{
  "best_global_step": 33200,
  "best_metric": 0.31929609179496765,
  "best_model_checkpoint": "saves/prompt-tuning/llama-3-8b-instruct/train_multirc_1745950264/checkpoint-33200",
  "epoch": 6.525328330206379,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008157272208173587,
      "grad_norm": 0.0964740738272667,
      "learning_rate": 0.29999999259779675,
      "loss": 12.5785,
      "num_input_tokens_seen": 7920,
      "step": 5
    },
    {
      "epoch": 0.0016314544416347174,
      "grad_norm": 0.2672581374645233,
      "learning_rate": 0.29999996252634736,
      "loss": 10.8637,
      "num_input_tokens_seen": 16848,
      "step": 10
    },
    {
      "epoch": 0.002447181662452076,
      "grad_norm": 0.17059220373630524,
      "learning_rate": 0.2999999093230187,
      "loss": 9.4478,
      "num_input_tokens_seen": 26592,
      "step": 15
    },
    {
      "epoch": 0.0032629088832694347,
      "grad_norm": 0.13515207171440125,
      "learning_rate": 0.299999832987819,
      "loss": 8.3337,
      "num_input_tokens_seen": 36864,
      "step": 20
    },
    {
      "epoch": 0.004078636104086793,
      "grad_norm": 0.07681969553232193,
      "learning_rate": 0.29999973352076004,
      "loss": 6.5187,
      "num_input_tokens_seen": 45360,
      "step": 25
    },
    {
      "epoch": 0.004894363324904152,
      "grad_norm": 0.04445548728108406,
      "learning_rate": 0.2999996109218572,
      "loss": 5.4382,
      "num_input_tokens_seen": 54992,
      "step": 30
    },
    {
      "epoch": 0.005710090545721511,
      "grad_norm": 0.07895876467227936,
      "learning_rate": 0.2999994651911293,
      "loss": 4.5904,
      "num_input_tokens_seen": 64752,
      "step": 35
    },
    {
      "epoch": 0.0065258177665388694,
      "grad_norm": 0.03524782881140709,
      "learning_rate": 0.2999992963285989,
      "loss": 3.87,
      "num_input_tokens_seen": 74960,
      "step": 40
    },
    {
      "epoch": 0.007341544987356228,
      "grad_norm": 0.1545085906982422,
      "learning_rate": 0.29999910433429194,
      "loss": 2.9895,
      "num_input_tokens_seen": 84576,
      "step": 45
    },
    {
      "epoch": 0.008157272208173586,
      "grad_norm": 0.0675736591219902,
      "learning_rate": 0.29999888920823814,
      "loss": 0.8504,
      "num_input_tokens_seen": 94896,
      "step": 50
    },
    {
      "epoch": 0.008972999428990946,
      "grad_norm": 0.009622232988476753,
      "learning_rate": 0.29999865095047057,
      "loss": 0.4766,
      "num_input_tokens_seen": 104352,
      "step": 55
    },
    {
      "epoch": 0.009788726649808304,
      "grad_norm": 0.008583979681134224,
      "learning_rate": 0.29999838956102604,
      "loss": 0.5108,
      "num_input_tokens_seen": 113296,
      "step": 60
    },
    {
      "epoch": 0.010604453870625663,
      "grad_norm": 0.009859999641776085,
      "learning_rate": 0.29999810503994484,
      "loss": 0.4518,
      "num_input_tokens_seen": 122144,
      "step": 65
    },
    {
      "epoch": 0.011420181091443021,
      "grad_norm": 0.025364574044942856,
      "learning_rate": 0.29999779738727084,
      "loss": 0.3967,
      "num_input_tokens_seen": 131168,
      "step": 70
    },
    {
      "epoch": 0.012235908312260381,
      "grad_norm": 0.012271604500710964,
      "learning_rate": 0.29999746660305154,
      "loss": 0.3796,
      "num_input_tokens_seen": 140704,
      "step": 75
    },
    {
      "epoch": 0.013051635533077739,
      "grad_norm": 0.008448345586657524,
      "learning_rate": 0.2999971126873379,
      "loss": 0.3874,
      "num_input_tokens_seen": 149328,
      "step": 80
    },
    {
      "epoch": 0.013867362753895097,
      "grad_norm": 0.005632409825921059,
      "learning_rate": 0.2999967356401845,
      "loss": 0.4082,
      "num_input_tokens_seen": 158576,
      "step": 85
    },
    {
      "epoch": 0.014683089974712456,
      "grad_norm": 0.009609322994947433,
      "learning_rate": 0.29999633546164944,
      "loss": 0.5083,
      "num_input_tokens_seen": 168576,
      "step": 90
    },
    {
      "epoch": 0.015498817195529814,
      "grad_norm": 0.004775932990014553,
      "learning_rate": 0.29999591215179444,
      "loss": 0.4153,
      "num_input_tokens_seen": 177440,
      "step": 95
    },
    {
      "epoch": 0.016314544416347172,
      "grad_norm": 0.004568667151033878,
      "learning_rate": 0.2999954657106849,
      "loss": 0.3606,
      "num_input_tokens_seen": 187888,
      "step": 100
    },
    {
      "epoch": 0.017130271637164532,
      "grad_norm": 0.008726774714887142,
      "learning_rate": 0.2999949961383896,
      "loss": 0.3757,
      "num_input_tokens_seen": 197824,
      "step": 105
    },
    {
      "epoch": 0.01794599885798189,
      "grad_norm": 0.019331641495227814,
      "learning_rate": 0.2999945034349809,
      "loss": 0.3914,
      "num_input_tokens_seen": 208352,
      "step": 110
    },
    {
      "epoch": 0.01876172607879925,
      "grad_norm": 0.016507068648934364,
      "learning_rate": 0.2999939876005348,
      "loss": 0.394,
      "num_input_tokens_seen": 216624,
      "step": 115
    },
    {
      "epoch": 0.019577453299616607,
      "grad_norm": 0.005690138321369886,
      "learning_rate": 0.29999344863513094,
      "loss": 0.3939,
      "num_input_tokens_seen": 225616,
      "step": 120
    },
    {
      "epoch": 0.020393180520433967,
      "grad_norm": 0.007653899025171995,
      "learning_rate": 0.2999928865388523,
      "loss": 0.3594,
      "num_input_tokens_seen": 235312,
      "step": 125
    },
    {
      "epoch": 0.021208907741251327,
      "grad_norm": 0.008164326660335064,
      "learning_rate": 0.29999230131178567,
      "loss": 0.3435,
      "num_input_tokens_seen": 244688,
      "step": 130
    },
    {
      "epoch": 0.022024634962068683,
      "grad_norm": 0.0038212703075259924,
      "learning_rate": 0.2999916929540212,
      "loss": 0.3956,
      "num_input_tokens_seen": 254720,
      "step": 135
    },
    {
      "epoch": 0.022840362182886043,
      "grad_norm": 0.03265831992030144,
      "learning_rate": 0.29999106146565285,
      "loss": 0.5454,
      "num_input_tokens_seen": 263616,
      "step": 140
    },
    {
      "epoch": 0.023656089403703402,
      "grad_norm": 0.020685728639364243,
      "learning_rate": 0.29999040684677786,
      "loss": 0.5033,
      "num_input_tokens_seen": 273120,
      "step": 145
    },
    {
      "epoch": 0.024471816624520762,
      "grad_norm": 0.01128391083329916,
      "learning_rate": 0.2999897290974972,
      "loss": 0.4017,
      "num_input_tokens_seen": 282016,
      "step": 150
    },
    {
      "epoch": 0.025287543845338118,
      "grad_norm": 0.01761569082736969,
      "learning_rate": 0.2999890282179155,
      "loss": 0.4116,
      "num_input_tokens_seen": 292416,
      "step": 155
    },
    {
      "epoch": 0.026103271066155478,
      "grad_norm": 0.01905200444161892,
      "learning_rate": 0.29998830420814077,
      "loss": 0.4238,
      "num_input_tokens_seen": 302544,
      "step": 160
    },
    {
      "epoch": 0.026918998286972837,
      "grad_norm": 0.014092384837567806,
      "learning_rate": 0.2999875570682846,
      "loss": 0.4144,
      "num_input_tokens_seen": 312704,
      "step": 165
    },
    {
      "epoch": 0.027734725507790194,
      "grad_norm": 0.004906647372990847,
      "learning_rate": 0.2999867867984623,
      "loss": 0.3996,
      "num_input_tokens_seen": 322768,
      "step": 170
    },
    {
      "epoch": 0.028550452728607553,
      "grad_norm": 0.01971416361629963,
      "learning_rate": 0.29998599339879267,
      "loss": 0.4299,
      "num_input_tokens_seen": 331408,
      "step": 175
    },
    {
      "epoch": 0.029366179949424913,
      "grad_norm": 0.007234855554997921,
      "learning_rate": 0.29998517686939796,
      "loss": 0.3385,
      "num_input_tokens_seen": 340736,
      "step": 180
    },
    {
      "epoch": 0.030181907170242273,
      "grad_norm": 0.00811337772756815,
      "learning_rate": 0.29998433721040413,
      "loss": 0.3903,
      "num_input_tokens_seen": 350416,
      "step": 185
    },
    {
      "epoch": 0.03099763439105963,
      "grad_norm": 0.004179325886070728,
      "learning_rate": 0.29998347442194073,
      "loss": 0.3505,
      "num_input_tokens_seen": 359632,
      "step": 190
    },
    {
      "epoch": 0.031813361611876985,
      "grad_norm": 0.01716136373579502,
      "learning_rate": 0.2999825885041407,
      "loss": 0.3897,
      "num_input_tokens_seen": 369600,
      "step": 195
    },
    {
      "epoch": 0.032629088832694345,
      "grad_norm": 0.025716662406921387,
      "learning_rate": 0.29998167945714077,
      "loss": 0.4564,
      "num_input_tokens_seen": 378944,
      "step": 200
    },
    {
      "epoch": 0.032629088832694345,
      "eval_loss": 0.4853101074695587,
      "eval_runtime": 151.8613,
      "eval_samples_per_second": 17.944,
      "eval_steps_per_second": 8.975,
      "num_input_tokens_seen": 378944,
      "step": 200
    },
    {
      "epoch": 0.033444816053511704,
      "grad_norm": 0.012503623962402344,
      "learning_rate": 0.2999807472810811,
      "loss": 0.519,
      "num_input_tokens_seen": 388816,
      "step": 205
    },
    {
      "epoch": 0.034260543274329064,
      "grad_norm": 0.0040049380622804165,
      "learning_rate": 0.29997979197610536,
      "loss": 0.4014,
      "num_input_tokens_seen": 399392,
      "step": 210
    },
    {
      "epoch": 0.035076270495146424,
      "grad_norm": 0.005801488645374775,
      "learning_rate": 0.299978813542361,
      "loss": 0.3297,
      "num_input_tokens_seen": 408832,
      "step": 215
    },
    {
      "epoch": 0.03589199771596378,
      "grad_norm": 0.010006844997406006,
      "learning_rate": 0.2999778119799988,
      "loss": 0.6463,
      "num_input_tokens_seen": 419456,
      "step": 220
    },
    {
      "epoch": 0.03670772493678114,
      "grad_norm": 0.008373760618269444,
      "learning_rate": 0.29997678728917326,
      "loss": 0.4877,
      "num_input_tokens_seen": 429280,
      "step": 225
    },
    {
      "epoch": 0.0375234521575985,
      "grad_norm": 0.010749702341854572,
      "learning_rate": 0.2999757394700424,
      "loss": 0.3861,
      "num_input_tokens_seen": 438416,
      "step": 230
    },
    {
      "epoch": 0.038339179378415855,
      "grad_norm": 0.02445523999631405,
      "learning_rate": 0.29997466852276783,
      "loss": 0.7511,
      "num_input_tokens_seen": 447600,
      "step": 235
    },
    {
      "epoch": 0.039154906599233215,
      "grad_norm": 0.011703121475875378,
      "learning_rate": 0.29997357444751466,
      "loss": 0.4873,
      "num_input_tokens_seen": 457680,
      "step": 240
    },
    {
      "epoch": 0.039970633820050575,
      "grad_norm": 0.011629808694124222,
      "learning_rate": 0.2999724572444516,
      "loss": 0.4332,
      "num_input_tokens_seen": 466496,
      "step": 245
    },
    {
      "epoch": 0.040786361040867934,
      "grad_norm": 0.012991552241146564,
      "learning_rate": 0.29997131691375095,
      "loss": 0.4573,
      "num_input_tokens_seen": 476400,
      "step": 250
    },
    {
      "epoch": 0.041602088261685294,
      "grad_norm": 0.004157013725489378,
      "learning_rate": 0.2999701534555886,
      "loss": 0.3673,
      "num_input_tokens_seen": 484960,
      "step": 255
    },
    {
      "epoch": 0.042417815482502653,
      "grad_norm": 0.0120658278465271,
      "learning_rate": 0.2999689668701439,
      "loss": 0.4138,
      "num_input_tokens_seen": 493760,
      "step": 260
    },
    {
      "epoch": 0.04323354270332001,
      "grad_norm": 0.010738413780927658,
      "learning_rate": 0.29996775715759993,
      "loss": 0.4389,
      "num_input_tokens_seen": 504400,
      "step": 265
    },
    {
      "epoch": 0.044049269924137366,
      "grad_norm": 0.008300382643938065,
      "learning_rate": 0.2999665243181432,
      "loss": 0.4302,
      "num_input_tokens_seen": 513264,
      "step": 270
    },
    {
      "epoch": 0.044864997144954726,
      "grad_norm": 0.015253307297825813,
      "learning_rate": 0.2999652683519638,
      "loss": 0.512,
      "num_input_tokens_seen": 523216,
      "step": 275
    },
    {
      "epoch": 0.045680724365772085,
      "grad_norm": 0.0037195361219346523,
      "learning_rate": 0.29996398925925544,
      "loss": 0.3932,
      "num_input_tokens_seen": 532832,
      "step": 280
    },
    {
      "epoch": 0.046496451586589445,
      "grad_norm": 0.00521418871358037,
      "learning_rate": 0.2999626870402154,
      "loss": 0.349,
      "num_input_tokens_seen": 542272,
      "step": 285
    },
    {
      "epoch": 0.047312178807406804,
      "grad_norm": 0.006327154114842415,
      "learning_rate": 0.29996136169504445,
      "loss": 0.379,
      "num_input_tokens_seen": 551152,
      "step": 290
    },
    {
      "epoch": 0.048127906028224164,
      "grad_norm": 0.012042248621582985,
      "learning_rate": 0.29996001322394694,
      "loss": 0.3636,
      "num_input_tokens_seen": 560880,
      "step": 295
    },
    {
      "epoch": 0.048943633249041524,
      "grad_norm": 0.009656425565481186,
      "learning_rate": 0.29995864162713093,
      "loss": 0.3995,
      "num_input_tokens_seen": 570320,
      "step": 300
    },
    {
      "epoch": 0.049759360469858877,
      "grad_norm": 0.012226720340549946,
      "learning_rate": 0.2999572469048079,
      "loss": 0.4136,
      "num_input_tokens_seen": 581328,
      "step": 305
    },
    {
      "epoch": 0.050575087690676236,
      "grad_norm": 0.012298260815441608,
      "learning_rate": 0.29995582905719287,
      "loss": 0.4253,
      "num_input_tokens_seen": 590080,
      "step": 310
    },
    {
      "epoch": 0.051390814911493596,
      "grad_norm": 0.00600745202973485,
      "learning_rate": 0.2999543880845046,
      "loss": 0.3929,
      "num_input_tokens_seen": 600064,
      "step": 315
    },
    {
      "epoch": 0.052206542132310955,
      "grad_norm": 0.005030220840126276,
      "learning_rate": 0.2999529239869652,
      "loss": 0.3514,
      "num_input_tokens_seen": 609920,
      "step": 320
    },
    {
      "epoch": 0.053022269353128315,
      "grad_norm": 0.007103821728378534,
      "learning_rate": 0.2999514367648005,
      "loss": 0.397,
      "num_input_tokens_seen": 618960,
      "step": 325
    },
    {
      "epoch": 0.053837996573945675,
      "grad_norm": 0.013804675079882145,
      "learning_rate": 0.29994992641823987,
      "loss": 0.3429,
      "num_input_tokens_seen": 627680,
      "step": 330
    },
    {
      "epoch": 0.054653723794763034,
      "grad_norm": 0.0052307723090052605,
      "learning_rate": 0.29994839294751613,
      "loss": 0.4069,
      "num_input_tokens_seen": 637520,
      "step": 335
    },
    {
      "epoch": 0.05546945101558039,
      "grad_norm": 0.0027780886739492416,
      "learning_rate": 0.29994683635286584,
      "loss": 0.3528,
      "num_input_tokens_seen": 646928,
      "step": 340
    },
    {
      "epoch": 0.05628517823639775,
      "grad_norm": 0.00396367721259594,
      "learning_rate": 0.2999452566345291,
      "loss": 0.3853,
      "num_input_tokens_seen": 655344,
      "step": 345
    },
    {
      "epoch": 0.057100905457215106,
      "grad_norm": 0.01068114209920168,
      "learning_rate": 0.2999436537927494,
      "loss": 0.3335,
      "num_input_tokens_seen": 664368,
      "step": 350
    },
    {
      "epoch": 0.057916632678032466,
      "grad_norm": 0.003915432374924421,
      "learning_rate": 0.299942027827774,
      "loss": 0.3335,
      "num_input_tokens_seen": 675056,
      "step": 355
    },
    {
      "epoch": 0.058732359898849826,
      "grad_norm": 0.0053165671415627,
      "learning_rate": 0.29994037873985363,
      "loss": 0.4959,
      "num_input_tokens_seen": 683872,
      "step": 360
    },
    {
      "epoch": 0.059548087119667185,
      "grad_norm": 0.01039879024028778,
      "learning_rate": 0.29993870652924254,
      "loss": 0.7145,
      "num_input_tokens_seen": 694576,
      "step": 365
    },
    {
      "epoch": 0.060363814340484545,
      "grad_norm": 0.011419568210840225,
      "learning_rate": 0.29993701119619876,
      "loss": 0.5279,
      "num_input_tokens_seen": 704272,
      "step": 370
    },
    {
      "epoch": 0.0611795415613019,
      "grad_norm": 0.005438313819468021,
      "learning_rate": 0.2999352927409835,
      "loss": 0.3219,
      "num_input_tokens_seen": 713984,
      "step": 375
    },
    {
      "epoch": 0.06199526878211926,
      "grad_norm": 0.007106679491698742,
      "learning_rate": 0.29993355116386194,
      "loss": 0.3867,
      "num_input_tokens_seen": 723360,
      "step": 380
    },
    {
      "epoch": 0.06281099600293662,
      "grad_norm": 0.005728765390813351,
      "learning_rate": 0.29993178646510266,
      "loss": 0.3696,
      "num_input_tokens_seen": 732064,
      "step": 385
    },
    {
      "epoch": 0.06362672322375397,
      "grad_norm": 0.017348576337099075,
      "learning_rate": 0.2999299986449777,
      "loss": 0.4109,
      "num_input_tokens_seen": 740336,
      "step": 390
    },
    {
      "epoch": 0.06444245044457134,
      "grad_norm": 0.00967331975698471,
      "learning_rate": 0.29992818770376284,
      "loss": 0.4358,
      "num_input_tokens_seen": 748960,
      "step": 395
    },
    {
      "epoch": 0.06525817766538869,
      "grad_norm": 0.003294999012723565,
      "learning_rate": 0.29992635364173725,
      "loss": 0.3723,
      "num_input_tokens_seen": 758192,
      "step": 400
    },
    {
      "epoch": 0.06525817766538869,
      "eval_loss": 0.37038955092430115,
      "eval_runtime": 152.4516,
      "eval_samples_per_second": 17.875,
      "eval_steps_per_second": 8.941,
      "num_input_tokens_seen": 758192,
      "step": 400
    },
    {
      "epoch": 0.06607390488620606,
      "grad_norm": 0.00816957838833332,
      "learning_rate": 0.2999244964591839,
      "loss": 0.3828,
      "num_input_tokens_seen": 767184,
      "step": 405
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 0.0032555856741964817,
      "learning_rate": 0.2999226161563891,
      "loss": 0.3614,
      "num_input_tokens_seen": 776784,
      "step": 410
    },
    {
      "epoch": 0.06770535932784078,
      "grad_norm": 0.006234012544155121,
      "learning_rate": 0.2999207127336429,
      "loss": 0.3754,
      "num_input_tokens_seen": 786640,
      "step": 415
    },
    {
      "epoch": 0.06852108654865813,
      "grad_norm": 0.006208541803061962,
      "learning_rate": 0.2999187861912387,
      "loss": 0.3916,
      "num_input_tokens_seen": 796624,
      "step": 420
    },
    {
      "epoch": 0.06933681376947548,
      "grad_norm": 0.007273564580827951,
      "learning_rate": 0.2999168365294737,
      "loss": 0.4407,
      "num_input_tokens_seen": 805920,
      "step": 425
    },
    {
      "epoch": 0.07015254099029285,
      "grad_norm": 0.003539619268849492,
      "learning_rate": 0.29991486374864856,
      "loss": 0.3302,
      "num_input_tokens_seen": 815664,
      "step": 430
    },
    {
      "epoch": 0.0709682682111102,
      "grad_norm": 0.004886243492364883,
      "learning_rate": 0.29991286784906745,
      "loss": 0.3709,
      "num_input_tokens_seen": 825328,
      "step": 435
    },
    {
      "epoch": 0.07178399543192757,
      "grad_norm": 0.005425168666988611,
      "learning_rate": 0.2999108488310382,
      "loss": 0.3675,
      "num_input_tokens_seen": 834336,
      "step": 440
    },
    {
      "epoch": 0.07259972265274492,
      "grad_norm": 0.016032058745622635,
      "learning_rate": 0.29990880669487213,
      "loss": 0.3664,
      "num_input_tokens_seen": 843584,
      "step": 445
    },
    {
      "epoch": 0.07341544987356229,
      "grad_norm": 0.0021686165127903223,
      "learning_rate": 0.29990674144088425,
      "loss": 0.3502,
      "num_input_tokens_seen": 853040,
      "step": 450
    },
    {
      "epoch": 0.07423117709437964,
      "grad_norm": 0.006131706293672323,
      "learning_rate": 0.299904653069393,
      "loss": 0.408,
      "num_input_tokens_seen": 862992,
      "step": 455
    },
    {
      "epoch": 0.075046904315197,
      "grad_norm": 0.010553373955190182,
      "learning_rate": 0.29990254158072044,
      "loss": 0.3839,
      "num_input_tokens_seen": 872048,
      "step": 460
    },
    {
      "epoch": 0.07586263153601436,
      "grad_norm": 0.0022507242392748594,
      "learning_rate": 0.2999004069751921,
      "loss": 0.3622,
      "num_input_tokens_seen": 882272,
      "step": 465
    },
    {
      "epoch": 0.07667835875683171,
      "grad_norm": 0.005361245013773441,
      "learning_rate": 0.2998982492531373,
      "loss": 0.3657,
      "num_input_tokens_seen": 892688,
      "step": 470
    },
    {
      "epoch": 0.07749408597764908,
      "grad_norm": 0.005466254893690348,
      "learning_rate": 0.2998960684148887,
      "loss": 0.3359,
      "num_input_tokens_seen": 902144,
      "step": 475
    },
    {
      "epoch": 0.07830981319846643,
      "grad_norm": 0.002678121905773878,
      "learning_rate": 0.29989386446078264,
      "loss": 0.3368,
      "num_input_tokens_seen": 911024,
      "step": 480
    },
    {
      "epoch": 0.0791255404192838,
      "grad_norm": 0.004106923472136259,
      "learning_rate": 0.299891637391159,
      "loss": 0.3886,
      "num_input_tokens_seen": 921408,
      "step": 485
    },
    {
      "epoch": 0.07994126764010115,
      "grad_norm": 0.006402906496077776,
      "learning_rate": 0.2998893872063612,
      "loss": 0.471,
      "num_input_tokens_seen": 931440,
      "step": 490
    },
    {
      "epoch": 0.08075699486091852,
      "grad_norm": 0.00266858609393239,
      "learning_rate": 0.2998871139067363,
      "loss": 0.393,
      "num_input_tokens_seen": 939280,
      "step": 495
    },
    {
      "epoch": 0.08157272208173587,
      "grad_norm": 0.0058729625307023525,
      "learning_rate": 0.2998848174926348,
      "loss": 0.3668,
      "num_input_tokens_seen": 949360,
      "step": 500
    },
    {
      "epoch": 0.08238844930255322,
      "grad_norm": 0.0034706909209489822,
      "learning_rate": 0.2998824979644109,
      "loss": 0.3538,
      "num_input_tokens_seen": 958672,
      "step": 505
    },
    {
      "epoch": 0.08320417652337059,
      "grad_norm": 0.006641618441790342,
      "learning_rate": 0.29988015532242224,
      "loss": 0.3562,
      "num_input_tokens_seen": 967456,
      "step": 510
    },
    {
      "epoch": 0.08401990374418794,
      "grad_norm": 0.005162003915756941,
      "learning_rate": 0.29987778956703015,
      "loss": 0.3669,
      "num_input_tokens_seen": 977360,
      "step": 515
    },
    {
      "epoch": 0.08483563096500531,
      "grad_norm": 0.002799984999001026,
      "learning_rate": 0.2998754006985994,
      "loss": 0.4737,
      "num_input_tokens_seen": 986368,
      "step": 520
    },
    {
      "epoch": 0.08565135818582266,
      "grad_norm": 0.016831301152706146,
      "learning_rate": 0.29987298871749846,
      "loss": 0.409,
      "num_input_tokens_seen": 995776,
      "step": 525
    },
    {
      "epoch": 0.08646708540664003,
      "grad_norm": 0.010214020498096943,
      "learning_rate": 0.2998705536240992,
      "loss": 0.4357,
      "num_input_tokens_seen": 1005280,
      "step": 530
    },
    {
      "epoch": 0.08728281262745738,
      "grad_norm": 0.0039898911491036415,
      "learning_rate": 0.2998680954187772,
      "loss": 0.4823,
      "num_input_tokens_seen": 1015600,
      "step": 535
    },
    {
      "epoch": 0.08809853984827473,
      "grad_norm": 0.0034917898010462523,
      "learning_rate": 0.2998656141019115,
      "loss": 0.4158,
      "num_input_tokens_seen": 1024912,
      "step": 540
    },
    {
      "epoch": 0.0889142670690921,
      "grad_norm": 0.004948853049427271,
      "learning_rate": 0.2998631096738848,
      "loss": 0.4457,
      "num_input_tokens_seen": 1034848,
      "step": 545
    },
    {
      "epoch": 0.08972999428990945,
      "grad_norm": 0.004330082330852747,
      "learning_rate": 0.29986058213508326,
      "loss": 0.4001,
      "num_input_tokens_seen": 1045648,
      "step": 550
    },
    {
      "epoch": 0.09054572151072682,
      "grad_norm": 0.010022643953561783,
      "learning_rate": 0.29985803148589674,
      "loss": 0.3361,
      "num_input_tokens_seen": 1056416,
      "step": 555
    },
    {
      "epoch": 0.09136144873154417,
      "grad_norm": 0.004089437425136566,
      "learning_rate": 0.2998554577267185,
      "loss": 0.3956,
      "num_input_tokens_seen": 1066048,
      "step": 560
    },
    {
      "epoch": 0.09217717595236154,
      "grad_norm": 0.007980486378073692,
      "learning_rate": 0.2998528608579455,
      "loss": 0.3405,
      "num_input_tokens_seen": 1075072,
      "step": 565
    },
    {
      "epoch": 0.09299290317317889,
      "grad_norm": 0.008167603053152561,
      "learning_rate": 0.2998502408799781,
      "loss": 0.3707,
      "num_input_tokens_seen": 1083968,
      "step": 570
    },
    {
      "epoch": 0.09380863039399624,
      "grad_norm": 0.017487971112132072,
      "learning_rate": 0.2998475977932205,
      "loss": 0.4031,
      "num_input_tokens_seen": 1093728,
      "step": 575
    },
    {
      "epoch": 0.09462435761481361,
      "grad_norm": 0.006021643057465553,
      "learning_rate": 0.29984493159808023,
      "loss": 0.4428,
      "num_input_tokens_seen": 1103280,
      "step": 580
    },
    {
      "epoch": 0.09544008483563096,
      "grad_norm": 0.0030144548509269953,
      "learning_rate": 0.29984224229496836,
      "loss": 0.3596,
      "num_input_tokens_seen": 1111744,
      "step": 585
    },
    {
      "epoch": 0.09625581205644833,
      "grad_norm": 0.0024890827480703592,
      "learning_rate": 0.2998395298842998,
      "loss": 0.3581,
      "num_input_tokens_seen": 1122080,
      "step": 590
    },
    {
      "epoch": 0.09707153927726568,
      "grad_norm": 0.004157871939241886,
      "learning_rate": 0.29983679436649263,
      "loss": 0.3805,
      "num_input_tokens_seen": 1132096,
      "step": 595
    },
    {
      "epoch": 0.09788726649808305,
      "grad_norm": 0.006914137862622738,
      "learning_rate": 0.2998340357419689,
      "loss": 0.3559,
      "num_input_tokens_seen": 1141408,
      "step": 600
    },
    {
      "epoch": 0.09788726649808305,
      "eval_loss": 0.37731045484542847,
      "eval_runtime": 152.2159,
      "eval_samples_per_second": 17.902,
      "eval_steps_per_second": 8.954,
      "num_input_tokens_seen": 1141408,
      "step": 600
    },
    {
      "epoch": 0.0987029937189004,
      "grad_norm": 0.005334604997187853,
      "learning_rate": 0.29983125401115385,
      "loss": 0.4028,
      "num_input_tokens_seen": 1150912,
      "step": 605
    },
    {
      "epoch": 0.09951872093971775,
      "grad_norm": 0.004185666795819998,
      "learning_rate": 0.29982844917447654,
      "loss": 0.3435,
      "num_input_tokens_seen": 1160672,
      "step": 610
    },
    {
      "epoch": 0.10033444816053512,
      "grad_norm": 0.00487235514447093,
      "learning_rate": 0.2998256212323695,
      "loss": 0.4057,
      "num_input_tokens_seen": 1170672,
      "step": 615
    },
    {
      "epoch": 0.10115017538135247,
      "grad_norm": 0.002710686530917883,
      "learning_rate": 0.29982277018526887,
      "loss": 0.3809,
      "num_input_tokens_seen": 1179792,
      "step": 620
    },
    {
      "epoch": 0.10196590260216984,
      "grad_norm": 0.009217911399900913,
      "learning_rate": 0.2998198960336143,
      "loss": 0.4043,
      "num_input_tokens_seen": 1189120,
      "step": 625
    },
    {
      "epoch": 0.10278162982298719,
      "grad_norm": 0.0071515836752951145,
      "learning_rate": 0.299816998777849,
      "loss": 0.3442,
      "num_input_tokens_seen": 1198272,
      "step": 630
    },
    {
      "epoch": 0.10359735704380456,
      "grad_norm": 0.0024489748757332563,
      "learning_rate": 0.2998140784184197,
      "loss": 0.3962,
      "num_input_tokens_seen": 1207216,
      "step": 635
    },
    {
      "epoch": 0.10441308426462191,
      "grad_norm": 0.01095146406441927,
      "learning_rate": 0.2998111349557769,
      "loss": 0.3533,
      "num_input_tokens_seen": 1215216,
      "step": 640
    },
    {
      "epoch": 0.10522881148543926,
      "grad_norm": 0.009696271270513535,
      "learning_rate": 0.29980816839037444,
      "loss": 0.4229,
      "num_input_tokens_seen": 1224672,
      "step": 645
    },
    {
      "epoch": 0.10604453870625663,
      "grad_norm": 0.00854188110679388,
      "learning_rate": 0.2998051787226698,
      "loss": 0.3355,
      "num_input_tokens_seen": 1234192,
      "step": 650
    },
    {
      "epoch": 0.10686026592707398,
      "grad_norm": 0.007341011893004179,
      "learning_rate": 0.29980216595312403,
      "loss": 0.3709,
      "num_input_tokens_seen": 1244192,
      "step": 655
    },
    {
      "epoch": 0.10767599314789135,
      "grad_norm": 0.007101545576006174,
      "learning_rate": 0.29979913008220177,
      "loss": 0.3637,
      "num_input_tokens_seen": 1253024,
      "step": 660
    },
    {
      "epoch": 0.1084917203687087,
      "grad_norm": 0.010782870464026928,
      "learning_rate": 0.2997960711103711,
      "loss": 0.3369,
      "num_input_tokens_seen": 1261120,
      "step": 665
    },
    {
      "epoch": 0.10930744758952607,
      "grad_norm": 0.0036527460906654596,
      "learning_rate": 0.29979298903810386,
      "loss": 0.4154,
      "num_input_tokens_seen": 1271168,
      "step": 670
    },
    {
      "epoch": 0.11012317481034342,
      "grad_norm": 0.005170115735381842,
      "learning_rate": 0.29978988386587524,
      "loss": 0.3457,
      "num_input_tokens_seen": 1280848,
      "step": 675
    },
    {
      "epoch": 0.11093890203116077,
      "grad_norm": 0.008761486038565636,
      "learning_rate": 0.2997867555941642,
      "loss": 0.4133,
      "num_input_tokens_seen": 1290784,
      "step": 680
    },
    {
      "epoch": 0.11175462925197814,
      "grad_norm": 0.0024575255811214447,
      "learning_rate": 0.299783604223453,
      "loss": 0.3437,
      "num_input_tokens_seen": 1299568,
      "step": 685
    },
    {
      "epoch": 0.1125703564727955,
      "grad_norm": 0.006418189965188503,
      "learning_rate": 0.29978042975422786,
      "loss": 0.351,
      "num_input_tokens_seen": 1308960,
      "step": 690
    },
    {
      "epoch": 0.11338608369361286,
      "grad_norm": 0.009774851612746716,
      "learning_rate": 0.29977723218697816,
      "loss": 0.3525,
      "num_input_tokens_seen": 1319184,
      "step": 695
    },
    {
      "epoch": 0.11420181091443021,
      "grad_norm": 0.0048735118471086025,
      "learning_rate": 0.299774011522197,
      "loss": 0.3722,
      "num_input_tokens_seen": 1328304,
      "step": 700
    },
    {
      "epoch": 0.11501753813524758,
      "grad_norm": 0.007406812161207199,
      "learning_rate": 0.29977076776038114,
      "loss": 0.4032,
      "num_input_tokens_seen": 1337648,
      "step": 705
    },
    {
      "epoch": 0.11583326535606493,
      "grad_norm": 0.003995394334197044,
      "learning_rate": 0.2997675009020307,
      "loss": 0.3562,
      "num_input_tokens_seen": 1348384,
      "step": 710
    },
    {
      "epoch": 0.11664899257688228,
      "grad_norm": 0.006450183689594269,
      "learning_rate": 0.2997642109476496,
      "loss": 0.4174,
      "num_input_tokens_seen": 1358272,
      "step": 715
    },
    {
      "epoch": 0.11746471979769965,
      "grad_norm": 0.0074956780299544334,
      "learning_rate": 0.299760897897745,
      "loss": 0.3067,
      "num_input_tokens_seen": 1366864,
      "step": 720
    },
    {
      "epoch": 0.118280447018517,
      "grad_norm": 0.002177286194637418,
      "learning_rate": 0.29975756175282803,
      "loss": 0.3052,
      "num_input_tokens_seen": 1376384,
      "step": 725
    },
    {
      "epoch": 0.11909617423933437,
      "grad_norm": 0.005742892622947693,
      "learning_rate": 0.29975420251341306,
      "loss": 0.5991,
      "num_input_tokens_seen": 1385088,
      "step": 730
    },
    {
      "epoch": 0.11991190146015172,
      "grad_norm": 0.0040392279624938965,
      "learning_rate": 0.29975082018001814,
      "loss": 0.5776,
      "num_input_tokens_seen": 1394816,
      "step": 735
    },
    {
      "epoch": 0.12072762868096909,
      "grad_norm": 0.0032477870117872953,
      "learning_rate": 0.2997474147531648,
      "loss": 0.4054,
      "num_input_tokens_seen": 1404912,
      "step": 740
    },
    {
      "epoch": 0.12154335590178644,
      "grad_norm": 0.008668714202940464,
      "learning_rate": 0.29974398623337833,
      "loss": 0.4602,
      "num_input_tokens_seen": 1415376,
      "step": 745
    },
    {
      "epoch": 0.1223590831226038,
      "grad_norm": 0.008527824655175209,
      "learning_rate": 0.2997405346211873,
      "loss": 0.4503,
      "num_input_tokens_seen": 1425456,
      "step": 750
    },
    {
      "epoch": 0.12317481034342116,
      "grad_norm": 0.006697394419461489,
      "learning_rate": 0.2997370599171241,
      "loss": 0.3863,
      "num_input_tokens_seen": 1435344,
      "step": 755
    },
    {
      "epoch": 0.12399053756423851,
      "grad_norm": 0.00727938674390316,
      "learning_rate": 0.2997335621217246,
      "loss": 0.3809,
      "num_input_tokens_seen": 1445584,
      "step": 760
    },
    {
      "epoch": 0.12480626478505588,
      "grad_norm": 0.0026301878970116377,
      "learning_rate": 0.29973004123552816,
      "loss": 0.3625,
      "num_input_tokens_seen": 1455056,
      "step": 765
    },
    {
      "epoch": 0.12562199200587323,
      "grad_norm": 0.007823913358151913,
      "learning_rate": 0.2997264972590777,
      "loss": 0.4856,
      "num_input_tokens_seen": 1464448,
      "step": 770
    },
    {
      "epoch": 0.1264377192266906,
      "grad_norm": 0.0031371598597615957,
      "learning_rate": 0.29972293019291973,
      "loss": 0.4568,
      "num_input_tokens_seen": 1472672,
      "step": 775
    },
    {
      "epoch": 0.12725344644750794,
      "grad_norm": 0.0035831257700920105,
      "learning_rate": 0.2997193400376045,
      "loss": 0.3422,
      "num_input_tokens_seen": 1481712,
      "step": 780
    },
    {
      "epoch": 0.12806917366832532,
      "grad_norm": 0.009325150400400162,
      "learning_rate": 0.2997157267936854,
      "loss": 0.4854,
      "num_input_tokens_seen": 1490704,
      "step": 785
    },
    {
      "epoch": 0.12888490088914267,
      "grad_norm": 0.008324564434587955,
      "learning_rate": 0.2997120904617199,
      "loss": 0.4312,
      "num_input_tokens_seen": 1500688,
      "step": 790
    },
    {
      "epoch": 0.12970062810996003,
      "grad_norm": 0.007723478600382805,
      "learning_rate": 0.29970843104226863,
      "loss": 0.4299,
      "num_input_tokens_seen": 1508960,
      "step": 795
    },
    {
      "epoch": 0.13051635533077738,
      "grad_norm": 0.01428726315498352,
      "learning_rate": 0.2997047485358959,
      "loss": 0.3491,
      "num_input_tokens_seen": 1518336,
      "step": 800
    },
    {
      "epoch": 0.13051635533077738,
      "eval_loss": 0.3604404628276825,
      "eval_runtime": 152.2238,
      "eval_samples_per_second": 17.901,
      "eval_steps_per_second": 8.954,
      "num_input_tokens_seen": 1518336,
      "step": 800
    },
    {
      "epoch": 0.13133208255159476,
      "grad_norm": 0.005888365674763918,
      "learning_rate": 0.2997010429431697,
      "loss": 0.3657,
      "num_input_tokens_seen": 1527888,
      "step": 805
    },
    {
      "epoch": 0.1321478097724121,
      "grad_norm": 0.01580217480659485,
      "learning_rate": 0.29969731426466134,
      "loss": 0.3821,
      "num_input_tokens_seen": 1536976,
      "step": 810
    },
    {
      "epoch": 0.13296353699322946,
      "grad_norm": 0.017772966995835304,
      "learning_rate": 0.299693562500946,
      "loss": 0.4028,
      "num_input_tokens_seen": 1547152,
      "step": 815
    },
    {
      "epoch": 0.13377926421404682,
      "grad_norm": 0.005914117209613323,
      "learning_rate": 0.29968978765260207,
      "loss": 0.4635,
      "num_input_tokens_seen": 1556704,
      "step": 820
    },
    {
      "epoch": 0.13459499143486417,
      "grad_norm": 0.011772295460104942,
      "learning_rate": 0.2996859897202118,
      "loss": 0.4046,
      "num_input_tokens_seen": 1565360,
      "step": 825
    },
    {
      "epoch": 0.13541071865568155,
      "grad_norm": 0.00877909455448389,
      "learning_rate": 0.2996821687043609,
      "loss": 0.3772,
      "num_input_tokens_seen": 1574640,
      "step": 830
    },
    {
      "epoch": 0.1362264458764989,
      "grad_norm": 0.006152914837002754,
      "learning_rate": 0.2996783246056384,
      "loss": 0.3768,
      "num_input_tokens_seen": 1583936,
      "step": 835
    },
    {
      "epoch": 0.13704217309731626,
      "grad_norm": 0.00191108463332057,
      "learning_rate": 0.29967445742463744,
      "loss": 0.3186,
      "num_input_tokens_seen": 1592992,
      "step": 840
    },
    {
      "epoch": 0.1378579003181336,
      "grad_norm": 0.0013458931352943182,
      "learning_rate": 0.29967056716195417,
      "loss": 0.3224,
      "num_input_tokens_seen": 1602160,
      "step": 845
    },
    {
      "epoch": 0.13867362753895096,
      "grad_norm": 0.0026225245092064142,
      "learning_rate": 0.2996666538181885,
      "loss": 0.4273,
      "num_input_tokens_seen": 1613280,
      "step": 850
    },
    {
      "epoch": 0.13948935475976834,
      "grad_norm": 0.002339197089895606,
      "learning_rate": 0.29966271739394407,
      "loss": 0.4423,
      "num_input_tokens_seen": 1623360,
      "step": 855
    },
    {
      "epoch": 0.1403050819805857,
      "grad_norm": 0.005090198013931513,
      "learning_rate": 0.29965875788982776,
      "loss": 0.3837,
      "num_input_tokens_seen": 1632960,
      "step": 860
    },
    {
      "epoch": 0.14112080920140305,
      "grad_norm": 0.010216495022177696,
      "learning_rate": 0.2996547753064503,
      "loss": 0.3398,
      "num_input_tokens_seen": 1641936,
      "step": 865
    },
    {
      "epoch": 0.1419365364222204,
      "grad_norm": 0.004175249021500349,
      "learning_rate": 0.29965076964442583,
      "loss": 0.4102,
      "num_input_tokens_seen": 1652896,
      "step": 870
    },
    {
      "epoch": 0.14275226364303778,
      "grad_norm": 0.013822285458445549,
      "learning_rate": 0.299646740904372,
      "loss": 0.4452,
      "num_input_tokens_seen": 1664416,
      "step": 875
    },
    {
      "epoch": 0.14356799086385513,
      "grad_norm": 0.004669289104640484,
      "learning_rate": 0.29964268908691016,
      "loss": 0.3868,
      "num_input_tokens_seen": 1673680,
      "step": 880
    },
    {
      "epoch": 0.14438371808467249,
      "grad_norm": 0.01090817991644144,
      "learning_rate": 0.29963861419266513,
      "loss": 0.3636,
      "num_input_tokens_seen": 1682256,
      "step": 885
    },
    {
      "epoch": 0.14519944530548984,
      "grad_norm": 0.0068641966208815575,
      "learning_rate": 0.29963451622226533,
      "loss": 0.3634,
      "num_input_tokens_seen": 1691264,
      "step": 890
    },
    {
      "epoch": 0.1460151725263072,
      "grad_norm": 0.005702511873096228,
      "learning_rate": 0.29963039517634277,
      "loss": 0.5298,
      "num_input_tokens_seen": 1699872,
      "step": 895
    },
    {
      "epoch": 0.14683089974712457,
      "grad_norm": 0.003164441091939807,
      "learning_rate": 0.2996262510555328,
      "loss": 0.3822,
      "num_input_tokens_seen": 1709008,
      "step": 900
    },
    {
      "epoch": 0.14764662696794192,
      "grad_norm": 0.009363356046378613,
      "learning_rate": 0.2996220838604746,
      "loss": 0.4623,
      "num_input_tokens_seen": 1719216,
      "step": 905
    },
    {
      "epoch": 0.14846235418875928,
      "grad_norm": 0.0028236659709364176,
      "learning_rate": 0.29961789359181085,
      "loss": 0.3386,
      "num_input_tokens_seen": 1729984,
      "step": 910
    },
    {
      "epoch": 0.14927808140957663,
      "grad_norm": 0.008778288960456848,
      "learning_rate": 0.29961368025018764,
      "loss": 0.4392,
      "num_input_tokens_seen": 1739424,
      "step": 915
    },
    {
      "epoch": 0.150093808630394,
      "grad_norm": 0.005201797001063824,
      "learning_rate": 0.2996094438362548,
      "loss": 0.3346,
      "num_input_tokens_seen": 1748496,
      "step": 920
    },
    {
      "epoch": 0.15090953585121136,
      "grad_norm": 0.00862154271453619,
      "learning_rate": 0.2996051843506657,
      "loss": 0.4758,
      "num_input_tokens_seen": 1758336,
      "step": 925
    },
    {
      "epoch": 0.15172526307202872,
      "grad_norm": 0.007905401289463043,
      "learning_rate": 0.299600901794077,
      "loss": 0.433,
      "num_input_tokens_seen": 1768208,
      "step": 930
    },
    {
      "epoch": 0.15254099029284607,
      "grad_norm": 0.00849554780870676,
      "learning_rate": 0.29959659616714923,
      "loss": 0.409,
      "num_input_tokens_seen": 1778672,
      "step": 935
    },
    {
      "epoch": 0.15335671751366342,
      "grad_norm": 0.009579714387655258,
      "learning_rate": 0.2995922674705464,
      "loss": 0.3739,
      "num_input_tokens_seen": 1788016,
      "step": 940
    },
    {
      "epoch": 0.1541724447344808,
      "grad_norm": 0.0050884741358459,
      "learning_rate": 0.2995879157049361,
      "loss": 0.3873,
      "num_input_tokens_seen": 1797072,
      "step": 945
    },
    {
      "epoch": 0.15498817195529815,
      "grad_norm": 0.005865721497684717,
      "learning_rate": 0.2995835408709893,
      "loss": 0.4369,
      "num_input_tokens_seen": 1807296,
      "step": 950
    },
    {
      "epoch": 0.1558038991761155,
      "grad_norm": 0.0039458805695176125,
      "learning_rate": 0.29957914296938076,
      "loss": 0.4532,
      "num_input_tokens_seen": 1816160,
      "step": 955
    },
    {
      "epoch": 0.15661962639693286,
      "grad_norm": 0.01110230851918459,
      "learning_rate": 0.2995747220007886,
      "loss": 0.4277,
      "num_input_tokens_seen": 1827152,
      "step": 960
    },
    {
      "epoch": 0.1574353536177502,
      "grad_norm": 0.0028490694239735603,
      "learning_rate": 0.2995702779658947,
      "loss": 0.3904,
      "num_input_tokens_seen": 1835968,
      "step": 965
    },
    {
      "epoch": 0.1582510808385676,
      "grad_norm": 0.0015328226145356894,
      "learning_rate": 0.29956581086538425,
      "loss": 0.3843,
      "num_input_tokens_seen": 1844800,
      "step": 970
    },
    {
      "epoch": 0.15906680805938495,
      "grad_norm": 0.00885944627225399,
      "learning_rate": 0.2995613206999462,
      "loss": 0.3802,
      "num_input_tokens_seen": 1854256,
      "step": 975
    },
    {
      "epoch": 0.1598825352802023,
      "grad_norm": 0.002515081549063325,
      "learning_rate": 0.29955680747027297,
      "loss": 0.3126,
      "num_input_tokens_seen": 1863904,
      "step": 980
    },
    {
      "epoch": 0.16069826250101965,
      "grad_norm": 0.004624353721737862,
      "learning_rate": 0.2995522711770607,
      "loss": 0.3848,
      "num_input_tokens_seen": 1873168,
      "step": 985
    },
    {
      "epoch": 0.16151398972183703,
      "grad_norm": 0.0041796560399234295,
      "learning_rate": 0.2995477118210087,
      "loss": 0.4115,
      "num_input_tokens_seen": 1883328,
      "step": 990
    },
    {
      "epoch": 0.16232971694265438,
      "grad_norm": 0.007343667559325695,
      "learning_rate": 0.29954312940282024,
      "loss": 0.3478,
      "num_input_tokens_seen": 1892528,
      "step": 995
    },
    {
      "epoch": 0.16314544416347174,
      "grad_norm": 0.016488511115312576,
      "learning_rate": 0.29953852392320196,
      "loss": 0.3909,
      "num_input_tokens_seen": 1901264,
      "step": 1000
    },
    {
      "epoch": 0.16314544416347174,
      "eval_loss": 0.36068928241729736,
      "eval_runtime": 152.2388,
      "eval_samples_per_second": 17.9,
      "eval_steps_per_second": 8.953,
      "num_input_tokens_seen": 1901264,
      "step": 1000
    },
    {
      "epoch": 0.1639611713842891,
      "grad_norm": 0.00794228259474039,
      "learning_rate": 0.2995338953828641,
      "loss": 0.4056,
      "num_input_tokens_seen": 1911680,
      "step": 1005
    },
    {
      "epoch": 0.16477689860510644,
      "grad_norm": 0.012136194854974747,
      "learning_rate": 0.2995292437825204,
      "loss": 0.4305,
      "num_input_tokens_seen": 1919872,
      "step": 1010
    },
    {
      "epoch": 0.16559262582592382,
      "grad_norm": 0.0019780087750405073,
      "learning_rate": 0.29952456912288816,
      "loss": 0.34,
      "num_input_tokens_seen": 1930480,
      "step": 1015
    },
    {
      "epoch": 0.16640835304674118,
      "grad_norm": 0.005421936511993408,
      "learning_rate": 0.2995198714046884,
      "loss": 0.3885,
      "num_input_tokens_seen": 1940048,
      "step": 1020
    },
    {
      "epoch": 0.16722408026755853,
      "grad_norm": 0.0012693708995357156,
      "learning_rate": 0.2995151506286454,
      "loss": 0.3717,
      "num_input_tokens_seen": 1949808,
      "step": 1025
    },
    {
      "epoch": 0.16803980748837588,
      "grad_norm": 0.006867701653391123,
      "learning_rate": 0.2995104067954873,
      "loss": 0.3799,
      "num_input_tokens_seen": 1959008,
      "step": 1030
    },
    {
      "epoch": 0.16885553470919323,
      "grad_norm": 0.0023599781561642885,
      "learning_rate": 0.2995056399059456,
      "loss": 0.3477,
      "num_input_tokens_seen": 1968384,
      "step": 1035
    },
    {
      "epoch": 0.16967126193001061,
      "grad_norm": 0.00698376540094614,
      "learning_rate": 0.2995008499607554,
      "loss": 0.3949,
      "num_input_tokens_seen": 1977904,
      "step": 1040
    },
    {
      "epoch": 0.17048698915082797,
      "grad_norm": 0.004345533438026905,
      "learning_rate": 0.2994960369606554,
      "loss": 0.348,
      "num_input_tokens_seen": 1987648,
      "step": 1045
    },
    {
      "epoch": 0.17130271637164532,
      "grad_norm": 0.0022358871065080166,
      "learning_rate": 0.2994912009063878,
      "loss": 0.3503,
      "num_input_tokens_seen": 1997568,
      "step": 1050
    },
    {
      "epoch": 0.17211844359246267,
      "grad_norm": 0.008815680630505085,
      "learning_rate": 0.29948634179869843,
      "loss": 0.3869,
      "num_input_tokens_seen": 2005408,
      "step": 1055
    },
    {
      "epoch": 0.17293417081328005,
      "grad_norm": 0.0023909390438348055,
      "learning_rate": 0.29948145963833656,
      "loss": 0.3422,
      "num_input_tokens_seen": 2015232,
      "step": 1060
    },
    {
      "epoch": 0.1737498980340974,
      "grad_norm": 0.0059565152041614056,
      "learning_rate": 0.29947655442605514,
      "loss": 0.3444,
      "num_input_tokens_seen": 2024800,
      "step": 1065
    },
    {
      "epoch": 0.17456562525491476,
      "grad_norm": 0.003388647921383381,
      "learning_rate": 0.2994716261626106,
      "loss": 0.3454,
      "num_input_tokens_seen": 2033712,
      "step": 1070
    },
    {
      "epoch": 0.1753813524757321,
      "grad_norm": 0.012863386422395706,
      "learning_rate": 0.2994666748487629,
      "loss": 0.3929,
      "num_input_tokens_seen": 2043264,
      "step": 1075
    },
    {
      "epoch": 0.17619707969654946,
      "grad_norm": 0.00301080453209579,
      "learning_rate": 0.2994617004852756,
      "loss": 0.374,
      "num_input_tokens_seen": 2052720,
      "step": 1080
    },
    {
      "epoch": 0.17701280691736684,
      "grad_norm": 0.007384104188531637,
      "learning_rate": 0.2994567030729159,
      "loss": 0.495,
      "num_input_tokens_seen": 2062064,
      "step": 1085
    },
    {
      "epoch": 0.1778285341381842,
      "grad_norm": 0.0015171031700447202,
      "learning_rate": 0.29945168261245436,
      "loss": 0.3811,
      "num_input_tokens_seen": 2071504,
      "step": 1090
    },
    {
      "epoch": 0.17864426135900155,
      "grad_norm": 0.01235247403383255,
      "learning_rate": 0.29944663910466524,
      "loss": 0.4241,
      "num_input_tokens_seen": 2083440,
      "step": 1095
    },
    {
      "epoch": 0.1794599885798189,
      "grad_norm": 0.005002574995160103,
      "learning_rate": 0.2994415725503263,
      "loss": 0.3932,
      "num_input_tokens_seen": 2092912,
      "step": 1100
    },
    {
      "epoch": 0.18027571580063625,
      "grad_norm": 0.003510650945827365,
      "learning_rate": 0.29943648295021885,
      "loss": 0.3917,
      "num_input_tokens_seen": 2102160,
      "step": 1105
    },
    {
      "epoch": 0.18109144302145364,
      "grad_norm": 0.0042625307105481625,
      "learning_rate": 0.2994313703051278,
      "loss": 0.3241,
      "num_input_tokens_seen": 2111808,
      "step": 1110
    },
    {
      "epoch": 0.181907170242271,
      "grad_norm": 0.001441487343981862,
      "learning_rate": 0.29942623461584156,
      "loss": 0.4398,
      "num_input_tokens_seen": 2122704,
      "step": 1115
    },
    {
      "epoch": 0.18272289746308834,
      "grad_norm": 0.007405006792396307,
      "learning_rate": 0.29942107588315214,
      "loss": 0.3745,
      "num_input_tokens_seen": 2132128,
      "step": 1120
    },
    {
      "epoch": 0.1835386246839057,
      "grad_norm": 0.004630859941244125,
      "learning_rate": 0.29941589410785513,
      "loss": 0.3877,
      "num_input_tokens_seen": 2141584,
      "step": 1125
    },
    {
      "epoch": 0.18435435190472307,
      "grad_norm": 0.0026371076237410307,
      "learning_rate": 0.29941068929074954,
      "loss": 0.3723,
      "num_input_tokens_seen": 2151168,
      "step": 1130
    },
    {
      "epoch": 0.18517007912554043,
      "grad_norm": 0.0023467629216611385,
      "learning_rate": 0.2994054614326381,
      "loss": 0.3889,
      "num_input_tokens_seen": 2159744,
      "step": 1135
    },
    {
      "epoch": 0.18598580634635778,
      "grad_norm": 0.0023145785089582205,
      "learning_rate": 0.29940021053432686,
      "loss": 0.3444,
      "num_input_tokens_seen": 2168464,
      "step": 1140
    },
    {
      "epoch": 0.18680153356717513,
      "grad_norm": 0.0045668999664485455,
      "learning_rate": 0.29939493659662575,
      "loss": 0.3471,
      "num_input_tokens_seen": 2179104,
      "step": 1145
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 0.005898769944906235,
      "learning_rate": 0.299389639620348,
      "loss": 0.3122,
      "num_input_tokens_seen": 2186992,
      "step": 1150
    },
    {
      "epoch": 0.18843298800880987,
      "grad_norm": 0.004837413318455219,
      "learning_rate": 0.29938431960631046,
      "loss": 0.354,
      "num_input_tokens_seen": 2196080,
      "step": 1155
    },
    {
      "epoch": 0.18924871522962722,
      "grad_norm": 0.004089660476893187,
      "learning_rate": 0.2993789765553335,
      "loss": 0.3253,
      "num_input_tokens_seen": 2204832,
      "step": 1160
    },
    {
      "epoch": 0.19006444245044457,
      "grad_norm": 0.0042502256110310555,
      "learning_rate": 0.2993736104682412,
      "loss": 0.37,
      "num_input_tokens_seen": 2214736,
      "step": 1165
    },
    {
      "epoch": 0.19088016967126192,
      "grad_norm": 0.0027505874168127775,
      "learning_rate": 0.299368221345861,
      "loss": 0.3955,
      "num_input_tokens_seen": 2224800,
      "step": 1170
    },
    {
      "epoch": 0.19169589689207928,
      "grad_norm": 0.0058053117245435715,
      "learning_rate": 0.29936280918902397,
      "loss": 0.35,
      "num_input_tokens_seen": 2234848,
      "step": 1175
    },
    {
      "epoch": 0.19251162411289666,
      "grad_norm": 0.00931461714208126,
      "learning_rate": 0.2993573739985648,
