236,
      "loss": 0.3659,
      "num_input_tokens_seen": 53872336,
      "step": 39175
    },
    {
      "epoch": 1.2540810447474553,
      "grad_norm": 0.004798342008143663,
      "learning_rate": 0.00031172875030117676,
      "loss": 0.4614,
      "num_input_tokens_seen": 53878880,
      "step": 39180
    },
    {
      "epoch": 1.2542410857179438,
      "grad_norm": 0.004145441111177206,
      "learning_rate": 0.0003079446682686726,
      "loss": 0.4595,
      "num_input_tokens_seen": 53885216,
      "step": 39185
    },
    {
      "epoch": 1.2544011266884323,
      "grad_norm": 0.00650867959484458,
      "learning_rate": 0.0003041836706322465,
      "loss": 0.5052,
      "num_input_tokens_seen": 53892096,
      "step": 39190
    },
    {
      "epoch": 1.2545611676589208,
      "grad_norm": 0.005561514757573605,
      "learning_rate": 0.0003004457579719011,
      "loss": 0.5801,
      "num_input_tokens_seen": 53898592,
      "step": 39195
    },
    {
      "epoch": 1.254721208629409,
      "grad_norm": 0.006450118497014046,
      "learning_rate": 0.00029673093086405867,
      "loss": 0.4386,
      "num_input_tokens_seen": 53905216,
      "step": 39200
    },
    {
      "epoch": 1.254721208629409,
      "eval_loss": 0.49421054124832153,
      "eval_runtime": 332.0386,
      "eval_samples_per_second": 41.82,
      "eval_steps_per_second": 20.91,
      "num_input_tokens_seen": 53905216,
      "step": 39200
    },
    {
      "epoch": 1.2548812495998976,
      "grad_norm": 0.007501902058720589,
      "learning_rate": 0.00029303918988159426,
      "loss": 0.3303,
      "num_input_tokens_seen": 53911904,
      "step": 39205
    },
    {
      "epoch": 1.255041290570386,
      "grad_norm": 0.00576972309499979,
      "learning_rate": 0.0002893705355938192,
      "loss": 0.5305,
      "num_input_tokens_seen": 53918640,
      "step": 39210
    },
    {
      "epoch": 1.2552013315408745,
      "grad_norm": 0.003709105309098959,
      "learning_rate": 0.0002857249685664975,
      "loss": 0.4047,
      "num_input_tokens_seen": 53925536,
      "step": 39215
    },
    {
      "epoch": 1.2553613725113628,
      "grad_norm": 0.00447639636695385,
      "learning_rate": 0.0002821024893618129,
      "loss": 0.6122,
      "num_input_tokens_seen": 53932144,
      "step": 39220
    },
    {
      "epoch": 1.2555214134818513,
      "grad_norm": 0.0042331404983997345,
      "learning_rate": 0.0002785030985383852,
      "loss": 0.4424,
      "num_input_tokens_seen": 53938880,
      "step": 39225
    },
    {
      "epoch": 1.2556814544523398,
      "grad_norm": 0.004385132342576981,
      "learning_rate": 0.00027492679665130356,
      "loss": 0.3999,
      "num_input_tokens_seen": 53945744,
      "step": 39230
    },
    {
      "epoch": 1.2558414954228283,
      "grad_norm": 0.005672778934240341,
      "learning_rate": 0.000271373584252077,
      "loss": 0.5619,
      "num_input_tokens_seen": 53952576,
      "step": 39235
    },
    {
      "epoch": 1.2560015363933168,
      "grad_norm": 0.004331821575760841,
      "learning_rate": 0.00026784346188865046,
      "loss": 0.3854,
      "num_input_tokens_seen": 53958944,
      "step": 39240
    },
    {
      "epoch": 1.256161577363805,
      "grad_norm": 0.0025435425341129303,
      "learning_rate": 0.0002643364301054218,
      "loss": 0.3236,
      "num_input_tokens_seen": 53966000,
      "step": 39245
    },
    {
      "epoch": 1.2563216183342936,
      "grad_norm": 0.005649274215102196,
      "learning_rate": 0.0002608524894431918,
      "loss": 0.4841,
      "num_input_tokens_seen": 53972784,
      "step": 39250
    },
    {
      "epoch": 1.256481659304782,
      "grad_norm": 0.00490552419796586,
      "learning_rate": 0.000257391640439264,
      "loss": 0.5052,
      "num_input_tokens_seen": 53979904,
      "step": 39255
    },
    {
      "epoch": 1.2566417002752703,
      "grad_norm": 0.006660494953393936,
      "learning_rate": 0.00025395388362732806,
      "loss": 0.5899,
      "num_input_tokens_seen": 53986928,
      "step": 39260
    },
    {
      "epoch": 1.2568017412457588,
      "grad_norm": 0.004187269136309624,
      "learning_rate": 0.00025053921953751,
      "loss": 0.4355,
      "num_input_tokens_seen": 53993904,
      "step": 39265
    },
    {
      "epoch": 1.2569617822162473,
      "grad_norm": 0.004852038808166981,
      "learning_rate": 0.00024714764869643855,
      "loss": 0.4186,
      "num_input_tokens_seen": 54000832,
      "step": 39270
    },
    {
      "epoch": 1.2571218231867358,
      "grad_norm": 0.008993199095129967,
      "learning_rate": 0.0002437791716270954,
      "loss": 0.5138,
      "num_input_tokens_seen": 54007616,
      "step": 39275