      "loss": 0.4324,
      "num_input_tokens_seen": 2244544,
      "step": 1180
    },
    {
      "epoch": 0.193327351333714,
      "grad_norm": 0.004373089876025915,
      "learning_rate": 0.2993519157753216,
      "loss": 0.4007,
      "num_input_tokens_seen": 2254480,
      "step": 1185
    },
    {
      "epoch": 0.19414307855453136,
      "grad_norm": 0.00404549902305007,
      "learning_rate": 0.2993464345201361,
      "loss": 0.3958,
      "num_input_tokens_seen": 2262656,
      "step": 1190
    },
    {
      "epoch": 0.19495880577534871,
      "grad_norm": 0.007270243018865585,
      "learning_rate": 0.2993409302338536,
      "loss": 0.3311,
      "num_input_tokens_seen": 2271184,
      "step": 1195
    },
    {
      "epoch": 0.1957745329961661,
      "grad_norm": 0.0034173361491411924,
      "learning_rate": 0.2993354029173229,
      "loss": 0.3915,
      "num_input_tokens_seen": 2279552,
      "step": 1200
    },
    {
      "epoch": 0.1957745329961661,
      "eval_loss": 0.3487788438796997,
      "eval_runtime": 152.2406,
      "eval_samples_per_second": 17.899,
      "eval_steps_per_second": 8.953,
      "num_input_tokens_seen": 2279552,
      "step": 1200
    },
    {
      "epoch": 0.19659026021698345,
      "grad_norm": 0.011265984736382961,
      "learning_rate": 0.2993298525713965,
      "loss": 0.3814,
      "num_input_tokens_seen": 2289280,
      "step": 1205
    },
    {
      "epoch": 0.1974059874378008,
      "grad_norm": 0.004943955224007368,
      "learning_rate": 0.29932427919693017,
      "loss": 0.3778,
      "num_input_tokens_seen": 2298480,
      "step": 1210
    },
    {
      "epoch": 0.19822171465861815,
      "grad_norm": 0.0021051359362900257,
      "learning_rate": 0.2993186827947834,
      "loss": 0.4766,
      "num_input_tokens_seen": 2308784,
      "step": 1215
    },
    {
      "epoch": 0.1990374418794355,
      "grad_norm": 0.00616390211507678,
      "learning_rate": 0.2993130633658194,
      "loss": 0.3443,
      "num_input_tokens_seen": 2318640,
      "step": 1220
    },
    {
      "epoch": 0.1998531691002529,
      "grad_norm": 0.004766480531543493,
      "learning_rate": 0.29930742091090456,
      "loss": 0.2252,
      "num_input_tokens_seen": 2327664,
      "step": 1225
    },
    {
      "epoch": 0.20066889632107024,
      "grad_norm": 0.004504763521254063,
      "learning_rate": 0.29930175543090914,
      "loss": 0.4701,
      "num_input_tokens_seen": 2338432,
      "step": 1230
    },
    {
      "epoch": 0.2014846235418876,
      "grad_norm": 0.008573617786169052,
      "learning_rate": 0.2992960669267068,
      "loss": 0.4292,
      "num_input_tokens_seen": 2349072,
      "step": 1235
    },
    {
      "epoch": 0.20230035076270494,
      "grad_norm": 0.005004954524338245,
      "learning_rate": 0.29929035539917476,
      "loss": 0.4192,
      "num_input_tokens_seen": 2358576,
      "step": 1240
    },
    {
      "epoch": 0.2031160779835223,
      "grad_norm": 0.019199885427951813,
      "learning_rate": 0.2992846208491938,
      "loss": 0.5199,
      "num_input_tokens_seen": 2367456,
      "step": 1245
    },
    {
      "epoch": 0.20393180520433968,
      "grad_norm": 0.005560137331485748,
      "learning_rate": 0.2992788632776483,
      "loss": 0.3488,
      "num_input_tokens_seen": 2377104,
      "step": 1250
    },
    {
      "epoch": 0.20474753242515703,
      "grad_norm": 0.008261569775640965,
      "learning_rate": 0.29927308268542613,
      "loss": 0.4263,
      "num_input_tokens_seen": 2385824,
      "step": 1255
    },
    {
      "epoch": 0.20556325964597438,
      "grad_norm": 0.002704968210309744,
      "learning_rate": 0.2992672790734187,
      "loss": 0.3348,
      "num_input_tokens_seen": 2396528,
      "step": 1260
    },
    {
      "epoch": 0.20637898686679174,
      "grad_norm": 0.002233333420008421,
      "learning_rate": 0.299261452442521,
      "loss": 0.386,
      "num_input_tokens_seen": 2406288,
      "step": 1265
    },
    {
      "epoch": 0.20719471408760912,
      "grad_norm": 0.0030942847952246666,
      "learning_rate": 0.29925560279363167,
      "loss": 0.3782,
      "num_input_tokens_seen": 2415296,
      "step": 1270
    },
    {
      "epoch": 0.20801044130842647,
      "grad_norm": 0.007090728264302015,
      "learning_rate": 0.29924973012765266,
      "loss": 0.3807,
      "num_input_tokens_seen": 2425168,
      "step": 1275
    },
    {
      "epoch": 0.20882616852924382,
      "grad_norm": 0.004465394653379917,
      "learning_rate": 0.29924383444548974,
      "loss": 0.3874,
      "num_input_tokens_seen": 2435040,
      "step": 1280
    },
    {
      "epoch": 0.20964189575006117,
      "grad_norm": 0.0067972964607179165,
      "learning_rate": 0.299237915748052,
      "loss": 0.3894,
      "num_input_tokens_seen": 2444224,
      "step": 1285
    },
    {
      "epoch": 0.21045762297087853,
      "grad_norm": 0.0009855479001998901,
      "learning_rate": 0.2992319740362522,
      "loss": 0.3315,
      "num_input_tokens_seen": 2453984,
      "step": 1290
    },
    {
      "epoch": 0.2112733501916959,
      "grad_norm": 0.008886980824172497,
      "learning_rate": 0.2992260093110066,
      "loss": 0.3452,
      "num_input_tokens_seen": 2464336,
      "step": 1295
    },
    {
      "epoch": 0.21208907741251326,
      "grad_norm": 0.0049349586479365826,
      "learning_rate": 0.2992200215732352,
      "loss": 0.3611,
      "num_input_tokens_seen": 2472688,
      "step": 1300
    },
    {
      "epoch": 0.2129048046333306,
      "grad_norm": 0.005696780513972044,
      "learning_rate": 0.2992140108238611,
      "loss": 0.3721,
      "num_input_tokens_seen": 2483168,
      "step": 1305
    },
    {
      "epoch": 0.21372053185414797,
      "grad_norm": 0.0028131951112300158,
      "learning_rate": 0.2992079770638115,
      "loss": 0.3547,
      "num_input_tokens_seen": 2492752,
      "step": 1310
    },
    {
      "epoch": 0.21453625907496532,
      "grad_norm": 0.0026537240482866764,
      "learning_rate": 0.29920192029401677,
      "loss": 0.3663,
      "num_input_tokens_seen": 2501424,
      "step": 1315
    },
    {
      "epoch": 0.2153519862957827,
      "grad_norm": 0.005100368056446314,
      "learning_rate": 0.2991958405154109,
      "loss": 0.3571,
      "num_input_tokens_seen": 2511648,
      "step": 1320
    },
    {
      "epoch": 0.21616771351660005,
      "grad_norm": 0.0017888896400108933,
      "learning_rate": 0.29918973772893154,
      "loss": 0.3521,
      "num_input_tokens_seen": 2521552,
      "step": 1325
    },
    {
      "epoch": 0.2169834407374174,
      "grad_norm": 0.002510588616132736,
      "learning_rate": 0.29918361193551973,
      "loss": 0.3471,
      "num_input_tokens_seen": 2530816,
      "step": 1330
    },
    {
      "epoch": 0.21779916795823476,
      "grad_norm": 0.0017670984379947186,
      "learning_rate": 0.29917746313612026,
      "loss": 0.4046,
      "num_input_tokens_seen": 2540880,
      "step": 1335
    },
    {
      "epoch": 0.21861489517905214,
      "grad_norm": 0.009466194547712803,
      "learning_rate": 0.29917129133168124,
      "loss": 0.3919,
      "num_input_tokens_seen": 2550160,
      "step": 1340
    },
    {
      "epoch": 0.2194306223998695,
      "grad_norm": 0.005079388152807951,
      "learning_rate": 0.2991650965231546,
      "loss": 0.4357,
      "num_input_tokens_seen": 2560704,
      "step": 1345
    },
    {
      "epoch": 0.22024634962068684,
      "grad_norm": 0.005486093927174807,
      "learning_rate": 0.29915887871149544,
      "loss": 0.3773,
      "num_input_tokens_seen": 2570592,
      "step": 1350
    },
    {
      "epoch": 0.2210620768415042,
      "grad_norm": 0.00134351197630167,
      "learning_rate": 0.2991526378976628,
      "loss": 0.3556,
      "num_input_tokens_seen": 2579472,
      "step": 1355
    },
    {
      "epoch": 0.22187780406232155,
      "grad_norm": 0.006212689448148012,
      "learning_rate": 0.29914637408261896,
      "loss": 0.3274,
      "num_input_tokens_seen": 2590016,
      "step": 1360
    },
    {
      "epoch": 0.22269353128313893,
      "grad_norm": 0.002838835120201111,
      "learning_rate": 0.29914008726733,
      "loss": 0.4257,
      "num_input_tokens_seen": 2599472,
      "step": 1365
    },
    {
      "epoch": 0.22350925850395628,
      "grad_norm": 0.004661644343286753,
      "learning_rate": 0.2991337774527653,
      "loss": 0.3776,
      "num_input_tokens_seen": 2609632,
      "step": 1370
    },
    {
      "epoch": 0.22432498572477363,
      "grad_norm": 0.0037247049622237682,
      "learning_rate": 0.2991274446398981,
      "loss": 0.3422,
      "num_input_tokens_seen": 2620560,
      "step": 1375
    },
    {
      "epoch": 0.225140712945591,
      "grad_norm": 0.004265011288225651,
      "learning_rate": 0.29912108882970484,
      "loss": 0.323,
      "num_input_tokens_seen": 2630192,
      "step": 1380
    },
    {
      "epoch": 0.22595644016640834,
      "grad_norm": 0.0020254570990800858,
      "learning_rate": 0.2991147100231657,
      "loss": 0.3469,
      "num_input_tokens_seen": 2639824,
      "step": 1385
    },
    {
      "epoch": 0.22677216738722572,
      "grad_norm": 0.0049719237722456455,
      "learning_rate": 0.2991083082212644,
      "loss": 0.385,
      "num_input_tokens_seen": 2648992,
      "step": 1390
    },
    {
      "epoch": 0.22758789460804307,
      "grad_norm": 0.005219262558966875,
      "learning_rate": 0.2991018834249881,
      "loss": 0.3544,
      "num_input_tokens_seen": 2659296,
      "step": 1395
    },
    {
      "epoch": 0.22840362182886043,
      "grad_norm": 0.0047143809497356415,
      "learning_rate": 0.29909543563532764,
      "loss": 0.3655,
      "num_input_tokens_seen": 2668256,
      "step": 1400
    },
    {
      "epoch": 0.22840362182886043,
      "eval_loss": 0.35670074820518494,
      "eval_runtime": 152.2256,
      "eval_samples_per_second": 17.901,
      "eval_steps_per_second": 8.954,
      "num_input_tokens_seen": 2668256,
      "step": 1400
    },
    {
      "epoch": 0.22921934904967778,
      "grad_norm": 0.002215733751654625,
      "learning_rate": 0.29908896485327746,
      "loss": 0.4124,
      "num_input_tokens_seen": 2679536,
      "step": 1405
    },
    {
      "epoch": 0.23003507627049516,
      "grad_norm": 0.011058218777179718,
      "learning_rate": 0.29908247107983527,
      "loss": 0.3587,
      "num_input_tokens_seen": 2688768,
      "step": 1410
    },
    {
      "epoch": 0.2308508034913125,
      "grad_norm": 0.00772440992295742,
      "learning_rate": 0.29907595431600253,
      "loss": 0.4562,
      "num_input_tokens_seen": 2697808,
      "step": 1415
    },
    {
      "epoch": 0.23166653071212986,
      "grad_norm": 0.0023707968648523092,
      "learning_rate": 0.29906941456278424,
      "loss": 0.3914,
      "num_input_tokens_seen": 2706848,
      "step": 1420
    },
    {
      "epoch": 0.23248225793294722,
      "grad_norm": 0.008985593914985657,
      "learning_rate": 0.2990628518211889,
      "loss": 0.3941,
      "num_input_tokens_seen": 2714816,
      "step": 1425
    },
    {
      "epoch": 0.23329798515376457,
      "grad_norm": 0.005025862716138363,
      "learning_rate": 0.2990562660922286,
      "loss": 0.3922,
      "num_input_tokens_seen": 2722512,
      "step": 1430
    },
    {
      "epoch": 0.23411371237458195,
      "grad_norm": 0.0023013073951005936,
      "learning_rate": 0.2990496573769189,
      "loss": 0.3921,
      "num_input_tokens_seen": 2732192,
      "step": 1435
    },
    {
      "epoch": 0.2349294395953993,
      "grad_norm": 0.004968327935785055,
      "learning_rate": 0.29904302567627894,
      "loss": 0.3549,
      "num_input_tokens_seen": 2741856,
      "step": 1440
    },
    {
      "epoch": 0.23574516681621666,
      "grad_norm": 0.008110670372843742,
      "learning_rate": 0.2990363709913314,
      "loss": 0.364,
      "num_input_tokens_seen": 2751104,
      "step": 1445
    },
    {
      "epoch": 0.236560894037034,
      "grad_norm": 0.0039255996234714985,
      "learning_rate": 0.29902969332310264,
      "loss": 0.3513,
      "num_input_tokens_seen": 2759840,
      "step": 1450
    },
    {
      "epoch": 0.23737662125785136,
      "grad_norm": 0.0036703196819871664,
      "learning_rate": 0.2990229926726223,
      "loss": 0.411,
      "num_input_tokens_seen": 2769072,
      "step": 1455
    },
    {
      "epoch": 0.23819234847866874,
      "grad_norm": 0.004289097618311644,
      "learning_rate": 0.29901626904092365,
      "loss": 0.3686,
      "num_input_tokens_seen": 2778848,
      "step": 1460
    },
    {
      "epoch": 0.2390080756994861,
      "grad_norm": 0.0013660286786034703,
      "learning_rate": 0.2990095224290438,
      "loss": 0.3706,
      "num_input_tokens_seen": 2789136,
      "step": 1465
    },
    {
      "epoch": 0.23982380292030345,
      "grad_norm": 0.00263753579929471,
      "learning_rate": 0.29900275283802297,
      "loss": 0.3787,
      "num_input_tokens_seen": 2800352,
      "step": 1470
    },
    {
      "epoch": 0.2406395301411208,
      "grad_norm": 0.003740750951692462,
      "learning_rate": 0.2989959602689051,
      "loss": 0.3387,
      "num_input_tokens_seen": 2810080,
      "step": 1475
    },
    {
      "epoch": 0.24145525736193818,
      "grad_norm": 0.00556564936414361,
      "learning_rate": 0.2989891447227379,
      "loss": 0.3653,
      "num_input_tokens_seen": 2819664,
      "step": 1480
    },
    {
      "epoch": 0.24227098458275553,
      "grad_norm": 0.0023764390498399734,
      "learning_rate": 0.29898230620057215,
      "loss": 0.3584,
      "num_input_tokens_seen": 2828304,
      "step": 1485
    },
    {
      "epoch": 0.24308671180357289,
      "grad_norm": 0.006198955699801445,
      "learning_rate": 0.2989754447034626,
      "loss": 0.3155,
      "num_input_tokens_seen": 2836992,
      "step": 1490
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.009568502195179462,
      "learning_rate": 0.2989685602324673,
      "loss": 0.4923,
      "num_input_tokens_seen": 2847584,
      "step": 1495
    },
    {
      "epoch": 0.2447181662452076,
      "grad_norm": 0.003244684310629964,
      "learning_rate": 0.298961652788648,
      "loss": 0.3702,
      "num_input_tokens_seen": 2856752,
      "step": 1500
    },
    {
      "epoch": 0.24553389346602497,
      "grad_norm": 0.003748510731384158,
      "learning_rate": 0.29895472237306986,
      "loss": 0.4023,
      "num_input_tokens_seen": 2866656,
      "step": 1505
    },
    {
      "epoch": 0.24634962068684232,
      "grad_norm": 0.00914808176457882,
      "learning_rate": 0.29894776898680164,
      "loss": 0.3818,
      "num_input_tokens_seen": 2875824,
      "step": 1510
    },
    {
      "epoch": 0.24716534790765968,
      "grad_norm": 0.002177808666601777,
      "learning_rate": 0.29894079263091566,
      "loss": 0.3914,
      "num_input_tokens_seen": 2885568,
      "step": 1515
    },
    {
      "epoch": 0.24798107512847703,
      "grad_norm": 0.001407710718922317,
      "learning_rate": 0.2989337933064877,
      "loss": 0.3823,
      "num_input_tokens_seen": 2895536,
      "step": 1520
    },
    {
      "epoch": 0.24879680234929438,
      "grad_norm": 0.006154419854283333,
      "learning_rate": 0.29892677101459725,
      "loss": 0.3957,
      "num_input_tokens_seen": 2905184,
      "step": 1525
    },
    {
      "epoch": 0.24961252957011176,
      "grad_norm": 0.003924441523849964,
      "learning_rate": 0.2989197257563272,
      "loss": 0.3273,
      "num_input_tokens_seen": 2913808,
      "step": 1530
    },
    {
      "epoch": 0.2504282567909291,
      "grad_norm": 0.0018241216894239187,
      "learning_rate": 0.2989126575327639,
      "loss": 0.3868,
      "num_input_tokens_seen": 2924064,
      "step": 1535
    },
    {
      "epoch": 0.25124398401174647,
      "grad_norm": 0.0037726215086877346,
      "learning_rate": 0.29890556634499754,
      "loss": 0.3991,
      "num_input_tokens_seen": 2934576,
      "step": 1540
    },
    {
      "epoch": 0.2520597112325638,
      "grad_norm": 0.005316824186593294,
      "learning_rate": 0.2988984521941216,
      "loss": 0.3763,
      "num_input_tokens_seen": 2944304,
      "step": 1545
    },
    {
      "epoch": 0.2528754384533812,
      "grad_norm": 0.002481879200786352,
      "learning_rate": 0.29889131508123307,
      "loss": 0.4463,
      "num_input_tokens_seen": 2953216,
      "step": 1550
    },
    {
      "epoch": 0.2536911656741985,
      "grad_norm": 0.0055935559794306755,
      "learning_rate": 0.2988841550074327,
      "loss": 0.3785,
      "num_input_tokens_seen": 2962832,
      "step": 1555
    },
    {
      "epoch": 0.2545068928950159,
      "grad_norm": 0.0067648557014763355,
      "learning_rate": 0.2988769719738246,
      "loss": 0.3373,
      "num_input_tokens_seen": 2971952,
      "step": 1560
    },
    {
      "epoch": 0.2553226201158333,
      "grad_norm": 0.005537833087146282,
      "learning_rate": 0.29886976598151666,
      "loss": 0.3758,
      "num_input_tokens_seen": 2981328,
      "step": 1565
    },
    {
      "epoch": 0.25613834733665064,
      "grad_norm": 0.0027685267850756645,
      "learning_rate": 0.29886253703161986,
      "loss": 0.3732,
      "num_input_tokens_seen": 2990848,
      "step": 1570
    },
    {
      "epoch": 0.256954074557468,
      "grad_norm": 0.0023493070621043444,
      "learning_rate": 0.29885528512524917,
      "loss": 0.367,
      "num_input_tokens_seen": 2998880,
      "step": 1575
    },
    {
      "epoch": 0.25776980177828535,
      "grad_norm": 0.0029412421863526106,
      "learning_rate": 0.29884801026352287,
      "loss": 0.3586,
      "num_input_tokens_seen": 3006864,
      "step": 1580
    },
    {
      "epoch": 0.2585855289991027,
      "grad_norm": 0.006063900887966156,
      "learning_rate": 0.2988407124475629,
      "loss": 0.3887,
      "num_input_tokens_seen": 3017408,
      "step": 1585
    },
    {
      "epoch": 0.25940125621992005,
      "grad_norm": 0.007010831497609615,
      "learning_rate": 0.2988333916784945,
      "loss": 0.3261,
      "num_input_tokens_seen": 3027968,
      "step": 1590
    },
    {
      "epoch": 0.2602169834407374,
      "grad_norm": 0.008332242257893085,
      "learning_rate": 0.2988260479574468,
      "loss": 0.3774,
      "num_input_tokens_seen": 3038480,
      "step": 1595
    },
    {
      "epoch": 0.26103271066155476,
      "grad_norm": 0.001946312258951366,
      "learning_rate": 0.2988186812855523,
      "loss": 0.375,
      "num_input_tokens_seen": 3047328,
      "step": 1600
    },
    {
      "epoch": 0.26103271066155476,
      "eval_loss": 0.3504978120326996,
      "eval_runtime": 152.1413,
      "eval_samples_per_second": 17.911,
      "eval_steps_per_second": 8.959,
      "num_input_tokens_seen": 3047328,
      "step": 1600
    },
    {
      "epoch": 0.2618484378823721,
      "grad_norm": 0.0016187162837013602,
      "learning_rate": 0.29881129166394693,
      "loss": 0.3554,
      "num_input_tokens_seen": 3057072,
      "step": 1605
    },
    {
      "epoch": 0.2626641651031895,
      "grad_norm": 0.0021581288892775774,
      "learning_rate": 0.29880387909377026,
      "loss": 0.3675,
      "num_input_tokens_seen": 3067920,
      "step": 1610
    },
    {
      "epoch": 0.26347989232400687,
      "grad_norm": 0.0012368966126814485,
      "learning_rate": 0.2987964435761655,
      "loss": 0.3382,
      "num_input_tokens_seen": 3077808,
      "step": 1615
    },
    {
      "epoch": 0.2642956195448242,
      "grad_norm": 0.005025389604270458,
      "learning_rate": 0.29878898511227925,
      "loss": 0.4006,
      "num_input_tokens_seen": 3086832,
      "step": 1620
    },
    {
      "epoch": 0.2651113467656416,
      "grad_norm": 0.0024944101460278034,
      "learning_rate": 0.2987815037032617,
      "loss": 0.3641,
      "num_input_tokens_seen": 3096928,
      "step": 1625
    },
    {
      "epoch": 0.26592707398645893,
      "grad_norm": 0.006220890674740076,
      "learning_rate": 0.29877399935026655,
      "loss": 0.3846,
      "num_input_tokens_seen": 3107120,
      "step": 1630
    },
    {
      "epoch": 0.2667428012072763,
      "grad_norm": 0.0020659391302615404,
      "learning_rate": 0.2987664720544511,
      "loss": 0.4095,
      "num_input_tokens_seen": 3115824,
      "step": 1635
    },
    {
      "epoch": 0.26755852842809363,
      "grad_norm": 0.0011669313535094261,
      "learning_rate": 0.2987589218169761,
      "loss": 0.4,
      "num_input_tokens_seen": 3125824,
      "step": 1640
    },
    {
      "epoch": 0.268374255648911,
      "grad_norm": 0.006907966919243336,
      "learning_rate": 0.29875134863900604,
      "loss": 0.3566,
      "num_input_tokens_seen": 3135616,
      "step": 1645
    },
    {
      "epoch": 0.26918998286972834,
      "grad_norm": 0.0044199079275131226,
      "learning_rate": 0.29874375252170865,
      "loss": 0.3861,
      "num_input_tokens_seen": 3146112,
      "step": 1650
    },
    {
      "epoch": 0.27000571009054575,
      "grad_norm": 0.0030392457265406847,
      "learning_rate": 0.2987361334662553,
      "loss": 0.3396,
      "num_input_tokens_seen": 3155648,
      "step": 1655
    },
    {
      "epoch": 0.2708214373113631,
      "grad_norm": 0.0024916944094002247,
      "learning_rate": 0.29872849147382113,
      "loss": 0.3479,
      "num_input_tokens_seen": 3165408,
      "step": 1660
    },
    {
      "epoch": 0.27163716453218045,
      "grad_norm": 0.004469271283596754,
      "learning_rate": 0.2987208265455845,
      "loss": 0.3681,
      "num_input_tokens_seen": 3175088,
      "step": 1665
    },
    {
      "epoch": 0.2724528917529978,
      "grad_norm": 0.012682132422924042,
      "learning_rate": 0.29871313868272753,
      "loss": 0.4173,
      "num_input_tokens_seen": 3183984,
      "step": 1670
    },
    {
      "epoch": 0.27326861897381516,
      "grad_norm": 0.009097849950194359,
      "learning_rate": 0.29870542788643567,
      "loss": 0.4474,
      "num_input_tokens_seen": 3194048,
      "step": 1675
    },
    {
      "epoch": 0.2740843461946325,
      "grad_norm": 0.003489701310172677,
      "learning_rate": 0.2986976941578981,
      "loss": 0.3741,
      "num_input_tokens_seen": 3203840,
      "step": 1680
    },
    {
      "epoch": 0.27490007341544986,
      "grad_norm": 0.0034544463269412518,
      "learning_rate": 0.29868993749830747,
      "loss": 0.3568,
      "num_input_tokens_seen": 3213856,
      "step": 1685
    },
    {
      "epoch": 0.2757158006362672,
      "grad_norm": 0.008346685208380222,
      "learning_rate": 0.2986821579088598,
      "loss": 0.3978,
      "num_input_tokens_seen": 3222816,
      "step": 1690
    },
    {
      "epoch": 0.27653152785708457,
      "grad_norm": 0.003913578111678362,
      "learning_rate": 0.29867435539075504,
      "loss": 0.3768,
      "num_input_tokens_seen": 3232832,
      "step": 1695
    },
    {
      "epoch": 0.2773472550779019,
      "grad_norm": 0.0011969923507422209,
      "learning_rate": 0.2986665299451963,
      "loss": 0.357,
      "num_input_tokens_seen": 3242016,
      "step": 1700
    },
    {
      "epoch": 0.27816298229871933,
      "grad_norm": 0.002618670230731368,
      "learning_rate": 0.29865868157339037,
      "loss": 0.3444,
      "num_input_tokens_seen": 3251248,
      "step": 1705
    },
    {
      "epoch": 0.2789787095195367,
      "grad_norm": 0.0031732767820358276,
      "learning_rate": 0.2986508102765476,
      "loss": 0.3505,
      "num_input_tokens_seen": 3261264,
      "step": 1710
    },
    {
      "epoch": 0.27979443674035404,
      "grad_norm": 0.0022109521087259054,
      "learning_rate": 0.2986429160558818,
      "loss": 0.3603,
      "num_input_tokens_seen": 3271088,
      "step": 1715
    },
    {
      "epoch": 0.2806101639611714,
      "grad_norm": 0.0026958913076668978,
      "learning_rate": 0.2986349989126104,
      "loss": 0.3864,
      "num_input_tokens_seen": 3280176,
      "step": 1720
    },
    {
      "epoch": 0.28142589118198874,
      "grad_norm": 0.005903821438550949,
      "learning_rate": 0.29862705884795426,
      "loss": 0.3549,
      "num_input_tokens_seen": 3290048,
      "step": 1725
    },
    {
      "epoch": 0.2822416184028061,
      "grad_norm": 0.0030801226384937763,
      "learning_rate": 0.2986190958631379,
      "loss": 0.3323,
      "num_input_tokens_seen": 3299504,
      "step": 1730
    },
    {
      "epoch": 0.28305734562362345,
      "grad_norm": 0.004728138912469149,
      "learning_rate": 0.29861110995938933,
      "loss": 0.427,
      "num_input_tokens_seen": 3308560,
      "step": 1735
    },
    {
      "epoch": 0.2838730728444408,
      "grad_norm": 0.00453863013535738,
      "learning_rate": 0.29860310113794,
      "loss": 0.3111,
      "num_input_tokens_seen": 3316848,
      "step": 1740
    },
    {
      "epoch": 0.28468880006525815,
      "grad_norm": 0.012554382905364037,
      "learning_rate": 0.29859506940002506,
      "loss": 0.4867,
      "num_input_tokens_seen": 3325536,
      "step": 1745
    },
    {
      "epoch": 0.28550452728607556,
      "grad_norm": 0.007355751469731331,
      "learning_rate": 0.298587014746883,
      "loss": 0.4487,
      "num_input_tokens_seen": 3336176,
      "step": 1750
    },
    {
      "epoch": 0.2863202545068929,
      "grad_norm": 0.008678053505718708,
      "learning_rate": 0.298578937179756,
      "loss": 0.5197,
      "num_input_tokens_seen": 3345968,
      "step": 1755
    },
    {
      "epoch": 0.28713598172771027,
      "grad_norm": 0.004027821589261293,
      "learning_rate": 0.29857083669988976,
      "loss": 0.4438,
      "num_input_tokens_seen": 3354640,
      "step": 1760
    },
    {
      "epoch": 0.2879517089485276,
      "grad_norm": 0.0052729202434420586,
      "learning_rate": 0.29856271330853346,
      "loss": 0.4004,
      "num_input_tokens_seen": 3363936,
      "step": 1765
    },
    {
      "epoch": 0.28876743616934497,
      "grad_norm": 0.0030548658687621355,
      "learning_rate": 0.2985545670069398,
      "loss": 0.4451,
      "num_input_tokens_seen": 3373168,
      "step": 1770
    },
    {
      "epoch": 0.2895831633901623,
      "grad_norm": 0.00690540112555027,
      "learning_rate": 0.29854639779636505,
      "loss": 0.4082,
      "num_input_tokens_seen": 3383232,
      "step": 1775
    },
    {
      "epoch": 0.2903988906109797,
      "grad_norm": 0.005807578098028898,
      "learning_rate": 0.298538205678069,
      "loss": 0.3541,
      "num_input_tokens_seen": 3393328,
      "step": 1780
    },
    {
      "epoch": 0.29121461783179703,
      "grad_norm": 0.008807721547782421,
      "learning_rate": 0.298529990653315,
      "loss": 0.3803,
      "num_input_tokens_seen": 3402192,
      "step": 1785
    },
    {
      "epoch": 0.2920303450526144,
      "grad_norm": 0.010624181479215622,
      "learning_rate": 0.29852175272336984,
      "loss": 0.4336,
      "num_input_tokens_seen": 3410928,
      "step": 1790
    },
    {
      "epoch": 0.2928460722734318,
      "grad_norm": 0.003229788737371564,
      "learning_rate": 0.29851349188950405,
      "loss": 0.4271,
      "num_input_tokens_seen": 3420160,
      "step": 1795
    },
    {
      "epoch": 0.29366179949424914,
      "grad_norm": 0.0074231731705367565,
      "learning_rate": 0.2985052081529914,
      "loss": 0.4324,
      "num_input_tokens_seen": 3429984,
      "step": 1800
    },
    {
      "epoch": 0.29366179949424914,
      "eval_loss": 0.4502207338809967,
      "eval_runtime": 152.2567,
      "eval_samples_per_second": 17.897,
      "eval_steps_per_second": 8.952,
      "num_input_tokens_seen": 3429984,
      "step": 1800
    },
    {
      "epoch": 0.2944775267150665,
      "grad_norm": 0.007252194453030825,
      "learning_rate": 0.29849690151510944,
      "loss": 0.4247,
      "num_input_tokens_seen": 3440304,
      "step": 1805
    },
    {
      "epoch": 0.29529325393588385,
      "grad_norm": 0.0021448712795972824,
      "learning_rate": 0.2984885719771392,
      "loss": 0.3625,
      "num_input_tokens_seen": 3449136,
      "step": 1810
    },
    {
      "epoch": 0.2961089811567012,
      "grad_norm": 0.010619794949889183,
      "learning_rate": 0.2984802195403651,
      "loss": 0.4373,
      "num_input_tokens_seen": 3459568,
      "step": 1815
    },
    {
      "epoch": 0.29692470837751855,
      "grad_norm": 0.005656272638589144,
      "learning_rate": 0.2984718442060752,
      "loss": 0.384,
      "num_input_tokens_seen": 3466976,
      "step": 1820
    },
    {
      "epoch": 0.2977404355983359,
      "grad_norm": 0.007924988865852356,
      "learning_rate": 0.2984634459755611,
      "loss": 0.337,
      "num_input_tokens_seen": 3477360,
      "step": 1825
    },
    {
      "epoch": 0.29855616281915326,
      "grad_norm": 0.0020434444304555655,
      "learning_rate": 0.29845502485011793,
      "loss": 0.3565,
      "num_input_tokens_seen": 3488048,
      "step": 1830
    },
    {
      "epoch": 0.2993718900399706,
      "grad_norm": 0.0048957206308841705,
      "learning_rate": 0.2984465808310444,
      "loss": 0.3231,
      "num_input_tokens_seen": 3497120,
      "step": 1835
    },
    {
      "epoch": 0.300187617260788,
      "grad_norm": 0.0028270494658499956,
      "learning_rate": 0.29843811391964253,
      "loss": 0.3655,
      "num_input_tokens_seen": 3506656,
      "step": 1840
    },
    {
      "epoch": 0.3010033444816054,
      "grad_norm": 0.0018389547476544976,
      "learning_rate": 0.2984296241172182,
      "loss": 0.3394,
      "num_input_tokens_seen": 3515024,
      "step": 1845
    },
    {
      "epoch": 0.3018190717024227,
      "grad_norm": 0.005379370413720608,
      "learning_rate": 0.29842111142508043,
      "loss": 0.3955,
      "num_input_tokens_seen": 3524304,
      "step": 1850
    },
    {
      "epoch": 0.3026347989232401,
      "grad_norm": 0.006842643953859806,
      "learning_rate": 0.29841257584454217,
      "loss": 0.4531,
      "num_input_tokens_seen": 3533280,
      "step": 1855
    },
    {
      "epoch": 0.30345052614405743,
      "grad_norm": 0.006003137212246656,
      "learning_rate": 0.29840401737691963,
      "loss": 0.3701,
      "num_input_tokens_seen": 3541984,
      "step": 1860
    },
    {
      "epoch": 0.3042662533648748,
      "grad_norm": 0.011189298704266548,
      "learning_rate": 0.29839543602353263,
      "loss": 0.4885,
      "num_input_tokens_seen": 3551168,
      "step": 1865
    },
    {
      "epoch": 0.30508198058569214,
      "grad_norm": 0.008511626161634922,
      "learning_rate": 0.2983868317857046,
      "loss": 0.4124,
      "num_input_tokens_seen": 3560880,
      "step": 1870
    },
    {
      "epoch": 0.3058977078065095,
      "grad_norm": 0.003564175684005022,
      "learning_rate": 0.2983782046647623,
      "loss": 0.3707,
      "num_input_tokens_seen": 3570704,
      "step": 1875
    },
    {
      "epoch": 0.30671343502732684,
      "grad_norm": 0.00641641765832901,
      "learning_rate": 0.2983695546620362,
      "loss": 0.3856,
      "num_input_tokens_seen": 3580608,
      "step": 1880
    },
    {
      "epoch": 0.3075291622481442,
      "grad_norm": 0.0013851739931851625,
      "learning_rate": 0.2983608817788603,
      "loss": 0.4132,
      "num_input_tokens_seen": 3589808,
      "step": 1885
    },
    {
      "epoch": 0.3083448894689616,
      "grad_norm": 0.0018597590969875455,
      "learning_rate": 0.29835218601657193,
      "loss": 0.3549,
      "num_input_tokens_seen": 3599392,
      "step": 1890
    },
    {
      "epoch": 0.30916061668977896,
      "grad_norm": 0.005102145951241255,
      "learning_rate": 0.29834346737651224,
      "loss": 0.3774,
      "num_input_tokens_seen": 3608208,
      "step": 1895
    },
    {
      "epoch": 0.3099763439105963,
      "grad_norm": 0.003004404716193676,
      "learning_rate": 0.29833472586002563,
      "loss": 0.362,
      "num_input_tokens_seen": 3617920,
      "step": 1900
    },
    {
      "epoch": 0.31079207113141366,
      "grad_norm": 0.0013659382238984108,
      "learning_rate": 0.29832596146846024,
      "loss": 0.4158,
      "num_input_tokens_seen": 3628688,
      "step": 1905
    },
    {
      "epoch": 0.311607798352231,
      "grad_norm": 0.005409178324043751,
      "learning_rate": 0.2983171742031676,
      "loss": 0.3627,
      "num_input_tokens_seen": 3638784,
      "step": 1910
    },
    {
      "epoch": 0.31242352557304837,
      "grad_norm": 0.00479511171579361,
      "learning_rate": 0.2983083640655028,
      "loss": 0.33,
      "num_input_tokens_seen": 3648192,
      "step": 1915
    },
    {
      "epoch": 0.3132392527938657,
      "grad_norm": 0.0021138284355401993,
      "learning_rate": 0.29829953105682455,
      "loss": 0.4225,
      "num_input_tokens_seen": 3657936,
      "step": 1920
    },
    {
      "epoch": 0.31405498001468307,
      "grad_norm": 0.009133751504123211,
      "learning_rate": 0.29829067517849495,
      "loss": 0.4518,
      "num_input_tokens_seen": 3668384,
      "step": 1925
    },
    {
      "epoch": 0.3148707072355004,
      "grad_norm": 0.004217454232275486,
      "learning_rate": 0.2982817964318797,
      "loss": 0.3969,
      "num_input_tokens_seen": 3678480,
      "step": 1930
    },
    {
      "epoch": 0.31568643445631783,
      "grad_norm": 0.008121499791741371,
      "learning_rate": 0.298272894818348,
      "loss": 0.4067,
      "num_input_tokens_seen": 3688624,
      "step": 1935
    },
    {
      "epoch": 0.3165021616771352,
      "grad_norm": 0.010452664457261562,
      "learning_rate": 0.2982639703392726,
      "loss": 0.4396,
      "num_input_tokens_seen": 3699184,
      "step": 1940
    },
    {
      "epoch": 0.31731788889795254,
      "grad_norm": 0.003999076783657074,
      "learning_rate": 0.29825502299602974,
      "loss": 0.4462,
      "num_input_tokens_seen": 3707760,
      "step": 1945
    },
    {
      "epoch": 0.3181336161187699,
      "grad_norm": 0.006358052603900433,
      "learning_rate": 0.2982460527899993,
      "loss": 0.3973,
      "num_input_tokens_seen": 3717360,
      "step": 1950
    },
    {
      "epoch": 0.31894934333958724,
      "grad_norm": 0.018805144354701042,
      "learning_rate": 0.29823705972256453,
      "loss": 0.5689,
      "num_input_tokens_seen": 3727680,
      "step": 1955
    },
    {
      "epoch": 0.3197650705604046,
      "grad_norm": 0.0066353022120893,
      "learning_rate": 0.2982280437951123,
      "loss": 0.446,
      "num_input_tokens_seen": 3736720,
      "step": 1960
    },
    {
      "epoch": 0.32058079778122195,
      "grad_norm": 0.008509685285389423,
      "learning_rate": 0.298219005009033,
      "loss": 0.4283,
      "num_input_tokens_seen": 3745632,
      "step": 1965
    },
    {
      "epoch": 0.3213965250020393,
      "grad_norm": 0.005886917933821678,
      "learning_rate": 0.29820994336572043,
      "loss": 0.4432,
      "num_input_tokens_seen": 3755184,
      "step": 1970
    },
    {
      "epoch": 0.32221225222285665,
      "grad_norm": 0.004455719608813524,
      "learning_rate": 0.2982008588665721,
      "loss": 0.3408,
      "num_input_tokens_seen": 3764784,
      "step": 1975
    },
    {
      "epoch": 0.32302797944367406,
      "grad_norm": 0.004428755026310682,
      "learning_rate": 0.2981917515129889,
      "loss": 0.3926,
      "num_input_tokens_seen": 3774256,
      "step": 1980
    },
    {
      "epoch": 0.3238437066644914,
      "grad_norm": 0.0024548813235014677,
      "learning_rate": 0.2981826213063753,
      "loss": 0.3296,
      "num_input_tokens_seen": 3783552,
      "step": 1985
    },
    {
      "epoch": 0.32465943388530877,
      "grad_norm": 0.002247497672215104,
      "learning_rate": 0.2981734682481394,
      "loss": 0.3746,
      "num_input_tokens_seen": 3793216,
      "step": 1990
    },
    {
      "epoch": 0.3254751611061261,
      "grad_norm": 0.0031565672252327204,
      "learning_rate": 0.29816429233969255,
      "loss": 0.3416,
      "num_input_tokens_seen": 3803456,
      "step": 1995
    },
    {
      "epoch": 0.3262908883269435,
      "grad_norm": 0.0011205293703824282,
      "learning_rate": 0.2981550935824499,
      "loss": 0.3761,
      "num_input_tokens_seen": 3814576,
      "step": 2000
    },
    {
      "epoch": 0.3262908883269435,
      "eval_loss": 0.4696444272994995,
      "eval_runtime": 152.3028,
      "eval_samples_per_second": 17.892,
      "eval_steps_per_second": 8.949,
      "num_input_tokens_seen": 3814576,
      "step": 2000
    },
    {
      "epoch": 0.3271066155477608,
      "grad_norm": 0.004469241946935654,
      "learning_rate": 0.29814587197783,
      "loss": 0.3745,
      "num_input_tokens_seen": 3824384,
      "step": 2005
    },
    {
      "epoch": 0.3279223427685782,
      "grad_norm": 0.003528786823153496,
      "learning_rate": 0.29813662752725495,
      "loss": 0.3619,
      "num_input_tokens_seen": 3834048,
      "step": 2010
    },
    {
      "epoch": 0.32873806998939553,
      "grad_norm": 0.006096781697124243,
      "learning_rate": 0.29812736023215025,
      "loss": 0.3211,
      "num_input_tokens_seen": 3844048,
      "step": 2015
    },
    {
      "epoch": 0.3295537972102129,
      "grad_norm": 0.003979805391281843,
      "learning_rate": 0.29811807009394514,
      "loss": 0.3639,
      "num_input_tokens_seen": 3854544,
      "step": 2020
    },
    {
      "epoch": 0.33036952443103024,
      "grad_norm": 0.006987850647419691,
      "learning_rate": 0.2981087571140723,
      "loss": 0.3866,
      "num_input_tokens_seen": 3864240,
      "step": 2025
    },
    {
      "epoch": 0.33118525165184765,
      "grad_norm": 0.002891786163672805,
      "learning_rate": 0.2980994212939678,
      "loss": 0.3374,
      "num_input_tokens_seen": 3873776,
      "step": 2030
    },
    {
      "epoch": 0.332000978872665,
      "grad_norm": 0.0019130646251142025,
      "learning_rate": 0.2980900626350715,
      "loss": 0.3527,
      "num_input_tokens_seen": 3882560,
      "step": 2035
    },
    {
      "epoch": 0.33281670609348235,
      "grad_norm": 0.0020527858287096024,
      "learning_rate": 0.29808068113882646,
      "loss": 0.3556,
      "num_input_tokens_seen": 3892144,
      "step": 2040
    },
    {
      "epoch": 0.3336324333142997,
      "grad_norm": 0.0014304855139926076,
      "learning_rate": 0.2980712768066795,
      "loss": 0.3514,
      "num_input_tokens_seen": 3902000,
      "step": 2045
    },
    {
      "epoch": 0.33444816053511706,
      "grad_norm": 0.0038903735112398863,
      "learning_rate": 0.2980618496400809,
      "loss": 0.3705,
      "num_input_tokens_seen": 3910976,
      "step": 2050
    },
    {
      "epoch": 0.3352638877559344,
      "grad_norm": 0.002188608283177018,
      "learning_rate": 0.2980523996404844,
      "loss": 0.361,
      "num_input_tokens_seen": 3920608,
      "step": 2055
    },
    {
      "epoch": 0.33607961497675176,
      "grad_norm": 0.005785853136330843,
      "learning_rate": 0.2980429268093473,
      "loss": 0.3614,
      "num_input_tokens_seen": 3929408,
      "step": 2060
    },
    {
      "epoch": 0.3368953421975691,
      "grad_norm": 0.001320487353950739,
      "learning_rate": 0.29803343114813047,
      "loss": 0.3653,
      "num_input_tokens_seen": 3938928,
      "step": 2065
    },
    {
      "epoch": 0.33771106941838647,
      "grad_norm": 0.0028050311375409365,
      "learning_rate": 0.2980239126582983,
      "loss": 0.3592,
      "num_input_tokens_seen": 3949248,
      "step": 2070
    },
    {
      "epoch": 0.3385267966392039,
      "grad_norm": 0.00369961722753942,
      "learning_rate": 0.2980143713413186,
      "loss": 0.3471,
      "num_input_tokens_seen": 3959088,
      "step": 2075
    },
    {
      "epoch": 0.33934252386002123,
      "grad_norm": 0.007369568571448326,
      "learning_rate": 0.29800480719866274,
      "loss": 0.3722,
      "num_input_tokens_seen": 3967456,
      "step": 2080
    },
    {
      "epoch": 0.3401582510808386,
      "grad_norm": 0.0047465660609304905,
      "learning_rate": 0.2979952202318057,
      "loss": 0.3523,
      "num_input_tokens_seen": 3976544,
      "step": 2085
    },
    {
      "epoch": 0.34097397830165593,
      "grad_norm": 0.0017195851542055607,
      "learning_rate": 0.2979856104422259,
      "loss": 0.3445,
      "num_input_tokens_seen": 3986992,
      "step": 2090
    },
    {
      "epoch": 0.3417897055224733,
      "grad_norm": 0.002607115777209401,
      "learning_rate": 0.2979759778314052,
      "loss": 0.3045,
      "num_input_tokens_seen": 3996432,
      "step": 2095
    },
    {
      "epoch": 0.34260543274329064,
      "grad_norm": 0.002147168619558215,
      "learning_rate": 0.2979663224008292,
      "loss": 0.439,
      "num_input_tokens_seen": 4005472,
      "step": 2100
    },
    {
      "epoch": 0.343421159964108,
      "grad_norm": 0.004074858967214823,
      "learning_rate": 0.2979566441519868,
      "loss": 0.3508,
      "num_input_tokens_seen": 4015456,
      "step": 2105
    },
    {
      "epoch": 0.34423688718492534,
      "grad_norm": 0.0052202665247023106,
      "learning_rate": 0.29794694308637054,
      "loss": 0.4001,
      "num_input_tokens_seen": 4024336,
      "step": 2110
    },
    {
      "epoch": 0.3450526144057427,
      "grad_norm": 0.0034603767562657595,
      "learning_rate": 0.2979372192054764,
      "loss": 0.4352,
      "num_input_tokens_seen": 4033920,
      "step": 2115
    },
    {
      "epoch": 0.3458683416265601,
      "grad_norm": 0.001423075795173645,
      "learning_rate": 0.297927472510804,
      "loss": 0.369,
      "num_input_tokens_seen": 4043984,
      "step": 2120
    },
    {
      "epoch": 0.34668406884737746,
      "grad_norm": 0.003054780652746558,
      "learning_rate": 0.29791770300385634,
      "loss": 0.3577,
      "num_input_tokens_seen": 4053680,
      "step": 2125
    },
    {
      "epoch": 0.3474997960681948,
      "grad_norm": 0.0029128326568752527,
      "learning_rate": 0.29790791068614003,
      "loss": 0.3549,
      "num_input_tokens_seen": 4062864,
      "step": 2130
    },
    {
      "epoch": 0.34831552328901216,
      "grad_norm": 0.0038180118426680565,
      "learning_rate": 0.2978980955591652,
      "loss": 0.4117,
      "num_input_tokens_seen": 4071984,
      "step": 2135
    },
    {
      "epoch": 0.3491312505098295,
      "grad_norm": 0.001555353868752718,
      "learning_rate": 0.2978882576244454,
      "loss": 0.4148,
      "num_input_tokens_seen": 4080528,
      "step": 2140
    },
    {
      "epoch": 0.34994697773064687,
      "grad_norm": 0.0028892504051327705,
      "learning_rate": 0.2978783968834978,
      "loss": 0.3495,
      "num_input_tokens_seen": 4089056,
      "step": 2145
    },
    {
      "epoch": 0.3507627049514642,
      "grad_norm": 0.002077678916975856,
      "learning_rate": 0.29786851333784303,
      "loss": 0.2945,
      "num_input_tokens_seen": 4098544,
      "step": 2150
    },
    {
      "epoch": 0.3515784321722816,
      "grad_norm": 0.0049983058124780655,
      "learning_rate": 0.2978586069890053,
      "loss": 0.4042,
      "num_input_tokens_seen": 4108544,
      "step": 2155
    },
    {
      "epoch": 0.3523941593930989,
      "grad_norm": 0.002761354437097907,
      "learning_rate": 0.29784867783851227,
      "loss": 0.4065,
      "num_input_tokens_seen": 4117328,
      "step": 2160
    },
    {
      "epoch": 0.3532098866139163,
      "grad_norm": 0.005747241899371147,
      "learning_rate": 0.2978387258878951,
      "loss": 0.4976,
      "num_input_tokens_seen": 4125904,
      "step": 2165
    },
    {
      "epoch": 0.3540256138347337,
      "grad_norm": 0.0015330675523728132,
      "learning_rate": 0.29782875113868856,
      "loss": 0.424,
      "num_input_tokens_seen": 4136336,
      "step": 2170
    },
    {
      "epoch": 0.35484134105555104,
      "grad_norm": 0.005234689451754093,
      "learning_rate": 0.2978187535924309,
      "loss": 0.3805,
      "num_input_tokens_seen": 4145360,
      "step": 2175
    },
    {
      "epoch": 0.3556570682763684,
      "grad_norm": 0.0018219243502244353,
      "learning_rate": 0.29780873325066376,
      "loss": 0.3919,
      "num_input_tokens_seen": 4154144,
      "step": 2180
    },
    {
      "epoch": 0.35647279549718575,
      "grad_norm": 0.004686444066464901,
      "learning_rate": 0.2977986901149325,
      "loss": 0.4029,
      "num_input_tokens_seen": 4162880,
      "step": 2185
    },
    {
      "epoch": 0.3572885227180031,
      "grad_norm": 0.001147166476584971,
      "learning_rate": 0.29778862418678587,
      "loss": 0.3539,
      "num_input_tokens_seen": 4171488,
      "step": 2190
    },
    {
      "epoch": 0.35810424993882045,
      "grad_norm": 0.0013062090147286654,
      "learning_rate": 0.29777853546777616,
      "loss": 0.3877,
      "num_input_tokens_seen": 4181072,
      "step": 2195
    },
    {
      "epoch": 0.3589199771596378,
      "grad_norm": 0.0030500334687530994,
      "learning_rate": 0.2977684239594592,
      "loss": 0.3441,
      "num_input_tokens_seen": 4190352,
      "step": 2200
    },
    {
      "epoch": 0.3589199771596378,
      "eval_loss": 0.3639482855796814,
      "eval_runtime": 152.2907,
      "eval_samples_per_second": 17.893,
      "eval_steps_per_second": 8.95,
      "num_input_tokens_seen": 4190352,
      "step": 2200
    },
    {
      "epoch": 0.35973570438045516,
      "grad_norm": 0.0010440831538289785,
      "learning_rate": 0.29775828966339424,
      "loss": 0.3491,
      "num_input_tokens_seen": 4199200,
      "step": 2205
    },
    {
      "epoch": 0.3605514316012725,
      "grad_norm": 0.0026245340704917908,
      "learning_rate": 0.29774813258114424,
      "loss": 0.3609,
      "num_input_tokens_seen": 4209472,
      "step": 2210
    },
    {
      "epoch": 0.3613671588220899,
      "grad_norm": 0.004888047929853201,
      "learning_rate": 0.29773795271427544,
      "loss": 0.3275,
      "num_input_tokens_seen": 4218880,
      "step": 2215
    },
    {
      "epoch": 0.36218288604290727,
      "grad_norm": 0.008028345182538033,
      "learning_rate": 0.2977277500643577,
      "loss": 0.3763,
      "num_input_tokens_seen": 4229008,
      "step": 2220
    },
    {
      "epoch": 0.3629986132637246,
      "grad_norm": 0.006132546812295914,
      "learning_rate": 0.29771752463296447,
      "loss": 0.4078,
      "num_input_tokens_seen": 4238640,
      "step": 2225
    },
    {
      "epoch": 0.363814340484542,
      "grad_norm": 0.003254302078858018,
      "learning_rate": 0.29770727642167266,
      "loss": 0.4471,
      "num_input_tokens_seen": 4248400,
      "step": 2230
    },
    {
      "epoch": 0.36463006770535933,
      "grad_norm": 0.0009824344888329506,
      "learning_rate": 0.29769700543206257,
      "loss": 0.3232,
      "num_input_tokens_seen": 4258480,
      "step": 2235
    },
    {
      "epoch": 0.3654457949261767,
      "grad_norm": 0.0058533442206680775,
      "learning_rate": 0.2976867116657182,
      "loss": 0.5106,
      "num_input_tokens_seen": 4267712,
      "step": 2240
    },
    {
      "epoch": 0.36626152214699403,
      "grad_norm": 0.006911291275173426,
      "learning_rate": 0.2976763951242269,
      "loss": 0.37,
      "num_input_tokens_seen": 4277952,
      "step": 2245
    },
    {
      "epoch": 0.3670772493678114,
      "grad_norm": 0.003764176508411765,
      "learning_rate": 0.29766605580917965,
      "loss": 0.3652,
      "num_input_tokens_seen": 4287072,
      "step": 2250
    },
    {
      "epoch": 0.36789297658862874,
      "grad_norm": 0.0028835588600486517,
      "learning_rate": 0.29765569372217093,
      "loss": 0.3605,
      "num_input_tokens_seen": 4297088,
      "step": 2255
    },
    {
      "epoch": 0.36870870380944615,
      "grad_norm": 0.004202001728117466,
      "learning_rate": 0.2976453088647987,
      "loss": 0.3504,
      "num_input_tokens_seen": 4304000,
      "step": 2260
    },
    {
      "epoch": 0.3695244310302635,
      "grad_norm": 0.0012384483125060797,
      "learning_rate": 0.2976349012386644,
      "loss": 0.3808,
      "num_input_tokens_seen": 4313344,
      "step": 2265
    },
    {
      "epoch": 0.37034015825108085,
      "grad_norm": 0.001392557518556714,
      "learning_rate": 0.29762447084537297,
      "loss": 0.3428,
      "num_input_tokens_seen": 4322000,
      "step": 2270
    },
    {
      "epoch": 0.3711558854718982,
      "grad_norm": 0.0056200479157269,
      "learning_rate": 0.29761401768653306,
      "loss": 0.406,
      "num_input_tokens_seen": 4332064,
      "step": 2275
    },
    {
      "epoch": 0.37197161269271556,
      "grad_norm": 0.002295647282153368,
      "learning_rate": 0.29760354176375653,
      "loss": 0.3892,
      "num_input_tokens_seen": 4341520,
      "step": 2280
    },
    {
      "epoch": 0.3727873399135329,
      "grad_norm": 0.0024849004112184048,
      "learning_rate": 0.29759304307865897,
      "loss": 0.3872,
      "num_input_tokens_seen": 4351856,
      "step": 2285
    },
    {
      "epoch": 0.37360306713435026,
      "grad_norm": 0.0019309837371110916,
      "learning_rate": 0.2975825216328594,
      "loss": 0.4324,
      "num_input_tokens_seen": 4360528,
      "step": 2290
    },
    {
      "epoch": 0.3744187943551676,
      "grad_norm": 0.0057870629243552685,
      "learning_rate": 0.2975719774279804,
      "loss": 0.3537,
      "num_input_tokens_seen": 4368992,
      "step": 2295
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 0.0038359055761247873,
      "learning_rate": 0.29756141046564794,
      "loss": 0.4766,
      "num_input_tokens_seen": 4378192,
      "step": 2300
    },
    {
      "epoch": 0.3760502487968023,
      "grad_norm": 0.003964667208492756,
      "learning_rate": 0.2975508207474916,
      "loss": 0.4583,
      "num_input_tokens_seen": 4387328,
      "step": 2305
    },
    {
      "epoch": 0.37686597601761973,
      "grad_norm": 0.0017966694431379437,
      "learning_rate": 0.2975402082751445,
      "loss": 0.4405,
      "num_input_tokens_seen": 4397120,
      "step": 2310
    },
    {
      "epoch": 0.3776817032384371,
      "grad_norm": 0.0013711638748645782,
      "learning_rate": 0.29752957305024313,
      "loss": 0.3727,
      "num_input_tokens_seen": 4406832,
      "step": 2315
    },
    {
      "epoch": 0.37849743045925444,
      "grad_norm": 0.0011683336924761534,
      "learning_rate": 0.2975189150744277,
      "loss": 0.3762,
      "num_input_tokens_seen": 4416288,
      "step": 2320
    },
    {
      "epoch": 0.3793131576800718,
      "grad_norm": 0.0014909222954884171,
      "learning_rate": 0.29750823434934165,
      "loss": 0.3625,
      "num_input_tokens_seen": 4425344,
      "step": 2325
    },
    {
      "epoch": 0.38012888490088914,
      "grad_norm": 0.0012975013814866543,
      "learning_rate": 0.29749753087663217,
      "loss": 0.3498,
      "num_input_tokens_seen": 4434992,
      "step": 2330
    },
    {
      "epoch": 0.3809446121217065,
      "grad_norm": 0.0024512489326298237,
      "learning_rate": 0.29748680465794985,
      "loss": 0.3586,
      "num_input_tokens_seen": 4444384,
      "step": 2335
    },
    {
      "epoch": 0.38176033934252385,
      "grad_norm": 0.0032295365817844868,
      "learning_rate": 0.29747605569494884,
      "loss": 0.404,
      "num_input_tokens_seen": 4455584,
      "step": 2340
    },
    {
      "epoch": 0.3825760665633412,
      "grad_norm": 0.006518251728266478,
      "learning_rate": 0.29746528398928673,
      "loss": 0.3467,
      "num_input_tokens_seen": 4465168,
      "step": 2345
    },
    {
      "epoch": 0.38339179378415855,
      "grad_norm": 0.0068752397783100605,
      "learning_rate": 0.2974544895426247,
      "loss": 0.402,
      "num_input_tokens_seen": 4474464,
      "step": 2350
    },
    {
      "epoch": 0.38420752100497596,
      "grad_norm": 0.003243526443839073,
      "learning_rate": 0.29744367235662733,
      "loss": 0.3155,
      "num_input_tokens_seen": 4483424,
      "step": 2355
    },
    {
      "epoch": 0.3850232482257933,
      "grad_norm": 0.005158260930329561,
      "learning_rate": 0.29743283243296276,
      "loss": 0.4255,
      "num_input_tokens_seen": 4492144,
      "step": 2360
    },
    {
      "epoch": 0.38583897544661067,
      "grad_norm": 0.0014987627509981394,
      "learning_rate": 0.29742196977330276,
      "loss": 0.3773,
      "num_input_tokens_seen": 4500624,
      "step": 2365
    },
    {
      "epoch": 0.386654702667428,
      "grad_norm": 0.004030466079711914,
      "learning_rate": 0.2974110843793223,
      "loss": 0.3398,
      "num_input_tokens_seen": 4509920,
      "step": 2370
    },
    {
      "epoch": 0.38747042988824537,
      "grad_norm": 0.004650078248232603,
      "learning_rate": 0.2974001762527002,
      "loss": 0.3821,
      "num_input_tokens_seen": 4519440,
      "step": 2375
    },
    {
      "epoch": 0.3882861571090627,
      "grad_norm": 0.0026420133654028177,
      "learning_rate": 0.2973892453951186,
      "loss": 0.342,
      "num_input_tokens_seen": 4529472,
      "step": 2380
    },
    {
      "epoch": 0.3891018843298801,
      "grad_norm": 0.00436594570055604,
      "learning_rate": 0.2973782918082631,
      "loss": 0.3626,
      "num_input_tokens_seen": 4539104,
      "step": 2385
    },
    {
      "epoch": 0.38991761155069743,
      "grad_norm": 0.002571185352280736,
      "learning_rate": 0.29736731549382295,
      "loss": 0.3547,
      "num_input_tokens_seen": 4548592,
      "step": 2390
    },
    {
      "epoch": 0.3907333387715148,
      "grad_norm": 0.0013661339180544019,
      "learning_rate": 0.2973563164534908,
      "loss": 0.3742,
      "num_input_tokens_seen": 4558096,
      "step": 2395
    },
    {
      "epoch": 0.3915490659923322,
      "grad_norm": 0.0022274968214333057,
      "learning_rate": 0.29734529468896287,
      "loss": 0.3423,
      "num_input_tokens_seen": 4567440,
      "step": 2400
    },
    {
      "epoch": 0.3915490659923322,
      "eval_loss": 0.3467554450035095,
      "eval_runtime": 152.2877,
      "eval_samples_per_second": 17.894,
      "eval_steps_per_second": 8.95,
      "num_input_tokens_seen": 4567440,
      "step": 2400
    },
    {
      "epoch": 0.39236479321314954,
      "grad_norm": 0.0025103650987148285,
      "learning_rate": 0.2973342502019388,
      "loss": 0.3844,
      "num_input_tokens_seen": 4576832,
      "step": 2405
    },
    {
      "epoch": 0.3931805204339669,
      "grad_norm": 0.006885293871164322,
      "learning_rate": 0.2973231829941219,
      "loss": 0.354,
      "num_input_tokens_seen": 4585792,
      "step": 2410
    },
    {
      "epoch": 0.39399624765478425,
      "grad_norm": 0.002613081829622388,
      "learning_rate": 0.2973120930672188,
      "loss": 0.3548,
      "num_input_tokens_seen": 4595744,
      "step": 2415
    },
    {
      "epoch": 0.3948119748756016,
      "grad_norm": 0.0022372007369995117,
      "learning_rate": 0.2973009804229397,
      "loss": 0.3909,
      "num_input_tokens_seen": 4606160,
      "step": 2420
    },
    {
      "epoch": 0.39562770209641895,
      "grad_norm": 0.006052400451153517,
      "learning_rate": 0.29728984506299827,
      "loss": 0.3552,
      "num_input_tokens_seen": 4615360,
      "step": 2425
    },
    {
      "epoch": 0.3964434293172363,
      "grad_norm": 0.001876826281659305,
      "learning_rate": 0.2972786869891118,
      "loss": 0.3561,
      "num_input_tokens_seen": 4625216,
      "step": 2430
    },
    {
      "epoch": 0.39725915653805366,
      "grad_norm": 0.0020729934331029654,
      "learning_rate": 0.29726750620300096,
      "loss": 0.3463,
      "num_input_tokens_seen": 4634128,
      "step": 2435
    },
    {
      "epoch": 0.398074883758871,
      "grad_norm": 0.0015269238501787186,
      "learning_rate": 0.29725630270639003,
      "loss": 0.485,
      "num_input_tokens_seen": 4644032,
      "step": 2440
    },
    {
      "epoch": 0.39889061097968836,
      "grad_norm": 0.0018505153711885214,
      "learning_rate": 0.2972450765010067,
      "loss": 0.3369,
      "num_input_tokens_seen": 4654192,
      "step": 2445
    },
    {
      "epoch": 0.3997063382005058,
      "grad_norm": 0.008293497376143932,
      "learning_rate": 0.29723382758858213,
      "loss": 0.4431,
      "num_input_tokens_seen": 4664032,
      "step": 2450
    },
    {
      "epoch": 0.4005220654213231,
      "grad_norm": 0.0020600759889930487,
      "learning_rate": 0.29722255597085107,
      "loss": 0.3889,
      "num_input_tokens_seen": 4673888,
      "step": 2455
    },
    {
      "epoch": 0.4013377926421405,
      "grad_norm": 0.0017217779532074928,
      "learning_rate": 0.2972112616495518,
      "loss": 0.4291,
      "num_input_tokens_seen": 4684064,
      "step": 2460
    },
    {
      "epoch": 0.40215351986295783,
      "grad_norm": 0.001803829101845622,
      "learning_rate": 0.297199944626426,
      "loss": 0.3349,
      "num_input_tokens_seen": 4694560,
      "step": 2465
    },
    {
      "epoch": 0.4029692470837752,
      "grad_norm": 0.008520104922354221,
      "learning_rate": 0.2971886049032189,
      "loss": 0.4001,
      "num_input_tokens_seen": 4703856,
      "step": 2470
    },
    {
      "epoch": 0.40378497430459254,
      "grad_norm": 0.004303759429603815,
      "learning_rate": 0.29717724248167926,
      "loss": 0.3709,
      "num_input_tokens_seen": 4714528,
      "step": 2475
    },
    {
      "epoch": 0.4046007015254099,
      "grad_norm": 0.004198370035737753,
      "learning_rate": 0.29716585736355927,
      "loss": 0.4125,
      "num_input_tokens_seen": 4724480,
      "step": 2480
    },
    {
      "epoch": 0.40541642874622724,
      "grad_norm": 0.005811301525682211,
      "learning_rate": 0.2971544495506147,
      "loss": 0.4421,
      "num_input_tokens_seen": 4732448,
      "step": 2485
    },
    {
      "epoch": 0.4062321559670446,
      "grad_norm": 0.0017361313803121448,
      "learning_rate": 0.2971430190446048,
      "loss": 0.4327,
      "num_input_tokens_seen": 4740752,
      "step": 2490
    },
    {
      "epoch": 0.407047883187862,
      "grad_norm": 0.001237746444530785,
      "learning_rate": 0.2971315658472921,
      "loss": 0.4158,
      "num_input_tokens_seen": 4750336,
      "step": 2495
    },
    {
      "epoch": 0.40786361040867936,
      "grad_norm": 0.002860764740034938,
      "learning_rate": 0.2971200899604431,
      "loss": 0.3405,
      "num_input_tokens_seen": 4759168,
      "step": 2500
    },
    {
      "epoch": 0.4086793376294967,
      "grad_norm": 0.004641514737159014,
      "learning_rate": 0.29710859138582735,
      "loss": 0.433,
      "num_input_tokens_seen": 4768928,
      "step": 2505
    },
    {
      "epoch": 0.40949506485031406,
      "grad_norm": 0.004347140435129404,
      "learning_rate": 0.29709707012521813,
      "loss": 0.3441,
      "num_input_tokens_seen": 4778080,
      "step": 2510
    },
    {
      "epoch": 0.4103107920711314,
      "grad_norm": 0.004527359269559383,
      "learning_rate": 0.29708552618039213,
      "loss": 0.3973,
      "num_input_tokens_seen": 4786432,
      "step": 2515
    },
    {
      "epoch": 0.41112651929194877,
      "grad_norm": 0.001474986900575459,
      "learning_rate": 0.2970739595531296,
      "loss": 0.4455,
      "num_input_tokens_seen": 4795920,
      "step": 2520
    },
    {
      "epoch": 0.4119422465127661,
      "grad_norm": 0.002768178703263402,
      "learning_rate": 0.2970623702452143,
      "loss": 0.354,
      "num_input_tokens_seen": 4805072,
      "step": 2525
    },
    {
      "epoch": 0.41275797373358347,
      "grad_norm": 0.0045648603700101376,
      "learning_rate": 0.2970507582584334,
      "loss": 0.3517,
      "num_input_tokens_seen": 4814896,
      "step": 2530
    },
    {
      "epoch": 0.4135737009544008,
      "grad_norm": 0.0017723022028803825,
      "learning_rate": 0.2970391235945776,
      "loss": 0.3614,
      "num_input_tokens_seen": 4822288,
      "step": 2535
    },
    {
      "epoch": 0.41438942817521823,
      "grad_norm": 0.002291724318638444,
      "learning_rate": 0.2970274662554412,
      "loss": 0.3414,
      "num_input_tokens_seen": 4831152,
      "step": 2540
    },
    {
      "epoch": 0.4152051553960356,
      "grad_norm": 0.0019398248987272382,
      "learning_rate": 0.2970157862428218,
      "loss": 0.3625,
      "num_input_tokens_seen": 4840512,
      "step": 2545
    },
    {
      "epoch": 0.41602088261685294,
      "grad_norm": 0.0025102091021835804,
      "learning_rate": 0.2970040835585206,
      "loss": 0.3245,
      "num_input_tokens_seen": 4850176,
      "step": 2550
    },
    {
      "epoch": 0.4168366098376703,
      "grad_norm": 0.0013439897447824478,
      "learning_rate": 0.2969923582043424,
      "loss": 0.415,
      "num_input_tokens_seen": 4859520,
      "step": 2555
    },
    {
      "epoch": 0.41765233705848764,
      "grad_norm": 0.0023237084969878197,
      "learning_rate": 0.2969806101820953,
      "loss": 0.3838,
      "num_input_tokens_seen": 4869232,
      "step": 2560
    },
    {
      "epoch": 0.418468064279305,
      "grad_norm": 0.0034620079677551985,
      "learning_rate": 0.2969688394935911,
      "loss": 0.4724,
      "num_input_tokens_seen": 4879792,
      "step": 2565
    },
    {
      "epoch": 0.41928379150012235,
      "grad_norm": 0.002062286017462611,
      "learning_rate": 0.2969570461406449,
      "loss": 0.3284,
      "num_input_tokens_seen": 4888192,
      "step": 2570
    },
    {
      "epoch": 0.4200995187209397,
      "grad_norm": 0.004110024776309729,
      "learning_rate": 0.29694523012507534,
      "loss": 0.5284,
      "num_input_tokens_seen": 4898416,
      "step": 2575
    },
    {
      "epoch": 0.42091524594175705,
      "grad_norm": 0.0028117536567151546,
      "learning_rate": 0.2969333914487048,
      "loss": 0.4057,
      "num_input_tokens_seen": 4907280,
      "step": 2580
    },
    {
      "epoch": 0.4217309731625744,
      "grad_norm": 0.0038076965138316154,
      "learning_rate": 0.2969215301133587,
      "loss": 0.3751,
      "num_input_tokens_seen": 4916160,
      "step": 2585
    },
    {
      "epoch": 0.4225467003833918,
      "grad_norm": 0.0024494349490851164,
      "learning_rate": 0.29690964612086634,
      "loss": 0.3834,
      "num_input_tokens_seen": 4924720,
      "step": 2590
    },
    {
      "epoch": 0.42336242760420917,
      "grad_norm": 0.0054839421063661575,
      "learning_rate": 0.2968977394730604,
      "loss": 0.3593,
      "num_input_tokens_seen": 4934784,
      "step": 2595
    },
    {
      "epoch": 0.4241781548250265,
      "grad_norm": 0.0022189058363437653,
      "learning_rate": 0.296885810171777,
      "loss": 0.3596,
      "num_input_tokens_seen": 4944384,
      "step": 2600
    },
    {
      "epoch": 0.4241781548250265,
      "eval_loss": 0.3570125997066498,
      "eval_runtime": 152.2779,
      "eval_samples_per_second": 17.895,
      "eval_steps_per_second": 8.951,
      "num_input_tokens_seen": 4944384,
      "step": 2600
    },
    {
      "epoch": 0.4249938820458439,
      "grad_norm": 0.004021304193884134,
      "learning_rate": 0.2968738582188558,
      "loss": 0.3833,
      "num_input_tokens_seen": 4954240,
      "step": 2605
    },
    {
      "epoch": 0.4258096092666612,
      "grad_norm": 0.001359145506285131,
      "learning_rate": 0.2968618836161399,
      "loss": 0.3357,
      "num_input_tokens_seen": 4963168,
      "step": 2610
    },
    {
      "epoch": 0.4266253364874786,
      "grad_norm": 0.0014179986901581287,
      "learning_rate": 0.296849886365476,
      "loss": 0.4378,
      "num_input_tokens_seen": 4972016,
      "step": 2615
    },
    {
      "epoch": 0.42744106370829593,
      "grad_norm": 0.005261885467916727,
      "learning_rate": 0.2968378664687142,
      "loss": 0.3796,
      "num_input_tokens_seen": 4980992,
      "step": 2620
    },
    {
      "epoch": 0.4282567909291133,
      "grad_norm": 0.002160927513614297,
      "learning_rate": 0.296825823927708,
      "loss": 0.3368,
      "num_input_tokens_seen": 4990784,
      "step": 2625
    },
    {
      "epoch": 0.42907251814993064,
      "grad_norm": 0.002372114220634103,
      "learning_rate": 0.29681375874431476,
      "loss": 0.3333,
      "num_input_tokens_seen": 5000464,
      "step": 2630
    },
    {
      "epoch": 0.42988824537074805,
      "grad_norm": 0.004467688035219908,
      "learning_rate": 0.29680167092039483,
      "loss": 0.3272,
      "num_input_tokens_seen": 5008960,
      "step": 2635
    },
    {
      "epoch": 0.4307039725915654,
      "grad_norm": 0.00209323619492352,
      "learning_rate": 0.2967895604578125,
      "loss": 0.4241,
      "num_input_tokens_seen": 5018944,
      "step": 2640
    },
    {
      "epoch": 0.43151969981238275,
      "grad_norm": 0.0020430677104741335,
      "learning_rate": 0.2967774273584352,
      "loss": 0.3567,
      "num_input_tokens_seen": 5029088,
      "step": 2645
    },
    {
      "epoch": 0.4323354270332001,
      "grad_norm": 0.001688318676315248,
      "learning_rate": 0.2967652716241342,
      "loss": 0.3887,
      "num_input_tokens_seen": 5038528,
      "step": 2650
    },
    {
      "epoch": 0.43315115425401746,
      "grad_norm": 0.0010388713562861085,
      "learning_rate": 0.29675309325678384,
      "loss": 0.3487,
      "num_input_tokens_seen": 5047648,
      "step": 2655
    },
    {
      "epoch": 0.4339668814748348,
      "grad_norm": 0.004077430348843336,
      "learning_rate": 0.29674089225826233,
      "loss": 0.3632,
      "num_input_tokens_seen": 5057168,
      "step": 2660
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.0021537169814109802,
      "learning_rate": 0.29672866863045116,
      "loss": 0.3383,
      "num_input_tokens_seen": 5066560,
      "step": 2665
    },
    {
      "epoch": 0.4355983359164695,
      "grad_norm": 0.003891316708177328,
      "learning_rate": 0.2967164223752354,
      "loss": 0.2872,
      "num_input_tokens_seen": 5076400,
      "step": 2670
    },
    {
      "epoch": 0.43641406313728687,
      "grad_norm": 0.0055531952530145645,
      "learning_rate": 0.2967041534945035,
      "loss": 0.527,
      "num_input_tokens_seen": 5085632,
      "step": 2675
    },
    {
      "epoch": 0.4372297903581043,
      "grad_norm": 0.0009986027143895626,
      "learning_rate": 0.2966918619901476,
      "loss": 0.3156,
      "num_input_tokens_seen": 5094832,
      "step": 2680
    },
    {
      "epoch": 0.43804551757892163,
      "grad_norm": 0.003438690910115838,
      "learning_rate": 0.2966795478640631,
      "loss": 0.4842,
      "num_input_tokens_seen": 5105168,
      "step": 2685
    },
    {
      "epoch": 0.438861244799739,
      "grad_norm": 0.0061935121193528175,
      "learning_rate": 0.29666721111814903,
      "loss": 0.4538,
      "num_input_tokens_seen": 5114240,
      "step": 2690
    },
    {
      "epoch": 0.43967697202055633,
      "grad_norm": 0.004904442001134157,
      "learning_rate": 0.2966548517543079,
      "loss": 0.3647,
      "num_input_tokens_seen": 5122960,
      "step": 2695
    },
    {
      "epoch": 0.4404926992413737,
      "grad_norm": 0.004085817374289036,
      "learning_rate": 0.29664246977444564,
      "loss": 0.3858,
      "num_input_tokens_seen": 5130896,
      "step": 2700
    },
    {
      "epoch": 0.44130842646219104,
      "grad_norm": 0.0027817634399980307,
      "learning_rate": 0.2966300651804717,
      "loss": 0.321,
      "num_input_tokens_seen": 5141072,
      "step": 2705
    },
    {
      "epoch": 0.4421241536830084,
      "grad_norm": 0.002625055843964219,
      "learning_rate": 0.296617637974299,
      "loss": 0.3828,
      "num_input_tokens_seen": 5149536,
      "step": 2710
    },
    {
      "epoch": 0.44293988090382574,
      "grad_norm": 0.0018178208265453577,
      "learning_rate": 0.2966051881578441,
      "loss": 0.3686,
      "num_input_tokens_seen": 5158944,
      "step": 2715
    },
    {
      "epoch": 0.4437556081246431,
      "grad_norm": 0.0038589772302657366,
      "learning_rate": 0.29659271573302676,
      "loss": 0.3606,
      "num_input_tokens_seen": 5169472,
      "step": 2720
    },
    {
      "epoch": 0.4445713353454605,
      "grad_norm": 0.0006516375578939915,
      "learning_rate": 0.2965802207017705,
      "loss": 0.3439,
      "num_input_tokens_seen": 5179376,
      "step": 2725
    },
    {
      "epoch": 0.44538706256627786,
      "grad_norm": 0.0015188453253358603,
      "learning_rate": 0.2965677030660021,
      "loss": 0.3495,
      "num_input_tokens_seen": 5188192,
      "step": 2730
    },
    {
      "epoch": 0.4462027897870952,
      "grad_norm": 0.0005723736248910427,
      "learning_rate": 0.2965551628276521,
      "loss": 0.3567,
      "num_input_tokens_seen": 5198320,
      "step": 2735
    },
    {
      "epoch": 0.44701851700791256,
      "grad_norm": 0.002675766358152032,
      "learning_rate": 0.29654259998865423,
      "loss": 0.3631,
      "num_input_tokens_seen": 5207872,
      "step": 2740
    },
    {
      "epoch": 0.4478342442287299,
      "grad_norm": 0.0030829254537820816,
      "learning_rate": 0.2965300145509458,
      "loss": 0.2824,
      "num_input_tokens_seen": 5216432,
      "step": 2745
    },
    {
      "epoch": 0.44864997144954727,
      "grad_norm": 0.004724303726106882,
      "learning_rate": 0.2965174065164678,
      "loss": 0.406,
      "num_input_tokens_seen": 5227024,
      "step": 2750
    },
    {
      "epoch": 0.4494656986703646,
      "grad_norm": 0.003439545864239335,
      "learning_rate": 0.2965047758871644,
      "loss": 0.2729,
      "num_input_tokens_seen": 5236928,
      "step": 2755
    },
    {
      "epoch": 0.450281425891182,
      "grad_norm": 0.0029972284100949764,
      "learning_rate": 0.2964921226649835,
      "loss": 0.4798,
      "num_input_tokens_seen": 5247120,
      "step": 2760
    },
    {
      "epoch": 0.4510971531119993,
      "grad_norm": 0.004546004347503185,
      "learning_rate": 0.2964794468518763,
      "loss": 0.4511,
      "num_input_tokens_seen": 5257136,
      "step": 2765
    },
    {
      "epoch": 0.4519128803328167,
      "grad_norm": 0.0026837282348424196,
      "learning_rate": 0.2964667484497977,
      "loss": 0.3578,
      "num_input_tokens_seen": 5266528,
      "step": 2770
    },
    {
      "epoch": 0.4527286075536341,
      "grad_norm": 0.001055366825312376,
      "learning_rate": 0.29645402746070587,
      "loss": 0.365,
      "num_input_tokens_seen": 5276640,
      "step": 2775
    },
    {
      "epoch": 0.45354433477445144,
      "grad_norm": 0.0020235616248100996,
      "learning_rate": 0.2964412838865625,
      "loss": 0.3416,
      "num_input_tokens_seen": 5286816,
      "step": 2780
    },
    {
      "epoch": 0.4543600619952688,
      "grad_norm": 0.003819419536739588,
      "learning_rate": 0.29642851772933293,
      "loss": 0.3457,
      "num_input_tokens_seen": 5296752,
      "step": 2785
    },
    {
      "epoch": 0.45517578921608615,
      "grad_norm": 0.0034626913256943226,
      "learning_rate": 0.29641572899098567,
      "loss": 0.3545,
      "num_input_tokens_seen": 5305216,
      "step": 2790
    },
    {
      "epoch": 0.4559915164369035,
      "grad_norm": 0.002586331218481064,
      "learning_rate": 0.29640291767349314,
      "loss": 0.3475,
      "num_input_tokens_seen": 5316032,
      "step": 2795
    },
    {
      "epoch": 0.45680724365772085,
      "grad_norm": 0.0025320437271147966,
      "learning_rate": 0.2963900837788308,
      "loss": 0.3747,
      "num_input_tokens_seen": 5325216,
      "step": 2800
    },
    {
      "epoch": 0.45680724365772085,
      "eval_loss": 0.3450164794921875,
      "eval_runtime": 152.3119,
      "eval_samples_per_second": 17.891,
      "eval_steps_per_second": 8.949,
      "num_input_tokens_seen": 5325216,
      "step": 2800
    },
    {
      "epoch": 0.4576229708785382,
      "grad_norm": 0.0020419969223439693,
      "learning_rate": 0.2963772273089779,
      "loss": 0.394,
      "num_input_tokens_seen": 5334016,
      "step": 2805
    },
    {
      "epoch": 0.45843869809935556,
      "grad_norm": 0.004899526480585337,
      "learning_rate": 0.2963643482659171,
      "loss": 0.3497,
      "num_input_tokens_seen": 5342672,
      "step": 2810
    },
    {
      "epoch": 0.4592544253201729,
      "grad_norm": 0.005618786904960871,
      "learning_rate": 0.2963514466516345,
      "loss": 0.3869,
      "num_input_tokens_seen": 5351328,
      "step": 2815
    },
    {
      "epoch": 0.4600701525409903,
      "grad_norm": 0.0009065830381587148,
      "learning_rate": 0.2963385224681196,
      "loss": 0.3686,
      "num_input_tokens_seen": 5361520,
      "step": 2820
    },
    {
      "epoch": 0.46088587976180767,
      "grad_norm": 0.0021144705824553967,
      "learning_rate": 0.29632557571736556,
      "loss": 0.3619,
      "num_input_tokens_seen": 5371184,
      "step": 2825
    },
    {
      "epoch": 0.461701606982625,
      "grad_norm": 0.0017334030708298087,
      "learning_rate": 0.2963126064013689,
      "loss": 0.3511,
      "num_input_tokens_seen": 5380704,
      "step": 2830
    },
    {
      "epoch": 0.4625173342034424,
      "grad_norm": 0.0027768868021667004,
      "learning_rate": 0.29629961452212966,
      "loss": 0.509,
      "num_input_tokens_seen": 5389584,
      "step": 2835
    },
    {
      "epoch": 0.46333306142425973,
      "grad_norm": 0.0035777813754975796,
      "learning_rate": 0.2962866000816513,
      "loss": 0.4011,
      "num_input_tokens_seen": 5399312,
      "step": 2840
    },
    {
      "epoch": 0.4641487886450771,
      "grad_norm": 0.00408755661919713,
      "learning_rate": 0.2962735630819409,
      "loss": 0.3813,
      "num_input_tokens_seen": 5408912,
      "step": 2845
    },
    {
      "epoch": 0.46496451586589443,
      "grad_norm": 0.005223833955824375,
      "learning_rate": 0.2962605035250089,
      "loss": 0.4246,
      "num_input_tokens_seen": 5418512,
      "step": 2850
    },
    {
      "epoch": 0.4657802430867118,
      "grad_norm": 0.004365354310721159,
      "learning_rate": 0.29624742141286914,
      "loss": 0.4157,
      "num_input_tokens_seen": 5428288,
      "step": 2855
    },
    {
      "epoch": 0.46659597030752914,
      "grad_norm": 0.003029085695743561,
      "learning_rate": 0.29623431674753925,
      "loss": 0.3669,
      "num_input_tokens_seen": 5437888,
      "step": 2860
    },
    {
      "epoch": 0.46741169752834655,
      "grad_norm": 0.0015382225392386317,
      "learning_rate": 0.29622118953103993,
      "loss": 0.345,
      "num_input_tokens_seen": 5446656,
      "step": 2865
    },
    {
      "epoch": 0.4682274247491639,
      "grad_norm": 0.0007544420077465475,
      "learning_rate": 0.2962080397653957,
      "loss": 0.4004,
      "num_input_tokens_seen": 5455696,
      "step": 2870
    },
    {
      "epoch": 0.46904315196998125,
      "grad_norm": 0.0017960118129849434,
      "learning_rate": 0.29619486745263435,
      "loss": 0.315,
      "num_input_tokens_seen": 5465056,
      "step": 2875
    },
    {
      "epoch": 0.4698588791907986,
      "grad_norm": 0.0011463663540780544,
      "learning_rate": 0.2961816725947873,
      "loss": 0.4367,
      "num_input_tokens_seen": 5474560,
      "step": 2880
    },
    {
      "epoch": 0.47067460641161596,
      "grad_norm": 0.002609727904200554,
      "learning_rate": 0.29616845519388924,
      "loss": 0.4196,
      "num_input_tokens_seen": 5483584,
      "step": 2885
    },
    {
      "epoch": 0.4714903336324333,
      "grad_norm": 0.002055482240393758,
      "learning_rate": 0.2961552152519785,
      "loss": 0.3975,
      "num_input_tokens_seen": 5492112,
      "step": 2890
    },
    {
      "epoch": 0.47230606085325066,
      "grad_norm": 0.0016856398433446884,
      "learning_rate": 0.29614195277109695,
      "loss": 0.4517,
      "num_input_tokens_seen": 5501536,
      "step": 2895
    },
    {
      "epoch": 0.473121788074068,
      "grad_norm": 0.0036685203667730093,
      "learning_rate": 0.2961286677532897,
      "loss": 0.4028,
      "num_input_tokens_seen": 5511280,
      "step": 2900
    },
    {
      "epoch": 0.47393751529488537,
      "grad_norm": 0.004363181069493294,
      "learning_rate": 0.2961153602006055,
      "loss": 0.336,
      "num_input_tokens_seen": 5520016,
      "step": 2905
    },
    {
      "epoch": 0.4747532425157027,
      "grad_norm": 0.0034700087271630764,
      "learning_rate": 0.29610203011509656,
      "loss": 0.3626,
      "num_input_tokens_seen": 5529808,
      "step": 2910
    },
    {
      "epoch": 0.47556896973652013,
      "grad_norm": 0.0014970524935051799,
      "learning_rate": 0.29608867749881856,
      "loss": 0.3178,
      "num_input_tokens_seen": 5538896,
      "step": 2915
    },
    {
      "epoch": 0.4763846969573375,
      "grad_norm": 0.0018269852735102177,
      "learning_rate": 0.29607530235383067,
      "loss": 0.4484,
      "num_input_tokens_seen": 5548112,
      "step": 2920
    },
    {
      "epoch": 0.47720042417815484,
      "grad_norm": 0.0006858830456621945,
      "learning_rate": 0.2960619046821954,
      "loss": 0.3515,
      "num_input_tokens_seen": 5557504,
      "step": 2925
    },
    {
      "epoch": 0.4780161513989722,
      "grad_norm": 0.0040230173617601395,
      "learning_rate": 0.2960484844859789,
      "loss": 0.4887,
      "num_input_tokens_seen": 5566208,
      "step": 2930
    },
    {
      "epoch": 0.47883187861978954,
      "grad_norm": 0.0012979180319234729,
      "learning_rate": 0.29603504176725076,
      "loss": 0.4304,
      "num_input_tokens_seen": 5576176,
      "step": 2935
    },
    {
      "epoch": 0.4796476058406069,
      "grad_norm": 0.0021733329631388187,
      "learning_rate": 0.296021576528084,
      "loss": 0.3825,
      "num_input_tokens_seen": 5584688,
      "step": 2940
    },
    {
      "epoch": 0.48046333306142425,
      "grad_norm": 0.0009058294235728681,
      "learning_rate": 0.29600808877055507,
      "loss": 0.401,
      "num_input_tokens_seen": 5592784,
      "step": 2945
    },
    {
      "epoch": 0.4812790602822416,
      "grad_norm": 0.002800081390887499,
      "learning_rate": 0.29599457849674404,
      "loss": 0.3519,
      "num_input_tokens_seen": 5602336,
      "step": 2950
    },
    {
      "epoch": 0.48209478750305895,
      "grad_norm": 0.0031082837376743555,
      "learning_rate": 0.2959810457087343,
      "loss": 0.377,
      "num_input_tokens_seen": 5612160,
      "step": 2955
    },
    {
      "epoch": 0.48291051472387636,
      "grad_norm": 0.0022484466899186373,
      "learning_rate": 0.2959674904086128,
      "loss": 0.3391,
      "num_input_tokens_seen": 5621392,
      "step": 2960
    },
    {
      "epoch": 0.4837262419446937,
      "grad_norm": 0.0018040946451947093,
      "learning_rate": 0.2959539125984699,
      "loss": 0.3103,
      "num_input_tokens_seen": 5631696,
      "step": 2965
    },
    {
      "epoch": 0.48454196916551107,
      "grad_norm": 0.0015286856796592474,
      "learning_rate": 0.2959403122803996,
      "loss": 0.3487,
      "num_input_tokens_seen": 5640880,
      "step": 2970
    },
    {
      "epoch": 0.4853576963863284,
      "grad_norm": 0.002592851873487234,
      "learning_rate": 0.2959266894564991,
      "loss": 0.3887,
      "num_input_tokens_seen": 5650368,
      "step": 2975
    },
    {
      "epoch": 0.48617342360714577,
      "grad_norm": 0.0015772442566230893,
      "learning_rate": 0.2959130441288692,
      "loss": 0.3395,
      "num_input_tokens_seen": 5660960,
      "step": 2980
    },
    {
      "epoch": 0.4869891508279631,
      "grad_norm": 0.0016279423143714666,
      "learning_rate": 0.2958993762996143,
      "loss": 0.3549,
      "num_input_tokens_seen": 5670592,
      "step": 2985
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.002887737937271595,
      "learning_rate": 0.2958856859708421,
      "loss": 0.3318,
      "num_input_tokens_seen": 5679360,
      "step": 2990
    },
    {
      "epoch": 0.48862060526959783,
      "grad_norm": 0.002777362009510398,
      "learning_rate": 0.2958719731446638,
      "loss": 0.3692,
      "num_input_tokens_seen": 5689264,
      "step": 2995
    },
    {
      "epoch": 0.4894363324904152,
      "grad_norm": 0.0021030870266258717,
      "learning_rate": 0.29585823782319404,
      "loss": 0.3322,
      "num_input_tokens_seen": 5698896,
      "step": 3000
    },
    {
      "epoch": 0.4894363324904152,
      "eval_loss": 0.4611145555973053,
      "eval_runtime": 152.3676,
      "eval_samples_per_second": 17.884,
      "eval_steps_per_second": 8.945,
      "num_input_tokens_seen": 5698896,
      "step": 3000
    },
    {
      "epoch": 0.4902520597112326,
      "grad_norm": 0.0009577450691722333,
      "learning_rate": 0.2958444800085511,
      "loss": 0.4883,
      "num_input_tokens_seen": 5708816,
      "step": 3005
    },
    {
      "epoch": 0.49106778693204994,
      "grad_norm": 0.003624190576374531,
      "learning_rate": 0.2958306997028565,
      "loss": 0.3456,
      "num_input_tokens_seen": 5718384,
      "step": 3010
    },
    {
      "epoch": 0.4918835141528673,
      "grad_norm": 0.0037653623148798943,
      "learning_rate": 0.2958168969082354,
      "loss": 0.3456,
      "num_input_tokens_seen": 5727120,
      "step": 3015
    },
    {
      "epoch": 0.49269924137368465,
      "grad_norm": 0.004122159909456968,
      "learning_rate": 0.2958030716268164,
      "loss": 0.4117,
      "num_input_tokens_seen": 5735888,
      "step": 3020
    },
    {
      "epoch": 0.493514968594502,
      "grad_norm": 0.003987944684922695,
      "learning_rate": 0.2957892238607314,
      "loss": 0.3884,
      "num_input_tokens_seen": 5744672,
      "step": 3025
    },
    {
      "epoch": 0.49433069581531935,
      "grad_norm": 0.0013585133710876107,
      "learning_rate": 0.2957753536121161,
      "loss": 0.377,
      "num_input_tokens_seen": 5753984,
      "step": 3030
    },
    {
      "epoch": 0.4951464230361367,
      "grad_norm": 0.0029268248472362757,
      "learning_rate": 0.29576146088310923,
      "loss": 0.3745,
      "num_input_tokens_seen": 5763472,
      "step": 3035
    },
    {
      "epoch": 0.49596215025695406,
      "grad_norm": 0.002193065593019128,
      "learning_rate": 0.2957475456758533,
      "loss": 0.3794,
      "num_input_tokens_seen": 5773440,
      "step": 3040
    },
    {
      "epoch": 0.4967778774777714,
      "grad_norm": 0.0024910708889365196,
      "learning_rate": 0.2957336079924944,
      "loss": 0.3546,
      "num_input_tokens_seen": 5783664,
      "step": 3045
    },
    {
      "epoch": 0.49759360469858877,
      "grad_norm": 0.0026398906484246254,
      "learning_rate": 0.2957196478351816,
      "loss": 0.3476,
      "num_input_tokens_seen": 5793872,
      "step": 3050
    },
    {
      "epoch": 0.4984093319194062,
      "grad_norm": 0.0011744144139811397,
      "learning_rate": 0.295705665206068,
      "loss": 0.3513,
      "num_input_tokens_seen": 5803600,
      "step": 3055
    },
    {
      "epoch": 0.4992250591402235,
      "grad_norm": 0.0033658018801361322,
      "learning_rate": 0.2956916601073097,
      "loss": 0.2831,
      "num_input_tokens_seen": 5812320,
      "step": 3060
    },
    {
      "epoch": 0.5000407863610409,
      "grad_norm": 0.003561178455129266,
      "learning_rate": 0.29567763254106655,
      "loss": 0.367,
      "num_input_tokens_seen": 5821408,
      "step": 3065
    },
    {
      "epoch": 0.5008565135818582,
      "grad_norm": 0.0016328580677509308,
      "learning_rate": 0.29566358250950175,
      "loss": 0.4145,
      "num_input_tokens_seen": 5830752,
      "step": 3070
    },
    {
      "epoch": 0.5016722408026756,
      "grad_norm": 0.006045615300536156,
      "learning_rate": 0.295649510014782,
      "loss": 0.4288,
      "num_input_tokens_seen": 5839616,
      "step": 3075
    },
    {
      "epoch": 0.5024879680234929,
      "grad_norm": 0.0009149805991910398,
      "learning_rate": 0.2956354150590775,
      "loss": 0.3473,
      "num_input_tokens_seen": 5848800,
      "step": 3080
    },
    {
      "epoch": 0.5033036952443103,
      "grad_norm": 0.0008091272320598364,
      "learning_rate": 0.2956212976445618,
      "loss": 0.355,
      "num_input_tokens_seen": 5857920,
      "step": 3085
    },
    {
      "epoch": 0.5041194224651276,
      "grad_norm": 0.004103023558855057,
      "learning_rate": 0.295607157773412,
      "loss": 0.3414,
      "num_input_tokens_seen": 5866624,
      "step": 3090
    },
    {
      "epoch": 0.504935149685945,
      "grad_norm": 0.005884118843823671,
      "learning_rate": 0.2955929954478087,
      "loss": 0.3874,
      "num_input_tokens_seen": 5876240,
      "step": 3095
    },
    {
      "epoch": 0.5057508769067623,
      "grad_norm": 0.0030214532744139433,
      "learning_rate": 0.29557881066993585,
      "loss": 0.365,
      "num_input_tokens_seen": 5885472,
      "step": 3100
    },
    {
      "epoch": 0.5065666041275797,
      "grad_norm": 0.0028707999736070633,
      "learning_rate": 0.29556460344198093,
      "loss": 0.308,
      "num_input_tokens_seen": 5895488,
      "step": 3105
    },
    {
      "epoch": 0.507382331348397,
      "grad_norm": 0.003483733395114541,
      "learning_rate": 0.29555037376613486,
      "loss": 0.3943,
      "num_input_tokens_seen": 5905840,
      "step": 3110
    },
    {
      "epoch": 0.5081980585692144,
      "grad_norm": 0.0016888321842998266,
      "learning_rate": 0.29553612164459203,
      "loss": 0.3627,
      "num_input_tokens_seen": 5915280,
      "step": 3115
    },
    {
      "epoch": 0.5090137857900318,
      "grad_norm": 0.002153498586267233,
      "learning_rate": 0.29552184707955037,
      "loss": 0.3615,
      "num_input_tokens_seen": 5924784,
      "step": 3120
    },
    {
      "epoch": 0.5098295130108492,
      "grad_norm": 0.0009719410445541143,
      "learning_rate": 0.29550755007321117,
      "loss": 0.3583,
      "num_input_tokens_seen": 5934432,
      "step": 3125
    },
    {
      "epoch": 0.5106452402316666,
      "grad_norm": 0.003338263835757971,
      "learning_rate": 0.29549323062777916,
      "loss": 0.3681,
      "num_input_tokens_seen": 5943360,
      "step": 3130
    },
    {
      "epoch": 0.5114609674524839,
      "grad_norm": 0.0032980393152683973,
      "learning_rate": 0.29547888874546263,
      "loss": 0.3066,
      "num_input_tokens_seen": 5952016,
      "step": 3135
    },
    {
      "epoch": 0.5122766946733013,
      "grad_norm": 0.0015422884607687593,
      "learning_rate": 0.2954645244284732,
      "loss": 0.3828,
      "num_input_tokens_seen": 5961568,
      "step": 3140
    },
    {
      "epoch": 0.5130924218941186,
      "grad_norm": 0.0034292142372578382,
      "learning_rate": 0.2954501376790261,
      "loss": 0.3633,
      "num_input_tokens_seen": 5970544,
      "step": 3145
    },
    {
      "epoch": 0.513908149114936,
      "grad_norm": 0.0017311143456026912,
      "learning_rate": 0.29543572849933997,
      "loss": 0.3593,
      "num_input_tokens_seen": 5978896,
      "step": 3150
    },
    {
      "epoch": 0.5147238763357533,
      "grad_norm": 0.0012415040982887149,
      "learning_rate": 0.2954212968916368,
      "loss": 0.3517,
      "num_input_tokens_seen": 5986976,
      "step": 3155
    },
    {
      "epoch": 0.5155396035565707,
      "grad_norm": 0.0016910429112613201,
      "learning_rate": 0.29540684285814217,
      "loss": 0.3393,
      "num_input_tokens_seen": 5995632,
      "step": 3160
    },
    {
      "epoch": 0.516355330777388,
      "grad_norm": 0.0014186527114361525,
      "learning_rate": 0.2953923664010851,
      "loss": 0.3661,
      "num_input_tokens_seen": 6005728,
      "step": 3165
    },
    {
      "epoch": 0.5171710579982054,
      "grad_norm": 0.003047883976250887,
      "learning_rate": 0.295377867522698,
      "loss": 0.3656,
      "num_input_tokens_seen": 6015856,
      "step": 3170
    },
    {
      "epoch": 0.5179867852190227,
      "grad_norm": 0.007520120590925217,
      "learning_rate": 0.2953633462252168,
      "loss": 0.3859,
      "num_input_tokens_seen": 6025088,
      "step": 3175
    },
    {
      "epoch": 0.5188025124398401,
      "grad_norm": 0.0024808263406157494,
      "learning_rate": 0.2953488025108809,
      "loss": 0.3702,
      "num_input_tokens_seen": 6034352,
      "step": 3180
    },
    {
      "epoch": 0.5196182396606575,
      "grad_norm": 0.0031205685809254646,
      "learning_rate": 0.295334236381933,
      "loss": 0.3965,
      "num_input_tokens_seen": 6044320,
      "step": 3185
    },
    {
      "epoch": 0.5204339668814748,
      "grad_norm": 0.0024575567804276943,
      "learning_rate": 0.29531964784061954,
      "loss": 0.3621,
      "num_input_tokens_seen": 6055072,
      "step": 3190
    },
    {
      "epoch": 0.5212496941022922,
      "grad_norm": 0.0021379648242145777,
      "learning_rate": 0.2953050368891902,
      "loss": 0.3454,
      "num_input_tokens_seen": 6064688,
      "step": 3195
    },
    {
      "epoch": 0.5220654213231095,
      "grad_norm": 0.0008532819338142872,
      "learning_rate": 0.29529040352989805,
      "loss": 0.3631,
      "num_input_tokens_seen": 6074432,
      "step": 3200
    },
    {
      "epoch": 0.5220654213231095,
      "eval_loss": 0.35240375995635986,
      "eval_runtime": 152.4078,
      "eval_samples_per_second": 17.88,
      "eval_steps_per_second": 8.943,
      "num_input_tokens_seen": 6074432,
      "step": 3200
    },
    {
      "epoch": 0.5228811485439269,
      "grad_norm": 0.001258724951185286,
      "learning_rate": 0.29527574776499993,
      "loss": 0.3533,
      "num_input_tokens_seen": 6084256,
      "step": 3205
    },
    {
      "epoch": 0.5236968757647442,
      "grad_norm": 0.0015087767969816923,
      "learning_rate": 0.2952610695967558,
      "loss": 0.3702,
      "num_input_tokens_seen": 6093616,
      "step": 3210
    },
    {
      "epoch": 0.5245126029855617,
      "grad_norm": 0.0032248185016214848,
      "learning_rate": 0.29524636902742935,
      "loss": 0.3734,
      "num_input_tokens_seen": 6103184,
      "step": 3215
    },
    {
      "epoch": 0.525328330206379,
      "grad_norm": 0.0037195775657892227,
      "learning_rate": 0.2952316460592875,
      "loss": 0.3672,
      "num_input_tokens_seen": 6111728,
      "step": 3220
    },
    {
      "epoch": 0.5261440574271964,
      "grad_norm": 0.0005406916025094688,
      "learning_rate": 0.29521690069460066,
      "loss": 0.3801,
      "num_input_tokens_seen": 6121328,
      "step": 3225
    },
    {
      "epoch": 0.5269597846480137,
      "grad_norm": 0.0009773432975634933,
      "learning_rate": 0.29520213293564285,
      "loss": 0.3622,
      "num_input_tokens_seen": 6130496,
      "step": 3230
    },
    {
      "epoch": 0.5277755118688311,
      "grad_norm": 0.0034450890962034464,
      "learning_rate": 0.29518734278469144,
      "loss": 0.2799,
      "num_input_tokens_seen": 6139344,
      "step": 3235
    },
    {
      "epoch": 0.5285912390896484,
      "grad_norm": 0.003888080595061183,
      "learning_rate": 0.29517253024402723,
      "loss": 0.361,
      "num_input_tokens_seen": 6149280,
      "step": 3240
    },
    {
      "epoch": 0.5294069663104658,
      "grad_norm": 0.0034895832650363445,
      "learning_rate": 0.2951576953159345,
      "loss": 0.2914,
      "num_input_tokens_seen": 6158224,
      "step": 3245
    },
    {
      "epoch": 0.5302226935312832,
      "grad_norm": 0.004313875921070576,
      "learning_rate": 0.29514283800270097,
      "loss": 0.3156,
      "num_input_tokens_seen": 6167648,
      "step": 3250
    },
    {
      "epoch": 0.5310384207521005,
      "grad_norm": 0.002685636281967163,
      "learning_rate": 0.2951279583066179,
      "loss": 0.4166,
      "num_input_tokens_seen": 6177600,
      "step": 3255
    },
    {
      "epoch": 0.5318541479729179,
      "grad_norm": 0.0009063413599506021,
      "learning_rate": 0.2951130562299798,
      "loss": 0.3436,
      "num_input_tokens_seen": 6186688,
      "step": 3260
    },
    {
      "epoch": 0.5326698751937352,
      "grad_norm": 0.0011001609964296222,
      "learning_rate": 0.29509813177508487,
      "loss": 0.3708,
      "num_input_tokens_seen": 6196384,
      "step": 3265
    },
    {
      "epoch": 0.5334856024145526,
      "grad_norm": 0.0008024419657886028,
      "learning_rate": 0.2950831849442346,
      "loss": 0.3641,
      "num_input_tokens_seen": 6205760,
      "step": 3270
    },
    {
      "epoch": 0.5343013296353699,
      "grad_norm": 0.004846091847866774,
      "learning_rate": 0.2950682157397339,
      "loss": 0.3559,
      "num_input_tokens_seen": 6215584,
      "step": 3275
    },
    {
      "epoch": 0.5351170568561873,
      "grad_norm": 0.0014480192912742496,
      "learning_rate": 0.2950532241638914,
      "loss": 0.4135,
      "num_input_tokens_seen": 6226016,
      "step": 3280
    },
    {
      "epoch": 0.5359327840770046,
      "grad_norm": 0.0023437386844307184,
      "learning_rate": 0.2950382102190188,
      "loss": 0.3553,
      "num_input_tokens_seen": 6236080,
      "step": 3285
    },
    {
      "epoch": 0.536748511297822,
      "grad_norm": 0.004378148820251226,
      "learning_rate": 0.2950231739074316,
      "loss": 0.3383,
      "num_input_tokens_seen": 6245056,
      "step": 3290
    },
    {
      "epoch": 0.5375642385186393,
      "grad_norm": 0.00246453401632607,
      "learning_rate": 0.29500811523144843,
      "loss": 0.3657,
      "num_input_tokens_seen": 6254224,
      "step": 3295
    },
    {
      "epoch": 0.5383799657394567,
      "grad_norm": 0.0012909513898193836,
      "learning_rate": 0.2949930341933917,
      "loss": 0.3743,
      "num_input_tokens_seen": 6263952,
      "step": 3300
    },
    {
      "epoch": 0.539195692960274,
      "grad_norm": 0.002099535893648863,
      "learning_rate": 0.29497793079558693,
      "loss": 0.3427,
      "num_input_tokens_seen": 6274096,
      "step": 3305
    },
    {
      "epoch": 0.5400114201810915,
      "grad_norm": 0.0012734553311020136,
      "learning_rate": 0.2949628050403633,
      "loss": 0.304,
      "num_input_tokens_seen": 6283904,
      "step": 3310
    },
    {
      "epoch": 0.5408271474019088,
      "grad_norm": 0.0011824715184047818,
      "learning_rate": 0.2949476569300535,
      "loss": 0.1891,
      "num_input_tokens_seen": 6293648,
      "step": 3315
    },
    {
      "epoch": 0.5416428746227262,
      "grad_norm": 0.0034635793417692184,
      "learning_rate": 0.29493248646699344,
      "loss": 0.5279,
      "num_input_tokens_seen": 6302720,
      "step": 3320
    },
    {
      "epoch": 0.5424586018435436,
      "grad_norm": 0.0013978006318211555,
      "learning_rate": 0.29491729365352265,
      "loss": 0.3865,
      "num_input_tokens_seen": 6312240,
      "step": 3325
    },
    {
      "epoch": 0.5432743290643609,
      "grad_norm": 0.0033100643195211887,
      "learning_rate": 0.29490207849198397,
      "loss": 0.4361,
      "num_input_tokens_seen": 6321312,
      "step": 3330
    },
    {
      "epoch": 0.5440900562851783,
      "grad_norm": 0.0008348653791472316,
      "learning_rate": 0.29488684098472384,
      "loss": 0.3976,
      "num_input_tokens_seen": 6330512,
      "step": 3335
    },
    {
      "epoch": 0.5449057835059956,
      "grad_norm": 0.0010373081313446164,
      "learning_rate": 0.2948715811340921,
      "loss": 0.3684,
      "num_input_tokens_seen": 6339872,
      "step": 3340
    },
    {
      "epoch": 0.545721510726813,
      "grad_norm": 0.0012093806872144341,
      "learning_rate": 0.294856298942442,
      "loss": 0.4588,
      "num_input_tokens_seen": 6349488,
      "step": 3345
    },
    {
      "epoch": 0.5465372379476303,
      "grad_norm": 0.0009683164535090327,
      "learning_rate": 0.2948409944121302,
      "loss": 0.3208,
      "num_input_tokens_seen": 6358480,
      "step": 3350
    },
    {
      "epoch": 0.5473529651684477,
      "grad_norm": 0.003701448207721114,
      "learning_rate": 0.29482566754551687,
      "loss": 0.4819,
      "num_input_tokens_seen": 6367264,
      "step": 3355
    },
    {
      "epoch": 0.548168692389265,
      "grad_norm": 0.002605573972687125,
      "learning_rate": 0.2948103183449656,
      "loss": 0.404,
      "num_input_tokens_seen": 6376784,
      "step": 3360
    },
    {
      "epoch": 0.5489844196100824,
      "grad_norm": 0.006992770824581385,
      "learning_rate": 0.2947949468128435,
      "loss": 0.3468,
      "num_input_tokens_seen": 6386336,
      "step": 3365
    },
    {
      "epoch": 0.5498001468308997,
      "grad_norm": 0.0019361565355211496,
      "learning_rate": 0.2947795529515209,
      "loss": 0.4516,
      "num_input_tokens_seen": 6395600,
      "step": 3370
    },
    {
      "epoch": 0.5506158740517171,
      "grad_norm": 0.002226795768365264,
      "learning_rate": 0.29476413676337193,
      "loss": 0.3693,
      "num_input_tokens_seen": 6405280,
      "step": 3375
    },
    {
      "epoch": 0.5514316012725344,
      "grad_norm": 0.0013331485679373145,
      "learning_rate": 0.2947486982507738,
      "loss": 0.3668,
      "num_input_tokens_seen": 6415184,
      "step": 3380
    },
    {
      "epoch": 0.5522473284933518,
      "grad_norm": 0.001479955855756998,
      "learning_rate": 0.29473323741610735,
      "loss": 0.3621,
      "num_input_tokens_seen": 6425984,
      "step": 3385
    },
    {
      "epoch": 0.5530630557141691,
      "grad_norm": 0.0018132368568331003,
      "learning_rate": 0.2947177542617569,
      "loss": 0.3701,
      "num_input_tokens_seen": 6435952,
      "step": 3390
    },
    {
      "epoch": 0.5538787829349865,
      "grad_norm": 0.002014352008700371,
      "learning_rate": 0.2947022487901101,
      "loss": 0.3606,
      "num_input_tokens_seen": 6445472,
      "step": 3395
    },
    {
      "epoch": 0.5546945101558038,
      "grad_norm": 0.005138829350471497,
      "learning_rate": 0.2946867210035581,
      "loss": 0.357,
      "num_input_tokens_seen": 6454208,
      "step": 3400
    },
    {
      "epoch": 0.5546945101558038,
      "eval_loss": 0.35230737924575806,
      "eval_runtime": 152.3709,
      "eval_samples_per_second": 17.884,
      "eval_steps_per_second": 8.945,
      "num_input_tokens_seen": 6454208,
      "step": 3400
    },
    {
      "epoch": 0.5555102373766213,
      "grad_norm": 0.001262438134290278,
      "learning_rate": 0.2946711709044954,
      "loss": 0.3818,
      "num_input_tokens_seen": 6463312,
      "step": 3405
    },
    {
      "epoch": 0.5563259645974387,
      "grad_norm": 0.005211641080677509,
      "learning_rate": 0.2946555984953202,
      "loss": 0.3738,
      "num_input_tokens_seen": 6471632,
      "step": 3410
    },
    {
      "epoch": 0.557141691818256,
      "grad_norm": 0.0012375812511891127,
      "learning_rate": 0.2946400037784338,
      "loss": 0.3613,
      "num_input_tokens_seen": 6480976,
      "step": 3415
    },
    {
      "epoch": 0.5579574190390734,
      "grad_norm": 0.0034810530487447977,
      "learning_rate": 0.29462438675624114,
      "loss": 0.345,
      "num_input_tokens_seen": 6490832,
      "step": 3420
    },
    {
      "epoch": 0.5587731462598907,
      "grad_norm": 0.00196433044038713,
      "learning_rate": 0.2946087474311506,
      "loss": 0.3687,
      "num_input_tokens_seen": 6500112,
      "step": 3425
    },
    {
      "epoch": 0.5595888734807081,
      "grad_norm": 0.002418990945443511,
      "learning_rate": 0.294593085805574,
      "loss": 0.4318,
      "num_input_tokens_seen": 6508992,
      "step": 3430
    },
    {
      "epoch": 0.5604046007015254,
      "grad_norm": 0.0018725753761827946,
      "learning_rate": 0.2945774018819264,
      "loss": 0.4131,
      "num_input_tokens_seen": 6518976,
      "step": 3435
    },
    {
      "epoch": 0.5612203279223428,
      "grad_norm": 0.004699255805462599,
      "learning_rate": 0.2945616956626266,
      "loss": 0.4126,
      "num_input_tokens_seen": 6528160,
      "step": 3440
    },
    {
      "epoch": 0.5620360551431601,
      "grad_norm": 0.0011422637617215514,
      "learning_rate": 0.2945459671500966,
      "loss": 0.4026,
      "num_input_tokens_seen": 6536128,
      "step": 3445
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 0.0023495664354413748,
      "learning_rate": 0.2945302163467621,
      "loss": 0.3841,
      "num_input_tokens_seen": 6545504,
      "step": 3450
    },
    {
      "epoch": 0.5636675095847948,
      "grad_norm": 0.005133964121341705,
      "learning_rate": 0.2945144432550519,
      "loss": 0.3804,
      "num_input_tokens_seen": 6556048,
      "step": 3455
    },
    {
      "epoch": 0.5644832368056122,
      "grad_norm": 0.0016441956395283341,
      "learning_rate": 0.29449864787739843,
      "loss": 0.3773,
      "num_input_tokens_seen": 6565616,
      "step": 3460
    },
    {
      "epoch": 0.5652989640264295,
      "grad_norm": 0.001914552180096507,
      "learning_rate": 0.2944828302162376,
      "loss": 0.3338,
      "num_input_tokens_seen": 6575600,
      "step": 3465
    },
    {
      "epoch": 0.5661146912472469,
      "grad_norm": 0.001550885965116322,
      "learning_rate": 0.2944669902740087,
      "loss": 0.4655,
      "num_input_tokens_seen": 6585008,
      "step": 3470
    },
    {
      "epoch": 0.5669304184680642,
      "grad_norm": 0.0016064915107563138,
      "learning_rate": 0.2944511280531544,
      "loss": 0.3542,
      "num_input_tokens_seen": 6594896,
      "step": 3475
    },
    {
      "epoch": 0.5677461456888816,
      "grad_norm": 0.0030623993370682,
      "learning_rate": 0.29443524355612083,
      "loss": 0.3799,
      "num_input_tokens_seen": 6603632,
      "step": 3480
    },
    {
      "epoch": 0.568561872909699,
      "grad_norm": 0.0012749159941449761,
      "learning_rate": 0.29441933678535764,
      "loss": 0.4181,
      "num_input_tokens_seen": 6613264,
      "step": 3485
    },
    {
      "epoch": 0.5693776001305163,
      "grad_norm": 0.0012668089475482702,
      "learning_rate": 0.29440340774331786,
      "loss": 0.3566,
      "num_input_tokens_seen": 6623040,
      "step": 3490
    },
    {
      "epoch": 0.5701933273513338,
      "grad_norm": 0.002800096059218049,
      "learning_rate": 0.2943874564324579,
      "loss": 0.4392,
      "num_input_tokens_seen": 6633200,
      "step": 3495
    },
    {
      "epoch": 0.5710090545721511,
      "grad_norm": 0.0017305637011304498,
      "learning_rate": 0.2943714828552376,
      "loss": 0.3481,
      "num_input_tokens_seen": 6642576,
      "step": 3500
    },
    {
      "epoch": 0.5718247817929685,
      "grad_norm": 0.0011520858388394117,
      "learning_rate": 0.29435548701412045,
      "loss": 0.3549,
      "num_input_tokens_seen": 6650736,
      "step": 3505
    },
    {
      "epoch": 0.5726405090137858,
      "grad_norm": 0.003968354780226946,
      "learning_rate": 0.2943394689115731,
      "loss": 0.3467,
      "num_input_tokens_seen": 6660416,
      "step": 3510
    },
    {
      "epoch": 0.5734562362346032,
      "grad_norm": 0.00580957718193531,
      "learning_rate": 0.29432342855006577,
      "loss": 0.3564,
      "num_input_tokens_seen": 6670576,
      "step": 3515
    },
    {
      "epoch": 0.5742719634554205,
      "grad_norm": 0.0013415246503427625,
      "learning_rate": 0.294307365932072,
      "loss": 0.3635,
      "num_input_tokens_seen": 6679680,
      "step": 3520
    },
    {
      "epoch": 0.5750876906762379,
      "grad_norm": 0.004807541146874428,
      "learning_rate": 0.294291281060069,