    },
    {
      "epoch": 1.2572818641572243,
      "grad_norm": 0.003017691895365715,
      "learning_rate": 0.00024043378884896493,
      "loss": 0.3848,
      "num_input_tokens_seen": 54014304,
      "step": 39280
    },
    {
      "epoch": 1.2574419051277128,
      "grad_norm": 0.006423099897801876,
      "learning_rate": 0.00023711150087793453,
      "loss": 0.4209,
      "num_input_tokens_seen": 54021248,
      "step": 39285
    },
    {
      "epoch": 1.257601946098201,
      "grad_norm": 0.007025274448096752,
      "learning_rate": 0.000233812308226361,
      "loss": 0.4967,
      "num_input_tokens_seen": 54027936,
      "step": 39290
    },
    {
      "epoch": 1.2577619870686896,
      "grad_norm": 0.006387133151292801,
      "learning_rate": 0.00023053621140300406,
      "loss": 0.4195,
      "num_input_tokens_seen": 54035120,
      "step": 39295
    },
    {
      "epoch": 1.257922028039178,
      "grad_norm": 0.008068781346082687,
      "learning_rate": 0.00022728321091307623,
      "loss": 0.5619,
      "num_input_tokens_seen": 54042032,
      "step": 39300
    },
    {
      "epoch": 1.2580820690096663,
      "grad_norm": 0.008205531165003777,
      "learning_rate": 0.0002240533072582429,
      "loss": 0.4197,
      "num_input_tokens_seen": 54048784,
      "step": 39305
    },
    {
      "epoch": 1.2582421099801548,
      "grad_norm": 0.005941591691225767,
      "learning_rate": 0.00022084650093658897,
      "loss": 0.4996,
      "num_input_tokens_seen": 54055664,
      "step": 39310
    },
    {
      "epoch": 1.2584021509506433,
      "grad_norm": 0.01161158923059702,
      "learning_rate": 0.0002176627924426522,
      "loss": 0.3887,
      "num_input_tokens_seen": 54062720,
      "step": 39315
    },
    {
      "epoch": 1.2585621919211318,
      "grad_norm": 0.0037168359849601984,
      "learning_rate": 0.0002145021822673898,
      "loss": 0.4492,
      "num_input_tokens_seen": 54069408,
      "step": 39320
    },
    {
      "epoch": 1.2587222328916203,
      "grad_norm": 0.00366068328730762,
      "learning_rate": 0.00021136467089822862,
      "loss": 0.4898,
      "num_input_tokens_seen": 54076768,
      "step": 39325
    },
    {
      "epoch": 1.2588822738621088,
      "grad_norm": 0.00454922067001462,
      "learning_rate": 0.00020825025881898162,
      "loss": 0.6223,
      "num_input_tokens_seen": 54083520,
      "step": 39330
    },
    {
      "epoch": 1.259042314832597,
      "grad_norm": 0.003774384269490838,
      "learning_rate": 0.0002051589465099479,
      "loss": 0.5713,
      "num_input_tokens_seen": 54090448,
      "step": 39335
    },
    {
      "epoch": 1.2592023558030856,
      "grad_norm": 0.005633055232465267,
      "learning_rate": 0.0002020907344478462,
      "loss": 0.6313,
      "num_input_tokens_seen": 54097296,
      "step": 39340
    },
    {
      "epoch": 1.259362396773574,
      "grad_norm": 0.004522132687270641,
      "learning_rate": 0.0001990456231058313,
      "loss": 0.3966,
      "num_input_tokens_seen": 54104272,
      "step": 39345
    },
    {
      "epoch": 1.2595224377440624,
      "grad_norm": 0.005263231229037046,
      "learning_rate": 0.00019602361295349423,
      "loss": 0.5788,
      "num_input_tokens_seen": 54110848,
      "step": 39350
    },
    {
      "epoch": 1.2596824787145509,
      "grad_norm": 0.004403846804052591,
      "learning_rate": 0.0001930247044568789,
      "loss": 0.3557,
      "num_input_tokens_seen": 54117584,
      "step": 39355
    },
    {
      "epoch": 1.2598425196850394,
      "grad_norm": 0.006946584675461054,
      "learning_rate": 0.00019004889807843205,
      "loss": 0.6297,
      "num_input_tokens_seen": 54124336,
      "step": 39360
    },
    {
      "epoch": 1.2600025606555278,
      "grad_norm": 0.004454789217561483,
      "learning_rate": 0.00018709619427708656,
      "loss": 0.4699,
      "num_input_tokens_seen": 54130992,
      "step": 39365
    },
    {
      "epoch": 1.2601626016260163,
      "grad_norm": 0.004535517655313015,
      "learning_rate": 0.00018416659350817822,
      "loss": 0.3811,
      "num_input_tokens_seen": 54137584,
      "step": 39370
    },
    {
      "epoch": 1.2603226425965048,
      "grad_norm": 0.002171353902667761,
      "learning_rate": 0.00018126009622346229,
      "loss": 0.3279,
      "num_input_tokens_seen": 54144400,
      "step": 39375
    },
    {
      "epoch": 1.2604826835669931,
      "grad_norm": 0.005494707264006138,
      "learning_rate": 0.00017837670287119687,
      "loss": 0.3849,
      "num_input_tokens_seen": 54151392,
      "step": 39380
    },
    {
      "epoch": 1.2606427245374816,
      "grad_norm": 0.004797439090907574,
      "learning_rate": 0.00017551641389602633,
      "loss": 0.398,
      "num_input_tokens_seen": 54157952,
      "step": 39385
    },
    {
      "epoch": 1.26080276550797,
      "grad_norm": 0.0036161504685878754,
      "learning_rate": 0.00017267922973903115,
      "loss": 0.4449,
      "num_input_tokens_seen": 54165008,
      "step": 39390
    },
    {
      "epoch": 1.2609628064784584,
      "grad_norm": 0.00565301114693284,
      "learning_rate": 0.00016986515083774467,
      "loss": 0.741,
      "num_input_tokens_seen": 54171920,
      "step": 39395
    },
    {
      "epoch": 1.2611228474489469,
      "grad_norm": 0.0046753655187785625,
      "learning_rate": 0.00016707417762611975,
      "loss": 0.4696,
      "num_input_tokens_seen": 54178832,
      "step": 39400
    },
    {
      "epoch": 1.2611228474489469,
      "eval_loss": 0.49426010251045227,
      "eval_runtime": 331.8173,
      "eval_samples_per_second": 41.848,
      "eval_steps_per_second": 20.924,
      "num_input_tokens_seen": 54178832,
      "step": 39400
    },
    {
      "epoch": 1.2612828884194354,
      "grad_norm": 0.002416584175080061,
      "learning_rate": 0.00016430631053459543,
      "loss": 0.3978,
      "num_input_tokens_seen": 54185312,
      "step": 39405
    },
    {
      "epoch": 1.2614429293899239,
      "grad_norm": 0.005246045999228954,
      "learning_rate": 0.0001615615499899803,
      "loss": 0.588,
      "num_input_tokens_seen": 54193008,
      "step": 39410
    },
    {
      "epoch": 1.2616029703604124,
      "grad_norm": 0.0040269773453474045,
      "learning_rate": 0.00015883989641556905,
      "loss": 0.5266,
      "num_input_tokens_seen": 54199984,
      "step": 39415
    },
    {
      "epoch": 1.2617630113309006,
      "grad_norm": 0.00325546832755208,
      "learning_rate": 0.00015614135023105934,
      "loss": 0.5782,
      "num_input_tokens_seen": 54206848,
      "step": 39420
    },
    {
      "epoch": 1.2619230523013891,
      "grad_norm": 0.006333900615572929,
      "learning_rate": 0.00015346591185261827,
      "loss": 0.7581,
      "num_input_tokens_seen": 54213600,
      "step": 39425
    },
    {
      "epoch": 1.2620830932718776,
      "grad_norm": 0.005635167472064495,
      "learning_rate": 0.00015081358169281576,
      "loss": 0.4581,
      "num_input_tokens_seen": 54220192,
      "step": 39430
    },
    {
      "epoch": 1.2622431342423661,
      "grad_norm": 0.004841622430831194,
      "learning_rate": 0.00014818436016069135,
      "loss": 0.6352,
      "num_input_tokens_seen": 54227040,
      "step": 39435
    },
    {
      "epoch": 1.2624031752128544,
      "grad_norm": 0.006485736463218927,
      "learning_rate": 0.00014557824766168735,
      "loss": 0.4045,
      "num_input_tokens_seen": 54233664,
      "step": 39440
    },
    {
      "epoch": 1.2625632161833429,
      "grad_norm": 0.005037694238126278,
      "learning_rate": 0.00014299524459769896,
      "loss": 0.6031,
      "num_input_tokens_seen": 54240480,
      "step": 39445
    },
    {
      "epoch": 1.2627232571538314,
      "grad_norm": 0.00869474746286869,
      "learning_rate": 0.0001404353513670742,
      "loss": 0.3759,
      "num_input_tokens_seen": 54247296,
      "step": 39450
    },
    {
      "epoch": 1.2628832981243199,
      "grad_norm": 0.004763994365930557,
      "learning_rate": 0.0001378985683645806,
      "loss": 0.28,
      "num_input_tokens_seen": 54254640,
      "step": 39455
    },
    {
      "epoch": 1.2630433390948084,
      "grad_norm": 0.005435759201645851,
      "learning_rate": 0.0001353848959813886,
      "loss": 0.4799,
      "num_input_tokens_seen": 54261312,
      "step": 39460
    },
    {
      "epoch": 1.2632033800652966,
      "grad_norm": 0.003765301313251257,
      "learning_rate": 0.00013289433460517142,
      "loss": 0.4302,
      "num_input_tokens_seen": 54267792,
      "step": 39465
    },
    {
      "epoch": 1.2633634210357851,
      "grad_norm": 0.004452323541045189,
      "learning_rate": 0.00013042688462000518,
      "loss": 0.4058,
      "num_input_tokens_seen": 54274736,
      "step": 39470
    },
    {
      "epoch": 1.2635234620062736,
      "grad_norm": 0.004734619054943323,
      "learning_rate": 0.0001279825464063855,
      "loss": 0.3603,
      "num_input_tokens_seen": 54281440,
      "step": 39475
    },
    {
      "epoch": 1.2636835029767621,
      "grad_norm": 0.004783754702657461,
      "learning_rate": 0.00012556132034126087,
      "loss": 0.4894,
      "num_input_tokens_seen": 54288320,
      "step": 39480
    },
    {
      "epoch": 1.2638435439472504,
      "grad_norm": 0.0068643647246062756,
      "learning_rate": 0.0001231632067980326,
      "loss": 0.4309,
      "num_input_tokens_seen": 54295392,