      "loss": 0.417,
      "num_input_tokens_seen": 6688544,
      "step": 3525
    },
    {
      "epoch": 0.5759034178970552,
      "grad_norm": 0.001967204501852393,
      "learning_rate": 0.29427517393653724,
      "loss": 0.3568,
      "num_input_tokens_seen": 6698560,
      "step": 3530
    },
    {
      "epoch": 0.5767191451178726,
      "grad_norm": 0.0054803406819701195,
      "learning_rate": 0.29425904456396046,
      "loss": 0.361,
      "num_input_tokens_seen": 6708816,
      "step": 3535
    },
    {
      "epoch": 0.5775348723386899,
      "grad_norm": 0.0011224254267290235,
      "learning_rate": 0.2942428929448262,
      "loss": 0.4543,
      "num_input_tokens_seen": 6717984,
      "step": 3540
    },
    {
      "epoch": 0.5783505995595073,
      "grad_norm": 0.002100761514157057,
      "learning_rate": 0.2942267190816252,
      "loss": 0.3389,
      "num_input_tokens_seen": 6727232,
      "step": 3545
    },
    {
      "epoch": 0.5791663267803246,
      "grad_norm": 0.0008706366061232984,
      "learning_rate": 0.2942105229768516,
      "loss": 0.3604,
      "num_input_tokens_seen": 6736832,
      "step": 3550
    },
    {
      "epoch": 0.579982054001142,
      "grad_norm": 0.0014993175864219666,
      "learning_rate": 0.29419430463300306,
      "loss": 0.3351,
      "num_input_tokens_seen": 6746640,
      "step": 3555
    },
    {
      "epoch": 0.5807977812219594,
      "grad_norm": 0.0035165126901119947,
      "learning_rate": 0.2941780640525808,
      "loss": 0.3742,
      "num_input_tokens_seen": 6756800,
      "step": 3560
    },
    {
      "epoch": 0.5816135084427767,
      "grad_norm": 0.0024434602819383144,
      "learning_rate": 0.2941618012380891,
      "loss": 0.3614,
      "num_input_tokens_seen": 6765840,
      "step": 3565
    },
    {
      "epoch": 0.5824292356635941,
      "grad_norm": 0.0006550165126100183,
      "learning_rate": 0.29414551619203605,
      "loss": 0.3506,
      "num_input_tokens_seen": 6774912,
      "step": 3570
    },
    {
      "epoch": 0.5832449628844114,
      "grad_norm": 0.0009631370194256306,
      "learning_rate": 0.29412920891693295,
      "loss": 0.3616,
      "num_input_tokens_seen": 6784832,
      "step": 3575
    },
    {
      "epoch": 0.5840606901052288,
      "grad_norm": 0.0007277146796695888,
      "learning_rate": 0.2941128794152946,
      "loss": 0.373,
      "num_input_tokens_seen": 6793504,
      "step": 3580
    },
    {
      "epoch": 0.5848764173260461,
      "grad_norm": 0.0038221306167542934,
      "learning_rate": 0.2940965276896392,
      "loss": 0.3777,
      "num_input_tokens_seen": 6802000,
      "step": 3585
    },
    {
      "epoch": 0.5856921445468636,
      "grad_norm": 0.0009058258146978915,
      "learning_rate": 0.2940801537424884,
      "loss": 0.3676,
      "num_input_tokens_seen": 6811680,
      "step": 3590
    },
    {
      "epoch": 0.5865078717676809,
      "grad_norm": 0.004695314913988113,
      "learning_rate": 0.2940637575763673,
      "loss": 0.3971,
      "num_input_tokens_seen": 6822112,
      "step": 3595
    },
    {
      "epoch": 0.5873235989884983,
      "grad_norm": 0.0031286869198083878,
      "learning_rate": 0.2940473391938043,
      "loss": 0.3337,
      "num_input_tokens_seen": 6831056,
      "step": 3600
    },
    {
      "epoch": 0.5873235989884983,
      "eval_loss": 0.3636453449726105,
      "eval_runtime": 152.4575,
      "eval_samples_per_second": 17.874,
      "eval_steps_per_second": 8.94,
      "num_input_tokens_seen": 6831056,
      "step": 3600
    },
    {
      "epoch": 0.5881393262093156,
      "grad_norm": 0.002094327239319682,
      "learning_rate": 0.29403089859733145,
      "loss": 0.5195,
      "num_input_tokens_seen": 6840528,
      "step": 3605
    },
    {
      "epoch": 0.588955053430133,
      "grad_norm": 0.002722780220210552,
      "learning_rate": 0.294014435789484,
      "loss": 0.3775,
      "num_input_tokens_seen": 6850016,
      "step": 3610
    },
    {
      "epoch": 0.5897707806509503,
      "grad_norm": 0.0029201386496424675,
      "learning_rate": 0.2939979507728007,
      "loss": 0.3831,
      "num_input_tokens_seen": 6859552,
      "step": 3615
    },
    {
      "epoch": 0.5905865078717677,
      "grad_norm": 0.0016360911540687084,
      "learning_rate": 0.2939814435498239,
      "loss": 0.354,
      "num_input_tokens_seen": 6868656,
      "step": 3620
    },
    {
      "epoch": 0.591402235092585,
      "grad_norm": 0.004151186905801296,
      "learning_rate": 0.29396491412309905,
      "loss": 0.3055,
      "num_input_tokens_seen": 6879600,
      "step": 3625
    },
    {
      "epoch": 0.5922179623134024,
      "grad_norm": 0.0016950840363278985,
      "learning_rate": 0.2939483624951753,
      "loss": 0.4318,
      "num_input_tokens_seen": 6888944,
      "step": 3630
    },
    {
      "epoch": 0.5930336895342198,
      "grad_norm": 0.0021833754144608974,
      "learning_rate": 0.2939317886686051,
      "loss": 0.3743,
      "num_input_tokens_seen": 6898288,
      "step": 3635
    },
    {
      "epoch": 0.5938494167550371,
      "grad_norm": 0.002403304213657975,
      "learning_rate": 0.2939151926459443,
      "loss": 0.3523,
      "num_input_tokens_seen": 6906880,
      "step": 3640
    },
    {
      "epoch": 0.5946651439758545,
      "grad_norm": 0.003935286775231361,
      "learning_rate": 0.2938985744297522,
      "loss": 0.436,
      "num_input_tokens_seen": 6915824,
      "step": 3645
    },
    {
      "epoch": 0.5954808711966718,
      "grad_norm": 0.001603191252797842,
      "learning_rate": 0.29388193402259166,
      "loss": 0.3535,
      "num_input_tokens_seen": 6925488,
      "step": 3650
    },
    {
      "epoch": 0.5962965984174892,
      "grad_norm": 0.0012254463508725166,
      "learning_rate": 0.29386527142702873,
      "loss": 0.3812,
      "num_input_tokens_seen": 6936112,
      "step": 3655
    },
    {
      "epoch": 0.5971123256383065,
      "grad_norm": 0.0008062535198405385,
      "learning_rate": 0.293848586645633,
      "loss": 0.3814,
      "num_input_tokens_seen": 6946480,
      "step": 3660
    },
    {
      "epoch": 0.5979280528591239,
      "grad_norm": 0.0016606647986918688,
      "learning_rate": 0.2938318796809775,
      "loss": 0.3579,
      "num_input_tokens_seen": 6955568,
      "step": 3665
    },
    {
      "epoch": 0.5987437800799412,
      "grad_norm": 0.0013946768594905734,
      "learning_rate": 0.29381515053563867,
      "loss": 0.3541,
      "num_input_tokens_seen": 6964400,
      "step": 3670
    },
    {
      "epoch": 0.5995595073007586,
      "grad_norm": 0.0015689725987613201,
      "learning_rate": 0.29379839921219636,
      "loss": 0.3777,
      "num_input_tokens_seen": 6974288,
      "step": 3675
    },
    {
      "epoch": 0.600375234521576,
      "grad_norm": 0.0009287587599828839,
      "learning_rate": 0.2937816257132338,
      "loss": 0.3505,
      "num_input_tokens_seen": 6985776,
      "step": 3680
    },
    {
      "epoch": 0.6011909617423934,
      "grad_norm": 0.0025504627265036106,
      "learning_rate": 0.2937648300413376,
      "loss": 0.3931,
      "num_input_tokens_seen": 6993440,
      "step": 3685
    },
    {
      "epoch": 0.6020066889632107,
      "grad_norm": 0.0018744500121101737,
      "learning_rate": 0.293748012199098,
      "loss": 0.347,
      "num_input_tokens_seen": 7003648,
      "step": 3690
    },
    {
      "epoch": 0.6028224161840281,
      "grad_norm": 0.0017623259918764234,
      "learning_rate": 0.29373117218910844,
      "loss": 0.3415,
      "num_input_tokens_seen": 7011776,
      "step": 3695
    },
    {
      "epoch": 0.6036381434048455,
      "grad_norm": 0.003454035148024559,
      "learning_rate": 0.2937143100139659,
      "loss": 0.387,
      "num_input_tokens_seen": 7020816,
      "step": 3700
    },
    {
      "epoch": 0.6044538706256628,
      "grad_norm": 0.000946946325711906,
      "learning_rate": 0.29369742567627083,
      "loss": 0.3434,
      "num_input_tokens_seen": 7031104,
      "step": 3705
    },
    {
      "epoch": 0.6052695978464802,
      "grad_norm": 0.004019375890493393,
      "learning_rate": 0.29368051917862675,
      "loss": 0.4466,
      "num_input_tokens_seen": 7039776,
      "step": 3710
    },
    {
      "epoch": 0.6060853250672975,
      "grad_norm": 0.0032454156316816807,
      "learning_rate": 0.2936635905236411,
      "loss": 0.4056,
      "num_input_tokens_seen": 7048976,
      "step": 3715
    },
    {
      "epoch": 0.6069010522881149,
      "grad_norm": 0.0016233330825343728,
      "learning_rate": 0.2936466397139244,
      "loss": 0.4353,
      "num_input_tokens_seen": 7057888,
      "step": 3720
    },
    {
      "epoch": 0.6077167795089322,
      "grad_norm": 0.0009374008513987064,
      "learning_rate": 0.2936296667520907,
      "loss": 0.3472,
      "num_input_tokens_seen": 7067632,
      "step": 3725
    },
    {
      "epoch": 0.6085325067297496,
      "grad_norm": 0.0007986014825291932,
      "learning_rate": 0.2936126716407574,
      "loss": 0.3459,
      "num_input_tokens_seen": 7076672,
      "step": 3730
    },
    {
      "epoch": 0.6093482339505669,
      "grad_norm": 0.005453506018966436,
      "learning_rate": 0.29359565438254537,
      "loss": 0.3975,
      "num_input_tokens_seen": 7087680,
      "step": 3735
    },
    {
      "epoch": 0.6101639611713843,
      "grad_norm": 0.004898997489362955,
      "learning_rate": 0.29357861498007887,
      "loss": 0.3902,
      "num_input_tokens_seen": 7097536,
      "step": 3740
    },
    {
      "epoch": 0.6109796883922016,
      "grad_norm": 0.0015428761253133416,
      "learning_rate": 0.29356155343598567,
      "loss": 0.3794,
      "num_input_tokens_seen": 7106768,
      "step": 3745
    },
    {
      "epoch": 0.611795415613019,
      "grad_norm": 0.0004903130466118455,
      "learning_rate": 0.2935444697528968,
      "loss": 0.3739,
      "num_input_tokens_seen": 7116320,
      "step": 3750
    },
    {
      "epoch": 0.6126111428338363,
      "grad_norm": 0.0014501658733934164,
      "learning_rate": 0.2935273639334468,
      "loss": 0.3331,
      "num_input_tokens_seen": 7125648,
      "step": 3755
    },
    {
      "epoch": 0.6134268700546537,
      "grad_norm": 0.0018250984139740467,
      "learning_rate": 0.29351023598027365,
      "loss": 0.3498,
      "num_input_tokens_seen": 7134352,
      "step": 3760
    },
    {
      "epoch": 0.614242597275471,
      "grad_norm": 0.001666666124947369,
      "learning_rate": 0.2934930858960186,
      "loss": 0.3638,
      "num_input_tokens_seen": 7144256,
      "step": 3765
    },
    {
      "epoch": 0.6150583244962884,
      "grad_norm": 0.0034105179365724325,
      "learning_rate": 0.29347591368332643,
      "loss": 0.3216,
      "num_input_tokens_seen": 7154592,
      "step": 3770
    },
    {
      "epoch": 0.6158740517171059,
      "grad_norm": 0.0024307342246174812,
      "learning_rate": 0.2934587193448454,
      "loss": 0.3726,
      "num_input_tokens_seen": 7164512,
      "step": 3775
    },
    {
      "epoch": 0.6166897789379232,
      "grad_norm": 0.001527345972135663,
      "learning_rate": 0.29344150288322696,
      "loss": 0.3804,
      "num_input_tokens_seen": 7173792,
      "step": 3780
    },
    {
      "epoch": 0.6175055061587406,
      "grad_norm": 0.0010351162636652589,
      "learning_rate": 0.2934242643011263,
      "loss": 0.3553,
      "num_input_tokens_seen": 7182480,
      "step": 3785
    },
    {
      "epoch": 0.6183212333795579,
      "grad_norm": 0.0015759880188852549,
      "learning_rate": 0.2934070036012016,
      "loss": 0.4098,
      "num_input_tokens_seen": 7191840,
      "step": 3790
    },
    {
      "epoch": 0.6191369606003753,
      "grad_norm": 0.0034929835237562656,
      "learning_rate": 0.29338972078611475,
      "loss": 0.363,
      "num_input_tokens_seen": 7201152,
      "step": 3795
    },
    {
      "epoch": 0.6199526878211926,
      "grad_norm": 0.0005707073141820729,
      "learning_rate": 0.2933724158585311,
      "loss": 0.3765,
      "num_input_tokens_seen": 7209536,
      "step": 3800
    },
    {
      "epoch": 0.6199526878211926,
      "eval_loss": 0.373830109834671,
      "eval_runtime": 152.3597,
      "eval_samples_per_second": 17.885,
      "eval_steps_per_second": 8.946,
      "num_input_tokens_seen": 7209536,
      "step": 3800
    },
    {
      "epoch": 0.62076841504201,
      "grad_norm": 0.0015859095146879554,
      "learning_rate": 0.29335508882111916,
      "loss": 0.3382,
      "num_input_tokens_seen": 7218736,
      "step": 3805
    },
    {
      "epoch": 0.6215841422628273,
      "grad_norm": 0.0017694768030196428,
      "learning_rate": 0.29333773967655097,
      "loss": 0.3172,
      "num_input_tokens_seen": 7227952,
      "step": 3810
    },
    {
      "epoch": 0.6223998694836447,
      "grad_norm": 0.0010319873690605164,
      "learning_rate": 0.2933203684275021,
      "loss": 0.3147,
      "num_input_tokens_seen": 7239056,
      "step": 3815
    },
    {
      "epoch": 0.623215596704462,
      "grad_norm": 0.0013685398735105991,
      "learning_rate": 0.2933029750766513,
      "loss": 0.375,
      "num_input_tokens_seen": 7248816,
      "step": 3820
    },
    {
      "epoch": 0.6240313239252794,
      "grad_norm": 0.006235698703676462,
      "learning_rate": 0.2932855596266809,
      "loss": 0.3773,
      "num_input_tokens_seen": 7258448,
      "step": 3825
    },
    {
      "epoch": 0.6248470511460967,
      "grad_norm": 0.0013814832782372832,
      "learning_rate": 0.2932681220802765,
      "loss": 0.3239,
      "num_input_tokens_seen": 7268128,
      "step": 3830
    },
    {
      "epoch": 0.6256627783669141,
      "grad_norm": 0.0008235547575168312,
      "learning_rate": 0.2932506624401274,
      "loss": 0.3703,
      "num_input_tokens_seen": 7276912,
      "step": 3835
    },
    {
      "epoch": 0.6264785055877314,
      "grad_norm": 0.0021575279533863068,
      "learning_rate": 0.29323318070892584,
      "loss": 0.3695,
      "num_input_tokens_seen": 7287328,
      "step": 3840
    },
    {
      "epoch": 0.6272942328085488,
      "grad_norm": 0.001990897348150611,
      "learning_rate": 0.29321567688936784,
      "loss": 0.3527,
      "num_input_tokens_seen": 7295568,
      "step": 3845
    },
    {
      "epoch": 0.6281099600293661,
      "grad_norm": 0.0017062280094251037,
      "learning_rate": 0.29319815098415275,
      "loss": 0.3419,
      "num_input_tokens_seen": 7305920,
      "step": 3850
    },
    {
      "epoch": 0.6289256872501835,
      "grad_norm": 0.007230832241475582,
      "learning_rate": 0.2931806029959832,
      "loss": 0.3711,
      "num_input_tokens_seen": 7313616,
      "step": 3855
    },
    {
      "epoch": 0.6297414144710008,
      "grad_norm": 0.0013346270425245166,
      "learning_rate": 0.29316303292756535,
      "loss": 0.3602,
      "num_input_tokens_seen": 7323328,
      "step": 3860
    },
    {
      "epoch": 0.6305571416918182,
      "grad_norm": 0.004037077538669109,
      "learning_rate": 0.29314544078160876,
      "loss": 0.3608,
      "num_input_tokens_seen": 7333248,
      "step": 3865
    },
    {
      "epoch": 0.6313728689126357,
      "grad_norm": 0.0012119965394958854,
      "learning_rate": 0.2931278265608263,
      "loss": 0.3375,
      "num_input_tokens_seen": 7343408,
      "step": 3870
    },
    {
      "epoch": 0.632188596133453,
      "grad_norm": 0.0033198921009898186,
      "learning_rate": 0.29311019026793433,
      "loss": 0.3423,
      "num_input_tokens_seen": 7353216,
      "step": 3875
    },
    {
      "epoch": 0.6330043233542704,
      "grad_norm": 0.0018325517885386944,
      "learning_rate": 0.29309253190565254,
      "loss": 0.3392,
      "num_input_tokens_seen": 7362768,
      "step": 3880
    },
    {
      "epoch": 0.6338200505750877,
      "grad_norm": 0.0023152800276875496,
      "learning_rate": 0.2930748514767042,
      "loss": 0.37,
      "num_input_tokens_seen": 7371488,
      "step": 3885
    },
    {
      "epoch": 0.6346357777959051,
      "grad_norm": 0.0005258062155917287,
      "learning_rate": 0.29305714898381574,
      "loss": 0.4172,
      "num_input_tokens_seen": 7380464,
      "step": 3890
    },
    {
      "epoch": 0.6354515050167224,
      "grad_norm": 0.0016227519372478127,
      "learning_rate": 0.29303942442971714,
      "loss": 0.3692,
      "num_input_tokens_seen": 7389792,
      "step": 3895
    },
    {
      "epoch": 0.6362672322375398,
      "grad_norm": 0.0025862164329737425,
      "learning_rate": 0.2930216778171417,
      "loss": 0.4073,
      "num_input_tokens_seen": 7400720,
      "step": 3900
    },
    {
      "epoch": 0.6370829594583571,
      "grad_norm": 0.00153203378431499,
      "learning_rate": 0.2930039091488263,
      "loss": 0.3681,
      "num_input_tokens_seen": 7409856,
      "step": 3905
    },
    {
      "epoch": 0.6378986866791745,
      "grad_norm": 0.001296021742746234,
      "learning_rate": 0.29298611842751093,
      "loss": 0.3366,
      "num_input_tokens_seen": 7420368,
      "step": 3910
    },
    {
      "epoch": 0.6387144138999918,
      "grad_norm": 0.002469644881784916,
      "learning_rate": 0.29296830565593923,
      "loss": 0.3692,
      "num_input_tokens_seen": 7429408,
      "step": 3915
    },
    {
      "epoch": 0.6395301411208092,
      "grad_norm": 0.001219552243128419,
      "learning_rate": 0.2929504708368582,
      "loss": 0.3425,
      "num_input_tokens_seen": 7438640,
      "step": 3920
    },
    {
      "epoch": 0.6403458683416265,
      "grad_norm": 0.0031776772812008858,
      "learning_rate": 0.29293261397301806,
      "loss": 0.3707,
      "num_input_tokens_seen": 7448768,
      "step": 3925
    },
    {
      "epoch": 0.6411615955624439,
      "grad_norm": 0.0032372423447668552,
      "learning_rate": 0.29291473506717275,
      "loss": 0.3767,
      "num_input_tokens_seen": 7458896,
      "step": 3930
    },
    {
      "epoch": 0.6419773227832613,
      "grad_norm": 0.0021738375071436167,
      "learning_rate": 0.29289683412207923,
      "loss": 0.3937,
      "num_input_tokens_seen": 7468528,
      "step": 3935
    },
    {
      "epoch": 0.6427930500040786,
      "grad_norm": 0.0020071386825293303,
      "learning_rate": 0.29287891114049813,
      "loss": 0.4119,
      "num_input_tokens_seen": 7478240,
      "step": 3940
    },
    {
      "epoch": 0.643608777224896,
      "grad_norm": 0.0016832530964165926,
      "learning_rate": 0.29286096612519347,
      "loss": 0.4041,
      "num_input_tokens_seen": 7487792,
      "step": 3945
    },
    {
      "epoch": 0.6444245044457133,
      "grad_norm": 0.00209226505830884,
      "learning_rate": 0.2928429990789325,
      "loss": 0.4269,
      "num_input_tokens_seen": 7498000,
      "step": 3950
    },
    {
      "epoch": 0.6452402316665307,
      "grad_norm": 0.003448059316724539,
      "learning_rate": 0.29282501000448596,
      "loss": 0.2997,
      "num_input_tokens_seen": 7508144,
      "step": 3955
    },
    {
      "epoch": 0.6460559588873481,
      "grad_norm": 0.0021540061570703983,
      "learning_rate": 0.2928069989046281,
      "loss": 0.4198,
      "num_input_tokens_seen": 7517984,
      "step": 3960
    },
    {
      "epoch": 0.6468716861081655,
      "grad_norm": 0.005968484096229076,
      "learning_rate": 0.2927889657821363,
      "loss": 0.5504,
      "num_input_tokens_seen": 7527216,
      "step": 3965
    },
    {
      "epoch": 0.6476874133289828,
      "grad_norm": 0.002804965479299426,
      "learning_rate": 0.2927709106397916,
      "loss": 0.3897,
      "num_input_tokens_seen": 7535968,
      "step": 3970
    },
    {
      "epoch": 0.6485031405498002,
      "grad_norm": 0.0006770302425138652,
      "learning_rate": 0.29275283348037834,
      "loss": 0.4611,
      "num_input_tokens_seen": 7546064,
      "step": 3975
    },
    {
      "epoch": 0.6493188677706175,
      "grad_norm": 0.0011480458779260516,
      "learning_rate": 0.29273473430668423,
      "loss": 0.3801,
      "num_input_tokens_seen": 7555408,
      "step": 3980
    },
    {
      "epoch": 0.6501345949914349,
      "grad_norm": 0.0012309905141592026,
      "learning_rate": 0.2927166131215003,
      "loss": 0.3542,
      "num_input_tokens_seen": 7564768,
      "step": 3985
    },
    {
      "epoch": 0.6509503222122522,
      "grad_norm": 0.002770074177533388,
      "learning_rate": 0.2926984699276212,
      "loss": 0.3078,
      "num_input_tokens_seen": 7574368,
      "step": 3990
    },
    {
      "epoch": 0.6517660494330696,
      "grad_norm": 0.001242046244442463,
      "learning_rate": 0.29268030472784473,
      "loss": 0.4329,
      "num_input_tokens_seen": 7583280,
      "step": 3995
    },
    {
      "epoch": 0.652581776653887,
      "grad_norm": 0.003057542024180293,
      "learning_rate": 0.2926621175249723,
      "loss": 0.3645,
      "num_input_tokens_seen": 7593024,
      "step": 4000
    },
    {
      "epoch": 0.652581776653887,
      "eval_loss": 0.4008273184299469,
      "eval_runtime": 152.3136,
      "eval_samples_per_second": 17.891,
      "eval_steps_per_second": 8.949,
      "num_input_tokens_seen": 7593024,
      "step": 4000
    },
    {
      "epoch": 0.6533975038747043,
      "grad_norm": 0.002349777612835169,
      "learning_rate": 0.29264390832180853,
      "loss": 0.4033,
      "num_input_tokens_seen": 7602256,
      "step": 4005
    },
    {
      "epoch": 0.6542132310955217,
      "grad_norm": 0.0009653707384131849,
      "learning_rate": 0.29262567712116144,
      "loss": 0.3615,
      "num_input_tokens_seen": 7612336,
      "step": 4010
    },
    {
      "epoch": 0.655028958316339,
      "grad_norm": 0.003401102963835001,
      "learning_rate": 0.29260742392584266,
      "loss": 0.3689,
      "num_input_tokens_seen": 7621952,
      "step": 4015
    },
    {
      "epoch": 0.6558446855371564,
      "grad_norm": 0.0008107854519039392,
      "learning_rate": 0.292589148738667,
      "loss": 0.3715,
      "num_input_tokens_seen": 7631056,
      "step": 4020
    },
    {
      "epoch": 0.6566604127579737,
      "grad_norm": 0.0016698449617251754,
      "learning_rate": 0.2925708515624527,
      "loss": 0.3597,
      "num_input_tokens_seen": 7640768,
      "step": 4025
    },
    {
      "epoch": 0.6574761399787911,
      "grad_norm": 0.0030649604741483927,
      "learning_rate": 0.29255253240002144,
      "loss": 0.3644,
      "num_input_tokens_seen": 7651248,
      "step": 4030
    },
    {
      "epoch": 0.6582918671996084,
      "grad_norm": 0.0038449291605502367,
      "learning_rate": 0.2925341912541983,
      "loss": 0.3774,
      "num_input_tokens_seen": 7661616,
      "step": 4035
    },
    {
      "epoch": 0.6591075944204258,
      "grad_norm": 0.0006964487838558853,
      "learning_rate": 0.2925158281278116,
      "loss": 0.3691,
      "num_input_tokens_seen": 7672656,
      "step": 4040
    },
    {
      "epoch": 0.6599233216412431,
      "grad_norm": 0.0031976664904505014,
      "learning_rate": 0.29249744302369324,
      "loss": 0.368,
      "num_input_tokens_seen": 7681728,
      "step": 4045
    },
    {
      "epoch": 0.6607390488620605,
      "grad_norm": 0.0032122842967510223,
      "learning_rate": 0.29247903594467844,
      "loss": 0.3399,
      "num_input_tokens_seen": 7690672,
      "step": 4050
    },
    {
      "epoch": 0.6615547760828779,
      "grad_norm": 0.0005480862455442548,
      "learning_rate": 0.2924606068936058,
      "loss": 0.3744,
      "num_input_tokens_seen": 7700544,
      "step": 4055
    },
    {
      "epoch": 0.6623705033036953,
      "grad_norm": 0.0025195013731718063,
      "learning_rate": 0.2924421558733173,
      "loss": 0.3702,
      "num_input_tokens_seen": 7710512,
      "step": 4060
    },
    {
      "epoch": 0.6631862305245126,
      "grad_norm": 0.0036428431048989296,
      "learning_rate": 0.2924236828866583,
      "loss": 0.3441,
      "num_input_tokens_seen": 7719744,
      "step": 4065
    },
    {
      "epoch": 0.66400195774533,
      "grad_norm": 0.001169126946479082,
      "learning_rate": 0.29240518793647763,
      "loss": 0.3256,
      "num_input_tokens_seen": 7729888,
      "step": 4070
    },
    {
      "epoch": 0.6648176849661473,
      "grad_norm": 0.0033220937475562096,
      "learning_rate": 0.29238667102562743,
      "loss": 0.3054,
      "num_input_tokens_seen": 7740992,
      "step": 4075
    },
    {
      "epoch": 0.6656334121869647,
      "grad_norm": 0.0036704738158732653,
      "learning_rate": 0.29236813215696317,
      "loss": 0.4552,
      "num_input_tokens_seen": 7749232,
      "step": 4080
    },
    {
      "epoch": 0.666449139407782,
      "grad_norm": 0.004460202530026436,
      "learning_rate": 0.2923495713333439,
      "loss": 0.3667,
      "num_input_tokens_seen": 7758672,
      "step": 4085
    },
    {
      "epoch": 0.6672648666285994,
      "grad_norm": 0.001284868223592639,
      "learning_rate": 0.29233098855763173,
      "loss": 0.3586,
      "num_input_tokens_seen": 7766896,
      "step": 4090
    },
    {
      "epoch": 0.6680805938494168,
      "grad_norm": 0.0029421241488307714,
      "learning_rate": 0.29231238383269254,
      "loss": 0.3802,
      "num_input_tokens_seen": 7777072,
      "step": 4095
    },
    {
      "epoch": 0.6688963210702341,
      "grad_norm": 0.0040875691920518875,
      "learning_rate": 0.2922937571613954,
      "loss": 0.392,
      "num_input_tokens_seen": 7787296,
      "step": 4100
    },
    {
      "epoch": 0.6697120482910515,
      "grad_norm": 0.0011924299178645015,
      "learning_rate": 0.29227510854661265,
      "loss": 0.3276,
      "num_input_tokens_seen": 7797072,
      "step": 4105
    },
    {
      "epoch": 0.6705277755118688,
      "grad_norm": 0.0007428276003338397,
      "learning_rate": 0.29225643799122025,
      "loss": 0.3956,
      "num_input_tokens_seen": 7806944,
      "step": 4110
    },
    {
      "epoch": 0.6713435027326862,
      "grad_norm": 0.002374313771724701,
      "learning_rate": 0.2922377454980974,
      "loss": 0.3613,
      "num_input_tokens_seen": 7815328,
      "step": 4115
    },
    {
      "epoch": 0.6721592299535035,
      "grad_norm": 0.0018511575181037188,
      "learning_rate": 0.29221903107012676,
      "loss": 0.3596,
      "num_input_tokens_seen": 7824976,
      "step": 4120
    },
    {
      "epoch": 0.6729749571743209,
      "grad_norm": 0.0010548952268436551,
      "learning_rate": 0.29220029471019426,
      "loss": 0.3215,
      "num_input_tokens_seen": 7834672,
      "step": 4125
    },
    {
      "epoch": 0.6737906843951382,
      "grad_norm": 0.001974851591512561,
      "learning_rate": 0.2921815364211893,
      "loss": 0.4078,
      "num_input_tokens_seen": 7844304,
      "step": 4130
    },
    {
      "epoch": 0.6746064116159556,
      "grad_norm": 0.003368524368852377,
      "learning_rate": 0.29216275620600474,
      "loss": 0.3008,
      "num_input_tokens_seen": 7853968,
      "step": 4135
    },
    {
      "epoch": 0.6754221388367729,
      "grad_norm": 0.002182536292821169,
      "learning_rate": 0.29214395406753657,
      "loss": 0.2513,
      "num_input_tokens_seen": 7862928,
      "step": 4140
    },
    {
      "epoch": 0.6762378660575903,
      "grad_norm": 0.0020522023551166058,
      "learning_rate": 0.2921251300086844,
      "loss": 0.3906,
      "num_input_tokens_seen": 7871840,
      "step": 4145
    },
    {
      "epoch": 0.6770535932784078,
      "grad_norm": 0.001176285557448864,
      "learning_rate": 0.2921062840323511,
      "loss": 0.4043,
      "num_input_tokens_seen": 7880912,
      "step": 4150
    },
    {
      "epoch": 0.6778693204992251,
      "grad_norm": 0.0009408268379047513,
      "learning_rate": 0.29208741614144307,
      "loss": 0.3641,
      "num_input_tokens_seen": 7890480,
      "step": 4155
    },
    {
      "epoch": 0.6786850477200425,
      "grad_norm": 0.0016013324493542314,
      "learning_rate": 0.2920685263388698,
      "loss": 0.369,
      "num_input_tokens_seen": 7900304,
      "step": 4160
    },
    {
      "epoch": 0.6795007749408598,
      "grad_norm": 0.0010581437963992357,
      "learning_rate": 0.2920496146275445,
      "loss": 0.2423,
      "num_input_tokens_seen": 7909440,
      "step": 4165
    },
    {
      "epoch": 0.6803165021616772,
      "grad_norm": 0.005362347234040499,
      "learning_rate": 0.29203068101038343,
      "loss": 0.5615,
      "num_input_tokens_seen": 7919248,
      "step": 4170
    },
    {
      "epoch": 0.6811322293824945,
      "grad_norm": 0.001870097708888352,
      "learning_rate": 0.2920117254903065,
      "loss": 0.3587,
      "num_input_tokens_seen": 7928592,
      "step": 4175
    },
    {
      "epoch": 0.6819479566033119,
      "grad_norm": 0.002797751920297742,
      "learning_rate": 0.29199274807023695,
      "loss": 0.4027,
      "num_input_tokens_seen": 7939040,
      "step": 4180
    },
    {
      "epoch": 0.6827636838241292,
      "grad_norm": 0.0008957167156040668,
      "learning_rate": 0.29197374875310117,
      "loss": 0.3622,
      "num_input_tokens_seen": 7949008,
      "step": 4185
    },
    {
      "epoch": 0.6835794110449466,
      "grad_norm": 0.0034165980760008097,
      "learning_rate": 0.2919547275418292,
      "loss": 0.3645,
      "num_input_tokens_seen": 7958528,
      "step": 4190
    },
    {
      "epoch": 0.6843951382657639,
      "grad_norm": 0.005705228075385094,
      "learning_rate": 0.29193568443935436,
      "loss": 0.4399,
      "num_input_tokens_seen": 7967424,
      "step": 4195
    },
    {
      "epoch": 0.6852108654865813,
      "grad_norm": 0.0019908170215785503,
      "learning_rate": 0.2919166194486133,
      "loss": 0.3501,
      "num_input_tokens_seen": 7977072,
      "step": 4200
    },
    {
      "epoch": 0.6852108654865813,
      "eval_loss": 0.3489711880683899,
      "eval_runtime": 152.4267,
      "eval_samples_per_second": 17.877,
      "eval_steps_per_second": 8.942,
      "num_input_tokens_seen": 7977072,
      "step": 4200
    },
    {
      "epoch": 0.6860265927073986,
      "grad_norm": 0.0007522680098190904,
      "learning_rate": 0.2918975325725461,
      "loss": 0.3549,
      "num_input_tokens_seen": 7987088,
      "step": 4205
    },
    {
      "epoch": 0.686842319928216,
      "grad_norm": 0.0009596769814379513,
      "learning_rate": 0.29187842381409607,
      "loss": 0.3906,
      "num_input_tokens_seen": 7996672,
      "step": 4210
    },
    {
      "epoch": 0.6876580471490333,
      "grad_norm": 0.0021668155677616596,
      "learning_rate": 0.29185929317621023,
      "loss": 0.3533,
      "num_input_tokens_seen": 8006704,
      "step": 4215
    },
    {
      "epoch": 0.6884737743698507,
      "grad_norm": 0.0022613664623349905,
      "learning_rate": 0.29184014066183867,
      "loss": 0.3669,
      "num_input_tokens_seen": 8016544,
      "step": 4220
    },
    {
      "epoch": 0.689289501590668,
      "grad_norm": 0.0007563804392702878,
      "learning_rate": 0.2918209662739349,
      "loss": 0.3617,
      "num_input_tokens_seen": 8026016,
      "step": 4225
    },
    {
      "epoch": 0.6901052288114854,
      "grad_norm": 0.000594273500610143,
      "learning_rate": 0.29180177001545593,
      "loss": 0.3451,
      "num_input_tokens_seen": 8037184,
      "step": 4230
    },
    {
      "epoch": 0.6909209560323027,
      "grad_norm": 0.0018053073436021805,
      "learning_rate": 0.29178255188936203,
      "loss": 0.3589,
      "num_input_tokens_seen": 8046336,
      "step": 4235
    },
    {
      "epoch": 0.6917366832531202,
      "grad_norm": 0.0017436012858524919,
      "learning_rate": 0.2917633118986169,
      "loss": 0.3408,
      "num_input_tokens_seen": 8054896,
      "step": 4240
    },
    {
      "epoch": 0.6925524104739376,
      "grad_norm": 0.0023730983957648277,
      "learning_rate": 0.2917440500461875,
      "loss": 0.3279,
      "num_input_tokens_seen": 8063632,
      "step": 4245
    },
    {
      "epoch": 0.6933681376947549,
      "grad_norm": 0.002799757057800889,
      "learning_rate": 0.29172476633504435,
      "loss": 0.4209,
      "num_input_tokens_seen": 8073072,
      "step": 4250
    },
    {
      "epoch": 0.6941838649155723,
      "grad_norm": 0.0016167842550203204,
      "learning_rate": 0.2917054607681612,
      "loss": 0.3378,
      "num_input_tokens_seen": 8081264,
      "step": 4255
    },
    {
      "epoch": 0.6949995921363896,
      "grad_norm": 0.001984634669497609,
      "learning_rate": 0.29168613334851523,
      "loss": 0.3757,
      "num_input_tokens_seen": 8090368,
      "step": 4260
    },
    {
      "epoch": 0.695815319357207,
      "grad_norm": 0.0014804551610723138,
      "learning_rate": 0.2916667840790869,
      "loss": 0.3291,
      "num_input_tokens_seen": 8100080,
      "step": 4265
    },
    {
      "epoch": 0.6966310465780243,
      "grad_norm": 0.0016950484132394195,
      "learning_rate": 0.2916474129628603,
      "loss": 0.3741,
      "num_input_tokens_seen": 8109056,
      "step": 4270
    },
    {
      "epoch": 0.6974467737988417,
      "grad_norm": 0.001427503302693367,
      "learning_rate": 0.29162802000282245,
      "loss": 0.3496,
      "num_input_tokens_seen": 8117616,
      "step": 4275
    },
    {
      "epoch": 0.698262501019659,
      "grad_norm": 0.002472704742103815,
      "learning_rate": 0.2916086052019642,
      "loss": 0.3456,
      "num_input_tokens_seen": 8127408,
      "step": 4280
    },
    {
      "epoch": 0.6990782282404764,
      "grad_norm": 0.0017054659547284245,
      "learning_rate": 0.2915891685632794,
      "loss": 0.3803,
      "num_input_tokens_seen": 8136624,
      "step": 4285
    },
    {
      "epoch": 0.6998939554612937,
      "grad_norm": 0.0036869074683636427,
      "learning_rate": 0.29156971008976545,
      "loss": 0.372,
      "num_input_tokens_seen": 8145904,
      "step": 4290
    },
    {
      "epoch": 0.7007096826821111,
      "grad_norm": 0.004232966806739569,
      "learning_rate": 0.2915502297844232,
      "loss": 0.3816,
      "num_input_tokens_seen": 8154064,
      "step": 4295
    },
    {
      "epoch": 0.7015254099029284,
      "grad_norm": 0.0034320936538279057,
      "learning_rate": 0.2915307276502566,
      "loss": 0.3771,
      "num_input_tokens_seen": 8162384,
      "step": 4300
    },
    {
      "epoch": 0.7023411371237458,
      "grad_norm": 0.0011885924031957984,
      "learning_rate": 0.29151120369027334,
      "loss": 0.4649,
      "num_input_tokens_seen": 8171456,
      "step": 4305
    },
    {
      "epoch": 0.7031568643445631,
      "grad_norm": 0.0027233175933361053,
      "learning_rate": 0.29149165790748405,
      "loss": 0.4118,
      "num_input_tokens_seen": 8180880,
      "step": 4310
    },
    {
      "epoch": 0.7039725915653805,
      "grad_norm": 0.0011091657215729356,
      "learning_rate": 0.291472090304903,
      "loss": 0.4166,
      "num_input_tokens_seen": 8189952,
      "step": 4315
    },
    {
      "epoch": 0.7047883187861979,
      "grad_norm": 0.0005107696051709354,
      "learning_rate": 0.2914525008855478,
      "loss": 0.3654,
      "num_input_tokens_seen": 8199968,
      "step": 4320
    },
    {
      "epoch": 0.7056040460070152,
      "grad_norm": 0.0010736582335084677,
      "learning_rate": 0.2914328896524394,
      "loss": 0.3656,
      "num_input_tokens_seen": 8209808,
      "step": 4325
    },
    {
      "epoch": 0.7064197732278326,
      "grad_norm": 0.0006038879510015249,
      "learning_rate": 0.291413256608602,
      "loss": 0.3804,
      "num_input_tokens_seen": 8220496,
      "step": 4330
    },
    {
      "epoch": 0.70723550044865,
      "grad_norm": 0.004338822793215513,
      "learning_rate": 0.29139360175706336,
      "loss": 0.3656,
      "num_input_tokens_seen": 8229600,
      "step": 4335
    },
    {
      "epoch": 0.7080512276694674,
      "grad_norm": 0.0017672310350462794,
      "learning_rate": 0.2913739251008544,
      "loss": 0.3452,
      "num_input_tokens_seen": 8240032,
      "step": 4340
    },
    {
      "epoch": 0.7088669548902847,
      "grad_norm": 0.0013943618396297097,
      "learning_rate": 0.29135422664300964,
      "loss": 0.3416,
      "num_input_tokens_seen": 8250624,
      "step": 4345
    },
    {
      "epoch": 0.7096826821111021,
      "grad_norm": 0.0019296559039503336,
      "learning_rate": 0.29133450638656677,
      "loss": 0.3551,
      "num_input_tokens_seen": 8260576,
      "step": 4350
    },
    {
      "epoch": 0.7104984093319194,
      "grad_norm": 0.002775868633762002,
      "learning_rate": 0.2913147643345669,
      "loss": 0.3934,
      "num_input_tokens_seen": 8270528,
      "step": 4355
    },
    {
      "epoch": 0.7113141365527368,
      "grad_norm": 0.0018735069315880537,
      "learning_rate": 0.29129500049005447,
      "loss": 0.3713,
      "num_input_tokens_seen": 8279280,
      "step": 4360
    },
    {
      "epoch": 0.7121298637735541,
      "grad_norm": 0.0014935629442334175,
      "learning_rate": 0.2912752148560773,
      "loss": 0.3548,
      "num_input_tokens_seen": 8287600,
      "step": 4365
    },
    {
      "epoch": 0.7129455909943715,
      "grad_norm": 0.00387400621548295,
      "learning_rate": 0.2912554074356866,
      "loss": 0.3911,
      "num_input_tokens_seen": 8297424,
      "step": 4370
    },
    {
      "epoch": 0.7137613182151888,
      "grad_norm": 0.0025730568449944258,
      "learning_rate": 0.2912355782319371,
      "loss": 0.3929,
      "num_input_tokens_seen": 8306512,
      "step": 4375
    },
    {
      "epoch": 0.7145770454360062,
      "grad_norm": 0.0014291424304246902,
      "learning_rate": 0.2912157272478864,
      "loss": 0.3295,
      "num_input_tokens_seen": 8315152,
      "step": 4380
    },
    {
      "epoch": 0.7153927726568236,
      "grad_norm": 0.0011761331697925925,
      "learning_rate": 0.291195854486596,
      "loss": 0.3311,
      "num_input_tokens_seen": 8324624,
      "step": 4385
    },
    {
      "epoch": 0.7162084998776409,
      "grad_norm": 0.0014208044158294797,
      "learning_rate": 0.2911759599511305,
      "loss": 0.3693,
      "num_input_tokens_seen": 8334528,
      "step": 4390
    },
    {
      "epoch": 0.7170242270984583,
      "grad_norm": 0.0017312612617388368,
      "learning_rate": 0.29115604364455777,
      "loss": 0.3477,
      "num_input_tokens_seen": 8343680,
      "step": 4395
    },
    {
      "epoch": 0.7178399543192756,
      "grad_norm": 0.0020438428036868572,
      "learning_rate": 0.2911361055699493,
      "loss": 0.3648,
      "num_input_tokens_seen": 8353296,
      "step": 4400
    },
    {
      "epoch": 0.7178399543192756,
      "eval_loss": 0.3518945574760437,
      "eval_runtime": 152.4668,
      "eval_samples_per_second": 17.873,
      "eval_steps_per_second": 8.94,
      "num_input_tokens_seen": 8353296,
      "step": 4400
    },
    {
      "epoch": 0.718655681540093,
      "grad_norm": 0.0037959045730531216,
      "learning_rate": 0.2911161457303797,
      "loss": 0.3589,
      "num_input_tokens_seen": 8362112,
      "step": 4405
    },
    {
      "epoch": 0.7194714087609103,
      "grad_norm": 0.0012153822463005781,
      "learning_rate": 0.291096164128927,
      "loss": 0.356,
      "num_input_tokens_seen": 8372016,
      "step": 4410
    },
    {
      "epoch": 0.7202871359817277,
      "grad_norm": 0.0025697401724755764,
      "learning_rate": 0.2910761607686727,
      "loss": 0.3545,
      "num_input_tokens_seen": 8380784,
      "step": 4415
    },
    {
      "epoch": 0.721102863202545,
      "grad_norm": 0.0020550964400172234,
      "learning_rate": 0.2910561356527016,
      "loss": 0.374,
      "num_input_tokens_seen": 8389792,
      "step": 4420
    },
    {
      "epoch": 0.7219185904233625,
      "grad_norm": 0.0013090346474200487,
      "learning_rate": 0.2910360887841017,
      "loss": 0.3648,
      "num_input_tokens_seen": 8399072,
      "step": 4425
    },
    {
      "epoch": 0.7227343176441798,
      "grad_norm": 0.0019328080816194415,
      "learning_rate": 0.2910160201659645,
      "loss": 0.356,
      "num_input_tokens_seen": 8407824,
      "step": 4430
    },
    {
      "epoch": 0.7235500448649972,
      "grad_norm": 0.0005362496012821794,
      "learning_rate": 0.29099592980138494,
      "loss": 0.3707,
      "num_input_tokens_seen": 8417648,
      "step": 4435
    },
    {
      "epoch": 0.7243657720858145,
      "grad_norm": 0.0004924011882394552,
      "learning_rate": 0.29097581769346115,
      "loss": 0.3631,
      "num_input_tokens_seen": 8427488,
      "step": 4440
    },
    {
      "epoch": 0.7251814993066319,
      "grad_norm": 0.00048287789104506373,
      "learning_rate": 0.29095568384529463,
      "loss": 0.3548,
      "num_input_tokens_seen": 8437072,
      "step": 4445
    },
    {
      "epoch": 0.7259972265274492,
      "grad_norm": 0.0014886732678860426,
      "learning_rate": 0.2909355282599903,
      "loss": 0.3345,
      "num_input_tokens_seen": 8446176,
      "step": 4450
    },
    {
      "epoch": 0.7268129537482666,
      "grad_norm": 0.0020935062784701586,
      "learning_rate": 0.29091535094065635,
      "loss": 0.3827,
      "num_input_tokens_seen": 8455072,
      "step": 4455
    },
    {
      "epoch": 0.727628680969084,
      "grad_norm": 0.003909138031303883,
      "learning_rate": 0.2908951518904045,
      "loss": 0.3487,
      "num_input_tokens_seen": 8464224,
      "step": 4460
    },
    {
      "epoch": 0.7284444081899013,
      "grad_norm": 0.000869783281814307,
      "learning_rate": 0.29087493111234963,
      "loss": 0.4097,
      "num_input_tokens_seen": 8474992,
      "step": 4465
    },
    {
      "epoch": 0.7292601354107187,
      "grad_norm": 0.0020814889576286077,
      "learning_rate": 0.29085468860961,
      "loss": 0.361,
      "num_input_tokens_seen": 8485680,
      "step": 4470
    },
    {
      "epoch": 0.730075862631536,
      "grad_norm": 0.0005736874300055206,
      "learning_rate": 0.2908344243853073,
      "loss": 0.367,
      "num_input_tokens_seen": 8495184,
      "step": 4475
    },
    {
      "epoch": 0.7308915898523534,
      "grad_norm": 0.001396417384967208,
      "learning_rate": 0.2908141384425666,
      "loss": 0.3513,
      "num_input_tokens_seen": 8505216,
      "step": 4480
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.0013882755301892757,
      "learning_rate": 0.2907938307845161,
      "loss": 0.3497,
      "num_input_tokens_seen": 8514704,
      "step": 4485
    },
    {
      "epoch": 0.7325230442939881,
      "grad_norm": 0.002146926010027528,
      "learning_rate": 0.2907735014142876,
      "loss": 0.3568,
      "num_input_tokens_seen": 8524320,
      "step": 4490
    },
    {
      "epoch": 0.7333387715148054,
      "grad_norm": 0.0016323317540809512,
      "learning_rate": 0.2907531503350161,
      "loss": 0.3685,
      "num_input_tokens_seen": 8533744,
      "step": 4495
    },
    {
      "epoch": 0.7341544987356228,
      "grad_norm": 0.0026022542733699083,
      "learning_rate": 0.29073277754983995,
      "loss": 0.3511,
      "num_input_tokens_seen": 8542192,
      "step": 4500
    },
    {
      "epoch": 0.7349702259564401,
      "grad_norm": 0.00309329223819077,
      "learning_rate": 0.290712383061901,
      "loss": 0.3438,
      "num_input_tokens_seen": 8551568,
      "step": 4505
    },
    {
      "epoch": 0.7357859531772575,
      "grad_norm": 0.001800190075300634,
      "learning_rate": 0.2906919668743443,
      "loss": 0.39,
      "num_input_tokens_seen": 8559984,
      "step": 4510
    },
    {
      "epoch": 0.7366016803980748,
      "grad_norm": 0.0010213229106739163,
      "learning_rate": 0.29067152899031823,
      "loss": 0.3544,
      "num_input_tokens_seen": 8569504,
      "step": 4515
    },
    {
      "epoch": 0.7374174076188923,
      "grad_norm": 0.0017170957289636135,
      "learning_rate": 0.2906510694129746,
      "loss": 0.3548,
      "num_input_tokens_seen": 8577824,
      "step": 4520
    },
    {
      "epoch": 0.7382331348397096,
      "grad_norm": 0.0018268371932208538,
      "learning_rate": 0.2906305881454685,
      "loss": 0.3673,
      "num_input_tokens_seen": 8588448,
      "step": 4525
    },
    {
      "epoch": 0.739048862060527,
      "grad_norm": 0.001069854013621807,
      "learning_rate": 0.2906100851909585,
      "loss": 0.3645,
      "num_input_tokens_seen": 8598384,
      "step": 4530
    },
    {
      "epoch": 0.7398645892813444,
      "grad_norm": 0.001284595811739564,
      "learning_rate": 0.29058956055260626,
      "loss": 0.3497,
      "num_input_tokens_seen": 8608864,
      "step": 4535
    },
    {
      "epoch": 0.7406803165021617,
      "grad_norm": 0.0013803421752527356,
      "learning_rate": 0.2905690142335771,
      "loss": 0.4063,
      "num_input_tokens_seen": 8618704,
      "step": 4540
    },
    {
      "epoch": 0.7414960437229791,
      "grad_norm": 0.0012932083336636424,
      "learning_rate": 0.29054844623703946,
      "loss": 0.3517,
      "num_input_tokens_seen": 8627360,
      "step": 4545
    },
    {
      "epoch": 0.7423117709437964,
      "grad_norm": 0.004122758749872446,
      "learning_rate": 0.2905278565661651,
      "loss": 0.4292,
      "num_input_tokens_seen": 8637728,
      "step": 4550
    },
    {
      "epoch": 0.7431274981646138,
      "grad_norm": 0.0007229445618577302,
      "learning_rate": 0.2905072452241293,
      "loss": 0.3907,
      "num_input_tokens_seen": 8646704,
      "step": 4555
    },
    {
      "epoch": 0.7439432253854311,
      "grad_norm": 0.005011997185647488,
      "learning_rate": 0.2904866122141106,
      "loss": 0.4252,
      "num_input_tokens_seen": 8656832,
      "step": 4560
    },
    {
      "epoch": 0.7447589526062485,
      "grad_norm": 0.002477700123563409,
      "learning_rate": 0.2904659575392908,
      "loss": 0.3806,
      "num_input_tokens_seen": 8665280,
      "step": 4565
    },
    {
      "epoch": 0.7455746798270658,
      "grad_norm": 0.0014291646657511592,
      "learning_rate": 0.2904452812028551,
      "loss": 0.3368,
      "num_input_tokens_seen": 8674112,
      "step": 4570
    },
    {
      "epoch": 0.7463904070478832,
      "grad_norm": 0.00228709913790226,
      "learning_rate": 0.2904245832079922,
      "loss": 0.384,
      "num_input_tokens_seen": 8683792,
      "step": 4575
    },
    {
      "epoch": 0.7472061342687005,
      "grad_norm": 0.0014792719157412648,
      "learning_rate": 0.29040386355789377,
      "loss": 0.3563,
      "num_input_tokens_seen": 8693280,
      "step": 4580
    },
    {
      "epoch": 0.7480218614895179,
      "grad_norm": 0.000599401188082993,
      "learning_rate": 0.29038312225575524,
      "loss": 0.3744,
      "num_input_tokens_seen": 8702752,
      "step": 4585
    },
    {
      "epoch": 0.7488375887103352,
      "grad_norm": 0.0038711526431143284,
      "learning_rate": 0.29036235930477505,
      "loss": 0.3932,
      "num_input_tokens_seen": 8711728,
      "step": 4590
    },
    {
      "epoch": 0.7496533159311526,
      "grad_norm": 0.003059445647522807,
      "learning_rate": 0.29034157470815514,
      "loss": 0.355,
      "num_input_tokens_seen": 8723024,
      "step": 4595
    },
    {
      "epoch": 0.7504690431519699,
      "grad_norm": 0.0006431664223782718,
      "learning_rate": 0.2903207684691008,
      "loss": 0.3577,
      "num_input_tokens_seen": 8733232,
      "step": 4600
    },
    {
      "epoch": 0.7504690431519699,
      "eval_loss": 0.3438520133495331,
      "eval_runtime": 152.4394,
      "eval_samples_per_second": 17.876,
      "eval_steps_per_second": 8.941,
      "num_input_tokens_seen": 8733232,
      "step": 4600
    },
    {
      "epoch": 0.7512847703727873,
      "grad_norm": 0.0022723576985299587,
      "learning_rate": 0.29029994059082054,
      "loss": 0.3043,
      "num_input_tokens_seen": 8743008,
      "step": 4605
    },
    {
      "epoch": 0.7521004975936046,
      "grad_norm": 0.0026244574692100286,
      "learning_rate": 0.2902790910765264,
      "loss": 0.3293,
      "num_input_tokens_seen": 8753232,
      "step": 4610
    },
    {
      "epoch": 0.7529162248144221,
      "grad_norm": 0.0021810822654515505,
      "learning_rate": 0.29025821992943346,
      "loss": 0.3442,
      "num_input_tokens_seen": 8762240,
      "step": 4615
    },
    {
      "epoch": 0.7537319520352395,
      "grad_norm": 0.002708848100155592,
      "learning_rate": 0.29023732715276046,
      "loss": 0.3086,
      "num_input_tokens_seen": 8770992,
      "step": 4620
    },
    {
      "epoch": 0.7545476792560568,
      "grad_norm": 0.0011049455497413874,
      "learning_rate": 0.2902164127497293,
      "loss": 0.3721,
      "num_input_tokens_seen": 8780432,
      "step": 4625
    },
    {
      "epoch": 0.7553634064768742,
      "grad_norm": 0.002147401450201869,
      "learning_rate": 0.2901954767235652,
      "loss": 0.3438,
      "num_input_tokens_seen": 8788800,
      "step": 4630
    },
    {
      "epoch": 0.7561791336976915,
      "grad_norm": 0.0008978354744613171,
      "learning_rate": 0.2901745190774968,
      "loss": 0.3644,
      "num_input_tokens_seen": 8797840,
      "step": 4635
    },
    {
      "epoch": 0.7569948609185089,
      "grad_norm": 0.0003910457598976791,
      "learning_rate": 0.290153539814756,
      "loss": 0.3661,
      "num_input_tokens_seen": 8807552,
      "step": 4640
    },
    {
      "epoch": 0.7578105881393262,
      "grad_norm": 0.0007968462305143476,
      "learning_rate": 0.2901325389385781,
      "loss": 0.3485,
      "num_input_tokens_seen": 8817088,
      "step": 4645
    },
    {
      "epoch": 0.7586263153601436,
      "grad_norm": 0.005035059526562691,
      "learning_rate": 0.2901115164522016,
      "loss": 0.3969,
      "num_input_tokens_seen": 8827328,
      "step": 4650
    },
    {
      "epoch": 0.7594420425809609,
      "grad_norm": 0.0017420102376490831,
      "learning_rate": 0.29009047235886865,
      "loss": 0.357,
      "num_input_tokens_seen": 8836560,
      "step": 4655
    },
    {
      "epoch": 0.7602577698017783,
      "grad_norm": 0.0029947436414659023,
      "learning_rate": 0.2900694066618243,
      "loss": 0.3754,
      "num_input_tokens_seen": 8845856,
      "step": 4660
    },
    {
      "epoch": 0.7610734970225956,
      "grad_norm": 0.0008220989257097244,
      "learning_rate": 0.2900483193643172,
      "loss": 0.3472,
      "num_input_tokens_seen": 8855840,
      "step": 4665
    },
    {
      "epoch": 0.761889224243413,
      "grad_norm": 0.004376905504614115,
      "learning_rate": 0.29002721046959934,
      "loss": 0.3483,
      "num_input_tokens_seen": 8866160,
      "step": 4670
    },
    {
      "epoch": 0.7627049514642303,
      "grad_norm": 0.0031859928276389837,
      "learning_rate": 0.29000607998092587,
      "loss": 0.3635,
      "num_input_tokens_seen": 8875456,
      "step": 4675
    },
    {
      "epoch": 0.7635206786850477,
      "grad_norm": 0.0007231806521303952,
      "learning_rate": 0.2899849279015555,
      "loss": 0.3522,
      "num_input_tokens_seen": 8884544,
      "step": 4680
    },
    {
      "epoch": 0.764336405905865,
      "grad_norm": 0.000804596405941993,
      "learning_rate": 0.28996375423475007,
      "loss": 0.3928,
      "num_input_tokens_seen": 8893648,
      "step": 4685
    },
    {
      "epoch": 0.7651521331266824,
      "grad_norm": 0.0030325439292937517,
      "learning_rate": 0.28994255898377486,
      "loss": 0.3715,
      "num_input_tokens_seen": 8902144,
      "step": 4690
    },
    {
      "epoch": 0.7659678603474998,
      "grad_norm": 0.0011791259748861194,
      "learning_rate": 0.2899213421518984,
      "loss": 0.3132,
      "num_input_tokens_seen": 8912096,
      "step": 4695
    },
    {
      "epoch": 0.7667835875683171,
      "grad_norm": 0.0027616070583462715,
      "learning_rate": 0.2899001037423926,
      "loss": 0.3298,
      "num_input_tokens_seen": 8922928,
      "step": 4700
    },
    {
      "epoch": 0.7675993147891346,
      "grad_norm": 0.0015781436813995242,
      "learning_rate": 0.28987884375853273,
      "loss": 0.3995,
      "num_input_tokens_seen": 8932576,
      "step": 4705
    },
    {
      "epoch": 0.7684150420099519,
      "grad_norm": 0.0010241982527077198,
      "learning_rate": 0.2898575622035974,
      "loss": 0.3036,
      "num_input_tokens_seen": 8942048,
      "step": 4710
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.0007784279296174645,
      "learning_rate": 0.2898362590808683,
      "loss": 0.3571,
      "num_input_tokens_seen": 8951296,
      "step": 4715
    },
    {
      "epoch": 0.7700464964515866,
      "grad_norm": 0.0013254666700959206,
      "learning_rate": 0.2898149343936308,
      "loss": 0.3442,
      "num_input_tokens_seen": 8959968,
      "step": 4720
    },
    {
      "epoch": 0.770862223672404,
      "grad_norm": 0.0011382079683244228,
      "learning_rate": 0.2897935881451734,
      "loss": 0.3641,
      "num_input_tokens_seen": 8969424,
      "step": 4725
    },
    {
      "epoch": 0.7716779508932213,
      "grad_norm": 0.0023346086964011192,
      "learning_rate": 0.28977222033878797,
      "loss": 0.3435,
      "num_input_tokens_seen": 8979248,
      "step": 4730
    },
    {
      "epoch": 0.7724936781140387,
      "grad_norm": 0.002974454080685973,
      "learning_rate": 0.28975083097776966,
      "loss": 0.3161,
      "num_input_tokens_seen": 8989760,
      "step": 4735
    },
    {
      "epoch": 0.773309405334856,
      "grad_norm": 0.0047255209647119045,
      "learning_rate": 0.28972942006541696,
      "loss": 0.3834,
      "num_input_tokens_seen": 8997920,
      "step": 4740
    },
    {
      "epoch": 0.7741251325556734,
      "grad_norm": 0.0022184664849191904,
      "learning_rate": 0.2897079876050318,
      "loss": 0.3974,
      "num_input_tokens_seen": 9007488,
      "step": 4745
    },
    {
      "epoch": 0.7749408597764907,
      "grad_norm": 0.0009656011243350804,
      "learning_rate": 0.2896865335999192,
      "loss": 0.361,
      "num_input_tokens_seen": 9016576,
      "step": 4750
    },
    {
      "epoch": 0.7757565869973081,
      "grad_norm": 0.0006395948003046215,
      "learning_rate": 0.28966505805338777,
      "loss": 0.3757,
      "num_input_tokens_seen": 9026656,
      "step": 4755
    },
    {
      "epoch": 0.7765723142181254,
      "grad_norm": 0.0031855469569563866,
      "learning_rate": 0.2896435609687492,
      "loss": 0.3279,
      "num_input_tokens_seen": 9035872,
      "step": 4760
    },
    {
      "epoch": 0.7773880414389428,
      "grad_norm": 0.003414687467738986,
      "learning_rate": 0.2896220423493187,
      "loss": 0.3618,
      "num_input_tokens_seen": 9046240,
      "step": 4765
    },
    {
      "epoch": 0.7782037686597602,
      "grad_norm": 0.0013357216957956553,
      "learning_rate": 0.28960050219841466,
      "loss": 0.371,
      "num_input_tokens_seen": 9056128,
      "step": 4770
    },
    {
      "epoch": 0.7790194958805775,
      "grad_norm": 0.0010903386864811182,
      "learning_rate": 0.28957894051935884,
      "loss": 0.3644,
      "num_input_tokens_seen": 9064368,
      "step": 4775
    },
    {
      "epoch": 0.7798352231013949,
      "grad_norm": 0.00099807302467525,
      "learning_rate": 0.2895573573154764,
      "loss": 0.3737,
      "num_input_tokens_seen": 9074048,
      "step": 4780
    },
    {
      "epoch": 0.7806509503222122,
      "grad_norm": 0.0005974062951281667,
      "learning_rate": 0.28953575259009556,
      "loss": 0.403,
      "num_input_tokens_seen": 9084176,
      "step": 4785
    },
    {
      "epoch": 0.7814666775430296,
      "grad_norm": 0.0012839181581512094,
      "learning_rate": 0.2895141263465482,
      "loss": 0.3566,
      "num_input_tokens_seen": 9094016,
      "step": 4790
    },
    {
      "epoch": 0.7822824047638469,
      "grad_norm": 0.002980039920657873,
      "learning_rate": 0.28949247858816934,
      "loss": 0.3378,
      "num_input_tokens_seen": 9103680,
      "step": 4795
    },
    {
      "epoch": 0.7830981319846644,
      "grad_norm": 0.0014586973702535033,
      "learning_rate": 0.2894708093182973,
      "loss": 0.3691,
      "num_input_tokens_seen": 9113632,
      "step": 4800
    },
    {
      "epoch": 0.7830981319846644,
      "eval_loss": 0.3404499292373657,
      "eval_runtime": 152.7278,
      "eval_samples_per_second": 17.842,
      "eval_steps_per_second": 8.924,
      "num_input_tokens_seen": 9113632,
      "step": 4800
    },
    {
      "epoch": 0.7839138592054817,
      "grad_norm": 0.0016386915231123567,
      "learning_rate": 0.2894491185402737,
      "loss": 0.3497,
      "num_input_tokens_seen": 9122192,
      "step": 4805
    },
    {
      "epoch": 0.7847295864262991,
      "grad_norm": 0.0017900982638821006,
      "learning_rate": 0.2894274062574437,
      "loss": 0.3643,
      "num_input_tokens_seen": 9132512,
      "step": 4810
    },
    {
      "epoch": 0.7855453136471164,
      "grad_norm": 0.003160003339871764,
      "learning_rate": 0.2894056724731554,
      "loss": 0.3325,
      "num_input_tokens_seen": 9141264,
      "step": 4815
    },
    {
      "epoch": 0.7863610408679338,
      "grad_norm": 0.0007771363016217947,
      "learning_rate": 0.28938391719076056,
      "loss": 0.3495,
      "num_input_tokens_seen": 9150160,
      "step": 4820
    },
    {
      "epoch": 0.7871767680887511,
      "grad_norm": 0.001155200065113604,
      "learning_rate": 0.28936214041361413,
      "loss": 0.3493,
      "num_input_tokens_seen": 9160096,
      "step": 4825
    },
    {
      "epoch": 0.7879924953095685,
      "grad_norm": 0.0017050483729690313,
      "learning_rate": 0.2893403421450743,
      "loss": 0.3468,
      "num_input_tokens_seen": 9168864,
      "step": 4830
    },
    {
      "epoch": 0.7888082225303858,
      "grad_norm": 0.001130185672082007,
      "learning_rate": 0.2893185223885026,
      "loss": 0.3392,
      "num_input_tokens_seen": 9178608,
      "step": 4835
    },
    {
      "epoch": 0.7896239497512032,
      "grad_norm": 0.0028392954263836145,
      "learning_rate": 0.289296681147264,
      "loss": 0.3741,
      "num_input_tokens_seen": 9187824,
      "step": 4840
    },
    {
      "epoch": 0.7904396769720206,
      "grad_norm": 0.0025137681514024734,
      "learning_rate": 0.28927481842472663,
      "loss": 0.3049,
      "num_input_tokens_seen": 9196464,
      "step": 4845
    },
    {
      "epoch": 0.7912554041928379,
      "grad_norm": 0.0049722022376954556,
      "learning_rate": 0.28925293422426207,
      "loss": 0.3989,
      "num_input_tokens_seen": 9206544,
      "step": 4850
    },
    {
      "epoch": 0.7920711314136553,
      "grad_norm": 0.0021759839728474617,
      "learning_rate": 0.28923102854924504,
      "loss": 0.3995,
      "num_input_tokens_seen": 9214288,
      "step": 4855
    },
    {
      "epoch": 0.7928868586344726,
      "grad_norm": 0.0026012035086750984,
      "learning_rate": 0.2892091014030537,
      "loss": 0.3939,
      "num_input_tokens_seen": 9222496,
      "step": 4860
    },
    {
      "epoch": 0.79370258585529,
      "grad_norm": 0.001392912003211677,
      "learning_rate": 0.2891871527890696,
      "loss": 0.3296,
      "num_input_tokens_seen": 9232480,
      "step": 4865
    },
    {
      "epoch": 0.7945183130761073,
      "grad_norm": 0.004026112146675587,
      "learning_rate": 0.2891651827106773,
      "loss": 0.3719,
      "num_input_tokens_seen": 9241408,
      "step": 4870
    },
    {
      "epoch": 0.7953340402969247,
      "grad_norm": 0.0032388942781835794,
      "learning_rate": 0.2891431911712651,
      "loss": 0.2639,
      "num_input_tokens_seen": 9250880,
      "step": 4875
    },
    {
      "epoch": 0.796149767517742,
      "grad_norm": 0.0020919875241816044,
      "learning_rate": 0.2891211781742241,
      "loss": 0.2821,
      "num_input_tokens_seen": 9259808,
      "step": 4880
    },
    {
      "epoch": 0.7969654947385594,
      "grad_norm": 0.001343271811492741,
      "learning_rate": 0.2890991437229492,
      "loss": 0.382,
      "num_input_tokens_seen": 9269552,
      "step": 4885
    },
    {
      "epoch": 0.7977812219593767,
      "grad_norm": 0.0033893196377903223,
      "learning_rate": 0.2890770878208383,
      "loss": 0.3533,
      "num_input_tokens_seen": 9279104,
      "step": 4890
    },
    {
      "epoch": 0.7985969491801942,
      "grad_norm": 0.0005033938796259463,
      "learning_rate": 0.28905501047129273,
      "loss": 0.3546,
      "num_input_tokens_seen": 9288944,
      "step": 4895
    },
    {
      "epoch": 0.7994126764010115,
      "grad_norm": 0.0020487187430262566,
      "learning_rate": 0.289032911677717,
      "loss": 0.3712,
      "num_input_tokens_seen": 9297504,
      "step": 4900
    },
    {
      "epoch": 0.8002284036218289,
      "grad_norm": 0.001635025953873992,
      "learning_rate": 0.28901079144351915,
      "loss": 0.3556,
      "num_input_tokens_seen": 9307856,
      "step": 4905
    },
    {
      "epoch": 0.8010441308426463,
      "grad_norm": 0.0007499021012336016,
      "learning_rate": 0.2889886497721103,
      "loss": 0.3601,
      "num_input_tokens_seen": 9316432,
      "step": 4910
    },
    {
      "epoch": 0.8018598580634636,
      "grad_norm": 0.0020320601761341095,
      "learning_rate": 0.28896648666690505,
      "loss": 0.3413,
      "num_input_tokens_seen": 9326016,
      "step": 4915
    },
    {
      "epoch": 0.802675585284281,
      "grad_norm": 0.00068518309853971,
      "learning_rate": 0.2889443021313212,
      "loss": 0.3509,
      "num_input_tokens_seen": 9333776,
      "step": 4920
    },
    {
      "epoch": 0.8034913125050983,
      "grad_norm": 0.002355481730774045,
      "learning_rate": 0.28892209616877984,
      "loss": 0.341,
      "num_input_tokens_seen": 9342640,
      "step": 4925
    },
    {
      "epoch": 0.8043070397259157,
      "grad_norm": 0.005637557711452246,
      "learning_rate": 0.28889986878270546,
      "loss": 0.3891,
      "num_input_tokens_seen": 9352112,
      "step": 4930
    },
    {
      "epoch": 0.805122766946733,
      "grad_norm": 0.0013455436564981937,
      "learning_rate": 0.28887761997652583,
      "loss": 0.3974,
      "num_input_tokens_seen": 9361088,
      "step": 4935
    },
    {
      "epoch": 0.8059384941675504,
      "grad_norm": 0.0010872866259887815,
      "learning_rate": 0.2888553497536719,
      "loss": 0.3678,
      "num_input_tokens_seen": 9370768,
      "step": 4940
    },
    {
      "epoch": 0.8067542213883677,
      "grad_norm": 0.002567209303379059,
      "learning_rate": 0.2888330581175781,
      "loss": 0.4439,
      "num_input_tokens_seen": 9381104,
      "step": 4945
    },
    {
      "epoch": 0.8075699486091851,
      "grad_norm": 0.0008604391478002071,
      "learning_rate": 0.28881074507168203,
      "loss": 0.3614,
      "num_input_tokens_seen": 9391568,
      "step": 4950
    },
    {
      "epoch": 0.8083856758300024,
      "grad_norm": 0.0004942660452798009,
      "learning_rate": 0.2887884106194247,
      "loss": 0.3814,
      "num_input_tokens_seen": 9401568,
      "step": 4955
    },
    {
      "epoch": 0.8092014030508198,
      "grad_norm": 0.0036541735753417015,
      "learning_rate": 0.28876605476425027,
      "loss": 0.3878,
      "num_input_tokens_seen": 9409248,
      "step": 4960
    },
    {
      "epoch": 0.8100171302716371,
      "grad_norm": 0.0013117118505761027,
      "learning_rate": 0.2887436775096064,
      "loss": 0.376,
      "num_input_tokens_seen": 9418944,
      "step": 4965
    },
    {
      "epoch": 0.8108328574924545,
      "grad_norm": 0.0022050661500543356,
      "learning_rate": 0.2887212788589439,
      "loss": 0.4296,
      "num_input_tokens_seen": 9429168,
      "step": 4970
    },
    {
      "epoch": 0.8116485847132718,
      "grad_norm": 0.001376767992042005,
      "learning_rate": 0.2886988588157169,
      "loss": 0.3536,
      "num_input_tokens_seen": 9440128,
      "step": 4975
    },
    {
      "epoch": 0.8124643119340892,
      "grad_norm": 0.003497577738016844,
      "learning_rate": 0.28867641738338284,
      "loss": 0.3598,
      "num_input_tokens_seen": 9449712,
      "step": 4980
    },
    {
      "epoch": 0.8132800391549067,
      "grad_norm": 0.0011276291916146874,
      "learning_rate": 0.2886539545654026,
      "loss": 0.3381,
      "num_input_tokens_seen": 9459376,
      "step": 4985
    },
    {
      "epoch": 0.814095766375724,
      "grad_norm": 0.0015840701526030898,
      "learning_rate": 0.28863147036524006,
      "loss": 0.3984,
      "num_input_tokens_seen": 9469520,
      "step": 4990
    },
    {
      "epoch": 0.8149114935965414,
      "grad_norm": 0.00040190352592617273,
      "learning_rate": 0.2886089647863626,
      "loss": 0.357,
      "num_input_tokens_seen": 9478032,
      "step": 4995
    },
    {
      "epoch": 0.8157272208173587,
      "grad_norm": 0.0007532895542681217,
      "learning_rate": 0.288586437832241,
      "loss": 0.3779,
      "num_input_tokens_seen": 9487952,
      "step": 5000
    },
    {
      "epoch": 0.8157272208173587,
      "eval_loss": 0.3713279664516449,
      "eval_runtime": 152.405,
      "eval_samples_per_second": 17.88,
      "eval_steps_per_second": 8.943,
      "num_input_tokens_seen": 9487952,
      "step": 5000
    },
    {
      "epoch": 0.8165429480381761,
      "grad_norm": 0.001900658244267106,
      "learning_rate": 0.28856388950634904,
      "loss": 0.3712,
      "num_input_tokens_seen": 9497568,
      "step": 5005
    },
    {
      "epoch": 0.8173586752589934,
      "grad_norm": 0.0007896355236880481,
      "learning_rate": 0.288541319812164,
      "loss": 0.3518,
      "num_input_tokens_seen": 9507776,
      "step": 5010
    },
    {
      "epoch": 0.8181744024798108,
      "grad_norm": 0.0006070649833418429,
      "learning_rate": 0.2885187287531665,
      "loss": 0.3461,
      "num_input_tokens_seen": 9516320,
      "step": 5015
    },
    {
      "epoch": 0.8189901297006281,
      "grad_norm": 0.00261489930562675,
      "learning_rate": 0.2884961163328402,
      "loss": 0.3317,
      "num_input_tokens_seen": 9526640,
      "step": 5020
    },
    {
      "epoch": 0.8198058569214455,
      "grad_norm": 0.0013423545751720667,
      "learning_rate": 0.28847348255467237,
      "loss": 0.3944,
      "num_input_tokens_seen": 9535952,
      "step": 5025
    },
    {
      "epoch": 0.8206215841422628,
      "grad_norm": 0.0013658229727298021,
      "learning_rate": 0.28845082742215333,
      "loss": 0.3661,
      "num_input_tokens_seen": 9545328,
      "step": 5030
    },
    {
      "epoch": 0.8214373113630802,
      "grad_norm": 0.0023482756223529577,
      "learning_rate": 0.2884281509387769,
      "loss": 0.3829,
      "num_input_tokens_seen": 9555520,
      "step": 5035
    },
    {
      "epoch": 0.8222530385838975,
      "grad_norm": 0.0026246889028698206,
      "learning_rate": 0.2884054531080399,
      "loss": 0.3853,
      "num_input_tokens_seen": 9564752,
      "step": 5040
    },
    {
      "epoch": 0.8230687658047149,
      "grad_norm": 0.0021749241277575493,
      "learning_rate": 0.28838273393344277,
      "loss": 0.3217,
      "num_input_tokens_seen": 9573680,
      "step": 5045
    },
    {
      "epoch": 0.8238844930255322,
      "grad_norm": 0.0023490427993237972,
      "learning_rate": 0.288359993418489,
      "loss": 0.3317,
      "num_input_tokens_seen": 9583072,
      "step": 5050
    },
    {
      "epoch": 0.8247002202463496,
      "grad_norm": 0.0009606581879779696,
      "learning_rate": 0.28833723156668556,
      "loss": 0.359,
      "num_input_tokens_seen": 9592816,
      "step": 5055
    },
    {
      "epoch": 0.8255159474671669,
      "grad_norm": 0.0005416304338723421,
      "learning_rate": 0.2883144483815425,
      "loss": 0.3563,
      "num_input_tokens_seen": 9602000,
      "step": 5060
    },
    {
      "epoch": 0.8263316746879843,
      "grad_norm": 0.0015198822366073728,
      "learning_rate": 0.28829164386657335,
      "loss": 0.3475,
      "num_input_tokens_seen": 9610176,
      "step": 5065
    },
    {
      "epoch": 0.8271474019088016,
      "grad_norm": 0.0007706001051701605,
      "learning_rate": 0.28826881802529486,
      "loss": 0.3461,
      "num_input_tokens_seen": 9618752,
      "step": 5070
    },
    {
      "epoch": 0.827963129129619,
      "grad_norm": 0.0038988813757896423,
      "learning_rate": 0.28824597086122705,
      "loss": 0.3995,
      "num_input_tokens_seen": 9628128,
      "step": 5075
    },
    {
      "epoch": 0.8287788563504365,
      "grad_norm": 0.000644067709799856,
      "learning_rate": 0.28822310237789317,
      "loss": 0.3127,
      "num_input_tokens_seen": 9636960,
      "step": 5080
    },
    {
      "epoch": 0.8295945835712538,
      "grad_norm": 0.002834655810147524,
      "learning_rate": 0.2882002125788199,
      "loss": 0.4206,
      "num_input_tokens_seen": 9646928,
      "step": 5085
    },
    {
      "epoch": 0.8304103107920712,
      "grad_norm": 0.002722094999626279,
      "learning_rate": 0.2881773014675371,
      "loss": 0.4605,
      "num_input_tokens_seen": 9655216,
      "step": 5090
    },
    {
      "epoch": 0.8312260380128885,
      "grad_norm": 0.0010035240557044744,
      "learning_rate": 0.288154369047578,
      "loss": 0.3724,
      "num_input_tokens_seen": 9664544,
      "step": 5095
    },
    {
      "epoch": 0.8320417652337059,
      "grad_norm": 0.0014099710388109088,
      "learning_rate": 0.28813141532247905,
      "loss": 0.4033,
      "num_input_tokens_seen": 9673856,
      "step": 5100
    },
    {
      "epoch": 0.8328574924545232,
      "grad_norm": 0.002778296358883381,
      "learning_rate": 0.28810844029578,
      "loss": 0.3616,
      "num_input_tokens_seen": 9683440,
      "step": 5105
    },
    {
      "epoch": 0.8336732196753406,
      "grad_norm": 0.0029431069269776344,
      "learning_rate": 0.2880854439710238,
      "loss": 0.3904,
      "num_input_tokens_seen": 9693392,
      "step": 5110
    },
    {
      "epoch": 0.8344889468961579,
      "grad_norm": 0.0010649880860000849,
      "learning_rate": 0.28806242635175694,
      "loss": 0.4445,
      "num_input_tokens_seen": 9703104,
      "step": 5115
    },
    {
      "epoch": 0.8353046741169753,
      "grad_norm": 0.0026069285813719034,
      "learning_rate": 0.2880393874415289,
      "loss": 0.5087,
      "num_input_tokens_seen": 9711760,
      "step": 5120
    },
    {
      "epoch": 0.8361204013377926,
      "grad_norm": 0.0018362143309786916,
      "learning_rate": 0.2880163272438926,
      "loss": 0.3382,
      "num_input_tokens_seen": 9721072,
      "step": 5125
    },
    {
      "epoch": 0.83693612855861,
      "grad_norm": 0.004021104890853167,
      "learning_rate": 0.2879932457624042,
      "loss": 0.4393,
      "num_input_tokens_seen": 9730672,
      "step": 5130
    },
    {
      "epoch": 0.8377518557794273,
      "grad_norm": 0.0023897860664874315,
      "learning_rate": 0.2879701430006232,
      "loss": 0.4189,
      "num_input_tokens_seen": 9740208,
      "step": 5135
    },
    {
      "epoch": 0.8385675830002447,
      "grad_norm": 0.0008107746834866703,
      "learning_rate": 0.28794701896211233,
      "loss": 0.3961,
      "num_input_tokens_seen": 9749616,
      "step": 5140
    },
    {
      "epoch": 0.839383310221062,
      "grad_norm": 0.0014983352739363909,
      "learning_rate": 0.28792387365043753,
      "loss": 0.3499,
      "num_input_tokens_seen": 9757760,
      "step": 5145
    },
    {
      "epoch": 0.8401990374418794,
      "grad_norm": 0.0018834671936929226,
      "learning_rate": 0.28790070706916815,
      "loss": 0.3675,
      "num_input_tokens_seen": 9766016,
      "step": 5150
    },
    {
      "epoch": 0.8410147646626968,
      "grad_norm": 0.0015786646399646997,
      "learning_rate": 0.2878775192218768,
      "loss": 0.3788,
      "num_input_tokens_seen": 9774112,
      "step": 5155
    },
    {
      "epoch": 0.8418304918835141,
      "grad_norm": 0.0035498940851539373,
      "learning_rate": 0.2878543101121393,
      "loss": 0.3838,
      "num_input_tokens_seen": 9784512,
      "step": 5160
    },
    {
      "epoch": 0.8426462191043315,
      "grad_norm": 0.0024980437010526657,
      "learning_rate": 0.28783107974353483,
      "loss": 0.3595,
      "num_input_tokens_seen": 9794288,
      "step": 5165
    },
    {
      "epoch": 0.8434619463251488,
      "grad_norm": 0.0008053541532717645,
      "learning_rate": 0.2878078281196457,
      "loss": 0.3846,
      "num_input_tokens_seen": 9802848,
      "step": 5170
    },
    {
      "epoch": 0.8442776735459663,
      "grad_norm": 0.0006052999524399638,
      "learning_rate": 0.28778455524405777,
      "loss": 0.3822,
      "num_input_tokens_seen": 9812720,
      "step": 5175
    },
    {
      "epoch": 0.8450934007667836,
      "grad_norm": 0.000794540683273226,
      "learning_rate": 0.2877612611203598,
      "loss": 0.3521,
      "num_input_tokens_seen": 9822928,
      "step": 5180
    },
    {
      "epoch": 0.845909127987601,
      "grad_norm": 0.0027700935024768114,
      "learning_rate": 0.28773794575214423,
      "loss": 0.3368,
      "num_input_tokens_seen": 9832176,
      "step": 5185
    },
    {
      "epoch": 0.8467248552084183,
      "grad_norm": 0.0006579187465831637,
      "learning_rate": 0.28771460914300645,
      "loss": 0.4612,
      "num_input_tokens_seen": 9842352,
      "step": 5190
    },
    {
      "epoch": 0.8475405824292357,
      "grad_norm": 0.0017559112748131156,
      "learning_rate": 0.2876912512965454,
      "loss": 0.3451,
      "num_input_tokens_seen": 9852448,
      "step": 5195
    },
    {
      "epoch": 0.848356309650053,
      "grad_norm": 0.0005460568936541677,
      "learning_rate": 0.287667872216363,
      "loss": 0.3499,
      "num_input_tokens_seen": 9861104,
      "step": 5200
    },
    {
      "epoch": 0.848356309650053,
      "eval_loss": 0.3860589265823364,
      "eval_runtime": 152.5321,
      "eval_samples_per_second": 17.865,
      "eval_steps_per_second": 8.936,
      "num_input_tokens_seen": 9861104,
      "step": 5200
    },
    {
      "epoch": 0.8491720368708704,
      "grad_norm": 0.0010514190653339028,
      "learning_rate": 0.2876444719060647,
      "loss": 0.4023,
      "num_input_tokens_seen": 9870608,
      "step": 5205
    },
    {
      "epoch": 0.8499877640916877,
      "grad_norm": 0.0010875442530959845,
      "learning_rate": 0.287621050369259,
      "loss": 0.3468,
      "num_input_tokens_seen": 9878880,
      "step": 5210
    },
    {
      "epoch": 0.8508034913125051,
      "grad_norm": 0.000611509196460247,
      "learning_rate": 0.28759760760955794,
      "loss": 0.356,
      "num_input_tokens_seen": 9887856,
      "step": 5215
    },
    {
      "epoch": 0.8516192185333225,
      "grad_norm": 0.0029587317258119583,
      "learning_rate": 0.2875741436305766,
      "loss": 0.3718,
      "num_input_tokens_seen": 9898224,
      "step": 5220
    },
    {
      "epoch": 0.8524349457541398,
      "grad_norm": 0.0007331509259529412,
      "learning_rate": 0.28755065843593347,
      "loss": 0.3528,
      "num_input_tokens_seen": 9906864,
      "step": 5225
    },
    {
      "epoch": 0.8532506729749572,
      "grad_norm": 0.0012116518337279558,
      "learning_rate": 0.2875271520292502,
      "loss": 0.3372,
      "num_input_tokens_seen": 9916544,
      "step": 5230
    },
    {
      "epoch": 0.8540664001957745,
      "grad_norm": 0.0006405313033610582,
      "learning_rate": 0.28750362441415184,
      "loss": 0.3453,
      "num_input_tokens_seen": 9925456,
      "step": 5235
    },
    {
      "epoch": 0.8548821274165919,
      "grad_norm": 0.0016552532324567437,
      "learning_rate": 0.28748007559426664,
      "loss": 0.367,
      "num_input_tokens_seen": 9934224,
      "step": 5240
    },
    {
      "epoch": 0.8556978546374092,
      "grad_norm": 0.001214861636981368,
      "learning_rate": 0.2874565055732261,
      "loss": 0.3568,
      "num_input_tokens_seen": 9943072,
      "step": 5245
    },
    {
      "epoch": 0.8565135818582266,
      "grad_norm": 0.0029843247029930353,
      "learning_rate": 0.28743291435466495,
      "loss": 0.3741,
      "num_input_tokens_seen": 9953472,
      "step": 5250
    },
    {
      "epoch": 0.8573293090790439,
      "grad_norm": 0.0004847095115110278,
      "learning_rate": 0.2874093019422214,
      "loss": 0.3456,
      "num_input_tokens_seen": 9961792,
      "step": 5255
    },
    {
      "epoch": 0.8581450362998613,
      "grad_norm": 0.00236251181922853,
      "learning_rate": 0.28738566833953666,
      "loss": 0.3887,
      "num_input_tokens_seen": 9971040,
      "step": 5260
    },
    {
      "epoch": 0.8589607635206787,
      "grad_norm": 0.0009613488800823689,
      "learning_rate": 0.28736201355025537,
      "loss": 0.3434,
      "num_input_tokens_seen": 9980832,
      "step": 5265
    },
    {
      "epoch": 0.8597764907414961,
      "grad_norm": 0.0018506987253203988,
      "learning_rate": 0.28733833757802535,
      "loss": 0.3731,
      "num_input_tokens_seen": 9990368,
      "step": 5270
    },
    {
      "epoch": 0.8605922179623134,
      "grad_norm": 0.001428880263119936,
      "learning_rate": 0.28731464042649785,
      "loss": 0.3405,
      "num_input_tokens_seen": 9998544,
      "step": 5275
    },
    {
      "epoch": 0.8614079451831308,
      "grad_norm": 0.0027385444846004248,
      "learning_rate": 0.2872909220993271,
      "loss": 0.3573,
      "num_input_tokens_seen": 10008640,
      "step": 5280
    },
    {
      "epoch": 0.8622236724039481,
      "grad_norm": 0.0015561298932880163,
      "learning_rate": 0.287267182600171,
      "loss": 0.3768,
      "num_input_tokens_seen": 10018336,
      "step": 5285
    },
    {
      "epoch": 0.8630393996247655,
      "grad_norm": 0.0014420811785385013,
      "learning_rate": 0.2872434219326902,
      "loss": 0.3457,
      "num_input_tokens_seen": 10028464,
      "step": 5290
    },
    {
      "epoch": 0.8638551268455829,
      "grad_norm": 0.0017497777007520199,
      "learning_rate": 0.28721964010054907,
      "loss": 0.3923,
      "num_input_tokens_seen": 10038560,
      "step": 5295
    },
    {
      "epoch": 0.8646708540664002,
      "grad_norm": 0.00047054511378519237,
      "learning_rate": 0.28719583710741503,
      "loss": 0.3696,
      "num_input_tokens_seen": 10046560,
      "step": 5300
    },
    {
      "epoch": 0.8654865812872176,
      "grad_norm": 0.0010279109701514244,
      "learning_rate": 0.28717201295695877,
      "loss": 0.3512,
      "num_input_tokens_seen": 10055696,
      "step": 5305
    },
    {
      "epoch": 0.8663023085080349,
      "grad_norm": 0.0007802279433235526,
      "learning_rate": 0.28714816765285434,
      "loss": 0.3707,
      "num_input_tokens_seen": 10065136,
      "step": 5310
    },
    {
      "epoch": 0.8671180357288523,
      "grad_norm": 0.0014393238816410303,
      "learning_rate": 0.28712430119877896,
      "loss": 0.3723,
      "num_input_tokens_seen": 10075824,
      "step": 5315
    },
    {
      "epoch": 0.8679337629496696,
      "grad_norm": 0.0013941999059170485,
      "learning_rate": 0.28710041359841304,
      "loss": 0.4114,
      "num_input_tokens_seen": 10085712,
      "step": 5320
    },
    {
      "epoch": 0.868749490170487,
      "grad_norm": 0.0006195633322931826,
      "learning_rate": 0.28707650485544056,
      "loss": 0.3702,
      "num_input_tokens_seen": 10096272,
      "step": 5325
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.004015599377453327,
      "learning_rate": 0.28705257497354836,
      "loss": 0.3857,
      "num_input_tokens_seen": 10104864,
      "step": 5330
    },
    {
      "epoch": 0.8703809446121217,
      "grad_norm": 0.0010701881255954504,
      "learning_rate": 0.28702862395642675,
      "loss": 0.3204,
      "num_input_tokens_seen": 10114656,
      "step": 5335
    },
    {
      "epoch": 0.871196671832939,
      "grad_norm": 0.0005101431743241847,
      "learning_rate": 0.28700465180776935,
      "loss": 0.3935,
      "num_input_tokens_seen": 10124288,
      "step": 5340
    },
    {
      "epoch": 0.8720123990537564,
      "grad_norm": 0.0005955781089141965,
      "learning_rate": 0.2869806585312729,
      "loss": 0.3335,
      "num_input_tokens_seen": 10134224,
      "step": 5345
    },
    {
      "epoch": 0.8728281262745737,
      "grad_norm": 0.0034062154591083527,
      "learning_rate": 0.28695664413063754,
      "loss": 0.4058,
      "num_input_tokens_seen": 10143840,
      "step": 5350
    },
    {
      "epoch": 0.8736438534953911,
      "grad_norm": 0.0012935565318912268,
      "learning_rate": 0.28693260860956654,
      "loss": 0.3397,
      "num_input_tokens_seen": 10153296,
      "step": 5355
    },
    {
      "epoch": 0.8744595807162086,
      "grad_norm": 0.003446738701313734,
      "learning_rate": 0.2869085519717665,
      "loss": 0.3701,
      "num_input_tokens_seen": 10164160,
      "step": 5360
    },
    {
      "epoch": 0.8752753079370259,
      "grad_norm": 0.000722623139154166,
      "learning_rate": 0.28688447422094726,
      "loss": 0.3883,
      "num_input_tokens_seen": 10174336,
      "step": 5365
    },
    {
      "epoch": 0.8760910351578433,
      "grad_norm": 0.00103437015786767,
      "learning_rate": 0.2868603753608219,
      "loss": 0.355,
      "num_input_tokens_seen": 10184640,
      "step": 5370
    },
    {
      "epoch": 0.8769067623786606,
      "grad_norm": 0.001861593103967607,
      "learning_rate": 0.28683625539510665,
      "loss": 0.3867,
      "num_input_tokens_seen": 10194080,
      "step": 5375
    },
    {
      "epoch": 0.877722489599478,
      "grad_norm": 0.0012979793827980757,
      "learning_rate": 0.28681211432752135,
      "loss": 0.3466,
      "num_input_tokens_seen": 10203200,
      "step": 5380
    },
    {
      "epoch": 0.8785382168202953,
      "grad_norm": 0.005057348404079676,
      "learning_rate": 0.2867879521617887,
      "loss": 0.3894,
      "num_input_tokens_seen": 10211792,
      "step": 5385
    },
    {
      "epoch": 0.8793539440411127,
      "grad_norm": 0.0009248919086530805,
      "learning_rate": 0.28676376890163485,
      "loss": 0.2698,
      "num_input_tokens_seen": 10221712,
      "step": 5390
    },
    {
      "epoch": 0.88016967126193,
      "grad_norm": 0.0018403398571535945,
      "learning_rate": 0.2867395645507891,
      "loss": 0.272,
      "num_input_tokens_seen": 10230400,
      "step": 5395
    },
    {
      "epoch": 0.8809853984827474,
      "grad_norm": 0.0033119749277830124,
      "learning_rate": 0.2867153391129842,
      "loss": 0.3244,
      "num_input_tokens_seen": 10239088,
      "step": 5400
    },
    {
      "epoch": 0.8809853984827474,
      "eval_loss": 0.3939324617385864,
      "eval_runtime": 152.6307,
      "eval_samples_per_second": 17.854,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 10239088,
      "step": 5400
    },
    {
      "epoch": 0.8818011257035647,
      "grad_norm": 0.001619360176846385,
      "learning_rate": 0.28669109259195585,
      "loss": 0.431,
      "num_input_tokens_seen": 10249088,
      "step": 5405
    },
    {
      "epoch": 0.8826168529243821,
      "grad_norm": 0.0007635471993125975,
      "learning_rate": 0.2866668249914433,
      "loss": 0.3591,
      "num_input_tokens_seen": 10257088,
      "step": 5410
    },
    {
      "epoch": 0.8834325801451994,
      "grad_norm": 0.0005895787035115063,
      "learning_rate": 0.2866425363151889,
      "loss": 0.3833,
      "num_input_tokens_seen": 10266992,
      "step": 5415
    },
    {
      "epoch": 0.8842483073660168,
      "grad_norm": 0.002813079860061407,
      "learning_rate": 0.2866182265669382,
      "loss": 0.3888,
      "num_input_tokens_seen": 10276000,
      "step": 5420
    },
    {
      "epoch": 0.8850640345868341,
      "grad_norm": 0.0027916384860873222,
      "learning_rate": 0.28659389575044014,
      "loss": 0.3401,
      "num_input_tokens_seen": 10285024,
      "step": 5425
    },
    {
      "epoch": 0.8858797618076515,
      "grad_norm": 0.000943958293646574,
      "learning_rate": 0.28656954386944683,
      "loss": 0.3653,
      "num_input_tokens_seen": 10295472,
      "step": 5430
    },
    {
      "epoch": 0.8866954890284688,
      "grad_norm": 0.0007565248524770141,
      "learning_rate": 0.28654517092771353,
      "loss": 0.3054,
      "num_input_tokens_seen": 10304896,
      "step": 5435
    },
    {
      "epoch": 0.8875112162492862,
      "grad_norm": 0.00306252995505929,
      "learning_rate": 0.286520776928999,
      "loss": 0.386,
      "num_input_tokens_seen": 10314560,
      "step": 5440
    },
    {
      "epoch": 0.8883269434701035,
      "grad_norm": 0.0021661340724676847,
      "learning_rate": 0.286496361877065,
      "loss": 0.3377,
      "num_input_tokens_seen": 10324064,
      "step": 5445
    },
    {
      "epoch": 0.889142670690921,
      "grad_norm": 0.001096583902835846,
      "learning_rate": 0.28647192577567676,
      "loss": 0.3603,
      "num_input_tokens_seen": 10333968,
      "step": 5450
    },
    {
      "epoch": 0.8899583979117384,
      "grad_norm": 0.0020656022243201733,
      "learning_rate": 0.28644746862860254,
      "loss": 0.3969,
      "num_input_tokens_seen": 10342832,
      "step": 5455
    },
    {
      "epoch": 0.8907741251325557,
      "grad_norm": 0.0017311809351667762,
      "learning_rate": 0.2864229904396139,
      "loss": 0.4085,
      "num_input_tokens_seen": 10352384,
      "step": 5460
    },
    {
      "epoch": 0.8915898523533731,
      "grad_norm": 0.0014213499380275607,
      "learning_rate": 0.28639849121248573,
      "loss": 0.3453,
      "num_input_tokens_seen": 10360656,
      "step": 5465
    },
    {
      "epoch": 0.8924055795741904,
      "grad_norm": 0.0012513267574831843,
      "learning_rate": 0.28637397095099615,
      "loss": 0.3349,
      "num_input_tokens_seen": 10370016,
      "step": 5470
    },
    {
      "epoch": 0.8932213067950078,
      "grad_norm": 0.0012176830787211657,
      "learning_rate": 0.28634942965892646,
      "loss": 0.3672,
      "num_input_tokens_seen": 10380016,
      "step": 5475
    },
    {
      "epoch": 0.8940370340158251,
      "grad_norm": 0.001177880447357893,
      "learning_rate": 0.28632486734006124,
      "loss": 0.3155,
      "num_input_tokens_seen": 10389936,
      "step": 5480
    },
    {
      "epoch": 0.8948527612366425,
      "grad_norm": 0.0014895322965458035,
      "learning_rate": 0.28630028399818835,
      "loss": 0.3105,
      "num_input_tokens_seen": 10399376,
      "step": 5485
    },
    {
      "epoch": 0.8956684884574598,
      "grad_norm": 0.0021247577387839556,
      "learning_rate": 0.2862756796370987,
      "loss": 0.3569,
      "num_input_tokens_seen": 10408576,
      "step": 5490
    },
    {
      "epoch": 0.8964842156782772,
      "grad_norm": 0.000542016641702503,
      "learning_rate": 0.2862510542605868,
      "loss": 0.3759,
      "num_input_tokens_seen": 10419232,
      "step": 5495
    },
    {
      "epoch": 0.8972999428990945,
      "grad_norm": 0.0033777281641960144,
      "learning_rate": 0.2862264078724501,
      "loss": 0.3631,
      "num_input_tokens_seen": 10429360,
      "step": 5500
    },
    {
      "epoch": 0.8981156701199119,
      "grad_norm": 0.0005842338432557881,
      "learning_rate": 0.28620174047648933,
      "loss": 0.3552,
      "num_input_tokens_seen": 10438912,
      "step": 5505
    },
    {
      "epoch": 0.8989313973407292,
      "grad_norm": 0.00047058958443813026,
      "learning_rate": 0.2861770520765086,
      "loss": 0.302,
      "num_input_tokens_seen": 10448416,
      "step": 5510
    },
    {
      "epoch": 0.8997471245615466,
      "grad_norm": 0.0005435189232230186,
      "learning_rate": 0.2861523426763151,
      "loss": 0.3454,
      "num_input_tokens_seen": 10457792,
      "step": 5515
    },
    {
      "epoch": 0.900562851782364,
      "grad_norm": 0.0007525840774178505,
      "learning_rate": 0.2861276122797194,
      "loss": 0.3653,
      "num_input_tokens_seen": 10466752,
      "step": 5520
    },
    {
      "epoch": 0.9013785790031813,
      "grad_norm": 0.001163660199381411,
      "learning_rate": 0.28610286089053516,
      "loss": 0.3561,
      "num_input_tokens_seen": 10475776,
      "step": 5525
    },
    {
      "epoch": 0.9021943062239987,
      "grad_norm": 0.0005853150505572557,
      "learning_rate": 0.28607808851257943,
      "loss": 0.4703,
      "num_input_tokens_seen": 10485328,
      "step": 5530
    },
    {
      "epoch": 0.903010033444816,
      "grad_norm": 0.0005424285773187876,
      "learning_rate": 0.28605329514967237,
      "loss": 0.3494,
      "num_input_tokens_seen": 10493792,
      "step": 5535
    },
    {
      "epoch": 0.9038257606656334,
      "grad_norm": 0.0021025280002504587,
      "learning_rate": 0.2860284808056374,
      "loss": 0.314,
      "num_input_tokens_seen": 10503712,
      "step": 5540
    },
    {
      "epoch": 0.9046414878864508,
      "grad_norm": 0.001351650687865913,
      "learning_rate": 0.28600364548430135,
      "loss": 0.2904,
      "num_input_tokens_seen": 10513392,
      "step": 5545
    },
    {
      "epoch": 0.9054572151072682,
      "grad_norm": 0.001328136189840734,
      "learning_rate": 0.28597878918949393,
      "loss": 0.5161,
      "num_input_tokens_seen": 10523456,
      "step": 5550
    },
    {
      "epoch": 0.9062729423280855,
      "grad_norm": 0.0010782337049022317,
      "learning_rate": 0.2859539119250485,
      "loss": 0.3374,
      "num_input_tokens_seen": 10533104,
      "step": 5555
    },
    {
      "epoch": 0.9070886695489029,
      "grad_norm": 0.0009336102521046996,
      "learning_rate": 0.2859290136948013,
      "loss": 0.3342,
      "num_input_tokens_seen": 10543648,
      "step": 5560
    },
    {
      "epoch": 0.9079043967697202,
      "grad_norm": 0.0014293917920440435,
      "learning_rate": 0.28590409450259197,
      "loss": 0.3986,
      "num_input_tokens_seen": 10551888,
      "step": 5565
    },
    {
      "epoch": 0.9087201239905376,
      "grad_norm": 0.0014421608066186309,
      "learning_rate": 0.28587915435226346,
      "loss": 0.3486,
      "num_input_tokens_seen": 10561696,
      "step": 5570
    },
    {
      "epoch": 0.9095358512113549,
      "grad_norm": 0.0005961920251138508,
      "learning_rate": 0.2858541932476617,
      "loss": 0.3107,
      "num_input_tokens_seen": 10572000,
      "step": 5575
    },
    {
      "epoch": 0.9103515784321723,
      "grad_norm": 0.0037993374280631542,
      "learning_rate": 0.2858292111926361,
      "loss": 0.3805,
      "num_input_tokens_seen": 10581920,
      "step": 5580
    },
    {
      "epoch": 0.9111673056529896,
      "grad_norm": 0.0034899876918643713,
      "learning_rate": 0.28580420819103924,
      "loss": 0.3519,
      "num_input_tokens_seen": 10590800,
      "step": 5585
    },
    {
      "epoch": 0.911983032873807,
      "grad_norm": 0.002142581855878234,
      "learning_rate": 0.2857791842467269,
      "loss": 0.3774,
      "num_input_tokens_seen": 10600272,
      "step": 5590
    },
    {
      "epoch": 0.9127987600946244,
      "grad_norm": 0.0006423555896617472,
      "learning_rate": 0.2857541393635579,
      "loss": 0.3539,
      "num_input_tokens_seen": 10609808,
      "step": 5595
    },
    {
      "epoch": 0.9136144873154417,
      "grad_norm": 0.0006563177448697388,
      "learning_rate": 0.2857290735453948,
      "loss": 0.3561,
      "num_input_tokens_seen": 10619840,
      "step": 5600
    },
    {
      "epoch": 0.9136144873154417,
      "eval_loss": 0.3589949905872345,
      "eval_runtime": 152.3918,
      "eval_samples_per_second": 17.882,
      "eval_steps_per_second": 8.944,
      "num_input_tokens_seen": 10619840,
      "step": 5600
    },
    {
      "epoch": 0.9144302145362591,
      "grad_norm": 0.0006086084758862853,
      "learning_rate": 0.28570398679610276,
      "loss": 0.3568,
      "num_input_tokens_seen": 10629344,
      "step": 5605
    },
    {
      "epoch": 0.9152459417570764,
      "grad_norm": 0.0004796513239853084,
      "learning_rate": 0.2856788791195506,
      "loss": 0.3612,
      "num_input_tokens_seen": 10639136,
      "step": 5610
    },
    {
      "epoch": 0.9160616689778938,
      "grad_norm": 0.004800764378160238,
      "learning_rate": 0.28565375051961023,
      "loss": 0.3502,
      "num_input_tokens_seen": 10647632,
      "step": 5615
    },
    {
      "epoch": 0.9168773961987111,
      "grad_norm": 0.0030559322331100702,
      "learning_rate": 0.28562860100015686,
      "loss": 0.3293,
      "num_input_tokens_seen": 10656368,
      "step": 5620
    },
    {
      "epoch": 0.9176931234195285,
      "grad_norm": 0.0011761193163692951,
      "learning_rate": 0.2856034305650687,
      "loss": 0.3319,
      "num_input_tokens_seen": 10665424,
      "step": 5625
    },
    {
      "epoch": 0.9185088506403458,
      "grad_norm": 0.005335725843906403,
      "learning_rate": 0.28557823921822756,
      "loss": 0.3879,
      "num_input_tokens_seen": 10675312,
      "step": 5630
    },
    {
      "epoch": 0.9193245778611632,
      "grad_norm": 0.0026574889197945595,
      "learning_rate": 0.2855530269635181,
      "loss": 0.3729,
      "num_input_tokens_seen": 10684016,
      "step": 5635
    },
    {
      "epoch": 0.9201403050819806,
      "grad_norm": 0.0007815182325430214,
      "learning_rate": 0.2855277938048284,
      "loss": 0.33,
      "num_input_tokens_seen": 10693616,
      "step": 5640
    },
    {
      "epoch": 0.920956032302798,
      "grad_norm": 0.0012656296603381634,
      "learning_rate": 0.2855025397460498,
      "loss": 0.3532,
      "num_input_tokens_seen": 10702640,
      "step": 5645
    },
    {
      "epoch": 0.9217717595236153,
      "grad_norm": 0.00046850155922584236,
      "learning_rate": 0.28547726479107666,
      "loss": 0.3627,
      "num_input_tokens_seen": 10711488,
      "step": 5650
    },
    {
      "epoch": 0.9225874867444327,
      "grad_norm": 0.001531000598333776,
      "learning_rate": 0.2854519689438068,
      "loss": 0.3432,
      "num_input_tokens_seen": 10720992,
      "step": 5655
    },
    {
      "epoch": 0.92340321396525,
      "grad_norm": 0.0008916199440136552,
      "learning_rate": 0.2854266522081412,
      "loss": 0.3389,
      "num_input_tokens_seen": 10729872,
      "step": 5660
    },
    {
      "epoch": 0.9242189411860674,
      "grad_norm": 0.0017003664979711175,
      "learning_rate": 0.28540131458798385,
      "loss": 0.373,
      "num_input_tokens_seen": 10739968,
      "step": 5665
    },
    {
      "epoch": 0.9250346684068848,
      "grad_norm": 0.0005272745620459318,
      "learning_rate": 0.28537595608724226,
      "loss": 0.3531,
      "num_input_tokens_seen": 10748976,
      "step": 5670
    },
    {
      "epoch": 0.9258503956277021,
      "grad_norm": 0.003243827959522605,
      "learning_rate": 0.28535057670982705,
      "loss": 0.3651,
      "num_input_tokens_seen": 10757968,
      "step": 5675
    },
    {
      "epoch": 0.9266661228485195,
      "grad_norm": 0.00312876608222723,
      "learning_rate": 0.285325176459652,
      "loss": 0.3626,
      "num_input_tokens_seen": 10767888,
      "step": 5680
    },
    {
      "epoch": 0.9274818500693368,
      "grad_norm": 0.0008270757389254868,
      "learning_rate": 0.28529975534063406,
      "loss": 0.3368,
      "num_input_tokens_seen": 10777296,
      "step": 5685
    },
    {
      "epoch": 0.9282975772901542,
      "grad_norm": 0.0022298891562968493,
      "learning_rate": 0.2852743133566936,
      "loss": 0.3872,
      "num_input_tokens_seen": 10787216,
      "step": 5690
    },
    {
      "epoch": 0.9291133045109715,
      "grad_norm": 0.0017204564064741135,
      "learning_rate": 0.2852488505117541,
      "loss": 0.3341,
      "num_input_tokens_seen": 10797024,
      "step": 5695
    },
    {
      "epoch": 0.9299290317317889,
      "grad_norm": 0.0010059778578579426,
      "learning_rate": 0.28522336680974214,
      "loss": 0.3455,
      "num_input_tokens_seen": 10806240,
      "step": 5700
    },
    {
      "epoch": 0.9307447589526062,
      "grad_norm": 0.0007705741445533931,
      "learning_rate": 0.2851978622545877,
      "loss": 0.3967,
      "num_input_tokens_seen": 10815680,
      "step": 5705
    },
    {
      "epoch": 0.9315604861734236,
      "grad_norm": 0.0011142742587253451,
      "learning_rate": 0.285172336850224,
      "loss": 0.3271,
      "num_input_tokens_seen": 10825664,
      "step": 5710
    },
    {
      "epoch": 0.9323762133942409,
      "grad_norm": 0.0008796769543550909,
      "learning_rate": 0.2851467906005871,
      "loss": 0.3313,
      "num_input_tokens_seen": 10836224,
      "step": 5715
    },
    {
      "epoch": 0.9331919406150583,
      "grad_norm": 0.0015415727393701673,
      "learning_rate": 0.28512122350961683,
      "loss": 0.3609,
      "num_input_tokens_seen": 10844960,
      "step": 5720
    },
    {
      "epoch": 0.9340076678358756,
      "grad_norm": 0.0005630935775116086,
      "learning_rate": 0.2850956355812559,
      "loss": 0.3761,
      "num_input_tokens_seen": 10854496,
      "step": 5725
    },
    {
      "epoch": 0.9348233950566931,
      "grad_norm": 0.000578169769141823,
      "learning_rate": 0.28507002681945015,
      "loss": 0.3584,
      "num_input_tokens_seen": 10863040,
      "step": 5730
    },
    {
      "epoch": 0.9356391222775104,
      "grad_norm": 0.0005168988718651235,
      "learning_rate": 0.28504439722814895,
      "loss": 0.3507,
      "num_input_tokens_seen": 10872576,
      "step": 5735
    },
    {
      "epoch": 0.9364548494983278,
      "grad_norm": 0.0022060940973460674,
      "learning_rate": 0.28501874681130457,
      "loss": 0.3402,
      "num_input_tokens_seen": 10882032,
      "step": 5740
    },
    {
      "epoch": 0.9372705767191452,
      "grad_norm": 0.0027990450616925955,
      "learning_rate": 0.2849930755728727,
      "loss": 0.3913,
      "num_input_tokens_seen": 10891728,
      "step": 5745
    },
    {
      "epoch": 0.9380863039399625,
      "grad_norm": 0.0025032430421561003,
      "learning_rate": 0.28496738351681217,
      "loss": 0.5283,
      "num_input_tokens_seen": 10901760,
      "step": 5750
    },
    {
      "epoch": 0.9389020311607799,
      "grad_norm": 0.0009797286475077271,
      "learning_rate": 0.284941670647085,
      "loss": 0.3927,
      "num_input_tokens_seen": 10910736,
      "step": 5755
    },
    {
      "epoch": 0.9397177583815972,
      "grad_norm": 0.0017110701883211732,
      "learning_rate": 0.2849159369676563,
      "loss": 0.4017,
      "num_input_tokens_seen": 10921408,
      "step": 5760
    },
    {
      "epoch": 0.9405334856024146,
      "grad_norm": 0.0023090136237442493,
      "learning_rate": 0.2848901824824948,
      "loss": 0.3208,
      "num_input_tokens_seen": 10931056,
      "step": 5765
    },
    {
      "epoch": 0.9413492128232319,
      "grad_norm": 0.0030152476392686367,
      "learning_rate": 0.284864407195572,
      "loss": 0.3874,
      "num_input_tokens_seen": 10940736,
      "step": 5770
    },
    {
      "epoch": 0.9421649400440493,
      "grad_norm": 0.0013134775217622519,
      "learning_rate": 0.28483861111086284,
      "loss": 0.3299,
      "num_input_tokens_seen": 10950320,
      "step": 5775
    },
    {
      "epoch": 0.9429806672648666,
      "grad_norm": 0.0008717115269973874,
      "learning_rate": 0.2848127942323453,
      "loss": 0.3366,
      "num_input_tokens_seen": 10958592,
      "step": 5780
    },
    {
      "epoch": 0.943796394485684,
      "grad_norm": 0.0013893210561946034,
      "learning_rate": 0.2847869565640007,
      "loss": 0.351,
      "num_input_tokens_seen": 10967088,
      "step": 5785
    },
    {
      "epoch": 0.9446121217065013,
      "grad_norm": 0.0019179065711796284,
      "learning_rate": 0.2847610981098136,
      "loss": 0.3636,
      "num_input_tokens_seen": 10976416,
      "step": 5790
    },
    {
      "epoch": 0.9454278489273187,
      "grad_norm": 0.0006526840152218938,
      "learning_rate": 0.2847352188737716,
      "loss": 0.3568,
      "num_input_tokens_seen": 10985344,
      "step": 5795
    },
    {
      "epoch": 0.946243576148136,
      "grad_norm": 0.002432263921946287,
      "learning_rate": 0.2847093188598658,
      "loss": 0.4045,
      "num_input_tokens_seen": 10994720,
      "step": 5800
    },
    {
      "epoch": 0.946243576148136,
      "eval_loss": 0.34319034218788147,
      "eval_runtime": 152.422,
      "eval_samples_per_second": 17.878,
      "eval_steps_per_second": 8.942,
      "num_input_tokens_seen": 10994720,
      "step": 5800
    },
    {
      "epoch": 0.9470593033689534,
      "grad_norm": 0.0009990980615839362,
      "learning_rate": 0.28468339807209003,
      "loss": 0.3362,
      "num_input_tokens_seen": 11004464,
      "step": 5805
    },
    {
      "epoch": 0.9478750305897707,
      "grad_norm": 0.0025449832901358604,
      "learning_rate": 0.2846574565144418,
      "loss": 0.4194,
      "num_input_tokens_seen": 11013968,
      "step": 5810
    },
    {
      "epoch": 0.9486907578105881,
      "grad_norm": 0.0010781160090118647,
      "learning_rate": 0.28463149419092154,
      "loss": 0.3662,
      "num_input_tokens_seen": 11023632,
      "step": 5815
    },
    {
      "epoch": 0.9495064850314054,
      "grad_norm": 0.0008508307510055602,
      "learning_rate": 0.284605511105533,
      "loss": 0.3447,
      "num_input_tokens_seen": 11033840,
      "step": 5820
    },
    {
      "epoch": 0.9503222122522229,
      "grad_norm": 0.0018948784563690424,
      "learning_rate": 0.28457950726228315,
      "loss": 0.4563,
      "num_input_tokens_seen": 11044304,
      "step": 5825
    },
    {
      "epoch": 0.9511379394730403,
      "grad_norm": 0.0014500734396278858,
      "learning_rate": 0.28455348266518193,
      "loss": 0.3654,
      "num_input_tokens_seen": 11054320,
      "step": 5830
    },
    {
      "epoch": 0.9519536666938576,
      "grad_norm": 0.0007087372941896319,
      "learning_rate": 0.28452743731824287,
      "loss": 0.3615,
      "num_input_tokens_seen": 11063728,
      "step": 5835
    },
    {
      "epoch": 0.952769393914675,
      "grad_norm": 0.0020589367486536503,
      "learning_rate": 0.28450137122548236,
      "loss": 0.3474,
      "num_input_tokens_seen": 11072672,
      "step": 5840
    },
    {
      "epoch": 0.9535851211354923,
      "grad_norm": 0.000968937820289284,
      "learning_rate": 0.2844752843909201,
      "loss": 0.407,
      "num_input_tokens_seen": 11082336,
      "step": 5845
    },
    {
      "epoch": 0.9544008483563097,
      "grad_norm": 0.0009886762127280235,
      "learning_rate": 0.28444917681857923,
      "loss": 0.3281,
      "num_input_tokens_seen": 11091232,
      "step": 5850
    },
    {
      "epoch": 0.955216575577127,
      "grad_norm": 0.0013766591437160969,
      "learning_rate": 0.28442304851248557,
      "loss": 0.3457,
      "num_input_tokens_seen": 11100272,
      "step": 5855
    },
    {
      "epoch": 0.9560323027979444,
      "grad_norm": 0.0006167002720758319,
      "learning_rate": 0.2843968994766686,
      "loss": 0.3337,
      "num_input_tokens_seen": 11110304,
      "step": 5860
    },
    {
      "epoch": 0.9568480300187617,
      "grad_norm": 0.0027575958520174026,
      "learning_rate": 0.28437072971516075,
      "loss": 0.3662,
      "num_input_tokens_seen": 11119216,
      "step": 5865
    },
    {
      "epoch": 0.9576637572395791,