      "step": 39485
    },
    {
      "epoch": 1.264003584917739,
      "grad_norm": 0.0037192651070654392,
      "learning_rate": 0.00012078820614650486,
      "loss": 0.3749,
      "num_input_tokens_seen": 54302448,
      "step": 39490
    },
    {
      "epoch": 1.2641636258882274,
      "grad_norm": 0.005339960101991892,
      "learning_rate": 0.00011843631875291804,
      "loss": 0.4143,
      "num_input_tokens_seen": 54309584,
      "step": 39495
    },
    {
      "epoch": 1.2643236668587159,
      "grad_norm": 0.0038518698420375586,
      "learning_rate": 0.00011610754497999863,
      "loss": 0.3954,
      "num_input_tokens_seen": 54317040,
      "step": 39500
    },
    {
      "epoch": 1.2644837078292044,
      "grad_norm": 0.0033304444514214993,
      "learning_rate": 0.0001138018851868594,
      "loss": 0.4308,
      "num_input_tokens_seen": 54324208,
      "step": 39505
    },
    {
      "epoch": 1.2646437487996927,
      "grad_norm": 0.004306076094508171,
      "learning_rate": 0.0001115193397290326,
      "loss": 0.4857,
      "num_input_tokens_seen": 54330944,
      "step": 39510
    },
    {
      "epoch": 1.2648037897701812,
      "grad_norm": 0.004816857632249594,
      "learning_rate": 0.00010925990895856996,
      "loss": 0.4851,
      "num_input_tokens_seen": 54337440,
      "step": 39515
    },
    {
      "epoch": 1.2649638307406696,
      "grad_norm": 0.006422546226531267,
      "learning_rate": 0.00010702359322385946,
      "loss": 0.3515,
      "num_input_tokens_seen": 54344608,
      "step": 39520
    },
    {
      "epoch": 1.265123871711158,
      "grad_norm": 0.006676146294921637,
      "learning_rate": 0.00010481039286977523,
      "loss": 0.461,
      "num_input_tokens_seen": 54351472,
      "step": 39525
    },
    {
      "epoch": 1.2652839126816464,
      "grad_norm": 0.004387765657156706,
      "learning_rate": 0.00010262030823764423,
      "loss": 0.3344,
      "num_input_tokens_seen": 54358208,
      "step": 39530
    },
    {
      "epoch": 1.265443953652135,
      "grad_norm": 0.00472960714250803,
      "learning_rate": 0.00010045333966517966,
      "loss": 0.4538,
      "num_input_tokens_seen": 54364656,
      "step": 39535
    },
    {
      "epoch": 1.2656039946226234,
      "grad_norm": 0.003973979037255049,
      "learning_rate": 9.83094874865642e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 54371808,
      "step": 39540
    },
    {
      "epoch": 1.265764035593112,
      "grad_norm": 0.006642491091042757,
      "learning_rate": 9.618875203241672e-05,
      "loss": 0.5658,
      "num_input_tokens_seen": 54378752,
      "step": 39545
    },
    {
      "epoch": 1.2659240765636004,
      "grad_norm": 0.004995420575141907,
      "learning_rate": 9.409113362977561e-05,
      "loss": 0.615,
      "num_input_tokens_seen": 54385456,
      "step": 39550
    },
    {
      "epoch": 1.2660841175340887,
      "grad_norm": 0.006008084397763014,
      "learning_rate": 9.20166326020988e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 54392608,
      "step": 39555
    },
    {
      "epoch": 1.2662441585045772,
      "grad_norm": 0.013180388137698174,
      "learning_rate": 8.996524926933035e-05,
      "loss": 0.5957,
      "num_input_tokens_seen": 54399728,
      "step": 39560
    },
    {
      "epoch": 1.2664041994750657,
      "grad_norm": 0.0060635400004684925,
      "learning_rate": 8.793698394781723e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 54406448,
      "step": 39565
    },
    {
      "epoch": 1.266564240445554,
      "grad_norm": 0.005931040272116661,
      "learning_rate": 8.593183695030926e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 54413712,
      "step": 39570
    },
    {
      "epoch": 1.2667242814160424,
      "grad_norm": 0.006419721059501171,
      "learning_rate": 8.39498085860757e-05,
      "loss": 0.3804,
      "num_input_tokens_seen": 54420576,
      "step": 39575
    },
    {
      "epoch": 1.266884322386531,
      "grad_norm": 0.004587596748024225,
      "learning_rate": 8.199089916072211e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 54427824,
      "step": 39580
    },
    {
      "epoch": 1.2670443633570194,
      "grad_norm": 0.0041415211744606495,
      "learning_rate": 8.005510897637346e-05,
      "loss": 0.4226,
      "num_input_tokens_seen": 54434576,
      "step": 39585
    },
    {
      "epoch": 1.267204404327508,
      "grad_norm": 0.006177142728120089,
      "learning_rate": 7.8142438331541e-05,
      "loss": 0.5364,
      "num_input_tokens_seen": 54441392,
      "step": 39590
    },
    {
      "epoch": 1.2673644452979964,
      "grad_norm": 0.008940581232309341,
      "learning_rate": 7.625288752117209e-05,