      "grad_norm": 0.0022789514623582363,
      "learning_rate": 0.2843445392319979,
      "loss": 0.3738,
      "num_input_tokens_seen": 11128752,
      "step": 5870
    },
    {
      "epoch": 0.9584794844603964,
      "grad_norm": 0.0005525905289687216,
      "learning_rate": 0.28431832803121865,
      "loss": 0.3688,
      "num_input_tokens_seen": 11137664,
      "step": 5875
    },
    {
      "epoch": 0.9592952116812138,
      "grad_norm": 0.0015544556081295013,
      "learning_rate": 0.28429209611686534,
      "loss": 0.354,
      "num_input_tokens_seen": 11147200,
      "step": 5880
    },
    {
      "epoch": 0.9601109389020311,
      "grad_norm": 0.0013067532563582063,
      "learning_rate": 0.28426584349298323,
      "loss": 0.3274,
      "num_input_tokens_seen": 11156864,
      "step": 5885
    },
    {
      "epoch": 0.9609266661228485,
      "grad_norm": 0.004912674892693758,
      "learning_rate": 0.2842395701636207,
      "loss": 0.3701,
      "num_input_tokens_seen": 11165792,
      "step": 5890
    },
    {
      "epoch": 0.9617423933436658,
      "grad_norm": 0.0038959309458732605,
      "learning_rate": 0.28421327613282954,
      "loss": 0.3516,
      "num_input_tokens_seen": 11175040,
      "step": 5895
    },
    {
      "epoch": 0.9625581205644832,
      "grad_norm": 0.0023594882804900408,
      "learning_rate": 0.28418696140466454,
      "loss": 0.3952,
      "num_input_tokens_seen": 11183744,
      "step": 5900
    },
    {
      "epoch": 0.9633738477853006,
      "grad_norm": 0.000467836856842041,
      "learning_rate": 0.2841606259831838,
      "loss": 0.3555,
      "num_input_tokens_seen": 11193232,
      "step": 5905
    },
    {
      "epoch": 0.9641895750061179,
      "grad_norm": 0.0015818431274965405,
      "learning_rate": 0.2841342698724486,
      "loss": 0.3729,
      "num_input_tokens_seen": 11202512,
      "step": 5910
    },
    {
      "epoch": 0.9650053022269353,
      "grad_norm": 0.0010277889668941498,
      "learning_rate": 0.28410789307652334,
      "loss": 0.3631,
      "num_input_tokens_seen": 11212624,
      "step": 5915
    },
    {
      "epoch": 0.9658210294477527,
      "grad_norm": 0.0015026202891021967,
      "learning_rate": 0.2840814955994756,
      "loss": 0.3761,
      "num_input_tokens_seen": 11222512,
      "step": 5920
    },
    {
      "epoch": 0.9666367566685701,
      "grad_norm": 0.001572132809087634,
      "learning_rate": 0.2840550774453763,
      "loss": 0.3487,
      "num_input_tokens_seen": 11232144,
      "step": 5925
    },
    {
      "epoch": 0.9674524838893874,
      "grad_norm": 0.0009717716602608562,
      "learning_rate": 0.28402863861829947,
      "loss": 0.3519,
      "num_input_tokens_seen": 11241168,
      "step": 5930
    },
    {
      "epoch": 0.9682682111102048,
      "grad_norm": 0.0009195140446536243,
      "learning_rate": 0.2840021791223222,
      "loss": 0.383,
      "num_input_tokens_seen": 11251552,
      "step": 5935
    },
    {
      "epoch": 0.9690839383310221,
      "grad_norm": 0.001986399292945862,
      "learning_rate": 0.2839756989615249,
      "loss": 0.369,
      "num_input_tokens_seen": 11262240,
      "step": 5940
    },
    {
      "epoch": 0.9698996655518395,
      "grad_norm": 0.0014884062111377716,
      "learning_rate": 0.28394919813999125,
      "loss": 0.3542,
      "num_input_tokens_seen": 11272256,
      "step": 5945
    },
    {
      "epoch": 0.9707153927726568,
      "grad_norm": 0.003389996010810137,
      "learning_rate": 0.28392267666180787,
      "loss": 0.3603,
      "num_input_tokens_seen": 11281904,
      "step": 5950
    },
    {
      "epoch": 0.9715311199934742,
      "grad_norm": 0.0031861753668636084,
      "learning_rate": 0.2838961345310648,
      "loss": 0.3544,
      "num_input_tokens_seen": 11291184,
      "step": 5955
    },
    {
      "epoch": 0.9723468472142915,
      "grad_norm": 0.002978677861392498,
      "learning_rate": 0.2838695717518552,
      "loss": 0.3658,
      "num_input_tokens_seen": 11301312,
      "step": 5960
    },
    {
      "epoch": 0.9731625744351089,
      "grad_norm": 0.002634078962728381,
      "learning_rate": 0.28384298832827526,
      "loss": 0.3741,
      "num_input_tokens_seen": 11311840,
      "step": 5965
    },
    {
      "epoch": 0.9739783016559262,
      "grad_norm": 0.00347490725107491,
      "learning_rate": 0.28381638426442457,
      "loss": 0.3578,
      "num_input_tokens_seen": 11321264,
      "step": 5970
    },
    {
      "epoch": 0.9747940288767436,
      "grad_norm": 0.0017261584289371967,
      "learning_rate": 0.2837897595644057,
      "loss": 0.3801,
      "num_input_tokens_seen": 11330096,
      "step": 5975
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.003351157996803522,
      "learning_rate": 0.28376311423232475,
      "loss": 0.3467,
      "num_input_tokens_seen": 11340064,
      "step": 5980
    },
    {
      "epoch": 0.9764254833183783,
      "grad_norm": 0.0008548491168767214,
      "learning_rate": 0.2837364482722905,
      "loss": 0.331,
      "num_input_tokens_seen": 11348608,
      "step": 5985
    },
    {
      "epoch": 0.9772412105391957,
      "grad_norm": 0.003935778047889471,
      "learning_rate": 0.28370976168841533,
      "loss": 0.4299,
      "num_input_tokens_seen": 11358384,
      "step": 5990
    },
    {
      "epoch": 0.978056937760013,
      "grad_norm": 0.002690159482881427,
      "learning_rate": 0.2836830544848146,
      "loss": 0.3549,
      "num_input_tokens_seen": 11368608,
      "step": 5995
    },
    {
      "epoch": 0.9788726649808304,
      "grad_norm": 0.0038847611285746098,
      "learning_rate": 0.2836563266656069,
      "loss": 0.3709,
      "num_input_tokens_seen": 11376976,
      "step": 6000
    },
    {
      "epoch": 0.9788726649808304,
      "eval_loss": 0.35364213585853577,
      "eval_runtime": 152.4483,
      "eval_samples_per_second": 17.875,
      "eval_steps_per_second": 8.941,
      "num_input_tokens_seen": 11376976,
      "step": 6000
    },
    {
      "epoch": 0.9796883922016477,
      "grad_norm": 0.0016283513978123665,
      "learning_rate": 0.283629578234914,
      "loss": 0.3264,
      "num_input_tokens_seen": 11386688,
      "step": 6005
    },
    {
      "epoch": 0.9805041194224652,
      "grad_norm": 0.002402381505817175,
      "learning_rate": 0.2836028091968608,
      "loss": 0.3558,
      "num_input_tokens_seen": 11395552,
      "step": 6010
    },
    {
      "epoch": 0.9813198466432825,
      "grad_norm": 0.0012892300728708506,
      "learning_rate": 0.28357601955557554,
      "loss": 0.2848,
      "num_input_tokens_seen": 11403968,
      "step": 6015
    },
    {
      "epoch": 0.9821355738640999,
      "grad_norm": 0.005521154962480068,
      "learning_rate": 0.2835492093151894,
      "loss": 0.5678,
      "num_input_tokens_seen": 11413888,
      "step": 6020
    },
    {
      "epoch": 0.9829513010849172,
      "grad_norm": 0.0018002984579652548,
      "learning_rate": 0.2835223784798369,
      "loss": 0.3464,
      "num_input_tokens_seen": 11423904,
      "step": 6025
    },
    {
      "epoch": 0.9837670283057346,
      "grad_norm": 0.004042245447635651,
      "learning_rate": 0.2834955270536557,
      "loss": 0.3983,
      "num_input_tokens_seen": 11434176,
      "step": 6030
    },
    {
      "epoch": 0.984582755526552,
      "grad_norm": 0.0035628683399409056,
      "learning_rate": 0.2834686550407866,
      "loss": 0.4138,
      "num_input_tokens_seen": 11443792,
      "step": 6035
    },
    {
      "epoch": 0.9853984827473693,
      "grad_norm": 0.0017782557988539338,
      "learning_rate": 0.28344176244537367,
      "loss": 0.355,
      "num_input_tokens_seen": 11453264,
      "step": 6040
    },
    {
      "epoch": 0.9862142099681867,
      "grad_norm": 0.0018189616966992617,
      "learning_rate": 0.28341484927156396,
      "loss": 0.3378,
      "num_input_tokens_seen": 11462256,
      "step": 6045
    },
    {
      "epoch": 0.987029937189004,
      "grad_norm": 0.0011059087701141834,
      "learning_rate": 0.28338791552350795,
      "loss": 0.3686,
      "num_input_tokens_seen": 11470560,
      "step": 6050
    },
    {
      "epoch": 0.9878456644098214,
      "grad_norm": 0.0009797089733183384,
      "learning_rate": 0.28336096120535914,
      "loss": 0.3566,
      "num_input_tokens_seen": 11481088,
      "step": 6055
    },
    {
      "epoch": 0.9886613916306387,
      "grad_norm": 0.0007964239921420813,
      "learning_rate": 0.2833339863212741,
      "loss": 0.3658,
      "num_input_tokens_seen": 11490800,
      "step": 6060
    },
    {
      "epoch": 0.9894771188514561,
      "grad_norm": 0.0030644431244581938,
      "learning_rate": 0.28330699087541283,
      "loss": 0.3655,
      "num_input_tokens_seen": 11500720,
      "step": 6065
    },
    {
      "epoch": 0.9902928460722734,
      "grad_norm": 0.001432551071047783,
      "learning_rate": 0.2832799748719384,
      "loss": 0.3687,
      "num_input_tokens_seen": 11510592,
      "step": 6070
    },
    {
      "epoch": 0.9911085732930908,
      "grad_norm": 0.001151201780885458,
      "learning_rate": 0.28325293831501686,
      "loss": 0.388,
      "num_input_tokens_seen": 11520096,
      "step": 6075
    },
    {
      "epoch": 0.9919243005139081,
      "grad_norm": 0.0006971094408072531,
      "learning_rate": 0.2832258812088177,
      "loss": 0.3847,
      "num_input_tokens_seen": 11529824,
      "step": 6080
    },
    {
      "epoch": 0.9927400277347255,
      "grad_norm": 0.00149812875315547,
      "learning_rate": 0.2831988035575134,
      "loss": 0.4222,
      "num_input_tokens_seen": 11537984,
      "step": 6085
    },
    {
      "epoch": 0.9935557549555428,
      "grad_norm": 0.004421835765242577,
      "learning_rate": 0.28317170536527975,
      "loss": 0.4196,
      "num_input_tokens_seen": 11547232,
      "step": 6090
    },
    {
      "epoch": 0.9943714821763602,
      "grad_norm": 0.0013955890899524093,
      "learning_rate": 0.2831445866362956,
      "loss": 0.3148,
      "num_input_tokens_seen": 11557840,
      "step": 6095
    },
    {
      "epoch": 0.9951872093971775,
      "grad_norm": 0.001986656803637743,
      "learning_rate": 0.2831174473747429,
      "loss": 0.3313,
      "num_input_tokens_seen": 11566880,
      "step": 6100
    },
    {
      "epoch": 0.996002936617995,
      "grad_norm": 0.0011710282415151596,
      "learning_rate": 0.2830902875848071,
      "loss": 0.3584,
      "num_input_tokens_seen": 11575424,
      "step": 6105
    },
    {
      "epoch": 0.9968186638388123,
      "grad_norm": 0.0012971843825653195,
      "learning_rate": 0.28306310727067635,
      "loss": 0.3635,
      "num_input_tokens_seen": 11584480,
      "step": 6110
    },
    {
      "epoch": 0.9976343910596297,
      "grad_norm": 0.0010815002024173737,
      "learning_rate": 0.2830359064365423,
      "loss": 0.3411,
      "num_input_tokens_seen": 11594112,
      "step": 6115
    },
    {
      "epoch": 0.998450118280447,
      "grad_norm": 0.0016010531689971685,
      "learning_rate": 0.28300868508659965,
      "loss": 0.3668,
      "num_input_tokens_seen": 11604128,
      "step": 6120
    },
    {
      "epoch": 0.9992658455012644,
      "grad_norm": 0.0011716462904587388,
      "learning_rate": 0.28298144322504626,
      "loss": 0.3618,
      "num_input_tokens_seen": 11614416,
      "step": 6125
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0038796174339950085,
      "learning_rate": 0.2829541808560832,
      "loss": 0.3862,
      "num_input_tokens_seen": 11623536,
      "step": 6130
    },
    {
      "epoch": 1.0008157272208174,
      "grad_norm": 0.0019342534942552447,
      "learning_rate": 0.2829268979839146,
      "loss": 0.3643,
      "num_input_tokens_seen": 11633472,
      "step": 6135
    },
    {
      "epoch": 1.0016314544416347,
      "grad_norm": 0.0023980848491191864,
      "learning_rate": 0.2828995946127479,
      "loss": 0.3619,
      "num_input_tokens_seen": 11643408,
      "step": 6140
    },
    {
      "epoch": 1.002447181662452,
      "grad_norm": 0.000727245700545609,
      "learning_rate": 0.2828722707467936,
      "loss": 0.3493,
      "num_input_tokens_seen": 11652832,
      "step": 6145
    },
    {
      "epoch": 1.0032629088832694,
      "grad_norm": 0.0007635481306351721,
      "learning_rate": 0.2828449263902653,
      "loss": 0.3491,
      "num_input_tokens_seen": 11663344,
      "step": 6150
    },
    {
      "epoch": 1.0040786361040868,
      "grad_norm": 0.001499034813605249,
      "learning_rate": 0.28281756154738,
      "loss": 0.3641,
      "num_input_tokens_seen": 11672720,
      "step": 6155
    },
    {
      "epoch": 1.0048943633249041,
      "grad_norm": 0.0016528472770005465,
      "learning_rate": 0.28279017622235764,
      "loss": 0.3765,
      "num_input_tokens_seen": 11681664,
      "step": 6160
    },
    {
      "epoch": 1.0057100905457215,
      "grad_norm": 0.0014916122891008854,
      "learning_rate": 0.28276277041942127,
      "loss": 0.3679,
      "num_input_tokens_seen": 11691664,
      "step": 6165
    },
    {
      "epoch": 1.0065258177665388,
      "grad_norm": 0.0012438639532774687,
      "learning_rate": 0.2827353441427974,
      "loss": 0.3358,
      "num_input_tokens_seen": 11701792,
      "step": 6170
    },
    {
      "epoch": 1.0073415449873562,
      "grad_norm": 0.002364593558013439,
      "learning_rate": 0.2827078973967153,
      "loss": 0.4056,
      "num_input_tokens_seen": 11711296,
      "step": 6175
    },
    {
      "epoch": 1.0081572722081735,
      "grad_norm": 0.0006214702734723687,
      "learning_rate": 0.2826804301854078,
      "loss": 0.3374,
      "num_input_tokens_seen": 11721168,
      "step": 6180
    },
    {
      "epoch": 1.0089729994289909,
      "grad_norm": 0.0007328621577471495,
      "learning_rate": 0.2826529425131105,
      "loss": 0.3503,
      "num_input_tokens_seen": 11731344,
      "step": 6185
    },
    {
      "epoch": 1.0097887266498082,
      "grad_norm": 0.0025327028706669807,
      "learning_rate": 0.2826254343840625,
      "loss": 0.3797,
      "num_input_tokens_seen": 11740896,
      "step": 6190
    },
    {
      "epoch": 1.0106044538706256,
      "grad_norm": 0.0021511169616132975,
      "learning_rate": 0.2825979058025059,
      "loss": 0.3337,
      "num_input_tokens_seen": 11748640,
      "step": 6195
    },
    {
      "epoch": 1.011420181091443,
      "grad_norm": 0.0024862117134034634,
      "learning_rate": 0.2825703567726858,
      "loss": 0.3859,
      "num_input_tokens_seen": 11758656,
      "step": 6200
    },
    {
      "epoch": 1.011420181091443,
      "eval_loss": 0.34999966621398926,
      "eval_runtime": 152.5292,
      "eval_samples_per_second": 17.865,
      "eval_steps_per_second": 8.936,
      "num_input_tokens_seen": 11758656,
      "step": 6200
    },
    {
      "epoch": 1.0122359083122603,
      "grad_norm": 0.0016684059519320726,
      "learning_rate": 0.2825427872988508,
      "loss": 0.367,
      "num_input_tokens_seen": 11769376,
      "step": 6205
    },
    {
      "epoch": 1.0130516355330776,
      "grad_norm": 0.0006422972655855119,
      "learning_rate": 0.28251519738525227,
      "loss": 0.3384,
      "num_input_tokens_seen": 11778768,
      "step": 6210
    },
    {
      "epoch": 1.013867362753895,
      "grad_norm": 0.0008421011152677238,
      "learning_rate": 0.28248758703614507,
      "loss": 0.358,
      "num_input_tokens_seen": 11789360,
      "step": 6215
    },
    {
      "epoch": 1.0146830899747123,
      "grad_norm": 0.001049783662892878,
      "learning_rate": 0.28245995625578696,
      "loss": 0.3207,
      "num_input_tokens_seen": 11799376,
      "step": 6220
    },
    {
      "epoch": 1.01549881719553,
      "grad_norm": 0.0029823416844010353,
      "learning_rate": 0.282432305048439,
      "loss": 0.3236,
      "num_input_tokens_seen": 11807328,
      "step": 6225
    },
    {
      "epoch": 1.0163145444163473,
      "grad_norm": 0.0007051668944768608,
      "learning_rate": 0.28240463341836536,
      "loss": 0.3438,
      "num_input_tokens_seen": 11817120,
      "step": 6230
    },
    {
      "epoch": 1.0171302716371646,
      "grad_norm": 0.0007981456001289189,
      "learning_rate": 0.2823769413698334,
      "loss": 0.3414,
      "num_input_tokens_seen": 11826128,
      "step": 6235
    },
    {
      "epoch": 1.017945998857982,
      "grad_norm": 0.001953994622454047,
      "learning_rate": 0.2823492289071135,
      "loss": 0.3686,
      "num_input_tokens_seen": 11835424,
      "step": 6240
    },
    {
      "epoch": 1.0187617260787993,
      "grad_norm": 0.00034453539410606027,
      "learning_rate": 0.2823214960344793,
      "loss": 0.3858,
      "num_input_tokens_seen": 11845568,
      "step": 6245
    },
    {
      "epoch": 1.0195774532996167,
      "grad_norm": 0.0007954009925015271,
      "learning_rate": 0.28229374275620756,
      "loss": 0.3722,
      "num_input_tokens_seen": 11855552,
      "step": 6250
    },
    {
      "epoch": 1.020393180520434,
      "grad_norm": 0.0012469723587855697,
      "learning_rate": 0.28226596907657814,
      "loss": 0.3386,
      "num_input_tokens_seen": 11864880,
      "step": 6255
    },
    {
      "epoch": 1.0212089077412514,
      "grad_norm": 0.0005951329367235303,
      "learning_rate": 0.28223817499987414,
      "loss": 0.3432,
      "num_input_tokens_seen": 11874592,
      "step": 6260
    },
    {
      "epoch": 1.0220246349620687,
      "grad_norm": 0.0012715536868199706,
      "learning_rate": 0.2822103605303818,
      "loss": 0.3587,
      "num_input_tokens_seen": 11883792,
      "step": 6265
    },
    {
      "epoch": 1.022840362182886,
      "grad_norm": 0.0008621043525636196,
      "learning_rate": 0.2821825256723903,
      "loss": 0.3889,
      "num_input_tokens_seen": 11892032,
      "step": 6270
    },
    {
      "epoch": 1.0236560894037035,
      "grad_norm": 0.0021375934593379498,
      "learning_rate": 0.2821546704301923,
      "loss": 0.3674,
      "num_input_tokens_seen": 11901264,
      "step": 6275
    },
    {
      "epoch": 1.0244718166245208,
      "grad_norm": 0.001231366884894669,
      "learning_rate": 0.2821267948080834,
      "loss": 0.326,
      "num_input_tokens_seen": 11910784,
      "step": 6280
    },
    {
      "epoch": 1.0252875438453382,
      "grad_norm": 0.004765636287629604,
      "learning_rate": 0.28209889881036226,
      "loss": 0.4098,
      "num_input_tokens_seen": 11921504,
      "step": 6285
    },
    {
      "epoch": 1.0261032710661555,
      "grad_norm": 0.0011467041913419962,
      "learning_rate": 0.28207098244133094,
      "loss": 0.3512,
      "num_input_tokens_seen": 11930768,
      "step": 6290
    },
    {
      "epoch": 1.0269189982869729,
      "grad_norm": 0.002400917001068592,
      "learning_rate": 0.2820430457052943,
      "loss": 0.3766,
      "num_input_tokens_seen": 11941072,
      "step": 6295
    },
    {
      "epoch": 1.0277347255077902,
      "grad_norm": 0.0014695338904857635,
      "learning_rate": 0.28201508860656077,
      "loss": 0.415,
      "num_input_tokens_seen": 11950144,
      "step": 6300
    },
    {
      "epoch": 1.0285504527286076,
      "grad_norm": 0.001324929646216333,
      "learning_rate": 0.2819871111494415,
      "loss": 0.3641,
      "num_input_tokens_seen": 11959536,
      "step": 6305
    },
    {
      "epoch": 1.029366179949425,
      "grad_norm": 0.0006959355669096112,
      "learning_rate": 0.28195911333825113,
      "loss": 0.3035,
      "num_input_tokens_seen": 11968624,
      "step": 6310
    },
    {
      "epoch": 1.0301819071702423,
      "grad_norm": 0.0008117877878248692,
      "learning_rate": 0.28193109517730713,
      "loss": 0.3901,
      "num_input_tokens_seen": 11979248,
      "step": 6315
    },
    {
      "epoch": 1.0309976343910596,
      "grad_norm": 0.001297555398195982,
      "learning_rate": 0.2819030566709303,
      "loss": 0.3499,
      "num_input_tokens_seen": 11988912,
      "step": 6320
    },
    {
      "epoch": 1.031813361611877,
      "grad_norm": 0.0012802515411749482,
      "learning_rate": 0.2818749978234445,
      "loss": 0.3337,
      "num_input_tokens_seen": 11998656,
      "step": 6325
    },
    {
      "epoch": 1.0326290888326943,
      "grad_norm": 0.0011543090222403407,
      "learning_rate": 0.2818469186391768,
      "loss": 0.3713,
      "num_input_tokens_seen": 12008816,
      "step": 6330
    },
    {
      "epoch": 1.0334448160535117,
      "grad_norm": 0.001254644594155252,
      "learning_rate": 0.28181881912245743,
      "loss": 0.3642,
      "num_input_tokens_seen": 12017968,
      "step": 6335
    },
    {
      "epoch": 1.034260543274329,
      "grad_norm": 0.0032255486585199833,
      "learning_rate": 0.2817906992776195,
      "loss": 0.3855,
      "num_input_tokens_seen": 12027456,
      "step": 6340
    },
    {
      "epoch": 1.0350762704951464,
      "grad_norm": 0.0020100411493331194,
      "learning_rate": 0.28176255910899967,
      "loss": 0.3661,
      "num_input_tokens_seen": 12037424,
      "step": 6345
    },
    {
      "epoch": 1.0358919977159637,
      "grad_norm": 0.0013325712643563747,
      "learning_rate": 0.2817343986209373,
      "loss": 0.3467,
      "num_input_tokens_seen": 12048048,
      "step": 6350
    },
    {
      "epoch": 1.036707724936781,
      "grad_norm": 0.0012209364213049412,
      "learning_rate": 0.2817062178177753,
      "loss": 0.3609,
      "num_input_tokens_seen": 12058096,
      "step": 6355
    },
    {
      "epoch": 1.0375234521575984,
      "grad_norm": 0.0011221298482269049,
      "learning_rate": 0.2816780167038593,
      "loss": 0.3522,
      "num_input_tokens_seen": 12067360,
      "step": 6360
    },
    {
      "epoch": 1.0383391793784158,
      "grad_norm": 0.0008153236121870577,
      "learning_rate": 0.28164979528353834,
      "loss": 0.2974,
      "num_input_tokens_seen": 12077680,
      "step": 6365
    },
    {
      "epoch": 1.0391549065992332,
      "grad_norm": 0.0008263017516583204,
      "learning_rate": 0.28162155356116453,
      "loss": 0.2624,
      "num_input_tokens_seen": 12087552,
      "step": 6370
    },
    {
      "epoch": 1.0399706338200505,
      "grad_norm": 0.0005464631831273437,
      "learning_rate": 0.28159329154109314,
      "loss": 0.4276,
      "num_input_tokens_seen": 12096432,
      "step": 6375
    },
    {
      "epoch": 1.0407863610408679,
      "grad_norm": 0.00197921646758914,
      "learning_rate": 0.28156500922768246,
      "loss": 0.4951,
      "num_input_tokens_seen": 12105600,
      "step": 6380
    },
    {
      "epoch": 1.0416020882616852,
      "grad_norm": 0.0012917346321046352,
      "learning_rate": 0.28153670662529406,
      "loss": 0.4076,
      "num_input_tokens_seen": 12114992,
      "step": 6385
    },
    {
      "epoch": 1.0424178154825026,
      "grad_norm": 0.0006965642678551376,
      "learning_rate": 0.28150838373829246,
      "loss": 0.4353,
      "num_input_tokens_seen": 12124496,
      "step": 6390
    },
    {
      "epoch": 1.04323354270332,
      "grad_norm": 0.0010350315133109689,
      "learning_rate": 0.2814800405710455,
      "loss": 0.3455,
      "num_input_tokens_seen": 12135424,
      "step": 6395
    },
    {
      "epoch": 1.0440492699241373,
      "grad_norm": 0.0017463075928390026,
      "learning_rate": 0.2814516771279239,
      "loss": 0.4941,
      "num_input_tokens_seen": 12144016,
      "step": 6400
    },
    {
      "epoch": 1.0440492699241373,
      "eval_loss": 0.34525206685066223,
      "eval_runtime": 152.5457,
      "eval_samples_per_second": 17.863,
      "eval_steps_per_second": 8.935,
      "num_input_tokens_seen": 12144016,
      "step": 6400
    },
    {
      "epoch": 1.0448649971449546,
      "grad_norm": 0.0024186030495911837,
      "learning_rate": 0.28142329341330186,
      "loss": 0.377,
      "num_input_tokens_seen": 12153872,
      "step": 6405
    },
    {
      "epoch": 1.0456807243657722,
      "grad_norm": 0.0014675580896437168,
      "learning_rate": 0.2813948894315564,
      "loss": 0.4224,
      "num_input_tokens_seen": 12164064,
      "step": 6410
    },
    {
      "epoch": 1.0464964515865895,
      "grad_norm": 0.0008995191892609,
      "learning_rate": 0.2813664651870677,
      "loss": 0.3839,
      "num_input_tokens_seen": 12172144,
      "step": 6415
    },
    {
      "epoch": 1.047312178807407,
      "grad_norm": 0.0008644465124234557,
      "learning_rate": 0.28133802068421926,
      "loss": 0.3675,
      "num_input_tokens_seen": 12182048,
      "step": 6420
    },
    {
      "epoch": 1.0481279060282243,
      "grad_norm": 0.0013097099727019668,
      "learning_rate": 0.28130955592739754,
      "loss": 0.3213,
      "num_input_tokens_seen": 12192208,
      "step": 6425
    },
    {
      "epoch": 1.0489436332490416,
      "grad_norm": 0.0011089433683082461,
      "learning_rate": 0.2812810709209922,
      "loss": 0.4721,
      "num_input_tokens_seen": 12201472,
      "step": 6430
    },
    {
      "epoch": 1.049759360469859,
      "grad_norm": 0.0007481072680093348,
      "learning_rate": 0.2812525656693959,
      "loss": 0.3862,
      "num_input_tokens_seen": 12211136,
      "step": 6435
    },
    {
      "epoch": 1.0505750876906763,
      "grad_norm": 0.0006149276159703732,
      "learning_rate": 0.28122404017700453,
      "loss": 0.3843,
      "num_input_tokens_seen": 12219776,
      "step": 6440
    },
    {
      "epoch": 1.0513908149114937,
      "grad_norm": 0.0012350379256531596,
      "learning_rate": 0.2811954944482171,
      "loss": 0.3308,
      "num_input_tokens_seen": 12230384,
      "step": 6445
    },
    {
      "epoch": 1.052206542132311,
      "grad_norm": 0.0006450458895415068,
      "learning_rate": 0.2811669284874358,
      "loss": 0.4026,
      "num_input_tokens_seen": 12239984,
      "step": 6450
    },
    {
      "epoch": 1.0530222693531284,
      "grad_norm": 0.0007598024676553905,
      "learning_rate": 0.2811383422990657,
      "loss": 0.3328,
      "num_input_tokens_seen": 12248336,
      "step": 6455
    },
    {
      "epoch": 1.0538379965739457,
      "grad_norm": 0.001161051681265235,
      "learning_rate": 0.2811097358875152,
      "loss": 0.3325,
      "num_input_tokens_seen": 12259312,
      "step": 6460
    },
    {
      "epoch": 1.054653723794763,
      "grad_norm": 0.0009248855640180409,
      "learning_rate": 0.2810811092571959,
      "loss": 0.3395,
      "num_input_tokens_seen": 12268288,
      "step": 6465
    },
    {
      "epoch": 1.0554694510155804,
      "grad_norm": 0.0008558229310438037,
      "learning_rate": 0.28105246241252224,
      "loss": 0.3383,
      "num_input_tokens_seen": 12277824,
      "step": 6470
    },
    {
      "epoch": 1.0562851782363978,
      "grad_norm": 0.0010617132065817714,
      "learning_rate": 0.28102379535791194,
      "loss": 0.3193,
      "num_input_tokens_seen": 12287520,
      "step": 6475
    },
    {
      "epoch": 1.0571009054572151,
      "grad_norm": 0.00057863729307428,
      "learning_rate": 0.2809951080977859,
      "loss": 0.4169,
      "num_input_tokens_seen": 12297216,
      "step": 6480
    },
    {
      "epoch": 1.0579166326780325,
      "grad_norm": 0.0006416150717996061,
      "learning_rate": 0.28096640063656797,
      "loss": 0.3681,
      "num_input_tokens_seen": 12307264,
      "step": 6485
    },
    {
      "epoch": 1.0587323598988498,
      "grad_norm": 0.0027720415964722633,
      "learning_rate": 0.2809376729786852,
      "loss": 0.3759,
      "num_input_tokens_seen": 12317552,
      "step": 6490
    },
    {
      "epoch": 1.0595480871196672,
      "grad_norm": 0.0016932986909523606,
      "learning_rate": 0.28090892512856785,
      "loss": 0.298,
      "num_input_tokens_seen": 12326496,
      "step": 6495
    },
    {
      "epoch": 1.0603638143404845,
      "grad_norm": 0.0008107390603981912,
      "learning_rate": 0.2808801570906491,
      "loss": 0.3763,
      "num_input_tokens_seen": 12334848,
      "step": 6500
    },
    {
      "epoch": 1.061179541561302,
      "grad_norm": 0.0006944409688003361,
      "learning_rate": 0.2808513688693654,
      "loss": 0.3659,
      "num_input_tokens_seen": 12345120,
      "step": 6505
    },
    {
      "epoch": 1.0619952687821193,
      "grad_norm": 0.0011734100989997387,
      "learning_rate": 0.28082256046915627,
      "loss": 0.391,
      "num_input_tokens_seen": 12355264,
      "step": 6510
    },
    {
      "epoch": 1.0628109960029366,
      "grad_norm": 0.002431705594062805,
      "learning_rate": 0.28079373189446427,
      "loss": 0.3633,
      "num_input_tokens_seen": 12365408,
      "step": 6515
    },
    {
      "epoch": 1.063626723223754,
      "grad_norm": 0.001968038035556674,
      "learning_rate": 0.28076488314973513,
      "loss": 0.3869,
      "num_input_tokens_seen": 12374368,
      "step": 6520
    },
    {
      "epoch": 1.0644424504445713,
      "grad_norm": 0.002767318394035101,
      "learning_rate": 0.28073601423941774,
      "loss": 0.3663,
      "num_input_tokens_seen": 12383664,
      "step": 6525
    },
    {
      "epoch": 1.0652581776653887,
      "grad_norm": 0.0010681211715564132,
      "learning_rate": 0.28070712516796403,
      "loss": 0.3635,
      "num_input_tokens_seen": 12393952,
      "step": 6530
    },
    {
      "epoch": 1.066073904886206,
      "grad_norm": 0.002573194680735469,
      "learning_rate": 0.28067821593982906,
      "loss": 0.341,
      "num_input_tokens_seen": 12404464,
      "step": 6535
    },
    {
      "epoch": 1.0668896321070234,
      "grad_norm": 0.0006447941996157169,
      "learning_rate": 0.28064928655947097,
      "loss": 0.3555,
      "num_input_tokens_seen": 12412880,
      "step": 6540
    },
    {
      "epoch": 1.0677053593278407,
      "grad_norm": 0.0014747665263712406,
      "learning_rate": 0.28062033703135103,
      "loss": 0.3482,
      "num_input_tokens_seen": 12423600,
      "step": 6545
    },
    {
      "epoch": 1.068521086548658,
      "grad_norm": 0.0012199333868920803,
      "learning_rate": 0.2805913673599337,
      "loss": 0.3463,
      "num_input_tokens_seen": 12433296,
      "step": 6550
    },
    {
      "epoch": 1.0693368137694754,
      "grad_norm": 0.0005281802150420845,
      "learning_rate": 0.2805623775496864,
      "loss": 0.3413,
      "num_input_tokens_seen": 12443104,
      "step": 6555
    },
    {
      "epoch": 1.0701525409902928,
      "grad_norm": 0.0012942126486450434,
      "learning_rate": 0.2805333676050797,
      "loss": 0.3313,
      "num_input_tokens_seen": 12452800,
      "step": 6560
    },
    {
      "epoch": 1.0709682682111101,
      "grad_norm": 0.0010221637785434723,
      "learning_rate": 0.2805043375305873,
      "loss": 0.3592,
      "num_input_tokens_seen": 12463280,
      "step": 6565
    },
    {
      "epoch": 1.0717839954319275,
      "grad_norm": 0.0012327907606959343,
      "learning_rate": 0.2804752873306861,
      "loss": 0.3259,
      "num_input_tokens_seen": 12472992,
      "step": 6570
    },
    {
      "epoch": 1.0725997226527448,
      "grad_norm": 0.0013957293704152107,
      "learning_rate": 0.2804462170098559,
      "loss": 0.374,
      "num_input_tokens_seen": 12483488,
      "step": 6575
    },
    {
      "epoch": 1.0734154498735622,
      "grad_norm": 0.0018182756612077355,
      "learning_rate": 0.2804171265725797,
      "loss": 0.3724,
      "num_input_tokens_seen": 12493728,
      "step": 6580
    },
    {
      "epoch": 1.0742311770943795,
      "grad_norm": 0.0023061803076416254,
      "learning_rate": 0.28038801602334373,
      "loss": 0.3628,
      "num_input_tokens_seen": 12502880,
      "step": 6585
    },
    {
      "epoch": 1.075046904315197,
      "grad_norm": 0.0021340888924896717,
      "learning_rate": 0.28035888536663717,
      "loss": 0.4072,
      "num_input_tokens_seen": 12512160,
      "step": 6590
    },
    {
      "epoch": 1.0758626315360145,
      "grad_norm": 0.0009881602600216866,
      "learning_rate": 0.2803297346069522,
      "loss": 0.3617,
      "num_input_tokens_seen": 12522432,
      "step": 6595
    },
    {
      "epoch": 1.0766783587568316,
      "grad_norm": 0.0006467676139436662,
      "learning_rate": 0.28030056374878437,
      "loss": 0.373,
      "num_input_tokens_seen": 12531776,
      "step": 6600
    },
    {
      "epoch": 1.0766783587568316,
      "eval_loss": 0.3695012032985687,
      "eval_runtime": 152.5918,
      "eval_samples_per_second": 17.858,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 12531776,
      "step": 6600
    },
    {
      "epoch": 1.0774940859776492,
      "grad_norm": 0.0006966182263568044,
      "learning_rate": 0.2802713727966321,
      "loss": 0.3812,
      "num_input_tokens_seen": 12541728,
      "step": 6605
    },
    {
      "epoch": 1.0783098131984665,
      "grad_norm": 0.0009914040565490723,
      "learning_rate": 0.28024216175499717,
      "loss": 0.3754,
      "num_input_tokens_seen": 12551008,
      "step": 6610
    },
    {
      "epoch": 1.0791255404192839,
      "grad_norm": 0.0010491987923160195,
      "learning_rate": 0.2802129306283841,
      "loss": 0.405,
      "num_input_tokens_seen": 12560784,
      "step": 6615
    },
    {
      "epoch": 1.0799412676401012,
      "grad_norm": 0.0013637678930535913,
      "learning_rate": 0.28018367942130074,
      "loss": 0.3638,
      "num_input_tokens_seen": 12570064,
      "step": 6620
    },
    {
      "epoch": 1.0807569948609186,
      "grad_norm": 0.000548800453543663,
      "learning_rate": 0.28015440813825804,
      "loss": 0.3487,
      "num_input_tokens_seen": 12578400,
      "step": 6625
    },
    {
      "epoch": 1.081572722081736,
      "grad_norm": 0.0021167967934161425,
      "learning_rate": 0.28012511678377006,
      "loss": 0.3415,
      "num_input_tokens_seen": 12588784,
      "step": 6630
    },
    {
      "epoch": 1.0823884493025533,
      "grad_norm": 0.0010074232704937458,
      "learning_rate": 0.28009580536235373,
      "loss": 0.3815,
      "num_input_tokens_seen": 12597680,
      "step": 6635
    },
    {
      "epoch": 1.0832041765233706,
      "grad_norm": 0.0013437567977234721,
      "learning_rate": 0.28006647387852934,
      "loss": 0.3443,
      "num_input_tokens_seen": 12605952,
      "step": 6640
    },
    {
      "epoch": 1.084019903744188,
      "grad_norm": 0.000636246579233557,
      "learning_rate": 0.28003712233682015,
      "loss": 0.364,
      "num_input_tokens_seen": 12615984,
      "step": 6645
    },
    {
      "epoch": 1.0848356309650053,
      "grad_norm": 0.0013160206144675612,
      "learning_rate": 0.2800077507417526,
      "loss": 0.3556,
      "num_input_tokens_seen": 12625408,
      "step": 6650
    },
    {
      "epoch": 1.0856513581858227,
      "grad_norm": 0.0009334936621598899,
      "learning_rate": 0.2799783590978561,
      "loss": 0.3342,
      "num_input_tokens_seen": 12634480,
      "step": 6655
    },
    {
      "epoch": 1.08646708540664,
      "grad_norm": 0.0032011177390813828,
      "learning_rate": 0.2799489474096632,
      "loss": 0.4084,
      "num_input_tokens_seen": 12645120,
      "step": 6660
    },
    {
      "epoch": 1.0872828126274574,
      "grad_norm": 0.0006627660477533937,
      "learning_rate": 0.27991951568170953,
      "loss": 0.376,
      "num_input_tokens_seen": 12655856,
      "step": 6665
    },
    {
      "epoch": 1.0880985398482748,
      "grad_norm": 0.0017432889435440302,
      "learning_rate": 0.2798900639185339,
      "loss": 0.3739,
      "num_input_tokens_seen": 12665072,
      "step": 6670
    },
    {
      "epoch": 1.0889142670690921,
      "grad_norm": 0.0004541601228993386,
      "learning_rate": 0.2798605921246781,
      "loss": 0.3684,
      "num_input_tokens_seen": 12674656,
      "step": 6675
    },
    {
      "epoch": 1.0897299942899095,
      "grad_norm": 0.0009525239001959562,
      "learning_rate": 0.2798311003046871,
      "loss": 0.3572,
      "num_input_tokens_seen": 12684464,
      "step": 6680
    },
    {
      "epoch": 1.0905457215107268,
      "grad_norm": 0.0007871373672969639,
      "learning_rate": 0.2798015884631089,
      "loss": 0.4101,
      "num_input_tokens_seen": 12693232,
      "step": 6685
    },
    {
      "epoch": 1.0913614487315442,
      "grad_norm": 0.0026632375083863735,
      "learning_rate": 0.27977205660449445,
      "loss": 0.338,
      "num_input_tokens_seen": 12702976,
      "step": 6690
    },
    {
      "epoch": 1.0921771759523615,
      "grad_norm": 0.0017856592312455177,
      "learning_rate": 0.2797425047333981,
      "loss": 0.3664,
      "num_input_tokens_seen": 12711552,
      "step": 6695
    },
    {
      "epoch": 1.0929929031731789,
      "grad_norm": 0.0008778095943853259,
      "learning_rate": 0.27971293285437715,
      "loss": 0.3523,
      "num_input_tokens_seen": 12721472,
      "step": 6700
    },
    {
      "epoch": 1.0938086303939962,
      "grad_norm": 0.0006722378893755376,
      "learning_rate": 0.2796833409719918,
      "loss": 0.3573,
      "num_input_tokens_seen": 12731664,
      "step": 6705
    },
    {
      "epoch": 1.0946243576148136,
      "grad_norm": 0.002635650336742401,
      "learning_rate": 0.27965372909080566,
      "loss": 0.3381,
      "num_input_tokens_seen": 12741584,
      "step": 6710
    },
    {
      "epoch": 1.095440084835631,
      "grad_norm": 0.0013944749953225255,
      "learning_rate": 0.27962409721538506,
      "loss": 0.2677,
      "num_input_tokens_seen": 12750832,
      "step": 6715
    },
    {
      "epoch": 1.0962558120564483,
      "grad_norm": 0.001672710059210658,
      "learning_rate": 0.27959444535029976,
      "loss": 0.3776,
      "num_input_tokens_seen": 12760096,
      "step": 6720
    },
    {
      "epoch": 1.0970715392772656,
      "grad_norm": 0.0008067198214121163,
      "learning_rate": 0.27956477350012243,
      "loss": 0.3419,
      "num_input_tokens_seen": 12768912,
      "step": 6725
    },
    {
      "epoch": 1.097887266498083,
      "grad_norm": 0.001672719488851726,
      "learning_rate": 0.27953508166942875,
      "loss": 0.3717,
      "num_input_tokens_seen": 12777824,
      "step": 6730
    },
    {
      "epoch": 1.0987029937189003,
      "grad_norm": 0.0021619885228574276,
      "learning_rate": 0.27950536986279767,
      "loss": 0.3658,
      "num_input_tokens_seen": 12786400,
      "step": 6735
    },
    {
      "epoch": 1.0995187209397177,
      "grad_norm": 0.0005158812855370343,
      "learning_rate": 0.2794756380848111,
      "loss": 0.3506,
      "num_input_tokens_seen": 12796752,
      "step": 6740
    },
    {
      "epoch": 1.100334448160535,
      "grad_norm": 0.001374006737023592,
      "learning_rate": 0.279445886340054,
      "loss": 0.3584,
      "num_input_tokens_seen": 12804160,
      "step": 6745
    },
    {
      "epoch": 1.1011501753813524,
      "grad_norm": 0.001793332863599062,
      "learning_rate": 0.27941611463311455,
      "loss": 0.3578,
      "num_input_tokens_seen": 12813152,
      "step": 6750
    },
    {
      "epoch": 1.1019659026021698,
      "grad_norm": 0.0017084276769310236,
      "learning_rate": 0.2793863229685839,
      "loss": 0.3521,
      "num_input_tokens_seen": 12823552,
      "step": 6755
    },
    {
      "epoch": 1.102781629822987,
      "grad_norm": 0.0026512620970606804,
      "learning_rate": 0.27935651135105627,
      "loss": 0.3389,
      "num_input_tokens_seen": 12833056,
      "step": 6760
    },
    {
      "epoch": 1.1035973570438045,
      "grad_norm": 0.0007619087700732052,
      "learning_rate": 0.279326679785129,
      "loss": 0.3379,
      "num_input_tokens_seen": 12841008,
      "step": 6765
    },
    {
      "epoch": 1.1044130842646218,
      "grad_norm": 0.00041776610305532813,
      "learning_rate": 0.2792968282754024,
      "loss": 0.362,
      "num_input_tokens_seen": 12849920,
      "step": 6770
    },
    {
      "epoch": 1.1052288114854392,
      "grad_norm": 0.0017980198608711362,
      "learning_rate": 0.2792669568264801,
      "loss": 0.3863,
      "num_input_tokens_seen": 12858192,
      "step": 6775
    },
    {
      "epoch": 1.1060445387062567,
      "grad_norm": 0.0009035516413860023,
      "learning_rate": 0.27923706544296856,
      "loss": 0.3671,
      "num_input_tokens_seen": 12866720,
      "step": 6780
    },
    {
      "epoch": 1.1068602659270739,
      "grad_norm": 0.0015289551811292768,
      "learning_rate": 0.2792071541294775,
      "loss": 0.378,
      "num_input_tokens_seen": 12876064,
      "step": 6785
    },
    {
      "epoch": 1.1076759931478914,
      "grad_norm": 0.0006152919377200305,
      "learning_rate": 0.27917722289061947,
      "loss": 0.3586,
      "num_input_tokens_seen": 12886032,
      "step": 6790
    },
    {
      "epoch": 1.1084917203687088,
      "grad_norm": 0.001093339640647173,
      "learning_rate": 0.27914727173101034,
      "loss": 0.3355,
      "num_input_tokens_seen": 12895248,
      "step": 6795
    },
    {
      "epoch": 1.1093074475895262,
      "grad_norm": 0.0008834140608087182,
      "learning_rate": 0.279117300655269,
      "loss": 0.3413,
      "num_input_tokens_seen": 12905136,
      "step": 6800
    },
    {
      "epoch": 1.1093074475895262,
      "eval_loss": 0.33867526054382324,
      "eval_runtime": 152.5234,
      "eval_samples_per_second": 17.866,
      "eval_steps_per_second": 8.936,
      "num_input_tokens_seen": 12905136,
      "step": 6800
    },
    {
      "epoch": 1.1101231748103435,
      "grad_norm": 0.0016840314492583275,
      "learning_rate": 0.2790873096680173,
      "loss": 0.3638,
      "num_input_tokens_seen": 12913632,
      "step": 6805
    },
    {
      "epoch": 1.1109389020311609,
      "grad_norm": 0.003126170951873064,
      "learning_rate": 0.2790572987738802,
      "loss": 0.3551,
      "num_input_tokens_seen": 12924320,
      "step": 6810
    },
    {
      "epoch": 1.1117546292519782,
      "grad_norm": 0.0006801526178605855,
      "learning_rate": 0.27902726797748584,
      "loss": 0.2942,
      "num_input_tokens_seen": 12933872,
      "step": 6815
    },
    {
      "epoch": 1.1125703564727956,
      "grad_norm": 0.000634153198916465,
      "learning_rate": 0.2789972172834652,
      "loss": 0.2931,
      "num_input_tokens_seen": 12943104,
      "step": 6820
    },
    {
      "epoch": 1.113386083693613,
      "grad_norm": 0.0009406222961843014,
      "learning_rate": 0.2789671466964527,
      "loss": 0.3998,
      "num_input_tokens_seen": 12951888,
      "step": 6825
    },
    {
      "epoch": 1.1142018109144303,
      "grad_norm": 0.0027363982517272234,
      "learning_rate": 0.2789370562210854,
      "loss": 0.3595,
      "num_input_tokens_seen": 12960992,
      "step": 6830
    },
    {
      "epoch": 1.1150175381352476,
      "grad_norm": 0.0006936319987289608,
      "learning_rate": 0.27890694586200376,
      "loss": 0.3426,
      "num_input_tokens_seen": 12968944,
      "step": 6835
    },
    {
      "epoch": 1.115833265356065,
      "grad_norm": 0.001334614004008472,
      "learning_rate": 0.2788768156238511,
      "loss": 0.3831,
      "num_input_tokens_seen": 12978304,
      "step": 6840
    },
    {
      "epoch": 1.1166489925768823,
      "grad_norm": 0.0013548446586355567,
      "learning_rate": 0.27884666551127385,
      "loss": 0.3031,
      "num_input_tokens_seen": 12987280,
      "step": 6845
    },
    {
      "epoch": 1.1174647197976997,
      "grad_norm": 0.003245868254452944,
      "learning_rate": 0.2788164955289217,
      "loss": 0.385,
      "num_input_tokens_seen": 12997312,
      "step": 6850
    },
    {
      "epoch": 1.118280447018517,
      "grad_norm": 0.0010349185904487967,
      "learning_rate": 0.27878630568144697,
      "loss": 0.3342,
      "num_input_tokens_seen": 13006320,
      "step": 6855
    },
    {
      "epoch": 1.1190961742393344,
      "grad_norm": 0.0009269205038435757,
      "learning_rate": 0.2787560959735056,
      "loss": 0.3727,
      "num_input_tokens_seen": 13015968,
      "step": 6860
    },
    {
      "epoch": 1.1199119014601517,
      "grad_norm": 0.0031207718420773745,
      "learning_rate": 0.27872586640975616,
      "loss": 0.3523,
      "num_input_tokens_seen": 13025728,
      "step": 6865
    },
    {
      "epoch": 1.120727628680969,
      "grad_norm": 0.0018816185183823109,
      "learning_rate": 0.27869561699486045,
      "loss": 0.3868,
      "num_input_tokens_seen": 13036032,
      "step": 6870
    },
    {
      "epoch": 1.1215433559017864,
      "grad_norm": 0.0010791235836222768,
      "learning_rate": 0.2786653477334833,
      "loss": 0.3529,
      "num_input_tokens_seen": 13045088,
      "step": 6875
    },
    {
      "epoch": 1.1223590831226038,
      "grad_norm": 0.0015611290000379086,
      "learning_rate": 0.2786350586302926,
      "loss": 0.2979,
      "num_input_tokens_seen": 13053456,
      "step": 6880
    },
    {
      "epoch": 1.1231748103434211,
      "grad_norm": 0.0013622696278616786,
      "learning_rate": 0.27860474968995935,
      "loss": 0.3771,
      "num_input_tokens_seen": 13062800,
      "step": 6885
    },
    {
      "epoch": 1.1239905375642385,
      "grad_norm": 0.0012033062521368265,
      "learning_rate": 0.27857442091715756,
      "loss": 0.365,
      "num_input_tokens_seen": 13071328,
      "step": 6890
    },
    {
      "epoch": 1.1248062647850559,
      "grad_norm": 0.0006632860749959946,
      "learning_rate": 0.27854407231656425,
      "loss": 0.3347,
      "num_input_tokens_seen": 13081376,
      "step": 6895
    },
    {
      "epoch": 1.1256219920058732,
      "grad_norm": 0.0015995276626199484,
      "learning_rate": 0.2785137038928596,
      "loss": 0.3474,
      "num_input_tokens_seen": 13091408,
      "step": 6900
    },
    {
      "epoch": 1.1264377192266906,
      "grad_norm": 0.0026856313925236464,
      "learning_rate": 0.27848331565072687,
      "loss": 0.3955,
      "num_input_tokens_seen": 13101296,
      "step": 6905
    },
    {
      "epoch": 1.127253446447508,
      "grad_norm": 0.0016941586509346962,
      "learning_rate": 0.27845290759485225,
      "loss": 0.338,
      "num_input_tokens_seen": 13111664,
      "step": 6910
    },
    {
      "epoch": 1.1280691736683253,
      "grad_norm": 0.000621258863247931,
      "learning_rate": 0.278422479729925,
      "loss": 0.3802,
      "num_input_tokens_seen": 13121328,
      "step": 6915
    },
    {
      "epoch": 1.1288849008891426,
      "grad_norm": 0.0013127580750733614,
      "learning_rate": 0.2783920320606375,
      "loss": 0.3961,
      "num_input_tokens_seen": 13129776,
      "step": 6920
    },
    {
      "epoch": 1.12970062810996,
      "grad_norm": 0.002444059820845723,
      "learning_rate": 0.2783615645916852,
      "loss": 0.3896,
      "num_input_tokens_seen": 13138880,
      "step": 6925
    },
    {
      "epoch": 1.1305163553307773,
      "grad_norm": 0.0016429460374638438,
      "learning_rate": 0.2783310773277666,
      "loss": 0.3141,
      "num_input_tokens_seen": 13147584,
      "step": 6930
    },
    {
      "epoch": 1.1313320825515947,
      "grad_norm": 0.001606132136657834,
      "learning_rate": 0.2783005702735831,
      "loss": 0.3137,
      "num_input_tokens_seen": 13157600,
      "step": 6935
    },
    {
      "epoch": 1.132147809772412,
      "grad_norm": 0.0012217191979289055,
      "learning_rate": 0.2782700434338394,
      "loss": 0.3538,
      "num_input_tokens_seen": 13166032,
      "step": 6940
    },
    {
      "epoch": 1.1329635369932294,
      "grad_norm": 0.000982588971965015,
      "learning_rate": 0.278239496813243,
      "loss": 0.3662,
      "num_input_tokens_seen": 13174336,
      "step": 6945
    },
    {
      "epoch": 1.1337792642140467,
      "grad_norm": 0.001159683451987803,
      "learning_rate": 0.27820893041650463,
      "loss": 0.3349,
      "num_input_tokens_seen": 13184112,
      "step": 6950
    },
    {
      "epoch": 1.134594991434864,
      "grad_norm": 0.000883732340298593,
      "learning_rate": 0.27817834424833804,
      "loss": 0.3681,
      "num_input_tokens_seen": 13193616,
      "step": 6955
    },
    {
      "epoch": 1.1354107186556814,
      "grad_norm": 0.0010262469295412302,
      "learning_rate": 0.27814773831345996,
      "loss": 0.3407,
      "num_input_tokens_seen": 13203584,
      "step": 6960
    },
    {
      "epoch": 1.136226445876499,
      "grad_norm": 0.0013003754429519176,
      "learning_rate": 0.2781171126165902,
      "loss": 0.3748,
      "num_input_tokens_seen": 13213680,
      "step": 6965
    },
    {
      "epoch": 1.1370421730973161,
      "grad_norm": 0.0004934382741339505,
      "learning_rate": 0.2780864671624517,
      "loss": 0.3556,
      "num_input_tokens_seen": 13222368,
      "step": 6970
    },
    {
      "epoch": 1.1378579003181337,
      "grad_norm": 0.0015973102999851108,
      "learning_rate": 0.27805580195577034,
      "loss": 0.3201,
      "num_input_tokens_seen": 13230112,
      "step": 6975
    },
    {
      "epoch": 1.1386736275389508,
      "grad_norm": 0.0013240681728348136,
      "learning_rate": 0.2780251170012751,
      "loss": 0.2932,
      "num_input_tokens_seen": 13239520,
      "step": 6980
    },
    {
      "epoch": 1.1394893547597684,
      "grad_norm": 0.0006890271906740963,
      "learning_rate": 0.27799441230369787,
      "loss": 0.3701,
      "num_input_tokens_seen": 13249280,
      "step": 6985
    },
    {
      "epoch": 1.1403050819805858,
      "grad_norm": 0.0018725987756624818,
      "learning_rate": 0.27796368786777387,
      "loss": 0.3532,
      "num_input_tokens_seen": 13258640,
      "step": 6990
    },
    {
      "epoch": 1.1411208092014031,
      "grad_norm": 0.0010488536208868027,
      "learning_rate": 0.277932943698241,
      "loss": 0.3526,
      "num_input_tokens_seen": 13268064,
      "step": 6995
    },
    {
      "epoch": 1.1419365364222205,
      "grad_norm": 0.001485652755945921,
      "learning_rate": 0.2779021797998406,
      "loss": 0.4019,
      "num_input_tokens_seen": 13278096,
      "step": 7000
    },
    {
      "epoch": 1.1419365364222205,
      "eval_loss": 0.35444387793540955,
      "eval_runtime": 152.7391,
      "eval_samples_per_second": 17.841,
      "eval_steps_per_second": 8.924,
      "num_input_tokens_seen": 13278096,
      "step": 7000
    },
    {
      "epoch": 1.1427522636430378,
      "grad_norm": 0.003758337814360857,
      "learning_rate": 0.2778713961773167,
      "loss": 0.3885,
      "num_input_tokens_seen": 13287296,
      "step": 7005
    },
    {
      "epoch": 1.1435679908638552,
      "grad_norm": 0.0021785160060971975,
      "learning_rate": 0.2778405928354166,
      "loss": 0.4008,
      "num_input_tokens_seen": 13296256,
      "step": 7010
    },
    {
      "epoch": 1.1443837180846725,
      "grad_norm": 0.00046544725773856044,
      "learning_rate": 0.27780976977889055,
      "loss": 0.3517,
      "num_input_tokens_seen": 13305360,
      "step": 7015
    },
    {
      "epoch": 1.14519944530549,
      "grad_norm": 0.0013884884538128972,
      "learning_rate": 0.27777892701249185,
      "loss": 0.3389,
      "num_input_tokens_seen": 13315344,
      "step": 7020
    },
    {
      "epoch": 1.1460151725263072,
      "grad_norm": 0.0010952455922961235,
      "learning_rate": 0.2777480645409768,
      "loss": 0.3321,
      "num_input_tokens_seen": 13324240,
      "step": 7025
    },
    {
      "epoch": 1.1468308997471246,
      "grad_norm": 0.0014365895185619593,
      "learning_rate": 0.27771718236910486,
      "loss": 0.3634,
      "num_input_tokens_seen": 13333696,
      "step": 7030
    },
    {
      "epoch": 1.147646626967942,
      "grad_norm": 0.0028489793185144663,
      "learning_rate": 0.27768628050163835,
      "loss": 0.3049,
      "num_input_tokens_seen": 13344400,
      "step": 7035
    },
    {
      "epoch": 1.1484623541887593,
      "grad_norm": 0.0010984748369082808,
      "learning_rate": 0.2776553589433428,
      "loss": 0.4175,
      "num_input_tokens_seen": 13353760,
      "step": 7040
    },
    {
      "epoch": 1.1492780814095767,
      "grad_norm": 0.0029544257558882236,
      "learning_rate": 0.27762441769898666,
      "loss": 0.3966,
      "num_input_tokens_seen": 13362368,
      "step": 7045
    },
    {
      "epoch": 1.150093808630394,
      "grad_norm": 0.0004505177494138479,
      "learning_rate": 0.2775934567733415,
      "loss": 0.3631,
      "num_input_tokens_seen": 13371120,
      "step": 7050
    },
    {
      "epoch": 1.1509095358512114,
      "grad_norm": 0.0016926538664847612,
      "learning_rate": 0.2775624761711819,
      "loss": 0.3563,
      "num_input_tokens_seen": 13380880,
      "step": 7055
    },
    {
      "epoch": 1.1517252630720287,
      "grad_norm": 0.0017845648108050227,
      "learning_rate": 0.2775314758972854,
      "loss": 0.3405,
      "num_input_tokens_seen": 13389568,
      "step": 7060
    },
    {
      "epoch": 1.152540990292846,
      "grad_norm": 0.0024849334731698036,
      "learning_rate": 0.2775004559564327,
      "loss": 0.3551,
      "num_input_tokens_seen": 13400592,
      "step": 7065
    },
    {
      "epoch": 1.1533567175136634,
      "grad_norm": 0.0012031184742227197,
      "learning_rate": 0.2774694163534073,
      "loss": 0.3868,
      "num_input_tokens_seen": 13409488,
      "step": 7070
    },
    {
      "epoch": 1.1541724447344808,
      "grad_norm": 0.00291914539411664,
      "learning_rate": 0.27743835709299614,
      "loss": 0.3678,
      "num_input_tokens_seen": 13418176,
      "step": 7075
    },
    {
      "epoch": 1.1549881719552981,
      "grad_norm": 0.001031563151627779,
      "learning_rate": 0.2774072781799888,
      "loss": 0.457,
      "num_input_tokens_seen": 13427120,
      "step": 7080
    },
    {
      "epoch": 1.1558038991761155,
      "grad_norm": 0.0019767354242503643,
      "learning_rate": 0.27737617961917804,
      "loss": 0.3735,
      "num_input_tokens_seen": 13436784,
      "step": 7085
    },
    {
      "epoch": 1.1566196263969328,
      "grad_norm": 0.0011822794331237674,
      "learning_rate": 0.27734506141535964,
      "loss": 0.4421,
      "num_input_tokens_seen": 13446944,
      "step": 7090
    },
    {
      "epoch": 1.1574353536177502,
      "grad_norm": 0.000917378521990031,
      "learning_rate": 0.2773139235733325,
      "loss": 0.3147,
      "num_input_tokens_seen": 13456400,
      "step": 7095
    },
    {
      "epoch": 1.1582510808385675,
      "grad_norm": 0.0007886185776442289,
      "learning_rate": 0.2772827660978984,
      "loss": 0.3248,
      "num_input_tokens_seen": 13464512,
      "step": 7100
    },
    {
      "epoch": 1.159066808059385,
      "grad_norm": 0.0007839500322006643,
      "learning_rate": 0.27725158899386226,
      "loss": 0.369,
      "num_input_tokens_seen": 13473712,
      "step": 7105
    },
    {
      "epoch": 1.1598825352802022,
      "grad_norm": 0.0013919449411332607,
      "learning_rate": 0.27722039226603196,
      "loss": 0.3364,
      "num_input_tokens_seen": 13481984,
      "step": 7110
    },
    {
      "epoch": 1.1606982625010196,
      "grad_norm": 0.0007018260657787323,
      "learning_rate": 0.2771891759192184,
      "loss": 0.3409,
      "num_input_tokens_seen": 13490992,
      "step": 7115
    },
    {
      "epoch": 1.161513989721837,
      "grad_norm": 0.0011529062176123261,
      "learning_rate": 0.2771579399582355,
      "loss": 0.3339,
      "num_input_tokens_seen": 13500416,
      "step": 7120
    },
    {
      "epoch": 1.1623297169426543,
      "grad_norm": 0.0028723019640892744,
      "learning_rate": 0.2771266843879004,
      "loss": 0.3451,
      "num_input_tokens_seen": 13509920,
      "step": 7125
    },
    {
      "epoch": 1.1631454441634717,
      "grad_norm": 0.0015574999852105975,
      "learning_rate": 0.2770954092130329,
      "loss": 0.3725,
      "num_input_tokens_seen": 13520176,
      "step": 7130
    },
    {
      "epoch": 1.163961171384289,
      "grad_norm": 0.0007420693873427808,
      "learning_rate": 0.27706411443845613,
      "loss": 0.3136,
      "num_input_tokens_seen": 13530496,
      "step": 7135
    },
    {
      "epoch": 1.1647768986051064,
      "grad_norm": 0.000509401666931808,
      "learning_rate": 0.27703280006899617,
      "loss": 0.3431,
      "num_input_tokens_seen": 13539536,
      "step": 7140
    },
    {
      "epoch": 1.1655926258259237,
      "grad_norm": 0.0006980726611800492,
      "learning_rate": 0.277001466109482,
      "loss": 0.3571,
      "num_input_tokens_seen": 13548976,
      "step": 7145
    },
    {
      "epoch": 1.1664083530467413,
      "grad_norm": 0.001404101261869073,
      "learning_rate": 0.2769701125647458,
      "loss": 0.3483,
      "num_input_tokens_seen": 13559184,
      "step": 7150
    },
    {
      "epoch": 1.1672240802675584,
      "grad_norm": 0.00439103227108717,
      "learning_rate": 0.27693873943962266,
      "loss": 0.3877,
      "num_input_tokens_seen": 13567072,
      "step": 7155
    },
    {
      "epoch": 1.168039807488376,
      "grad_norm": 0.0008706832304596901,
      "learning_rate": 0.2769073467389506,
      "loss": 0.3727,
      "num_input_tokens_seen": 13575536,
      "step": 7160
    },
    {
      "epoch": 1.1688555347091931,
      "grad_norm": 0.0006327937007881701,
      "learning_rate": 0.2768759344675709,
      "loss": 0.3539,
      "num_input_tokens_seen": 13585056,
      "step": 7165
    },
    {
      "epoch": 1.1696712619300107,
      "grad_norm": 0.0018786301370710135,
      "learning_rate": 0.27684450263032767,
      "loss": 0.3528,
      "num_input_tokens_seen": 13594432,
      "step": 7170
    },
    {
      "epoch": 1.170486989150828,
      "grad_norm": 0.0035210042260587215,
      "learning_rate": 0.2768130512320682,
      "loss": 0.3047,
      "num_input_tokens_seen": 13603904,
      "step": 7175
    },
    {
      "epoch": 1.1713027163716454,
      "grad_norm": 0.0018821959383785725,
      "learning_rate": 0.27678158027764244,
      "loss": 0.3859,
      "num_input_tokens_seen": 13614000,
      "step": 7180
    },
    {
      "epoch": 1.1721184435924628,
      "grad_norm": 0.0016135943587869406,
      "learning_rate": 0.27675008977190385,
      "loss": 0.3228,
      "num_input_tokens_seen": 13622560,
      "step": 7185
    },
    {
      "epoch": 1.17293417081328,
      "grad_norm": 0.000694341491907835,
      "learning_rate": 0.2767185797197086,
      "loss": 0.3271,
      "num_input_tokens_seen": 13632560,
      "step": 7190
    },
    {
      "epoch": 1.1737498980340975,
      "grad_norm": 0.002029036870226264,
      "learning_rate": 0.2766870501259159,
      "loss": 0.3917,
      "num_input_tokens_seen": 13641664,
      "step": 7195
    },
    {
      "epoch": 1.1745656252549148,
      "grad_norm": 0.002461954951286316,
      "learning_rate": 0.276655500995388,
      "loss": 0.3835,
      "num_input_tokens_seen": 13651520,
      "step": 7200
    },
    {
      "epoch": 1.1745656252549148,
      "eval_loss": 0.4184120297431946,
      "eval_runtime": 152.5855,
      "eval_samples_per_second": 17.859,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 13651520,
      "step": 7200
    },
    {
      "epoch": 1.1753813524757322,
      "grad_norm": 0.0005532087525352836,
      "learning_rate": 0.27662393233299015,
      "loss": 0.3848,
      "num_input_tokens_seen": 13661328,
      "step": 7205
    },
    {
      "epoch": 1.1761970796965495,
      "grad_norm": 0.0014781320933252573,
      "learning_rate": 0.27659234414359074,
      "loss": 0.3309,
      "num_input_tokens_seen": 13671472,
      "step": 7210
    },
    {
      "epoch": 1.1770128069173669,
      "grad_norm": 0.0005892399349249899,
      "learning_rate": 0.27656073643206097,
      "loss": 0.3446,
      "num_input_tokens_seen": 13679856,
      "step": 7215
    },
    {
      "epoch": 1.1778285341381842,
      "grad_norm": 0.003454157616943121,
      "learning_rate": 0.27652910920327517,
      "loss": 0.4397,
      "num_input_tokens_seen": 13689088,
      "step": 7220
    },
    {
      "epoch": 1.1786442613590016,
      "grad_norm": 0.0006296744686551392,
      "learning_rate": 0.2764974624621107,
      "loss": 0.3254,
      "num_input_tokens_seen": 13698320,
      "step": 7225
    },
    {
      "epoch": 1.179459988579819,
      "grad_norm": 0.0024948534555733204,
      "learning_rate": 0.2764657962134479,
      "loss": 0.4963,
      "num_input_tokens_seen": 13707680,
      "step": 7230
    },
    {
      "epoch": 1.1802757158006363,
      "grad_norm": 0.0010752322850748897,
      "learning_rate": 0.27643411046217,
      "loss": 0.3986,
      "num_input_tokens_seen": 13716784,
      "step": 7235
    },
    {
      "epoch": 1.1810914430214536,
      "grad_norm": 0.0010079179191961884,
      "learning_rate": 0.27640240521316334,
      "loss": 0.3381,
      "num_input_tokens_seen": 13727296,
      "step": 7240
    },
    {
      "epoch": 1.181907170242271,
      "grad_norm": 0.0009920584270730615,
      "learning_rate": 0.2763706804713174,
      "loss": 0.358,
      "num_input_tokens_seen": 13737136,
      "step": 7245
    },
    {
      "epoch": 1.1827228974630883,
      "grad_norm": 0.0005514549557119608,
      "learning_rate": 0.2763389362415245,
      "loss": 0.3779,
      "num_input_tokens_seen": 13747088,
      "step": 7250
    },
    {
      "epoch": 1.1835386246839057,
      "grad_norm": 0.0029738713055849075,
      "learning_rate": 0.27630717252867987,
      "loss": 0.3663,
      "num_input_tokens_seen": 13756608,
      "step": 7255
    },
    {
      "epoch": 1.184354351904723,
      "grad_norm": 0.0007154115592129529,
      "learning_rate": 0.276275389337682,
      "loss": 0.3726,
      "num_input_tokens_seen": 13764720,
      "step": 7260
    },
    {
      "epoch": 1.1851700791255404,
      "grad_norm": 0.0023761307820677757,
      "learning_rate": 0.2762435866734322,
      "loss": 0.3314,
      "num_input_tokens_seen": 13774400,
      "step": 7265
    },
    {
      "epoch": 1.1859858063463578,
      "grad_norm": 0.0007787389331497252,
      "learning_rate": 0.27621176454083485,
      "loss": 0.3257,
      "num_input_tokens_seen": 13783824,
      "step": 7270
    },
    {
      "epoch": 1.186801533567175,
      "grad_norm": 0.0007606134749948978,
      "learning_rate": 0.2761799229447973,
      "loss": 0.3097,
      "num_input_tokens_seen": 13793984,
      "step": 7275
    },
    {
      "epoch": 1.1876172607879925,
      "grad_norm": 0.0029448969289660454,
      "learning_rate": 0.27614806189023006,
      "loss": 0.3474,
      "num_input_tokens_seen": 13805360,
      "step": 7280
    },
    {
      "epoch": 1.1884329880088098,
      "grad_norm": 0.0014692695112898946,
      "learning_rate": 0.27611618138204636,
      "loss": 0.3253,
      "num_input_tokens_seen": 13815008,
      "step": 7285
    },
    {
      "epoch": 1.1892487152296272,
      "grad_norm": 0.0025196652859449387,
      "learning_rate": 0.2760842814251626,
      "loss": 0.3694,
      "num_input_tokens_seen": 13825008,
      "step": 7290
    },
    {
      "epoch": 1.1900644424504445,
      "grad_norm": 0.0020189471542835236,
      "learning_rate": 0.2760523620244982,
      "loss": 0.3484,
      "num_input_tokens_seen": 13833664,
      "step": 7295
    },
    {
      "epoch": 1.1908801696712619,
      "grad_norm": 0.0010102532105520368,
      "learning_rate": 0.27602042318497544,
      "loss": 0.3696,
      "num_input_tokens_seen": 13843312,
      "step": 7300
    },
    {
      "epoch": 1.1916958968920792,
      "grad_norm": 0.0018897748086601496,
      "learning_rate": 0.2759884649115198,
      "loss": 0.414,
      "num_input_tokens_seen": 13851408,
      "step": 7305
    },
    {
      "epoch": 1.1925116241128966,
      "grad_norm": 0.0016456498997285962,
      "learning_rate": 0.2759564872090596,
      "loss": 0.3461,
      "num_input_tokens_seen": 13860928,
      "step": 7310
    },
    {
      "epoch": 1.193327351333714,
      "grad_norm": 0.0020902513060718775,
      "learning_rate": 0.2759244900825262,
      "loss": 0.332,
      "num_input_tokens_seen": 13871168,
      "step": 7315
    },
    {
      "epoch": 1.1941430785545313,
      "grad_norm": 0.0018765906570479274,
      "learning_rate": 0.2758924735368539,
      "loss": 0.3787,
      "num_input_tokens_seen": 13882448,
      "step": 7320
    },
    {
      "epoch": 1.1949588057753486,
      "grad_norm": 0.0008483424899168313,
      "learning_rate": 0.27586043757698014,
      "loss": 0.3596,
      "num_input_tokens_seen": 13891344,
      "step": 7325
    },
    {
      "epoch": 1.195774532996166,
      "grad_norm": 0.002608418930321932,
      "learning_rate": 0.27582838220784534,
      "loss": 0.3743,
      "num_input_tokens_seen": 13900736,
      "step": 7330
    },
    {
      "epoch": 1.1965902602169836,
      "grad_norm": 0.0007968625868670642,
      "learning_rate": 0.27579630743439265,
      "loss": 0.3386,
      "num_input_tokens_seen": 13910960,
      "step": 7335
    },
    {
      "epoch": 1.1974059874378007,
      "grad_norm": 0.0008965309243649244,
      "learning_rate": 0.2757642132615686,
      "loss": 0.3583,
      "num_input_tokens_seen": 13920304,
      "step": 7340
    },
    {
      "epoch": 1.1982217146586183,
      "grad_norm": 0.0011739319888874888,
      "learning_rate": 0.2757320996943223,
      "loss": 0.3336,
      "num_input_tokens_seen": 13930288,
      "step": 7345
    },
    {
      "epoch": 1.1990374418794354,
      "grad_norm": 0.00065523327793926,
      "learning_rate": 0.2756999667376062,
      "loss": 0.3244,
      "num_input_tokens_seen": 13939872,
      "step": 7350
    },
    {
      "epoch": 1.199853169100253,
      "grad_norm": 0.001180437975563109,
      "learning_rate": 0.2756678143963756,
      "loss": 0.4081,
      "num_input_tokens_seen": 13948784,
      "step": 7355
    },
    {
      "epoch": 1.2006688963210703,
      "grad_norm": 0.0021433094516396523,
      "learning_rate": 0.2756356426755888,
      "loss": 0.3165,
      "num_input_tokens_seen": 13957840,
      "step": 7360
    },
    {
      "epoch": 1.2014846235418877,
      "grad_norm": 0.0007768502691760659,
      "learning_rate": 0.27560345158020705,
      "loss": 0.3453,
      "num_input_tokens_seen": 13967392,
      "step": 7365
    },
    {
      "epoch": 1.202300350762705,
      "grad_norm": 0.0016342711169272661,
      "learning_rate": 0.27557124111519465,
      "loss": 0.3822,
      "num_input_tokens_seen": 13976144,
      "step": 7370
    },
    {
      "epoch": 1.2031160779835224,
      "grad_norm": 0.002692136447876692,
      "learning_rate": 0.27553901128551883,
      "loss": 0.3533,
      "num_input_tokens_seen": 13985952,
      "step": 7375
    },
    {
      "epoch": 1.2039318052043397,
      "grad_norm": 0.0010627576848492026,
      "learning_rate": 0.2755067620961498,
      "loss": 0.3635,
      "num_input_tokens_seen": 13995520,
      "step": 7380
    },
    {
      "epoch": 1.204747532425157,
      "grad_norm": 0.0025253870990127325,
      "learning_rate": 0.27547449355206094,
      "loss": 0.3394,
      "num_input_tokens_seen": 14004496,
      "step": 7385
    },
    {
      "epoch": 1.2055632596459744,
      "grad_norm": 0.001058046123944223,
      "learning_rate": 0.2754422056582283,
      "loss": 0.3719,
      "num_input_tokens_seen": 14014368,
      "step": 7390
    },
    {
      "epoch": 1.2063789868667918,
      "grad_norm": 0.0013596685603260994,
      "learning_rate": 0.27540989841963115,
      "loss": 0.3574,
      "num_input_tokens_seen": 14024464,
      "step": 7395
    },
    {
      "epoch": 1.2071947140876091,
      "grad_norm": 0.000922013889066875,
      "learning_rate": 0.27537757184125167,
      "loss": 0.3365,
      "num_input_tokens_seen": 14034784,
      "step": 7400
    },
    {
      "epoch": 1.2071947140876091,
      "eval_loss": 0.3426559269428253,
      "eval_runtime": 152.5751,
      "eval_samples_per_second": 17.86,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 14034784,
      "step": 7400
    },
    {
      "epoch": 1.2080104413084265,
      "grad_norm": 0.0037346086464822292,
      "learning_rate": 0.275345225928075,
      "loss": 0.4382,
      "num_input_tokens_seen": 14043504,
      "step": 7405
    },
    {
      "epoch": 1.2088261685292438,
      "grad_norm": 0.001248649088665843,
      "learning_rate": 0.2753128606850893,
      "loss": 0.3458,
      "num_input_tokens_seen": 14053552,
      "step": 7410
    },
    {
      "epoch": 1.2096418957500612,
      "grad_norm": 0.0033625790383666754,
      "learning_rate": 0.2752804761172858,
      "loss": 0.4412,
      "num_input_tokens_seen": 14063360,
      "step": 7415
    },
    {
      "epoch": 1.2104576229708786,
      "grad_norm": 0.0008967912290245295,
      "learning_rate": 0.27524807222965836,
      "loss": 0.4393,
      "num_input_tokens_seen": 14073088,
      "step": 7420
    },
    {
      "epoch": 1.211273350191696,
      "grad_norm": 0.001820297446101904,
      "learning_rate": 0.27521564902720436,
      "loss": 0.3296,
      "num_input_tokens_seen": 14083216,
      "step": 7425
    },
    {
      "epoch": 1.2120890774125133,
      "grad_norm": 0.0021982467733323574,
      "learning_rate": 0.2751832065149236,
      "loss": 0.3486,
      "num_input_tokens_seen": 14092448,
      "step": 7430
    },
    {
      "epoch": 1.2129048046333306,
      "grad_norm": 0.0036574818659573793,
      "learning_rate": 0.2751507446978193,
      "loss": 0.3949,
      "num_input_tokens_seen": 14102896,
      "step": 7435
    },
    {
      "epoch": 1.213720531854148,
      "grad_norm": 0.0018331363098695874,
      "learning_rate": 0.2751182635808974,
      "loss": 0.3304,
      "num_input_tokens_seen": 14111936,
      "step": 7440
    },
    {
      "epoch": 1.2145362590749653,
      "grad_norm": 0.000862038868945092,
      "learning_rate": 0.27508576316916694,
      "loss": 0.3644,
      "num_input_tokens_seen": 14122048,
      "step": 7445
    },
    {
      "epoch": 1.2153519862957827,
      "grad_norm": 0.0007994803017936647,
      "learning_rate": 0.2750532434676399,
      "loss": 0.3537,
      "num_input_tokens_seen": 14132288,
      "step": 7450
    },
    {
      "epoch": 1.2161677135166,
      "grad_norm": 0.0008962511201389134,
      "learning_rate": 0.27502070448133115,
      "loss": 0.3309,
      "num_input_tokens_seen": 14141728,
      "step": 7455
    },
    {
      "epoch": 1.2169834407374174,
      "grad_norm": 0.001891458174213767,
      "learning_rate": 0.2749881462152587,
      "loss": 0.298,
      "num_input_tokens_seen": 14151664,
      "step": 7460
    },
    {
      "epoch": 1.2177991679582347,
      "grad_norm": 0.003960195928812027,
      "learning_rate": 0.2749555686744434,
      "loss": 0.5186,
      "num_input_tokens_seen": 14161008,
      "step": 7465
    },
    {
      "epoch": 1.218614895179052,
      "grad_norm": 0.0015091205714270473,
      "learning_rate": 0.2749229718639091,
      "loss": 0.3247,
      "num_input_tokens_seen": 14170256,
      "step": 7470
    },
    {
      "epoch": 1.2194306223998694,
      "grad_norm": 0.0005804325337521732,
      "learning_rate": 0.27489035578868265,
      "loss": 0.3326,
      "num_input_tokens_seen": 14179136,
      "step": 7475
    },
    {
      "epoch": 1.2202463496206868,
      "grad_norm": 0.0019873687997460365,
      "learning_rate": 0.2748577204537939,
      "loss": 0.4827,
      "num_input_tokens_seen": 14188224,
      "step": 7480
    },
    {
      "epoch": 1.2210620768415041,
      "grad_norm": 0.0021346488501876593,
      "learning_rate": 0.2748250658642756,
      "loss": 0.3581,
      "num_input_tokens_seen": 14197440,
      "step": 7485
    },
    {
      "epoch": 1.2218778040623215,
      "grad_norm": 0.0025103269144892693,
      "learning_rate": 0.2747923920251634,
      "loss": 0.4585,
      "num_input_tokens_seen": 14206736,
      "step": 7490
    },
    {
      "epoch": 1.2226935312831388,
      "grad_norm": 0.0010144688421860337,
      "learning_rate": 0.27475969894149627,
      "loss": 0.3572,
      "num_input_tokens_seen": 14215200,
      "step": 7495
    },
    {
      "epoch": 1.2235092585039562,
      "grad_norm": 0.0007037764880806208,
      "learning_rate": 0.2747269866183156,
      "loss": 0.3471,
      "num_input_tokens_seen": 14224992,
      "step": 7500
    },
    {
      "epoch": 1.2243249857247736,
      "grad_norm": 0.0010320536093786359,
      "learning_rate": 0.27469425506066625,
      "loss": 0.3567,
      "num_input_tokens_seen": 14235520,
      "step": 7505
    },
    {
      "epoch": 1.225140712945591,
      "grad_norm": 0.002441722434014082,
      "learning_rate": 0.27466150427359576,
      "loss": 0.3409,
      "num_input_tokens_seen": 14244880,
      "step": 7510
    },
    {
      "epoch": 1.2259564401664083,
      "grad_norm": 0.0013171572936698794,
      "learning_rate": 0.2746287342621547,
      "loss": 0.3813,
      "num_input_tokens_seen": 14254352,
      "step": 7515
    },
    {
      "epoch": 1.2267721673872258,
      "grad_norm": 0.001110463635995984,
      "learning_rate": 0.2745959450313966,
      "loss": 0.3141,
      "num_input_tokens_seen": 14263600,
      "step": 7520
    },
    {
      "epoch": 1.227587894608043,
      "grad_norm": 0.0033513905946165323,
      "learning_rate": 0.27456313658637804,
      "loss": 0.3405,
      "num_input_tokens_seen": 14272400,
      "step": 7525
    },
    {
      "epoch": 1.2284036218288605,
      "grad_norm": 0.0011807181872427464,
      "learning_rate": 0.27453030893215846,
      "loss": 0.3204,
      "num_input_tokens_seen": 14281872,
      "step": 7530
    },
    {
      "epoch": 1.2292193490496777,
      "grad_norm": 0.0005789776332676411,
      "learning_rate": 0.2744974620738003,
      "loss": 0.4145,
      "num_input_tokens_seen": 14290656,
      "step": 7535
    },
    {
      "epoch": 1.2300350762704952,
      "grad_norm": 0.0008679148741066456,
      "learning_rate": 0.27446459601636897,
      "loss": 0.349,
      "num_input_tokens_seen": 14300048,
      "step": 7540
    },
    {
      "epoch": 1.2308508034913126,
      "grad_norm": 0.0014222976751625538,
      "learning_rate": 0.2744317107649328,
      "loss": 0.3246,
      "num_input_tokens_seen": 14309456,
      "step": 7545
    },
    {
      "epoch": 1.23166653071213,
      "grad_norm": 0.0012476559495553374,
      "learning_rate": 0.2743988063245631,
      "loss": 0.4677,
      "num_input_tokens_seen": 14318816,
      "step": 7550
    },
    {
      "epoch": 1.2324822579329473,
      "grad_norm": 0.0014097937382757664,
      "learning_rate": 0.2743658827003342,
      "loss": 0.3311,
      "num_input_tokens_seen": 14328752,
      "step": 7555
    },
    {
      "epoch": 1.2332979851537647,
      "grad_norm": 0.002829904668033123,
      "learning_rate": 0.27433293989732327,
      "loss": 0.3848,
      "num_input_tokens_seen": 14338336,
      "step": 7560
    },
    {
      "epoch": 1.234113712374582,
      "grad_norm": 0.0009233684395439923,
      "learning_rate": 0.27429997792061056,
      "loss": 0.3903,
      "num_input_tokens_seen": 14348464,
      "step": 7565
    },
    {
      "epoch": 1.2349294395953994,
      "grad_norm": 0.0003594121662899852,
      "learning_rate": 0.27426699677527927,
      "loss": 0.3512,
      "num_input_tokens_seen": 14357392,
      "step": 7570
    },
    {
      "epoch": 1.2357451668162167,
      "grad_norm": 0.0004608806630130857,
      "learning_rate": 0.2742339964664154,
      "loss": 0.357,
      "num_input_tokens_seen": 14368112,
      "step": 7575
    },
    {
      "epoch": 1.236560894037034,
      "grad_norm": 0.0019248670432716608,
      "learning_rate": 0.274200976999108,
      "loss": 0.3448,
      "num_input_tokens_seen": 14377360,
      "step": 7580
    },
    {
      "epoch": 1.2373766212578514,
      "grad_norm": 0.001458756742067635,
      "learning_rate": 0.27416793837844916,
      "loss": 0.3902,
      "num_input_tokens_seen": 14386528,
      "step": 7585
    },
    {
      "epoch": 1.2381923484786688,
      "grad_norm": 0.0018568841041997075,
      "learning_rate": 0.27413488060953384,
      "loss": 0.3371,
      "num_input_tokens_seen": 14396800,
      "step": 7590
    },
    {
      "epoch": 1.2390080756994861,
      "grad_norm": 0.001777553465217352,
      "learning_rate": 0.27410180369745996,
      "loss": 0.3685,
      "num_input_tokens_seen": 14406016,
      "step": 7595
    },
    {
      "epoch": 1.2398238029203035,
      "grad_norm": 0.0008271367405541241,
      "learning_rate": 0.27406870764732844,
      "loss": 0.3167,
      "num_input_tokens_seen": 14415120,
      "step": 7600
    },
    {
      "epoch": 1.2398238029203035,
      "eval_loss": 0.3401506841182709,
      "eval_runtime": 152.638,
      "eval_samples_per_second": 17.853,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 14415120,
      "step": 7600
    },
    {
      "epoch": 1.2406395301411208,
      "grad_norm": 0.001543251215480268,
      "learning_rate": 0.27403559246424297,
      "loss": 0.3759,
      "num_input_tokens_seen": 14424384,
      "step": 7605
    },
    {
      "epoch": 1.2414552573619382,
      "grad_norm": 0.0013567311689257622,
      "learning_rate": 0.2740024581533105,
      "loss": 0.3559,
      "num_input_tokens_seen": 14433456,
      "step": 7610
    },
    {
      "epoch": 1.2422709845827555,
      "grad_norm": 0.0015311470488086343,
      "learning_rate": 0.2739693047196406,
      "loss": 0.3806,
      "num_input_tokens_seen": 14441536,
      "step": 7615
    },
    {
      "epoch": 1.2430867118035729,
      "grad_norm": 0.0024011030327528715,
      "learning_rate": 0.27393613216834606,
      "loss": 0.3895,
      "num_input_tokens_seen": 14451152,
      "step": 7620
    },
    {
      "epoch": 1.2439024390243902,
      "grad_norm": 0.0015542153269052505,
      "learning_rate": 0.2739029405045424,
      "loss": 0.3861,
      "num_input_tokens_seen": 14460112,
      "step": 7625
    },
    {
      "epoch": 1.2447181662452076,
      "grad_norm": 0.0017498547676950693,
      "learning_rate": 0.2738697297333483,
      "loss": 0.3562,
      "num_input_tokens_seen": 14468928,
      "step": 7630
    },
    {
      "epoch": 1.245533893466025,
      "grad_norm": 0.0009821751154959202,
      "learning_rate": 0.2738364998598852,
      "loss": 0.3723,
      "num_input_tokens_seen": 14478736,
      "step": 7635
    },
    {
      "epoch": 1.2463496206868423,
      "grad_norm": 0.002326379995793104,
      "learning_rate": 0.27380325088927765,
      "loss": 0.3631,
      "num_input_tokens_seen": 14488944,
      "step": 7640
    },
    {
      "epoch": 1.2471653479076596,
      "grad_norm": 0.00193942675832659,
      "learning_rate": 0.27376998282665294,
      "loss": 0.3567,
      "num_input_tokens_seen": 14498928,
      "step": 7645
    },
    {
      "epoch": 1.247981075128477,
      "grad_norm": 0.0013505022507160902,
      "learning_rate": 0.27373669567714154,
      "loss": 0.3738,
      "num_input_tokens_seen": 14509232,
      "step": 7650
    },
    {
      "epoch": 1.2487968023492944,
      "grad_norm": 0.001674992498010397,
      "learning_rate": 0.27370338944587663,
      "loss": 0.4779,
      "num_input_tokens_seen": 14517984,
      "step": 7655
    },
    {
      "epoch": 1.2496125295701117,
      "grad_norm": 0.0012064436450600624,
      "learning_rate": 0.27367006413799455,
      "loss": 0.3494,
      "num_input_tokens_seen": 14527168,
      "step": 7660
    },
    {
      "epoch": 1.250428256790929,
      "grad_norm": 0.0014614287065342069,
      "learning_rate": 0.2736367197586345,
      "loss": 0.3974,
      "num_input_tokens_seen": 14536528,
      "step": 7665
    },
    {
      "epoch": 1.2512439840117464,
      "grad_norm": 0.0010425238870084286,
      "learning_rate": 0.2736033563129385,
      "loss": 0.3615,
      "num_input_tokens_seen": 14547456,
      "step": 7670
    },
    {
      "epoch": 1.2520597112325638,
      "grad_norm": 0.0006389397312887013,
      "learning_rate": 0.27356997380605164,
      "loss": 0.3272,
      "num_input_tokens_seen": 14556256,
      "step": 7675
    },
    {
      "epoch": 1.2528754384533811,
      "grad_norm": 0.0004111092712264508,
      "learning_rate": 0.27353657224312194,
      "loss": 0.3175,
      "num_input_tokens_seen": 14566160,
      "step": 7680
    },
    {
      "epoch": 1.2536911656741985,
      "grad_norm": 0.001630382495932281,
      "learning_rate": 0.2735031516293004,
      "loss": 0.3601,
      "num_input_tokens_seen": 14575328,
      "step": 7685
    },
    {
      "epoch": 1.2545068928950158,
      "grad_norm": 0.0007419736357405782,
      "learning_rate": 0.2734697119697408,
      "loss": 0.3385,
      "num_input_tokens_seen": 14586128,
      "step": 7690
    },
    {
      "epoch": 1.2553226201158334,
      "grad_norm": 0.0025057727470993996,
      "learning_rate": 0.27343625326959997,
      "loss": 0.3639,
      "num_input_tokens_seen": 14595552,
      "step": 7695
    },
    {
      "epoch": 1.2561383473366505,
      "grad_norm": 0.000984017737209797,
      "learning_rate": 0.27340277553403775,
      "loss": 0.3359,
      "num_input_tokens_seen": 14605328,
      "step": 7700
    },
    {
      "epoch": 1.256954074557468,
      "grad_norm": 0.00028428557561710477,
      "learning_rate": 0.2733692787682167,
      "loss": 0.357,
      "num_input_tokens_seen": 14615216,
      "step": 7705
    },
    {
      "epoch": 1.2577698017782852,
      "grad_norm": 0.0017617435660213232,
      "learning_rate": 0.27333576297730255,
      "loss": 0.3198,
      "num_input_tokens_seen": 14623888,
      "step": 7710
    },
    {
      "epoch": 1.2585855289991028,
      "grad_norm": 0.0026219624560326338,
      "learning_rate": 0.2733022281664638,
      "loss": 0.3925,
      "num_input_tokens_seen": 14632256,
      "step": 7715
    },
    {
      "epoch": 1.25940125621992,
      "grad_norm": 0.001473675831221044,
      "learning_rate": 0.273268674340872,
      "loss": 0.3252,
      "num_input_tokens_seen": 14642672,
      "step": 7720
    },
    {
      "epoch": 1.2602169834407375,
      "grad_norm": 0.0009979966562241316,
      "learning_rate": 0.27323510150570146,
      "loss": 0.3421,
      "num_input_tokens_seen": 14651536,
      "step": 7725
    },
    {
      "epoch": 1.2610327106615546,
      "grad_norm": 0.0006843790761195123,
      "learning_rate": 0.27320150966612966,
      "loss": 0.3622,
      "num_input_tokens_seen": 14660912,
      "step": 7730
    },
    {
      "epoch": 1.2618484378823722,
      "grad_norm": 0.0010568298166617751,
      "learning_rate": 0.2731678988273368,
      "loss": 0.3755,
      "num_input_tokens_seen": 14670864,
      "step": 7735
    },
    {
      "epoch": 1.2626641651031896,
      "grad_norm": 0.0017463455442339182,
      "learning_rate": 0.27313426899450605,
      "loss": 0.3778,
      "num_input_tokens_seen": 14680144,
      "step": 7740
    },
    {
      "epoch": 1.263479892324007,
      "grad_norm": 0.0006916412967257202,
      "learning_rate": 0.27310062017282366,
      "loss": 0.3443,
      "num_input_tokens_seen": 14690544,
      "step": 7745
    },
    {
      "epoch": 1.2642956195448243,
      "grad_norm": 0.00048372388118878007,
      "learning_rate": 0.2730669523674787,
      "loss": 0.3466,
      "num_input_tokens_seen": 14698480,
      "step": 7750
    },
    {
      "epoch": 1.2651113467656416,
      "grad_norm": 0.0024591851979494095,
      "learning_rate": 0.2730332655836631,
      "loss": 0.3469,
      "num_input_tokens_seen": 14708272,
      "step": 7755
    },
    {
      "epoch": 1.265927073986459,
      "grad_norm": 0.0009980143513530493,
      "learning_rate": 0.2729995598265718,
      "loss": 0.3313,
      "num_input_tokens_seen": 14717584,
      "step": 7760
    },
    {
      "epoch": 1.2667428012072763,
      "grad_norm": 0.0006806350429542363,
      "learning_rate": 0.2729658351014027,
      "loss": 0.3646,
      "num_input_tokens_seen": 14727248,
      "step": 7765
    },
    {
      "epoch": 1.2675585284280937,
      "grad_norm": 0.0009787739254534245,
      "learning_rate": 0.27293209141335656,
      "loss": 0.3528,
      "num_input_tokens_seen": 14736752,
      "step": 7770
    },
    {
      "epoch": 1.268374255648911,
      "grad_norm": 0.0009401346324011683,
      "learning_rate": 0.27289832876763703,
      "loss": 0.3327,
      "num_input_tokens_seen": 14748256,
      "step": 7775
    },
    {
      "epoch": 1.2691899828697284,
      "grad_norm": 0.001109839417040348,
      "learning_rate": 0.27286454716945074,
      "loss": 0.36,
      "num_input_tokens_seen": 14757280,
      "step": 7780
    },
    {
      "epoch": 1.2700057100905457,
      "grad_norm": 0.0017366833053529263,
      "learning_rate": 0.27283074662400725,
      "loss": 0.3412,
      "num_input_tokens_seen": 14766992,
      "step": 7785
    },
    {
      "epoch": 1.270821437311363,
      "grad_norm": 0.002166845602914691,
      "learning_rate": 0.2727969271365191,
      "loss": 0.3291,
      "num_input_tokens_seen": 14776592,
      "step": 7790
    },
    {
      "epoch": 1.2716371645321805,
      "grad_norm": 0.0024181711487472057,
      "learning_rate": 0.2727630887122016,
      "loss": 0.3588,
      "num_input_tokens_seen": 14785584,
      "step": 7795
    },
    {
      "epoch": 1.2724528917529978,
      "grad_norm": 0.0008806711994111538,
      "learning_rate": 0.27272923135627314,
      "loss": 0.3442,
      "num_input_tokens_seen": 14794784,
      "step": 7800
    },
    {
      "epoch": 1.2724528917529978,
      "eval_loss": 0.33385178446769714,
      "eval_runtime": 152.5995,
      "eval_samples_per_second": 17.857,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 14794784,
      "step": 7800
    },
    {
      "epoch": 1.2732686189738152,
      "grad_norm": 0.0033470341004431248,
      "learning_rate": 0.2726953550739548,
      "loss": 0.3498,
      "num_input_tokens_seen": 14804112,
      "step": 7805
    },
    {
      "epoch": 1.2740843461946325,
      "grad_norm": 0.0011281052138656378,
      "learning_rate": 0.27266145987047086,
      "loss": 0.3332,
      "num_input_tokens_seen": 14813360,
      "step": 7810
    },
    {
      "epoch": 1.2749000734154499,
      "grad_norm": 0.0011265557259321213,
      "learning_rate": 0.27262754575104836,
      "loss": 0.3282,
      "num_input_tokens_seen": 14822288,
      "step": 7815
    },
    {
      "epoch": 1.2757158006362672,
      "grad_norm": 0.000714506721124053,
      "learning_rate": 0.27259361272091726,
      "loss": 0.3348,
      "num_input_tokens_seen": 14832784,
      "step": 7820
    },
    {
      "epoch": 1.2765315278570846,
      "grad_norm": 0.0013406077632680535,
      "learning_rate": 0.27255966078531046,
      "loss": 0.3609,
      "num_input_tokens_seen": 14841600,
      "step": 7825
    },
    {
      "epoch": 1.277347255077902,
      "grad_norm": 0.0008062028209678829,
      "learning_rate": 0.2725256899494638,
      "loss": 0.3415,
      "num_input_tokens_seen": 14851984,
      "step": 7830
    },
    {
      "epoch": 1.2781629822987193,
      "grad_norm": 0.001729815499857068,
      "learning_rate": 0.272491700218616,
      "loss": 0.3684,
      "num_input_tokens_seen": 14862544,
      "step": 7835
    },
    {
      "epoch": 1.2789787095195366,
      "grad_norm": 0.0024457357358187437,
      "learning_rate": 0.27245769159800876,
      "loss": 0.3546,
      "num_input_tokens_seen": 14872560,
      "step": 7840
    },
    {
      "epoch": 1.279794436740354,
      "grad_norm": 0.0029819784685969353,
      "learning_rate": 0.2724236640928865,
      "loss": 0.3103,
      "num_input_tokens_seen": 14881872,
      "step": 7845
    },
    {
      "epoch": 1.2806101639611713,
      "grad_norm": 0.00134527578484267,
      "learning_rate": 0.27238961770849673,
      "loss": 0.4377,
      "num_input_tokens_seen": 14890992,
      "step": 7850
    },
    {
      "epoch": 1.2814258911819887,
      "grad_norm": 0.0011079877149313688,
      "learning_rate": 0.27235555245008997,
      "loss": 0.3522,
      "num_input_tokens_seen": 14899888,
      "step": 7855
    },
    {
      "epoch": 1.282241618402806,
      "grad_norm": 0.001740630716085434,
      "learning_rate": 0.2723214683229193,
      "loss": 0.3517,
      "num_input_tokens_seen": 14908944,
      "step": 7860
    },
    {
      "epoch": 1.2830573456236234,
      "grad_norm": 0.001616714638657868,
      "learning_rate": 0.27228736533224107,
      "loss": 0.3663,
      "num_input_tokens_seen": 14918624,
      "step": 7865
    },
    {
      "epoch": 1.2838730728444407,
      "grad_norm": 0.0018864829326048493,
      "learning_rate": 0.27225324348331437,
      "loss": 0.3856,
      "num_input_tokens_seen": 14927632,
      "step": 7870
    },
    {
      "epoch": 1.284688800065258,
      "grad_norm": 0.0014550258638337255,
      "learning_rate": 0.27221910278140116,
      "loss": 0.3721,
      "num_input_tokens_seen": 14936416,
      "step": 7875
    },
    {
      "epoch": 1.2855045272860757,
      "grad_norm": 0.0011493482161313295,
      "learning_rate": 0.2721849432317664,
      "loss": 0.3492,
      "num_input_tokens_seen": 14946496,
      "step": 7880
    },
    {
      "epoch": 1.2863202545068928,
      "grad_norm": 0.0008837906643748283,
      "learning_rate": 0.2721507648396779,
      "loss": 0.3522,
      "num_input_tokens_seen": 14956000,
      "step": 7885
    },
    {
      "epoch": 1.2871359817277104,
      "grad_norm": 0.0038232102524489164,
      "learning_rate": 0.27211656761040653,
      "loss": 0.3267,
      "num_input_tokens_seen": 14966096,
      "step": 7890
    },
    {
      "epoch": 1.2879517089485275,
      "grad_norm": 0.001907195313833654,
      "learning_rate": 0.2720823515492257,
      "loss": 0.3494,
      "num_input_tokens_seen": 14975856,
      "step": 7895
    },
    {
      "epoch": 1.288767436169345,
      "grad_norm": 0.0037104864604771137,
      "learning_rate": 0.27204811666141215,
      "loss": 0.3553,
      "num_input_tokens_seen": 14985120,
      "step": 7900
    },
    {
      "epoch": 1.2895831633901622,
      "grad_norm": 0.0016493836883455515,
      "learning_rate": 0.2720138629522452,
      "loss": 0.3339,
      "num_input_tokens_seen": 14994480,
      "step": 7905
    },
    {
      "epoch": 1.2903988906109798,
      "grad_norm": 0.0016654436476528645,
      "learning_rate": 0.2719795904270073,
      "loss": 0.346,
      "num_input_tokens_seen": 15004384,
      "step": 7910
    },
    {
      "epoch": 1.291214617831797,
      "grad_norm": 0.0015225897077471018,
      "learning_rate": 0.2719452990909837,
      "loss": 0.3232,
      "num_input_tokens_seen": 15014784,
      "step": 7915
    },
    {
      "epoch": 1.2920303450526145,
      "grad_norm": 0.0027936806436628103,
      "learning_rate": 0.2719109889494625,
      "loss": 0.3424,
      "num_input_tokens_seen": 15023936,
      "step": 7920
    },
    {
      "epoch": 1.2928460722734318,
      "grad_norm": 0.0015111372340470552,
      "learning_rate": 0.27187666000773475,
      "loss": 0.3427,
      "num_input_tokens_seen": 15033696,
      "step": 7925
    },
    {
      "epoch": 1.2936617994942492,
      "grad_norm": 0.0021270529832690954,
      "learning_rate": 0.2718423122710944,
      "loss": 0.3576,
      "num_input_tokens_seen": 15043104,
      "step": 7930
    },
    {
      "epoch": 1.2944775267150666,
      "grad_norm": 0.0027285991236567497,
      "learning_rate": 0.2718079457448384,
      "loss": 0.3359,
      "num_input_tokens_seen": 15052912,
      "step": 7935
    },
    {
      "epoch": 1.295293253935884,
      "grad_norm": 0.002705897204577923,
      "learning_rate": 0.27177356043426637,
      "loss": 0.3529,
      "num_input_tokens_seen": 15062032,
      "step": 7940
    },
    {
      "epoch": 1.2961089811567013,
      "grad_norm": 0.0008175402181223035,
      "learning_rate": 0.27173915634468104,
      "loss": 0.3394,
      "num_input_tokens_seen": 15070976,
      "step": 7945
    },
    {
      "epoch": 1.2969247083775186,
      "grad_norm": 0.0013237791135907173,
      "learning_rate": 0.27170473348138796,
      "loss": 0.3581,
      "num_input_tokens_seen": 15081264,
      "step": 7950
    },
    {
      "epoch": 1.297740435598336,
      "grad_norm": 0.001996150705963373,
      "learning_rate": 0.27167029184969554,
      "loss": 0.3384,
      "num_input_tokens_seen": 15090800,
      "step": 7955
    },
    {
      "epoch": 1.2985561628191533,
      "grad_norm": 0.003761867294088006,
      "learning_rate": 0.27163583145491504,
      "loss": 0.3647,
      "num_input_tokens_seen": 15100944,
      "step": 7960
    },
    {
      "epoch": 1.2993718900399707,
      "grad_norm": 0.002229970181360841,
      "learning_rate": 0.2716013523023608,
      "loss": 0.38,
      "num_input_tokens_seen": 15111552,
      "step": 7965
    },
    {
      "epoch": 1.300187617260788,
      "grad_norm": 0.0007579394732601941,
      "learning_rate": 0.27156685439734995,
      "loss": 0.3827,
      "num_input_tokens_seen": 15120816,
      "step": 7970
    },
    {
      "epoch": 1.3010033444816054,
      "grad_norm": 0.0014663906767964363,
      "learning_rate": 0.2715323377452024,
      "loss": 0.365,
      "num_input_tokens_seen": 15129600,
      "step": 7975
    },
    {
      "epoch": 1.3018190717024227,
      "grad_norm": 0.000680694414768368,
      "learning_rate": 0.2714978023512411,
      "loss": 0.3391,
      "num_input_tokens_seen": 15139408,
      "step": 7980
    },
    {
      "epoch": 1.30263479892324,
      "grad_norm": 0.0012336545623838902,
      "learning_rate": 0.2714632482207918,
      "loss": 0.3329,
      "num_input_tokens_seen": 15148944,
      "step": 7985
    },
    {
      "epoch": 1.3034505261440574,
      "grad_norm": 0.0009170203120447695,
      "learning_rate": 0.2714286753591833,
      "loss": 0.3239,
      "num_input_tokens_seen": 15158992,
      "step": 7990
    },
    {
      "epoch": 1.3042662533648748,
      "grad_norm": 0.000383903767215088,
      "learning_rate": 0.27139408377174706,
      "loss": 0.3548,
      "num_input_tokens_seen": 15167408,
      "step": 7995
    },
    {
      "epoch": 1.3050819805856921,
      "grad_norm": 0.0010837524896487594,
      "learning_rate": 0.27135947346381756,
      "loss": 0.3655,
      "num_input_tokens_seen": 15176240,
      "step": 8000
    },
    {
      "epoch": 1.3050819805856921,
      "eval_loss": 0.33565425872802734,
      "eval_runtime": 152.5979,
      "eval_samples_per_second": 17.857,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 15176240,
      "step": 8000
    },
    {
      "epoch": 1.3058977078065095,
      "grad_norm": 0.0012737183133140206,
      "learning_rate": 0.2713248444407322,
      "loss": 0.3481,
      "num_input_tokens_seen": 15185600,
      "step": 8005
    },
    {
      "epoch": 1.3067134350273268,
      "grad_norm": 0.0010291474172845483,
      "learning_rate": 0.27129019670783106,
      "loss": 0.3264,
      "num_input_tokens_seen": 15193840,
      "step": 8010
    },
    {
      "epoch": 1.3075291622481442,
      "grad_norm": 0.0019084069645032287,
      "learning_rate": 0.27125553027045746,
      "loss": 0.351,
      "num_input_tokens_seen": 15203664,
      "step": 8015
    },
    {
      "epoch": 1.3083448894689615,
      "grad_norm": 0.0028493942227214575,
      "learning_rate": 0.2712208451339572,
      "loss": 0.3492,
      "num_input_tokens_seen": 15213664,
      "step": 8020
    },
    {
      "epoch": 1.309160616689779,
      "grad_norm": 0.0009308592416346073,
      "learning_rate": 0.27118614130367935,
      "loss": 0.3446,
      "num_input_tokens_seen": 15223104,
      "step": 8025
    },
    {
      "epoch": 1.3099763439105963,
      "grad_norm": 0.0020288003142923117,
      "learning_rate": 0.2711514187849756,
      "loss": 0.3385,
      "num_input_tokens_seen": 15232224,
      "step": 8030
    },
    {
      "epoch": 1.3107920711314136,
      "grad_norm": 0.001960600493475795,
      "learning_rate": 0.27111667758320057,
      "loss": 0.3434,
      "num_input_tokens_seen": 15241264,
      "step": 8035
    },
    {
      "epoch": 1.311607798352231,
      "grad_norm": 0.003164214314892888,
      "learning_rate": 0.27108191770371176,
      "loss": 0.3261,
      "num_input_tokens_seen": 15248736,
      "step": 8040
    },
    {
      "epoch": 1.3124235255730483,
      "grad_norm": 0.00301061337813735,
      "learning_rate": 0.2710471391518697,
      "loss": 0.3905,
      "num_input_tokens_seen": 15257776,
      "step": 8045
    },
    {
      "epoch": 1.3132392527938657,
      "grad_norm": 0.0007341303862631321,
      "learning_rate": 0.2710123419330375,
      "loss": 0.3281,
      "num_input_tokens_seen": 15266128,
      "step": 8050
    },
    {
      "epoch": 1.314054980014683,
      "grad_norm": 0.0013504666276276112,
      "learning_rate": 0.2709775260525816,
      "loss": 0.3722,
      "num_input_tokens_seen": 15274720,
      "step": 8055
    },
    {
      "epoch": 1.3148707072355004,
      "grad_norm": 0.0010047564283013344,
      "learning_rate": 0.27094269151587075,
      "loss": 0.3683,
      "num_input_tokens_seen": 15283536,
      "step": 8060
    },
    {
      "epoch": 1.315686434456318,
      "grad_norm": 0.0007002020138315856,
      "learning_rate": 0.27090783832827703,
      "loss": 0.3275,
      "num_input_tokens_seen": 15293824,
      "step": 8065
    },
    {
      "epoch": 1.316502161677135,
      "grad_norm": 0.0007264278829097748,
      "learning_rate": 0.2708729664951753,
      "loss": 0.3425,
      "num_input_tokens_seen": 15302672,
      "step": 8070
    },
    {
      "epoch": 1.3173178888979526,
      "grad_norm": 0.00035634846426546574,
      "learning_rate": 0.27083807602194304,
      "loss": 0.2704,
      "num_input_tokens_seen": 15312432,
      "step": 8075
    },
    {
      "epoch": 1.3181336161187698,
      "grad_norm": 0.001459911116398871,
      "learning_rate": 0.270803166913961,
      "loss": 0.4151,
      "num_input_tokens_seen": 15321744,
      "step": 8080
    },
    {
      "epoch": 1.3189493433395874,
      "grad_norm": 0.0015965605853125453,
      "learning_rate": 0.27076823917661247,
      "loss": 0.3257,
      "num_input_tokens_seen": 15330624,
      "step": 8085
    },
    {
      "epoch": 1.3197650705604045,
      "grad_norm": 0.0006674563628621399,
      "learning_rate": 0.2707332928152838,
      "loss": 0.3499,
      "num_input_tokens_seen": 15340560,
      "step": 8090
    },
    {
      "epoch": 1.320580797781222,
      "grad_norm": 0.0009301419486291707,
      "learning_rate": 0.2706983278353641,
      "loss": 0.3819,
      "num_input_tokens_seen": 15349152,
      "step": 8095
    },
    {
      "epoch": 1.3213965250020392,
      "grad_norm": 0.0010409840615466237,
      "learning_rate": 0.27066334424224553,
      "loss": 0.3553,
      "num_input_tokens_seen": 15358608,
      "step": 8100
    },
    {
      "epoch": 1.3222122522228568,
      "grad_norm": 0.0016103923553600907,
      "learning_rate": 0.27062834204132297,
      "loss": 0.3458,
      "num_input_tokens_seen": 15368592,
      "step": 8105
    },
    {
      "epoch": 1.3230279794436741,
      "grad_norm": 0.0022003185003995895,
      "learning_rate": 0.27059332123799407,
      "loss": 0.3298,
      "num_input_tokens_seen": 15377104,
      "step": 8110
    },
    {
      "epoch": 1.3238437066644915,
      "grad_norm": 0.001945051015354693,
      "learning_rate": 0.27055828183765956,
      "loss": 0.3969,
      "num_input_tokens_seen": 15386368,
      "step": 8115
    },
    {
      "epoch": 1.3246594338853088,
      "grad_norm": 0.0011589191854000092,
      "learning_rate": 0.270523223845723,
      "loss": 0.3448,
      "num_input_tokens_seen": 15394736,
      "step": 8120
    },
    {
      "epoch": 1.3254751611061262,
      "grad_norm": 0.0007966370903886855,
      "learning_rate": 0.2704881472675907,
      "loss": 0.3774,
      "num_input_tokens_seen": 15404944,
      "step": 8125
    },
    {
      "epoch": 1.3262908883269435,
      "grad_norm": 0.0008655904675833881,
      "learning_rate": 0.270453052108672,
      "loss": 0.3304,
      "num_input_tokens_seen": 15413760,
      "step": 8130
    },
    {
      "epoch": 1.3271066155477609,
      "grad_norm": 0.002252976642921567,
      "learning_rate": 0.2704179383743789,
      "loss": 0.3265,
      "num_input_tokens_seen": 15424544,
      "step": 8135
    },
    {
      "epoch": 1.3279223427685782,
      "grad_norm": 0.0020291318651288748,
      "learning_rate": 0.27038280607012644,
      "loss": 0.3467,
      "num_input_tokens_seen": 15434368,
      "step": 8140
    },
    {
      "epoch": 1.3287380699893956,
      "grad_norm": 0.000875335477758199,
      "learning_rate": 0.27034765520133247,
      "loss": 0.3663,
      "num_input_tokens_seen": 15443296,
      "step": 8145
    },
    {
      "epoch": 1.329553797210213,
      "grad_norm": 0.0026770136319100857,
      "learning_rate": 0.2703124857734177,
      "loss": 0.3833,
      "num_input_tokens_seen": 15452656,
      "step": 8150
    },
    {
      "epoch": 1.3303695244310303,
      "grad_norm": 0.002557041123509407,
      "learning_rate": 0.27027729779180565,
      "loss": 0.3388,
      "num_input_tokens_seen": 15462848,
      "step": 8155
    },
    {
      "epoch": 1.3311852516518476,
      "grad_norm": 0.0019259932450950146,
      "learning_rate": 0.27024209126192283,
      "loss": 0.3653,
      "num_input_tokens_seen": 15472288,
      "step": 8160
    },
    {
      "epoch": 1.332000978872665,
      "grad_norm": 0.0006483773468062282,
      "learning_rate": 0.2702068661891984,
      "loss": 0.3281,
      "num_input_tokens_seen": 15481024,
      "step": 8165
    },
    {
      "epoch": 1.3328167060934824,
      "grad_norm": 0.0017941053956747055,
      "learning_rate": 0.2701716225790647,
      "loss": 0.391,
      "num_input_tokens_seen": 15490624,
      "step": 8170
    },
    {
      "epoch": 1.3336324333142997,
      "grad_norm": 0.0013511090073734522,
      "learning_rate": 0.27013636043695655,
      "loss": 0.3361,
      "num_input_tokens_seen": 15499680,
      "step": 8175
    },
    {
      "epoch": 1.334448160535117,
      "grad_norm": 0.0007949526188895106,
      "learning_rate": 0.27010107976831194,
      "loss": 0.3363,
      "num_input_tokens_seen": 15509280,
      "step": 8180
    },
    {
      "epoch": 1.3352638877559344,
      "grad_norm": 0.0011341277277097106,
      "learning_rate": 0.2700657805785715,
      "loss": 0.3649,
      "num_input_tokens_seen": 15519376,
      "step": 8185
    },
    {
      "epoch": 1.3360796149767518,
      "grad_norm": 0.0008935451041907072,
      "learning_rate": 0.2700304628731789,
      "loss": 0.3153,
      "num_input_tokens_seen": 15529120,
      "step": 8190
    },
    {
      "epoch": 1.3368953421975691,
      "grad_norm": 0.0007093923632055521,
      "learning_rate": 0.26999512665758046,
      "loss": 0.3282,
      "num_input_tokens_seen": 15538352,
      "step": 8195
    },
    {
      "epoch": 1.3377110694183865,
      "grad_norm": 0.0014283545315265656,
      "learning_rate": 0.2699597719372256,
      "loss": 0.3667,
      "num_input_tokens_seen": 15548080,
      "step": 8200
    },
    {
      "epoch": 1.3377110694183865,
      "eval_loss": 0.33593860268592834,
      "eval_runtime": 152.5703,
      "eval_samples_per_second": 17.861,
      "eval_steps_per_second": 8.934,
      "num_input_tokens_seen": 15548080,
      "step": 8200
    },
    {
      "epoch": 1.3385267966392038,
      "grad_norm": 0.0015726103447377682,