      "loss": 0.6025,
      "num_input_tokens_seen": 54447872,
      "step": 39595
    },
    {
      "epoch": 1.2675244862684847,
      "grad_norm": 0.0036927314940840006,
      "learning_rate": 7.4386456836667e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 54454880,
      "step": 39600
    },
    {
      "epoch": 1.2675244862684847,
      "eval_loss": 0.49425679445266724,
      "eval_runtime": 331.9948,
      "eval_samples_per_second": 41.826,
      "eval_steps_per_second": 20.913,
      "num_input_tokens_seen": 54454880,
      "step": 39600
    },
    {
      "epoch": 1.2676845272389732,
      "grad_norm": 0.0020353528670966625,
      "learning_rate": 7.254314656586214e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 54462032,
      "step": 39605
    },
    {
      "epoch": 1.2678445682094617,
      "grad_norm": 0.005065666511654854,
      "learning_rate": 7.07229569929968e-05,
      "loss": 0.5466,
      "num_input_tokens_seen": 54468704,
      "step": 39610
    },
    {
      "epoch": 1.26800460917995,
      "grad_norm": 0.004077026154845953,
      "learning_rate": 6.892588839879643e-05,
      "loss": 0.3139,
      "num_input_tokens_seen": 54475296,
      "step": 39615
    },
    {
      "epoch": 1.2681646501504384,
      "grad_norm": 0.00603028479963541,
      "learning_rate": 6.71519410603727e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 54481920,
      "step": 39620
    },
    {
      "epoch": 1.268324691120927,
      "grad_norm": 0.002498287009075284,
      "learning_rate": 6.540111525129011e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 54488688,
      "step": 39625
    },
    {
      "epoch": 1.2684847320914154,
      "grad_norm": 0.004017556086182594,
      "learning_rate": 6.367341124154934e-05,
      "loss": 0.2573,
      "num_input_tokens_seen": 54495296,
      "step": 39630
    },
    {
      "epoch": 1.268644773061904,
      "grad_norm": 0.0036966942716389894,
      "learning_rate": 6.19688292975873e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 54502112,
      "step": 39635
    },
    {
      "epoch": 1.2688048140323924,
      "grad_norm": 0.008614734746515751,
      "learning_rate": 6.0287369682260336e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 54508928,
      "step": 39640
    },
    {
      "epoch": 1.2689648550028807,
      "grad_norm": 0.015342490747570992,
      "learning_rate": 5.8629032654894384e-05,
      "loss": 0.6147,
      "num_input_tokens_seen": 54515536,
      "step": 39645
    },
    {
      "epoch": 1.2691248959733692,
      "grad_norm": 0.0040258741937577724,
      "learning_rate": 5.699381847120155e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 54522912,
      "step": 39650
    },
    {
      "epoch": 1.2692849369438577,
      "grad_norm": 0.005509316921234131,
      "learning_rate": 5.5381727383380094e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 54529600,
      "step": 39655
    },
    {
      "epoch": 1.269444977914346,
      "grad_norm": 0.004058588296175003,
      "learning_rate": 5.379275964001451e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 54536256,
      "step": 39660
    },
    {
      "epoch": 1.2696050188848345,
      "grad_norm": 0.005145275499671698,
      "learning_rate": 5.222691548614211e-05,
      "loss": 0.3254,
      "num_input_tokens_seen": 54543056,
      "step": 39665
    },
    {
      "epoch": 1.269765059855323,
      "grad_norm": 0.004237387329339981,
      "learning_rate": 5.068419516323641e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 54550016,
      "step": 39670
    },
    {
      "epoch": 1.2699251008258114,
      "grad_norm": 0.006733264308422804,
      "learning_rate": 4.91645989092071e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 54556960,
      "step": 39675
    },
    {
      "epoch": 1.2700851417963,
      "grad_norm": 0.005315643735229969,
      "learning_rate": 4.7668126958400056e-05,
      "loss": 0.5752,
      "num_input_tokens_seen": 54563760,
      "step": 39680
    },
    {
      "epoch": 1.2702451827667882,
      "grad_norm": 0.003766103647649288,
      "learning_rate": 4.619477954159734e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 54570768,
      "step": 39685
    },
    {
      "epoch": 1.2704052237372767,
      "grad_norm": 0.00521598057821393,
      "learning_rate": 4.4744556885983884e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 54577712,
      "step": 39690
    },
    {
      "epoch": 1.2705652647077652,
      "grad_norm": 0.005544449668377638,
      "learning_rate": 4.331745921523078e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 54584176,
      "step": 39695
    },
    {
      "epoch": 1.2707253056782537,
      "grad_norm": 0.0029632337391376495,
      "learning_rate": 4.191348674937867e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 54590512,
      "step": 39700
    },
    {
      "epoch": 1.270885346648742,
      "grad_norm": 0.008102604188024998,
      "learning_rate": 4.0532639704971006e-05,
      "loss": 0.6051,
      "num_input_tokens_seen": 54597568,
      "step": 39705
    },
    {
      "epoch": 1.2710453876192305,
      "grad_norm": 0.0049330564215779305,
      "learning_rate": 3.917491829493747e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 54604624,
      "step": 39710
    },
    {
      "epoch": 1.271205428589719,
      "grad_norm": 0.004166396800428629,
      "learning_rate": 3.78403227286439e-05,
      "loss": 0.5159,
      "num_input_tokens_seen": 54611008,
      "step": 39715
    },
    {
      "epoch": 1.2713654695602075,
      "grad_norm": 0.004125366453081369,
      "learning_rate": 3.652885321192567e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 54618160,
      "step": 39720
    },
    {
      "epoch": 1.271525510530696,
      "grad_norm": 0.008453198708593845,
      "learning_rate": 3.524050994702099e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 54624560,
      "step": 39725
    },
    {
      "epoch": 1.2716855515011842,
      "grad_norm": 0.007176497485488653,
      "learning_rate": 3.3975293132604276e-05,
      "loss": 0.3557,
      "num_input_tokens_seen": 54630960,
      "step": 39730
    },
    {
      "epoch": 1.2718455924716727,
      "grad_norm": 0.005038166418671608,
      "learning_rate": 3.2733202963786125e-05,
      "loss": 0.3694,
      "num_input_tokens_seen": 54637952,
      "step": 39735
    },
    {
      "epoch": 1.2720056334421612,
      "grad_norm": 0.005283446051180363,
      "learning_rate": 3.15142396321133e-05,
      "loss": 0.5741,
      "num_input_tokens_seen": 54644960,
      "step": 39740
    },
    {
      "epoch": 1.2721656744126497,
      "grad_norm": 0.003673679893836379,
      "learning_rate": 3.0318403325552132e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 54651872,
      "step": 39745
    },
    {
      "epoch": 1.272325715383138,
      "grad_norm": 0.005671453196555376,
      "learning_rate": 2.914569422855506e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 54658704,
      "step": 39750
    },
    {
      "epoch": 1.2724857563536265,
      "grad_norm": 0.005116924177855253,
      "learning_rate": 2.7996112521927462e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 54665632,
      "step": 39755
    },
    {
      "epoch": 1.272645797324115,
      "grad_norm": 0.004990240093320608,
      "learning_rate": 2.68696583829775e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 54672432,
      "step": 39760
    },
    {
      "epoch": 1.2728058382946035,
      "grad_norm": 0.004011854529380798,
      "learning_rate": 2.576633198539957e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 54679136,
      "step": 39765
    },
    {
      "epoch": 1.272965879265092,
      "grad_norm": 0.007876025512814522,
      "learning_rate": 2.46861334993409e-05,
      "loss": 0.371,
      "num_input_tokens_seen": 54685456,
      "step": 39770
    },
    {
      "epoch": 1.2731259202355802,
      "grad_norm": 0.004998389165848494,
      "learning_rate": 2.3629063091384903e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 54691936,
      "step": 39775
    },
    {
      "epoch": 1.2732859612060687,
      "grad_norm": 0.005232954863458872,
      "learning_rate": 2.2595120924567834e-05,
      "loss": 0.6072,
      "num_input_tokens_seen": 54698944,
      "step": 39780
    },
    {
      "epoch": 1.2734460021765572,
      "grad_norm": 0.004834949970245361,
      "learning_rate": 2.158430715829551e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 54706176,
      "step": 39785
    },
    {
      "epoch": 1.2736060431470455,
      "grad_norm": 0.0050677102990448475,
      "learning_rate": 2.059662194849321e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 54713360,
      "step": 39790
    },
    {
      "epoch": 1.273766084117534,
      "grad_norm": 0.009034112095832825,
      "learning_rate": 1.9632065447422463e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 54720384,
      "step": 39795
    },
    {
      "epoch": 1.2739261250880225,
      "grad_norm": 0.013339472934603691,
      "learning_rate": 1.8690637803880916e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 54727600,
      "step": 39800
    },
    {
      "epoch": 1.2739261250880225,
      "eval_loss": 0.49426430463790894,
      "eval_runtime": 332.0591,
      "eval_samples_per_second": 41.818,
      "eval_steps_per_second": 20.909,
      "num_input_tokens_seen": 54727600,
      "step": 39800
    },
    {
      "epoch": 1.274086166058511,
      "grad_norm": 0.004642052110284567,
      "learning_rate": 1.7772339163019123e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 54734624,
      "step": 39805
    },
    {
      "epoch": 1.2742462070289995,
      "grad_norm": 0.003326823702082038,
      "learning_rate": 1.6877169666457138e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 54741424,
      "step": 39810
    },
    {
      "epoch": 1.274406247999488,
      "grad_norm": 0.007395625114440918,
      "learning_rate": 1.6005129452234532e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 54747760,
      "step": 39815
    },
    {
      "epoch": 1.2745662889699763,
      "grad_norm": 0.00343105080537498,
      "learning_rate": 1.5156218654843733e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 54754880,
      "step": 39820
    },
    {
      "epoch": 1.2747263299404648,
      "grad_norm": 0.0038751361425966024,
      "learning_rate": 1.4330437405196683e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 54761840,
      "step": 39825
    },
    {
      "epoch": 1.2748863709109532,
      "grad_norm": 0.00366242416203022,
      "learning_rate": 1.352778583062486e-05,
      "loss": 0.5163,
      "num_input_tokens_seen": 54768896,
      "step": 39830
    },
    {
      "epoch": 1.2750464118814415,
      "grad_norm": 0.005101477261632681,
      "learning_rate": 1.2748264054929237e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 54775792,
      "step": 39835
    },
    {
      "epoch": 1.27520645285193,
      "grad_norm": 0.008035027422010899,
      "learning_rate": 1.1991872198297004e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 54783008,
      "step": 39840
    },
    {
      "epoch": 1.2753664938224185,
      "grad_norm": 0.00566979730501771,
      "learning_rate": 1.1258610377384847e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 54790336,
      "step": 39845
    },
    {
      "epoch": 1.275526534792907,
      "grad_norm": 0.006106571760028601,
      "learning_rate": 1.0548478705268982e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 54797264,
      "step": 39850
    },
    {
      "epoch": 1.2756865757633955,
      "grad_norm": 0.007692999672144651,
      "learning_rate": 9.86147729147846e-06,
      "loss": 0.6021,
      "num_input_tokens_seen": 54803952,
      "step": 39855
    },
    {
      "epoch": 1.275846616733884,
      "grad_norm": 0.0075046829879283905,
      "learning_rate": 9.197606241928557e-06,
      "loss": 0.5051,
      "num_input_tokens_seen": 54810688,
      "step": 39860
    },
    {
      "epoch": 1.2760066577043723,
      "grad_norm": 0.0051543014124035835,
      "learning_rate": 8.556865659004042e-06,
      "loss": 0.3728,
      "num_input_tokens_seen": 54817216,
      "step": 39865
    },
    {
      "epoch": 1.2761666986748608,
      "grad_norm": 0.008325561881065369,
      "learning_rate": 7.939255641525867e-06,
      "loss": 0.5634,
      "num_input_tokens_seen": 54824416,
      "step": 39870
    },
    {
      "epoch": 1.2763267396453493,
      "grad_norm": 0.003542707534506917,
      "learning_rate": 7.344776284751164e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 54831648,
      "step": 39875
    },
    {
      "epoch": 1.2764867806158375,
      "grad_norm": 0.005141942761838436,
      "learning_rate": 6.773427680323296e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 54839168,
      "step": 39880
    },
    {
      "epoch": 1.276646821586326,
      "grad_norm": 0.003757500322535634,
      "learning_rate": 6.225209916355112e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 54845904,
      "step": 39885
    },
    {
      "epoch": 1.2768068625568145,
      "grad_norm": 0.0047017671167850494,
      "learning_rate": 5.7001230774123e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 54852784,
      "step": 39890
    },
    {
      "epoch": 1.276966903527303,
      "grad_norm": 0.004946654662489891,
      "learning_rate": 5.198167244446772e-06,
      "loss": 0.5106,
      "num_input_tokens_seen": 54859440,
      "step": 39895
    },
    {
      "epoch": 1.2771269444977915,
      "grad_norm": 0.00405862694606185,
      "learning_rate": 4.71934249487993e-06,
      "loss": 0.6145,
      "num_input_tokens_seen": 54866608,
      "step": 39900
    },
    {
      "epoch": 1.27728698546828,
      "grad_norm": 0.003218887373805046,
      "learning_rate": 4.2636489025527075e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 54873104,
      "step": 39905
    },
    {
      "epoch": 1.2774470264387683,
      "grad_norm": 0.0038609218318015337,
      "learning_rate": 3.831086537742223e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 54879648,
      "step": 39910
    },
    {
      "epoch": 1.2776070674092568,
      "grad_norm": 0.002092010574415326,
      "learning_rate": 3.4216554671451236e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 54886688,
      "step": 39915
    },
    {
      "epoch": 1.2777671083797453,
      "grad_norm": 0.00404264684766531,
      "learning_rate": 3.035355753894242e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 54893328,
      "step": 39920
    },
    {
      "epoch": 1.2779271493502335,
      "grad_norm": 0.006586118135601282,
      "learning_rate": 2.6721874575752477e-06,
      "loss": 0.5776,
      "num_input_tokens_seen": 54899984,
      "step": 39925
    },
    {
      "epoch": 1.278087190320722,
      "grad_norm": 0.004300561733543873,
      "learning_rate": 2.3321506341933418e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 54906832,
      "step": 39930
    },
    {
      "epoch": 1.2782472312912105,
      "grad_norm": 0.0033757342025637627,
      "learning_rate": 2.0152453361732546e-06,
      "loss": 0.3766,
      "num_input_tokens_seen": 54913632,
      "step": 39935
    },
    {
      "epoch": 1.278407272261699,
      "grad_norm": 0.002653788309544325,
      "learning_rate": 1.7214716123925554e-06,
      "loss": 0.5157,
      "num_input_tokens_seen": 54920528,
      "step": 39940
    },
    {
      "epoch": 1.2785673132321875,
      "grad_norm": 0.008940167725086212,
      "learning_rate": 1.4508295081649968e-06,
      "loss": 0.5987,
      "num_input_tokens_seen": 54927584,
      "step": 39945
    },
    {
      "epoch": 1.2787273542026758,
      "grad_norm": 0.008837027475237846,
      "learning_rate": 1.2033190652238623e-06,
      "loss": 0.5308,
      "num_input_tokens_seen": 54934240,
      "step": 39950
    },
    {
      "epoch": 1.2788873951731643,
      "grad_norm": 0.004228559322655201,
      "learning_rate": 9.78940321721966e-07,
      "loss": 0.5319,
      "num_input_tokens_seen": 54941232,
      "step": 39955
    },
    {
      "epoch": 1.2790474361436528,
      "grad_norm": 0.009533025324344635,
      "learning_rate": 7.776933122816132e-07,
      "loss": 0.3879,
      "num_input_tokens_seen": 54947808,
      "step": 39960
    },
    {
      "epoch": 1.2792074771141413,
      "grad_norm": 0.00748869962990284,
      "learning_rate": 5.99578067927986e-07,
      "loss": 0.5018,
      "num_input_tokens_seen": 54954704,
      "step": 39965
    },
    {
      "epoch": 1.2793675180846296,
      "grad_norm": 0.003846898442134261,
      "learning_rate": 4.445946161224512e-07,
      "loss": 0.471,
      "num_input_tokens_seen": 54961680,
      "step": 39970
    },
    {
      "epoch": 1.279527559055118,
      "grad_norm": 0.0027124364860355854,
      "learning_rate": 3.127429807792126e-07,
      "loss": 0.5018,
      "num_input_tokens_seen": 54968656,
      "step": 39975
    },
    {
      "epoch": 1.2796876000256066,
      "grad_norm": 0.005146179813891649,
      "learning_rate": 2.040231822320049e-07,
      "loss": 0.4346,
      "num_input_tokens_seen": 54975280,
      "step": 39980
    },
    {
      "epoch": 1.279847640996095,
      "grad_norm": 0.004543966148048639,
      "learning_rate": 1.1843523723409354e-07,
      "loss": 0.4366,
      "num_input_tokens_seen": 54982032,
      "step": 39985
    },
    {
      "epoch": 1.2800076819665835,
      "grad_norm": 0.0042426371946930885,
      "learning_rate": 5.597915897492811e-08,
      "loss": 0.3981,
      "num_input_tokens_seen": 54988688,
      "step": 39990
    },
    {
      "epoch": 1.2801677229370718,
      "grad_norm": 0.0032367845997214317,
      "learning_rate": 1.6654957113448885e-08,
      "loss": 0.3084,
      "num_input_tokens_seen": 54995344,
      "step": 39995
    },
    {
      "epoch": 1.2803277639075603,
      "grad_norm": 0.004257837776094675,
      "learning_rate": 4.626377114735902e-10,
      "loss": 0.4098,
      "num_input_tokens_seen": 55002224,
      "step": 40000
    },
    {
      "epoch": 1.2803277639075603,
      "eval_loss": 0.49421823024749756,
      "eval_runtime": 332.1613,
      "eval_samples_per_second": 41.805,
      "eval_steps_per_second": 20.902,
      "num_input_tokens_seen": 55002224,
      "step": 40000
    },
    {
      "epoch": 1.2803277639075603,
      "num_input_tokens_seen": 55002224,
      "step": 40000,
      "total_flos": 2.30315013845078e+17,
      "train_loss": 0.6158369358956813,
      "train_runtime": 75158.795,
      "train_samples_per_second": 2.129,
      "train_steps_per_second": 0.532
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 55002224,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.30315013845078e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}