      "learning_rate": 0.26992439871756635,
      "loss": 0.3419,
      "num_input_tokens_seen": 15556992,
      "step": 8205
    },
    {
      "epoch": 1.3393425238600212,
      "grad_norm": 0.002585038309916854,
      "learning_rate": 0.2698890070040578,
      "loss": 0.3392,
      "num_input_tokens_seen": 15565600,
      "step": 8210
    },
    {
      "epoch": 1.3401582510808385,
      "grad_norm": 0.00099877396132797,
      "learning_rate": 0.2698535968021577,
      "loss": 0.4036,
      "num_input_tokens_seen": 15574960,
      "step": 8215
    },
    {
      "epoch": 1.3409739783016559,
      "grad_norm": 0.0014918571105226874,
      "learning_rate": 0.26981816811732684,
      "loss": 0.3304,
      "num_input_tokens_seen": 15586208,
      "step": 8220
    },
    {
      "epoch": 1.3417897055224732,
      "grad_norm": 0.0023422674275934696,
      "learning_rate": 0.26978272095502875,
      "loss": 0.4001,
      "num_input_tokens_seen": 15594960,
      "step": 8225
    },
    {
      "epoch": 1.3426054327432906,
      "grad_norm": 0.00237402506172657,
      "learning_rate": 0.26974725532072974,
      "loss": 0.394,
      "num_input_tokens_seen": 15604144,
      "step": 8230
    },
    {
      "epoch": 1.343421159964108,
      "grad_norm": 0.0005520775448530912,
      "learning_rate": 0.26971177121989914,
      "loss": 0.306,
      "num_input_tokens_seen": 15614192,
      "step": 8235
    },
    {
      "epoch": 1.3442368871849253,
      "grad_norm": 0.0012708721915259957,
      "learning_rate": 0.2696762686580091,
      "loss": 0.4699,
      "num_input_tokens_seen": 15624176,
      "step": 8240
    },
    {
      "epoch": 1.3450526144057426,
      "grad_norm": 0.0014721256447955966,
      "learning_rate": 0.26964074764053436,
      "loss": 0.4337,
      "num_input_tokens_seen": 15633440,
      "step": 8245
    },
    {
      "epoch": 1.3458683416265602,
      "grad_norm": 0.0007525411201640964,
      "learning_rate": 0.2696052081729529,
      "loss": 0.3353,
      "num_input_tokens_seen": 15641968,
      "step": 8250
    },
    {
      "epoch": 1.3466840688473773,
      "grad_norm": 0.002155542140826583,
      "learning_rate": 0.2695696502607453,
      "loss": 0.4243,
      "num_input_tokens_seen": 15651184,
      "step": 8255
    },
    {
      "epoch": 1.347499796068195,
      "grad_norm": 0.0006414417875930667,
      "learning_rate": 0.26953407390939504,
      "loss": 0.386,
      "num_input_tokens_seen": 15660816,
      "step": 8260
    },
    {
      "epoch": 1.348315523289012,
      "grad_norm": 0.0009825363522395492,
      "learning_rate": 0.26949847912438835,
      "loss": 0.3469,
      "num_input_tokens_seen": 15669152,
      "step": 8265
    },
    {
      "epoch": 1.3491312505098296,
      "grad_norm": 0.0005520020495168865,
      "learning_rate": 0.26946286591121454,
      "loss": 0.3862,
      "num_input_tokens_seen": 15679664,
      "step": 8270
    },
    {
      "epoch": 1.3499469777306468,
      "grad_norm": 0.0016719845589250326,
      "learning_rate": 0.2694272342753655,
      "loss": 0.4553,
      "num_input_tokens_seen": 15690352,
      "step": 8275
    },
    {
      "epoch": 1.3507627049514643,
      "grad_norm": 0.0016277785180136561,
      "learning_rate": 0.26939158422233617,
      "loss": 0.3252,
      "num_input_tokens_seen": 15699984,
      "step": 8280
    },
    {
      "epoch": 1.3515784321722815,
      "grad_norm": 0.0009056241251528263,
      "learning_rate": 0.26935591575762413,
      "loss": 0.3243,
      "num_input_tokens_seen": 15708960,
      "step": 8285
    },
    {
      "epoch": 1.352394159393099,
      "grad_norm": 0.002191969659179449,
      "learning_rate": 0.26932022888672996,
      "loss": 0.377,
      "num_input_tokens_seen": 15717984,
      "step": 8290
    },
    {
      "epoch": 1.3532098866139162,
      "grad_norm": 0.002184625482186675,
      "learning_rate": 0.26928452361515703,
      "loss": 0.3473,
      "num_input_tokens_seen": 15727264,
      "step": 8295
    },
    {
      "epoch": 1.3540256138347337,
      "grad_norm": 0.002861141925677657,
      "learning_rate": 0.26924879994841155,
      "loss": 0.3563,
      "num_input_tokens_seen": 15736288,
      "step": 8300
    },
    {
      "epoch": 1.354841341055551,
      "grad_norm": 0.0013937724288553,
      "learning_rate": 0.2692130578920025,
      "loss": 0.3276,
      "num_input_tokens_seen": 15745520,
      "step": 8305
    },
    {
      "epoch": 1.3556570682763684,
      "grad_norm": 0.0013792515965178609,
      "learning_rate": 0.26917729745144187,
      "loss": 0.2871,
      "num_input_tokens_seen": 15754576,
      "step": 8310
    },
    {
      "epoch": 1.3564727954971858,
      "grad_norm": 0.0030944799073040485,
      "learning_rate": 0.2691415186322443,
      "loss": 0.4562,
      "num_input_tokens_seen": 15764656,
      "step": 8315
    },
    {
      "epoch": 1.3572885227180032,
      "grad_norm": 0.0015749281737953424,
      "learning_rate": 0.2691057214399273,
      "loss": 0.3781,
      "num_input_tokens_seen": 15774016,
      "step": 8320
    },
    {
      "epoch": 1.3581042499388205,
      "grad_norm": 0.0011638016439974308,
      "learning_rate": 0.2690699058800113,
      "loss": 0.3648,
      "num_input_tokens_seen": 15783792,
      "step": 8325
    },
    {
      "epoch": 1.3589199771596379,
      "grad_norm": 0.0007159218657761812,
      "learning_rate": 0.2690340719580194,
      "loss": 0.3496,
      "num_input_tokens_seen": 15794160,
      "step": 8330
    },
    {
      "epoch": 1.3597357043804552,
      "grad_norm": 0.0012566579971462488,
      "learning_rate": 0.2689982196794778,
      "loss": 0.3674,
      "num_input_tokens_seen": 15804176,
      "step": 8335
    },
    {
      "epoch": 1.3605514316012726,
      "grad_norm": 0.0010097040794789791,
      "learning_rate": 0.2689623490499153,
      "loss": 0.3644,
      "num_input_tokens_seen": 15814640,
      "step": 8340
    },
    {
      "epoch": 1.36136715882209,
      "grad_norm": 0.0009934140834957361,
      "learning_rate": 0.2689264600748636,
      "loss": 0.3268,
      "num_input_tokens_seen": 15824528,
      "step": 8345
    },
    {
      "epoch": 1.3621828860429073,
      "grad_norm": 0.0020505059510469437,
      "learning_rate": 0.26889055275985724,
      "loss": 0.3721,
      "num_input_tokens_seen": 15833760,
      "step": 8350
    },
    {
      "epoch": 1.3629986132637246,
      "grad_norm": 0.001988539705052972,
      "learning_rate": 0.2688546271104335,
      "loss": 0.3562,
      "num_input_tokens_seen": 15841744,
      "step": 8355
    },
    {
      "epoch": 1.363814340484542,
      "grad_norm": 0.0008591380901634693,
      "learning_rate": 0.26881868313213275,
      "loss": 0.3527,
      "num_input_tokens_seen": 15851728,
      "step": 8360
    },
    {
      "epoch": 1.3646300677053593,
      "grad_norm": 0.0008895773789845407,
      "learning_rate": 0.2687827208304978,
      "loss": 0.3599,
      "num_input_tokens_seen": 15860896,
      "step": 8365
    },
    {
      "epoch": 1.3654457949261767,
      "grad_norm": 0.0008820080547593534,
      "learning_rate": 0.26874674021107464,
      "loss": 0.3416,
      "num_input_tokens_seen": 15870608,
      "step": 8370
    },
    {
      "epoch": 1.366261522146994,
      "grad_norm": 0.0016070220153778791,
      "learning_rate": 0.2687107412794118,
      "loss": 0.3544,
      "num_input_tokens_seen": 15879472,
      "step": 8375
    },
    {
      "epoch": 1.3670772493678114,
      "grad_norm": 0.0011707022786140442,
      "learning_rate": 0.26867472404106096,
      "loss": 0.35,
      "num_input_tokens_seen": 15888768,
      "step": 8380
    },
    {
      "epoch": 1.3678929765886287,
      "grad_norm": 0.00041988902376033366,
      "learning_rate": 0.26863868850157624,
      "loss": 0.3551,
      "num_input_tokens_seen": 15897616,
      "step": 8385
    },
    {
      "epoch": 1.368708703809446,
      "grad_norm": 0.0017826068215072155,
      "learning_rate": 0.26860263466651485,
      "loss": 0.3208,
      "num_input_tokens_seen": 15906960,
      "step": 8390
    },
    {
      "epoch": 1.3695244310302634,
      "grad_norm": 0.0021542184986174107,
      "learning_rate": 0.26856656254143674,
      "loss": 0.3788,
      "num_input_tokens_seen": 15916224,
      "step": 8395
    },
    {
      "epoch": 1.3703401582510808,
      "grad_norm": 0.0010267135221511126,
      "learning_rate": 0.2685304721319047,
      "loss": 0.3315,
      "num_input_tokens_seen": 15926832,
      "step": 8400
    },
    {
      "epoch": 1.3703401582510808,
      "eval_loss": 0.33547404408454895,
      "eval_runtime": 152.6423,
      "eval_samples_per_second": 17.852,
      "eval_steps_per_second": 8.929,
      "num_input_tokens_seen": 15926832,
      "step": 8400
    },
    {
      "epoch": 1.3711558854718982,
      "grad_norm": 0.0010137949138879776,
      "learning_rate": 0.2684943634434843,
      "loss": 0.304,
      "num_input_tokens_seen": 15936512,
      "step": 8405
    },
    {
      "epoch": 1.3719716126927155,
      "grad_norm": 0.0010081386426463723,
      "learning_rate": 0.268458236481744,
      "loss": 0.404,
      "num_input_tokens_seen": 15945104,
      "step": 8410
    },
    {
      "epoch": 1.3727873399135329,
      "grad_norm": 0.0012916034320369363,
      "learning_rate": 0.2684220912522549,
      "loss": 0.3599,
      "num_input_tokens_seen": 15954464,
      "step": 8415
    },
    {
      "epoch": 1.3736030671343502,
      "grad_norm": 0.0011621073354035616,
      "learning_rate": 0.2683859277605913,
      "loss": 0.3654,
      "num_input_tokens_seen": 15964096,
      "step": 8420
    },
    {
      "epoch": 1.3744187943551676,
      "grad_norm": 0.0016335274558514357,
      "learning_rate": 0.2683497460123298,
      "loss": 0.3458,
      "num_input_tokens_seen": 15974352,
      "step": 8425
    },
    {
      "epoch": 1.375234521575985,
      "grad_norm": 0.0009879317367449403,
      "learning_rate": 0.26831354601305013,
      "loss": 0.3302,
      "num_input_tokens_seen": 15983712,
      "step": 8430
    },
    {
      "epoch": 1.3760502487968023,
      "grad_norm": 0.001195727614685893,
      "learning_rate": 0.26827732776833496,
      "loss": 0.369,
      "num_input_tokens_seen": 15992464,
      "step": 8435
    },
    {
      "epoch": 1.3768659760176196,
      "grad_norm": 0.0012824099976569414,
      "learning_rate": 0.26824109128376944,
      "loss": 0.3381,
      "num_input_tokens_seen": 16002544,
      "step": 8440
    },
    {
      "epoch": 1.3776817032384372,
      "grad_norm": 0.0017129417974501848,
      "learning_rate": 0.2682048365649417,
      "loss": 0.3299,
      "num_input_tokens_seen": 16011440,
      "step": 8445
    },
    {
      "epoch": 1.3784974304592543,
      "grad_norm": 0.002552281366661191,
      "learning_rate": 0.2681685636174428,
      "loss": 0.3847,
      "num_input_tokens_seen": 16021600,
      "step": 8450
    },
    {
      "epoch": 1.379313157680072,
      "grad_norm": 0.0017608256312087178,
      "learning_rate": 0.2681322724468663,
      "loss": 0.3888,
      "num_input_tokens_seen": 16030704,
      "step": 8455
    },
    {
      "epoch": 1.380128884900889,
      "grad_norm": 0.002764645731076598,
      "learning_rate": 0.2680959630588089,
      "loss": 0.3798,
      "num_input_tokens_seen": 16040000,
      "step": 8460
    },
    {
      "epoch": 1.3809446121217066,
      "grad_norm": 0.0019998059142380953,
      "learning_rate": 0.26805963545886985,
      "loss": 0.3929,
      "num_input_tokens_seen": 16050816,
      "step": 8465
    },
    {
      "epoch": 1.3817603393425237,
      "grad_norm": 0.0018003095174208283,
      "learning_rate": 0.26802328965265143,
      "loss": 0.3303,
      "num_input_tokens_seen": 16059920,
      "step": 8470
    },
    {
      "epoch": 1.3825760665633413,
      "grad_norm": 0.0014364372473210096,
      "learning_rate": 0.26798692564575854,
      "loss": 0.3691,
      "num_input_tokens_seen": 16067968,
      "step": 8475
    },
    {
      "epoch": 1.3833917937841584,
      "grad_norm": 0.001686482923105359,
      "learning_rate": 0.26795054344379904,
      "loss": 0.3372,
      "num_input_tokens_seen": 16078032,
      "step": 8480
    },
    {
      "epoch": 1.384207521004976,
      "grad_norm": 0.0005514679942280054,
      "learning_rate": 0.2679141430523835,
      "loss": 0.3555,
      "num_input_tokens_seen": 16088464,
      "step": 8485
    },
    {
      "epoch": 1.3850232482257934,
      "grad_norm": 0.0010772118112072349,
      "learning_rate": 0.2678777244771252,
      "loss": 0.368,
      "num_input_tokens_seen": 16097792,
      "step": 8490
    },
    {
      "epoch": 1.3858389754466107,
      "grad_norm": 0.001254203380085528,
      "learning_rate": 0.2678412877236405,
      "loss": 0.3557,
      "num_input_tokens_seen": 16108400,
      "step": 8495
    },
    {
      "epoch": 1.386654702667428,
      "grad_norm": 0.0007747320923954248,
      "learning_rate": 0.2678048327975484,
      "loss": 0.2941,
      "num_input_tokens_seen": 16117504,
      "step": 8500
    },
    {
      "epoch": 1.3874704298882454,
      "grad_norm": 0.0007481608190573752,
      "learning_rate": 0.2677683597044706,
      "loss": 0.4465,
      "num_input_tokens_seen": 16126752,
      "step": 8505
    },
    {
      "epoch": 1.3882861571090628,
      "grad_norm": 0.0017178449779748917,
      "learning_rate": 0.2677318684500318,
      "loss": 0.3676,
      "num_input_tokens_seen": 16135648,
      "step": 8510
    },
    {
      "epoch": 1.3891018843298801,
      "grad_norm": 0.0015501261223107576,
      "learning_rate": 0.2676953590398593,
      "loss": 0.3754,
      "num_input_tokens_seen": 16146272,
      "step": 8515
    },
    {
      "epoch": 1.3899176115506975,
      "grad_norm": 0.0016387284267693758,
      "learning_rate": 0.2676588314795834,
      "loss": 0.3549,
      "num_input_tokens_seen": 16155248,
      "step": 8520
    },
    {
      "epoch": 1.3907333387715148,
      "grad_norm": 0.0008714684518054128,
      "learning_rate": 0.26762228577483715,
      "loss": 0.3309,
      "num_input_tokens_seen": 16164144,
      "step": 8525
    },
    {
      "epoch": 1.3915490659923322,
      "grad_norm": 0.0009711458114907146,
      "learning_rate": 0.2675857219312563,
      "loss": 0.3729,
      "num_input_tokens_seen": 16174304,
      "step": 8530
    },
    {
      "epoch": 1.3923647932131495,
      "grad_norm": 0.001130557619035244,
      "learning_rate": 0.2675491399544794,
      "loss": 0.3436,
      "num_input_tokens_seen": 16182032,
      "step": 8535
    },
    {
      "epoch": 1.393180520433967,
      "grad_norm": 0.001260851975530386,
      "learning_rate": 0.2675125398501479,
      "loss": 0.3375,
      "num_input_tokens_seen": 16190944,
      "step": 8540
    },
    {
      "epoch": 1.3939962476547842,
      "grad_norm": 0.0016170135932043195,
      "learning_rate": 0.26747592162390604,
      "loss": 0.3429,
      "num_input_tokens_seen": 16200816,
      "step": 8545
    },
    {
      "epoch": 1.3948119748756016,
      "grad_norm": 0.0012360552791506052,
      "learning_rate": 0.26743928528140076,
      "loss": 0.3263,
      "num_input_tokens_seen": 16210128,
      "step": 8550
    },
    {
      "epoch": 1.395627702096419,
      "grad_norm": 0.001046120305545628,
      "learning_rate": 0.26740263082828186,
      "loss": 0.3435,
      "num_input_tokens_seen": 16220416,
      "step": 8555
    },
    {
      "epoch": 1.3964434293172363,
      "grad_norm": 0.0010058339685201645,
      "learning_rate": 0.2673659582702019,
      "loss": 0.3703,
      "num_input_tokens_seen": 16229984,
      "step": 8560
    },
    {
      "epoch": 1.3972591565380537,
      "grad_norm": 0.0008207212085835636,
      "learning_rate": 0.2673292676128163,
      "loss": 0.338,
      "num_input_tokens_seen": 16240144,
      "step": 8565
    },
    {
      "epoch": 1.398074883758871,
      "grad_norm": 0.0010455019073560834,
      "learning_rate": 0.2672925588617831,
      "loss": 0.4178,
      "num_input_tokens_seen": 16249632,
      "step": 8570
    },
    {
      "epoch": 1.3988906109796884,
      "grad_norm": 0.0012633049627766013,
      "learning_rate": 0.2672558320227634,
      "loss": 0.3152,
      "num_input_tokens_seen": 16258496,
      "step": 8575
    },
    {
      "epoch": 1.3997063382005057,
      "grad_norm": 0.0011473449412733316,
      "learning_rate": 0.2672190871014209,
      "loss": 0.3744,
      "num_input_tokens_seen": 16268000,
      "step": 8580
    },
    {
      "epoch": 1.400522065421323,
      "grad_norm": 0.0014347417745739222,
      "learning_rate": 0.267182324103422,
      "loss": 0.3472,
      "num_input_tokens_seen": 16276800,
      "step": 8585
    },
    {
      "epoch": 1.4013377926421404,
      "grad_norm": 0.0025958833284676075,
      "learning_rate": 0.2671455430344362,
      "loss": 0.3487,
      "num_input_tokens_seen": 16286544,
      "step": 8590
    },
    {
      "epoch": 1.4021535198629578,
      "grad_norm": 0.0024140875320881605,
      "learning_rate": 0.2671087439001355,
      "loss": 0.3987,
      "num_input_tokens_seen": 16295856,
      "step": 8595
    },
    {
      "epoch": 1.4029692470837751,
      "grad_norm": 0.000652038783300668,
      "learning_rate": 0.2670719267061948,
      "loss": 0.3966,
      "num_input_tokens_seen": 16305344,
      "step": 8600
    },
    {
      "epoch": 1.4029692470837751,
      "eval_loss": 0.37761470675468445,
      "eval_runtime": 152.6905,
      "eval_samples_per_second": 17.847,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 16305344,
      "step": 8600
    },
    {
      "epoch": 1.4037849743045925,
      "grad_norm": 0.001967628952115774,
      "learning_rate": 0.2670350914582918,
      "loss": 0.3649,
      "num_input_tokens_seen": 16313648,
      "step": 8605
    },
    {
      "epoch": 1.4046007015254098,
      "grad_norm": 0.001882020733319223,
      "learning_rate": 0.26699823816210694,
      "loss": 0.3783,
      "num_input_tokens_seen": 16323568,
      "step": 8610
    },
    {
      "epoch": 1.4054164287462272,
      "grad_norm": 0.0009618583135306835,
      "learning_rate": 0.26696136682332344,
      "loss": 0.3481,
      "num_input_tokens_seen": 16333024,
      "step": 8615
    },
    {
      "epoch": 1.4062321559670445,
      "grad_norm": 0.0007063319790177047,
      "learning_rate": 0.2669244774476274,
      "loss": 0.3783,
      "num_input_tokens_seen": 16342256,
      "step": 8620
    },
    {
      "epoch": 1.407047883187862,
      "grad_norm": 0.000990750384517014,
      "learning_rate": 0.2668875700407075,
      "loss": 0.3568,
      "num_input_tokens_seen": 16351216,
      "step": 8625
    },
    {
      "epoch": 1.4078636104086795,
      "grad_norm": 0.0016798527212813497,
      "learning_rate": 0.26685064460825547,
      "loss": 0.3469,
      "num_input_tokens_seen": 16360768,
      "step": 8630
    },
    {
      "epoch": 1.4086793376294966,
      "grad_norm": 0.0007216762751340866,
      "learning_rate": 0.26681370115596553,
      "loss": 0.3713,
      "num_input_tokens_seen": 16369552,
      "step": 8635
    },
    {
      "epoch": 1.4094950648503142,
      "grad_norm": 0.0014862575335428119,
      "learning_rate": 0.26677673968953497,
      "loss": 0.3332,
      "num_input_tokens_seen": 16379168,
      "step": 8640
    },
    {
      "epoch": 1.4103107920711313,
      "grad_norm": 0.001282443292438984,
      "learning_rate": 0.2667397602146636,
      "loss": 0.3409,
      "num_input_tokens_seen": 16388592,
      "step": 8645
    },
    {
      "epoch": 1.4111265192919489,
      "grad_norm": 0.00525260204449296,
      "learning_rate": 0.2667027627370542,
      "loss": 0.3696,
      "num_input_tokens_seen": 16397696,
      "step": 8650
    },
    {
      "epoch": 1.411942246512766,
      "grad_norm": 0.0018251874716952443,
      "learning_rate": 0.26666574726241216,
      "loss": 0.3872,
      "num_input_tokens_seen": 16406544,
      "step": 8655
    },
    {
      "epoch": 1.4127579737335836,
      "grad_norm": 0.0006337693776004016,
      "learning_rate": 0.2666287137964458,
      "loss": 0.3176,
      "num_input_tokens_seen": 16415968,
      "step": 8660
    },
    {
      "epoch": 1.4135737009544007,
      "grad_norm": 0.0006092751282267272,
      "learning_rate": 0.26659166234486614,
      "loss": 0.3178,
      "num_input_tokens_seen": 16426416,
      "step": 8665
    },
    {
      "epoch": 1.4143894281752183,
      "grad_norm": 0.0008232222753576934,
      "learning_rate": 0.2665545929133869,
      "loss": 0.279,
      "num_input_tokens_seen": 16435664,
      "step": 8670
    },
    {
      "epoch": 1.4152051553960356,
      "grad_norm": 0.0016273082001134753,
      "learning_rate": 0.2665175055077248,
      "loss": 0.4661,
      "num_input_tokens_seen": 16445328,
      "step": 8675
    },
    {
      "epoch": 1.416020882616853,
      "grad_norm": 0.0013893797295168042,
      "learning_rate": 0.2664804001335991,
      "loss": 0.3212,
      "num_input_tokens_seen": 16456016,
      "step": 8680
    },
    {
      "epoch": 1.4168366098376703,
      "grad_norm": 0.0005273137940093875,
      "learning_rate": 0.26644327679673185,
      "loss": 0.3346,
      "num_input_tokens_seen": 16466384,
      "step": 8685
    },
    {
      "epoch": 1.4176523370584877,
      "grad_norm": 0.0009471720550209284,
      "learning_rate": 0.26640613550284803,
      "loss": 0.3431,
      "num_input_tokens_seen": 16475264,
      "step": 8690
    },
    {
      "epoch": 1.418468064279305,
      "grad_norm": 0.0012557455338537693,
      "learning_rate": 0.26636897625767525,
      "loss": 0.338,
      "num_input_tokens_seen": 16485824,
      "step": 8695
    },
    {
      "epoch": 1.4192837915001224,
      "grad_norm": 0.0018762864638119936,
      "learning_rate": 0.266331799066944,
      "loss": 0.2873,
      "num_input_tokens_seen": 16494000,
      "step": 8700
    },
    {
      "epoch": 1.4200995187209398,
      "grad_norm": 0.0013094347668811679,
      "learning_rate": 0.2662946039363874,
      "loss": 0.324,
      "num_input_tokens_seen": 16502032,
      "step": 8705
    },
    {
      "epoch": 1.420915245941757,
      "grad_norm": 0.0029715029522776604,
      "learning_rate": 0.2662573908717414,
      "loss": 0.3664,
      "num_input_tokens_seen": 16511872,
      "step": 8710
    },
    {
      "epoch": 1.4217309731625745,
      "grad_norm": 0.0008763913647271693,
      "learning_rate": 0.2662201598787447,
      "loss": 0.3507,
      "num_input_tokens_seen": 16522064,
      "step": 8715
    },
    {
      "epoch": 1.4225467003833918,
      "grad_norm": 0.0014476038049906492,
      "learning_rate": 0.2661829109631389,
      "loss": 0.413,
      "num_input_tokens_seen": 16531712,
      "step": 8720
    },
    {
      "epoch": 1.4233624276042092,
      "grad_norm": 0.0017495026113465428,
      "learning_rate": 0.26614564413066816,
      "loss": 0.3893,
      "num_input_tokens_seen": 16541584,
      "step": 8725
    },
    {
      "epoch": 1.4241781548250265,
      "grad_norm": 0.0014511013869196177,
      "learning_rate": 0.2661083593870795,
      "loss": 0.4009,
      "num_input_tokens_seen": 16550144,
      "step": 8730
    },
    {
      "epoch": 1.4249938820458439,
      "grad_norm": 0.0018471790244802833,
      "learning_rate": 0.26607105673812276,
      "loss": 0.3201,
      "num_input_tokens_seen": 16561104,
      "step": 8735
    },
    {
      "epoch": 1.4258096092666612,
      "grad_norm": 0.0013866206863895059,
      "learning_rate": 0.2660337361895504,
      "loss": 0.3585,
      "num_input_tokens_seen": 16572304,
      "step": 8740
    },
    {
      "epoch": 1.4266253364874786,
      "grad_norm": 0.0009537318837828934,
      "learning_rate": 0.26599639774711775,
      "loss": 0.3398,
      "num_input_tokens_seen": 16582416,
      "step": 8745
    },
    {
      "epoch": 1.427441063708296,
      "grad_norm": 0.0007443674840033054,
      "learning_rate": 0.2659590414165829,
      "loss": 0.3482,
      "num_input_tokens_seen": 16591280,
      "step": 8750
    },
    {
      "epoch": 1.4282567909291133,
      "grad_norm": 0.0012281580129638314,
      "learning_rate": 0.2659216672037066,
      "loss": 0.3259,
      "num_input_tokens_seen": 16602320,
      "step": 8755
    },
    {
      "epoch": 1.4290725181499306,
      "grad_norm": 0.0007252118084579706,
      "learning_rate": 0.26588427511425244,
      "loss": 0.3441,
      "num_input_tokens_seen": 16611296,
      "step": 8760
    },
    {
      "epoch": 1.429888245370748,
      "grad_norm": 0.0009929138468578458,
      "learning_rate": 0.26584686515398676,
      "loss": 0.3256,
      "num_input_tokens_seen": 16620544,
      "step": 8765
    },
    {
      "epoch": 1.4307039725915653,
      "grad_norm": 0.0007764738984405994,
      "learning_rate": 0.2658094373286787,
      "loss": 0.3266,
      "num_input_tokens_seen": 16630816,
      "step": 8770
    },
    {
      "epoch": 1.4315196998123827,
      "grad_norm": 0.0015781824477016926,
      "learning_rate": 0.2657719916441,
      "loss": 0.3529,
      "num_input_tokens_seen": 16639680,
      "step": 8775
    },
    {
      "epoch": 1.4323354270332,
      "grad_norm": 0.0012183137005195022,
      "learning_rate": 0.2657345281060253,
      "loss": 0.3356,
      "num_input_tokens_seen": 16649504,
      "step": 8780
    },
    {
      "epoch": 1.4331511542540174,
      "grad_norm": 0.002383400918915868,
      "learning_rate": 0.26569704672023203,
      "loss": 0.3577,
      "num_input_tokens_seen": 16658448,
      "step": 8785
    },
    {
      "epoch": 1.4339668814748348,
      "grad_norm": 0.004179631359875202,
      "learning_rate": 0.26565954749250015,
      "loss": 0.3202,
      "num_input_tokens_seen": 16667712,
      "step": 8790
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 0.00423602806404233,
      "learning_rate": 0.2656220304286126,
      "loss": 0.3363,
      "num_input_tokens_seen": 16677600,
      "step": 8795
    },
    {
      "epoch": 1.4355983359164695,
      "grad_norm": 0.003157505067065358,
      "learning_rate": 0.265584495534355,
      "loss": 0.3021,
      "num_input_tokens_seen": 16686528,
      "step": 8800
    },
    {
      "epoch": 1.4355983359164695,
      "eval_loss": 0.3356603682041168,
      "eval_runtime": 152.5989,
      "eval_samples_per_second": 17.857,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 16686528,
      "step": 8800
    },
    {
      "epoch": 1.4364140631372868,
      "grad_norm": 0.001327692880295217,
      "learning_rate": 0.2655469428155156,
      "loss": 0.3125,
      "num_input_tokens_seen": 16696240,
      "step": 8805
    },
    {
      "epoch": 1.4372297903581042,
      "grad_norm": 0.00048622567555867136,
      "learning_rate": 0.2655093722778856,
      "loss": 0.3065,
      "num_input_tokens_seen": 16706432,
      "step": 8810
    },
    {
      "epoch": 1.4380455175789217,
      "grad_norm": 0.016503941267728806,
      "learning_rate": 0.2654717839272588,
      "loss": 0.4389,
      "num_input_tokens_seen": 16716272,
      "step": 8815
    },
    {
      "epoch": 1.4388612447997389,
      "grad_norm": 0.0007920844363979995,
      "learning_rate": 0.2654341777694318,
      "loss": 0.3867,
      "num_input_tokens_seen": 16726064,
      "step": 8820
    },
    {
      "epoch": 1.4396769720205564,
      "grad_norm": 0.0009791050106287003,
      "learning_rate": 0.265396553810204,
      "loss": 0.3707,
      "num_input_tokens_seen": 16734688,
      "step": 8825
    },
    {
      "epoch": 1.4404926992413736,
      "grad_norm": 0.0023462509270757437,
      "learning_rate": 0.26535891205537737,
      "loss": 0.408,
      "num_input_tokens_seen": 16745008,
      "step": 8830
    },
    {
      "epoch": 1.4413084264621912,
      "grad_norm": 0.0010360275628045201,
      "learning_rate": 0.26532125251075683,
      "loss": 0.3189,
      "num_input_tokens_seen": 16755344,
      "step": 8835
    },
    {
      "epoch": 1.4421241536830083,
      "grad_norm": 0.001338123343884945,
      "learning_rate": 0.26528357518214996,
      "loss": 0.5041,
      "num_input_tokens_seen": 16764528,
      "step": 8840
    },
    {
      "epoch": 1.4429398809038259,
      "grad_norm": 0.001414402388036251,
      "learning_rate": 0.26524588007536704,
      "loss": 0.3669,
      "num_input_tokens_seen": 16772752,
      "step": 8845
    },
    {
      "epoch": 1.443755608124643,
      "grad_norm": 0.0014296716544777155,
      "learning_rate": 0.26520816719622115,
      "loss": 0.331,
      "num_input_tokens_seen": 16781984,
      "step": 8850
    },
    {
      "epoch": 1.4445713353454606,
      "grad_norm": 0.0012468144996091723,
      "learning_rate": 0.2651704365505281,
      "loss": 0.3586,
      "num_input_tokens_seen": 16791776,
      "step": 8855
    },
    {
      "epoch": 1.445387062566278,
      "grad_norm": 0.0012889007339254022,
      "learning_rate": 0.26513268814410634,
      "loss": 0.3516,
      "num_input_tokens_seen": 16798544,
      "step": 8860
    },
    {
      "epoch": 1.4462027897870953,
      "grad_norm": 0.0006073064869269729,
      "learning_rate": 0.2650949219827773,
      "loss": 0.3468,
      "num_input_tokens_seen": 16808736,
      "step": 8865
    },
    {
      "epoch": 1.4470185170079126,
      "grad_norm": 0.0009052992099896073,
      "learning_rate": 0.26505713807236486,
      "loss": 0.3315,
      "num_input_tokens_seen": 16819232,
      "step": 8870
    },
    {
      "epoch": 1.44783424422873,
      "grad_norm": 0.0007111437153071165,
      "learning_rate": 0.26501933641869585,
      "loss": 0.3657,
      "num_input_tokens_seen": 16828960,
      "step": 8875
    },
    {
      "epoch": 1.4486499714495473,
      "grad_norm": 0.0018618903122842312,
      "learning_rate": 0.26498151702759976,
      "loss": 0.3814,
      "num_input_tokens_seen": 16838272,
      "step": 8880
    },
    {
      "epoch": 1.4494656986703647,
      "grad_norm": 0.0006037238636054099,
      "learning_rate": 0.2649436799049088,
      "loss": 0.3171,
      "num_input_tokens_seen": 16848464,
      "step": 8885
    },
    {
      "epoch": 1.450281425891182,
      "grad_norm": 0.00077297433745116,
      "learning_rate": 0.2649058250564579,
      "loss": 0.332,
      "num_input_tokens_seen": 16857168,
      "step": 8890
    },
    {
      "epoch": 1.4510971531119994,
      "grad_norm": 0.0013454882428050041,
      "learning_rate": 0.26486795248808476,
      "loss": 0.3647,
      "num_input_tokens_seen": 16867024,
      "step": 8895
    },
    {
      "epoch": 1.4519128803328167,
      "grad_norm": 0.0014495630748569965,
      "learning_rate": 0.2648300622056298,
      "loss": 0.3504,
      "num_input_tokens_seen": 16876544,
      "step": 8900
    },
    {
      "epoch": 1.452728607553634,
      "grad_norm": 0.0006013906095176935,
      "learning_rate": 0.2647921542149363,
      "loss": 0.356,
      "num_input_tokens_seen": 16886608,
      "step": 8905
    },
    {
      "epoch": 1.4535443347744514,
      "grad_norm": 0.00136878015473485,
      "learning_rate": 0.26475422852185,
      "loss": 0.3702,
      "num_input_tokens_seen": 16897120,
      "step": 8910
    },
    {
      "epoch": 1.4543600619952688,
      "grad_norm": 0.0006951249670237303,
      "learning_rate": 0.2647162851322196,
      "loss": 0.3187,
      "num_input_tokens_seen": 16906384,
      "step": 8915
    },
    {
      "epoch": 1.4551757892160861,
      "grad_norm": 0.0006716210045851767,
      "learning_rate": 0.2646783240518964,
      "loss": 0.3383,
      "num_input_tokens_seen": 16915664,
      "step": 8920
    },
    {
      "epoch": 1.4559915164369035,
      "grad_norm": 0.002091946778818965,
      "learning_rate": 0.26464034528673447,
      "loss": 0.3612,
      "num_input_tokens_seen": 16925824,
      "step": 8925
    },
    {
      "epoch": 1.4568072436577209,
      "grad_norm": 0.0005281685735099018,
      "learning_rate": 0.26460234884259065,
      "loss": 0.3729,
      "num_input_tokens_seen": 16936320,
      "step": 8930
    },
    {
      "epoch": 1.4576229708785382,
      "grad_norm": 0.0024389622267335653,
      "learning_rate": 0.2645643347253245,
      "loss": 0.37,
      "num_input_tokens_seen": 16946144,
      "step": 8935
    },
    {
      "epoch": 1.4584386980993556,
      "grad_norm": 0.0008511905325576663,
      "learning_rate": 0.2645263029407982,
      "loss": 0.3431,
      "num_input_tokens_seen": 16956608,
      "step": 8940
    },
    {
      "epoch": 1.459254425320173,
      "grad_norm": 0.0014303147327154875,
      "learning_rate": 0.2644882534948767,
      "loss": 0.3295,
      "num_input_tokens_seen": 16966432,
      "step": 8945
    },
    {
      "epoch": 1.4600701525409903,
      "grad_norm": 0.0010319697903469205,
      "learning_rate": 0.2644501863934278,
      "loss": 0.2557,
      "num_input_tokens_seen": 16976352,
      "step": 8950
    },
    {
      "epoch": 1.4608858797618076,
      "grad_norm": 0.0008665916975587606,
      "learning_rate": 0.26441210164232193,
      "loss": 0.4608,
      "num_input_tokens_seen": 16985760,
      "step": 8955
    },
    {
      "epoch": 1.461701606982625,
      "grad_norm": 0.0017868743743747473,
      "learning_rate": 0.26437399924743216,
      "loss": 0.3531,
      "num_input_tokens_seen": 16995312,
      "step": 8960
    },
    {
      "epoch": 1.4625173342034423,
      "grad_norm": 0.0011109436163678765,
      "learning_rate": 0.26433587921463436,
      "loss": 0.3549,
      "num_input_tokens_seen": 17005776,
      "step": 8965
    },
    {
      "epoch": 1.4633330614242597,
      "grad_norm": 0.0008748316904529929,
      "learning_rate": 0.2642977415498072,
      "loss": 0.3981,
      "num_input_tokens_seen": 17014848,
      "step": 8970
    },
    {
      "epoch": 1.464148788645077,
      "grad_norm": 0.0009561990154907107,
      "learning_rate": 0.26425958625883195,
      "loss": 0.3101,
      "num_input_tokens_seen": 17024064,
      "step": 8975
    },
    {
      "epoch": 1.4649645158658944,
      "grad_norm": 0.0017958336975425482,
      "learning_rate": 0.2642214133475926,
      "loss": 0.3872,
      "num_input_tokens_seen": 17033584,
      "step": 8980
    },
    {
      "epoch": 1.4657802430867117,
      "grad_norm": 0.0006303873960860074,
      "learning_rate": 0.26418322282197587,
      "loss": 0.3677,
      "num_input_tokens_seen": 17043712,
      "step": 8985
    },
    {
      "epoch": 1.466595970307529,
      "grad_norm": 0.0019445159705355763,
      "learning_rate": 0.2641450146878714,
      "loss": 0.3417,
      "num_input_tokens_seen": 17052848,
      "step": 8990
    },
    {
      "epoch": 1.4674116975283464,
      "grad_norm": 0.001223896280862391,
      "learning_rate": 0.26410678895117107,
      "loss": 0.3326,
      "num_input_tokens_seen": 17063280,
      "step": 8995
    },
    {
      "epoch": 1.468227424749164,
      "grad_norm": 0.0006340033723972738,
      "learning_rate": 0.26406854561777,
      "loss": 0.3295,
      "num_input_tokens_seen": 17073648,
      "step": 9000
    },
    {
      "epoch": 1.468227424749164,
      "eval_loss": 0.3456483483314514,
      "eval_runtime": 152.5751,
      "eval_samples_per_second": 17.86,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 17073648,
      "step": 9000
    },
    {
      "epoch": 1.4690431519699811,
      "grad_norm": 0.0004031471617054194,
      "learning_rate": 0.26403028469356576,
      "loss": 0.2711,
      "num_input_tokens_seen": 17082832,
      "step": 9005
    },
    {
      "epoch": 1.4698588791907987,
      "grad_norm": 0.0011994115775451064,
      "learning_rate": 0.2639920061844585,
      "loss": 0.3292,
      "num_input_tokens_seen": 17091104,
      "step": 9010
    },
    {
      "epoch": 1.4706746064116158,
      "grad_norm": 0.0006951419054530561,
      "learning_rate": 0.2639537100963515,
      "loss": 0.4019,
      "num_input_tokens_seen": 17098544,
      "step": 9015
    },
    {
      "epoch": 1.4714903336324334,
      "grad_norm": 0.0010427339002490044,
      "learning_rate": 0.26391539643515033,
      "loss": 0.3415,
      "num_input_tokens_seen": 17108640,
      "step": 9020
    },
    {
      "epoch": 1.4723060608532506,
      "grad_norm": 0.0007075723842717707,
      "learning_rate": 0.26387706520676346,
      "loss": 0.3512,
      "num_input_tokens_seen": 17117504,
      "step": 9025
    },
    {
      "epoch": 1.4731217880740681,
      "grad_norm": 0.0020205839537084103,
      "learning_rate": 0.26383871641710205,
      "loss": 0.3592,
      "num_input_tokens_seen": 17125792,
      "step": 9030
    },
    {
      "epoch": 1.4739375152948853,
      "grad_norm": 0.0028163676615804434,
      "learning_rate": 0.26380035007208,
      "loss": 0.3788,
      "num_input_tokens_seen": 17135040,
      "step": 9035
    },
    {
      "epoch": 1.4747532425157028,
      "grad_norm": 0.0004929215065203607,
      "learning_rate": 0.26376196617761394,
      "loss": 0.3297,
      "num_input_tokens_seen": 17143472,
      "step": 9040
    },
    {
      "epoch": 1.4755689697365202,
      "grad_norm": 0.0013056587195023894,
      "learning_rate": 0.263723564739623,
      "loss": 0.3815,
      "num_input_tokens_seen": 17153472,
      "step": 9045
    },
    {
      "epoch": 1.4763846969573375,
      "grad_norm": 0.0010595122585073113,
      "learning_rate": 0.2636851457640293,
      "loss": 0.353,
      "num_input_tokens_seen": 17163584,
      "step": 9050
    },
    {
      "epoch": 1.477200424178155,
      "grad_norm": 0.0013513054000213742,
      "learning_rate": 0.26364670925675737,
      "loss": 0.3417,
      "num_input_tokens_seen": 17174704,
      "step": 9055
    },
    {
      "epoch": 1.4780161513989722,
      "grad_norm": 0.000946579675655812,
      "learning_rate": 0.2636082552237347,
      "loss": 0.3491,
      "num_input_tokens_seen": 17183232,
      "step": 9060
    },
    {
      "epoch": 1.4788318786197896,
      "grad_norm": 0.0012505605118349195,
      "learning_rate": 0.26356978367089146,
      "loss": 0.3339,
      "num_input_tokens_seen": 17192640,
      "step": 9065
    },
    {
      "epoch": 1.479647605840607,
      "grad_norm": 0.0010741421720013022,
      "learning_rate": 0.26353129460416036,
      "loss": 0.3516,
      "num_input_tokens_seen": 17202432,
      "step": 9070
    },
    {
      "epoch": 1.4804633330614243,
      "grad_norm": 0.0009389688493683934,
      "learning_rate": 0.2634927880294769,
      "loss": 0.3307,
      "num_input_tokens_seen": 17212496,
      "step": 9075
    },
    {
      "epoch": 1.4812790602822417,
      "grad_norm": 0.001322324969805777,
      "learning_rate": 0.26345426395277927,
      "loss": 0.2866,
      "num_input_tokens_seen": 17223056,
      "step": 9080
    },
    {
      "epoch": 1.482094787503059,
      "grad_norm": 0.002141987206414342,
      "learning_rate": 0.2634157223800084,
      "loss": 0.3673,
      "num_input_tokens_seen": 17233760,
      "step": 9085
    },
    {
      "epoch": 1.4829105147238764,
      "grad_norm": 0.0022376494016498327,
      "learning_rate": 0.26337716331710787,
      "loss": 0.3798,
      "num_input_tokens_seen": 17242592,
      "step": 9090
    },
    {
      "epoch": 1.4837262419446937,
      "grad_norm": 0.002616268116980791,
      "learning_rate": 0.2633385867700239,
      "loss": 0.2814,
      "num_input_tokens_seen": 17252512,
      "step": 9095
    },
    {
      "epoch": 1.484541969165511,
      "grad_norm": 0.0020434882026165724,
      "learning_rate": 0.2632999927447056,
      "loss": 0.3397,
      "num_input_tokens_seen": 17262784,
      "step": 9100
    },
    {
      "epoch": 1.4853576963863284,
      "grad_norm": 0.0017344941152259707,
      "learning_rate": 0.2632613812471046,
      "loss": 0.3307,
      "num_input_tokens_seen": 17272224,
      "step": 9105
    },
    {
      "epoch": 1.4861734236071458,
      "grad_norm": 0.0017721086042001843,
      "learning_rate": 0.2632227522831753,
      "loss": 0.2723,
      "num_input_tokens_seen": 17281984,
      "step": 9110
    },
    {
      "epoch": 1.4869891508279631,
      "grad_norm": 0.001961593981832266,
      "learning_rate": 0.26318410585887475,
      "loss": 0.3982,
      "num_input_tokens_seen": 17290848,
      "step": 9115
    },
    {
      "epoch": 1.4878048780487805,
      "grad_norm": 0.0036692447029054165,
      "learning_rate": 0.2631454419801627,
      "loss": 0.3217,
      "num_input_tokens_seen": 17301184,
      "step": 9120
    },
    {
      "epoch": 1.4886206052695978,
      "grad_norm": 0.0011879249941557646,
      "learning_rate": 0.2631067606530016,
      "loss": 0.3266,
      "num_input_tokens_seen": 17311632,
      "step": 9125
    },
    {
      "epoch": 1.4894363324904152,
      "grad_norm": 0.0029554986394941807,
      "learning_rate": 0.2630680618833567,
      "loss": 0.3143,
      "num_input_tokens_seen": 17321360,
      "step": 9130
    },
    {
      "epoch": 1.4902520597112325,
      "grad_norm": 0.0009699404472485185,
      "learning_rate": 0.26302934567719566,
      "loss": 0.3531,
      "num_input_tokens_seen": 17331680,
      "step": 9135
    },
    {
      "epoch": 1.4910677869320499,
      "grad_norm": 0.002019468229264021,
      "learning_rate": 0.2629906120404892,
      "loss": 0.3728,
      "num_input_tokens_seen": 17342320,
      "step": 9140
    },
    {
      "epoch": 1.4918835141528672,
      "grad_norm": 0.0017747626407071948,
      "learning_rate": 0.26295186097921036,
      "loss": 0.3624,
      "num_input_tokens_seen": 17353024,
      "step": 9145
    },
    {
      "epoch": 1.4926992413736846,
      "grad_norm": 0.002410839544609189,
      "learning_rate": 0.2629130924993351,
      "loss": 0.3276,
      "num_input_tokens_seen": 17364560,
      "step": 9150
    },
    {
      "epoch": 1.493514968594502,
      "grad_norm": 0.0006312234327197075,
      "learning_rate": 0.2628743066068421,
      "loss": 0.3817,
      "num_input_tokens_seen": 17373424,
      "step": 9155
    },
    {
      "epoch": 1.4943306958153193,
      "grad_norm": 0.0005809601861983538,
      "learning_rate": 0.26283550330771244,
      "loss": 0.3111,
      "num_input_tokens_seen": 17382992,
      "step": 9160
    },
    {
      "epoch": 1.4951464230361367,
      "grad_norm": 0.001785491593182087,
      "learning_rate": 0.2627966826079303,
      "loss": 0.3081,
      "num_input_tokens_seen": 17392160,
      "step": 9165
    },
    {
      "epoch": 1.495962150256954,
      "grad_norm": 0.0006466325139626861,
      "learning_rate": 0.26275784451348216,
      "loss": 0.3626,
      "num_input_tokens_seen": 17400592,
      "step": 9170
    },
    {
      "epoch": 1.4967778774777714,
      "grad_norm": 0.0004489612183533609,
      "learning_rate": 0.2627189890303574,
      "loss": 0.3085,
      "num_input_tokens_seen": 17409744,
      "step": 9175
    },
    {
      "epoch": 1.4975936046985887,
      "grad_norm": 0.00106246757786721,
      "learning_rate": 0.262680116164548,
      "loss": 0.3365,
      "num_input_tokens_seen": 17420688,
      "step": 9180
    },
    {
      "epoch": 1.4984093319194063,
      "grad_norm": 0.001301855081692338,
      "learning_rate": 0.2626412259220487,
      "loss": 0.2624,
      "num_input_tokens_seen": 17430288,
      "step": 9185
    },
    {
      "epoch": 1.4992250591402234,
      "grad_norm": 0.0009507809299975634,
      "learning_rate": 0.2626023183088568,
      "loss": 0.3491,
      "num_input_tokens_seen": 17439216,
      "step": 9190
    },
    {
      "epoch": 1.500040786361041,
      "grad_norm": 0.0013338524149730802,
      "learning_rate": 0.26256339333097234,
      "loss": 0.3418,
      "num_input_tokens_seen": 17448000,
      "step": 9195
    },
    {
      "epoch": 1.5008565135818581,
      "grad_norm": 0.0005267634405754507,
      "learning_rate": 0.2625244509943981,
      "loss": 0.3563,
      "num_input_tokens_seen": 17457952,
      "step": 9200
    },
    {
      "epoch": 1.5008565135818581,
      "eval_loss": 0.3691868484020233,
      "eval_runtime": 152.57,
      "eval_samples_per_second": 17.861,
      "eval_steps_per_second": 8.934,
      "num_input_tokens_seen": 17457952,
      "step": 9200
    },
    {
      "epoch": 1.5016722408026757,
      "grad_norm": 0.003926859237253666,
      "learning_rate": 0.2624854913051395,
      "loss": 0.4273,
      "num_input_tokens_seen": 17467488,
      "step": 9205
    },
    {
      "epoch": 1.5024879680234928,
      "grad_norm": 0.0013655901420861483,
      "learning_rate": 0.26244651426920446,
      "loss": 0.3848,
      "num_input_tokens_seen": 17477440,
      "step": 9210
    },
    {
      "epoch": 1.5033036952443104,
      "grad_norm": 0.0009567333618178964,
      "learning_rate": 0.26240751989260386,
      "loss": 0.3257,
      "num_input_tokens_seen": 17486448,
      "step": 9215
    },
    {
      "epoch": 1.5041194224651275,
      "grad_norm": 0.0017014300683513284,
      "learning_rate": 0.2623685081813511,
      "loss": 0.3467,
      "num_input_tokens_seen": 17496656,
      "step": 9220
    },
    {
      "epoch": 1.504935149685945,
      "grad_norm": 0.0025567093398422003,
      "learning_rate": 0.2623294791414623,
      "loss": 0.3736,
      "num_input_tokens_seen": 17505248,
      "step": 9225
    },
    {
      "epoch": 1.5057508769067622,
      "grad_norm": 0.0011596762342378497,
      "learning_rate": 0.26229043277895614,
      "loss": 0.3279,
      "num_input_tokens_seen": 17514912,
      "step": 9230
    },
    {
      "epoch": 1.5065666041275798,
      "grad_norm": 0.0021334595512598753,
      "learning_rate": 0.2622513690998542,
      "loss": 0.3652,
      "num_input_tokens_seen": 17525248,
      "step": 9235
    },
    {
      "epoch": 1.507382331348397,
      "grad_norm": 0.0022818001452833414,
      "learning_rate": 0.26221228811018044,
      "loss": 0.3236,
      "num_input_tokens_seen": 17534288,
      "step": 9240
    },
    {
      "epoch": 1.5081980585692145,
      "grad_norm": 0.0016742820153012872,
      "learning_rate": 0.2621731898159617,
      "loss": 0.3586,
      "num_input_tokens_seen": 17544992,
      "step": 9245
    },
    {
      "epoch": 1.5090137857900316,
      "grad_norm": 0.0025149076245725155,
      "learning_rate": 0.26213407422322743,
      "loss": 0.3461,
      "num_input_tokens_seen": 17553744,
      "step": 9250
    },
    {
      "epoch": 1.5098295130108492,
      "grad_norm": 0.0031052858103066683,
      "learning_rate": 0.2620949413380098,
      "loss": 0.3775,
      "num_input_tokens_seen": 17563792,
      "step": 9255
    },
    {
      "epoch": 1.5106452402316666,
      "grad_norm": 0.0013315140968188643,
      "learning_rate": 0.26205579116634353,
      "loss": 0.39,
      "num_input_tokens_seen": 17572512,
      "step": 9260
    },
    {
      "epoch": 1.511460967452484,
      "grad_norm": 0.0014842812670394778,
      "learning_rate": 0.26201662371426604,
      "loss": 0.3539,
      "num_input_tokens_seen": 17580544,
      "step": 9265
    },
    {
      "epoch": 1.5122766946733013,
      "grad_norm": 0.0008977788384072483,
      "learning_rate": 0.2619774389878175,
      "loss": 0.3347,
      "num_input_tokens_seen": 17589264,
      "step": 9270
    },
    {
      "epoch": 1.5130924218941186,
      "grad_norm": 0.0015977637376636267,
      "learning_rate": 0.2619382369930407,
      "loss": 0.3769,
      "num_input_tokens_seen": 17599024,
      "step": 9275
    },
    {
      "epoch": 1.513908149114936,
      "grad_norm": 0.0010257791727781296,
      "learning_rate": 0.261899017735981,
      "loss": 0.2765,
      "num_input_tokens_seen": 17608544,
      "step": 9280
    },
    {
      "epoch": 1.5147238763357533,
      "grad_norm": 0.001112426514737308,
      "learning_rate": 0.2618597812226866,
      "loss": 0.4023,
      "num_input_tokens_seen": 17617424,
      "step": 9285
    },
    {
      "epoch": 1.5155396035565707,
      "grad_norm": 0.001576905488036573,
      "learning_rate": 0.2618205274592082,
      "loss": 0.3967,
      "num_input_tokens_seen": 17625776,
      "step": 9290
    },
    {
      "epoch": 1.516355330777388,
      "grad_norm": 0.0017468073638156056,
      "learning_rate": 0.2617812564515992,
      "loss": 0.3903,
      "num_input_tokens_seen": 17634752,
      "step": 9295
    },
    {
      "epoch": 1.5171710579982054,
      "grad_norm": 0.0012307509314268827,
      "learning_rate": 0.2617419682059158,
      "loss": 0.3397,
      "num_input_tokens_seen": 17644256,
      "step": 9300
    },
    {
      "epoch": 1.5179867852190227,
      "grad_norm": 0.0010559820802882314,
      "learning_rate": 0.26170266272821663,
      "loss": 0.3288,
      "num_input_tokens_seen": 17653328,
      "step": 9305
    },
    {
      "epoch": 1.51880251243984,
      "grad_norm": 0.0008046905277296901,
      "learning_rate": 0.26166334002456315,
      "loss": 0.3404,
      "num_input_tokens_seen": 17663408,
      "step": 9310
    },
    {
      "epoch": 1.5196182396606575,
      "grad_norm": 0.0017667179927229881,
      "learning_rate": 0.2616240001010194,
      "loss": 0.3468,
      "num_input_tokens_seen": 17673296,
      "step": 9315
    },
    {
      "epoch": 1.5204339668814748,
      "grad_norm": 0.0013709714403375983,
      "learning_rate": 0.26158464296365197,
      "loss": 0.3199,
      "num_input_tokens_seen": 17682736,
      "step": 9320
    },
    {
      "epoch": 1.5212496941022922,
      "grad_norm": 0.0009973326232284307,
      "learning_rate": 0.2615452686185304,
      "loss": 0.295,
      "num_input_tokens_seen": 17692304,
      "step": 9325
    },
    {
      "epoch": 1.5220654213231095,
      "grad_norm": 0.0014448050642386079,
      "learning_rate": 0.26150587707172673,
      "loss": 0.372,
      "num_input_tokens_seen": 17701536,
      "step": 9330
    },
    {
      "epoch": 1.5228811485439269,
      "grad_norm": 0.001248083310201764,
      "learning_rate": 0.2614664683293154,
      "loss": 0.3135,
      "num_input_tokens_seen": 17710720,
      "step": 9335
    },
    {
      "epoch": 1.5236968757647442,
      "grad_norm": 0.0017098065000027418,
      "learning_rate": 0.26142704239737397,
      "loss": 0.3918,
      "num_input_tokens_seen": 17719568,
      "step": 9340
    },
    {
      "epoch": 1.5245126029855616,
      "grad_norm": 0.0018517187563702464,
      "learning_rate": 0.26138759928198235,
      "loss": 0.3761,
      "num_input_tokens_seen": 17727904,
      "step": 9345
    },
    {
      "epoch": 1.5253283302063791,
      "grad_norm": 0.0012525420170277357,
      "learning_rate": 0.26134813898922304,
      "loss": 0.3528,
      "num_input_tokens_seen": 17735936,
      "step": 9350
    },
    {
      "epoch": 1.5261440574271963,
      "grad_norm": 0.0012874515959993005,
      "learning_rate": 0.26130866152518145,
      "loss": 0.3074,
      "num_input_tokens_seen": 17745520,
      "step": 9355
    },
    {
      "epoch": 1.5269597846480139,
      "grad_norm": 0.0027694462332874537,
      "learning_rate": 0.2612691668959455,
      "loss": 0.4162,
      "num_input_tokens_seen": 17755600,
      "step": 9360
    },
    {
      "epoch": 1.527775511868831,
      "grad_norm": 0.0007836173172108829,
      "learning_rate": 0.2612296551076057,
      "loss": 0.3846,
      "num_input_tokens_seen": 17765296,
      "step": 9365
    },
    {
      "epoch": 1.5285912390896486,
      "grad_norm": 0.0014653103426098824,
      "learning_rate": 0.26119012616625525,
      "loss": 0.3506,
      "num_input_tokens_seen": 17774240,
      "step": 9370
    },
    {
      "epoch": 1.5294069663104657,
      "grad_norm": 0.001266359118744731,
      "learning_rate": 0.26115058007799,
      "loss": 0.383,
      "num_input_tokens_seen": 17783600,
      "step": 9375
    },
    {
      "epoch": 1.5302226935312833,
      "grad_norm": 0.0029248392675071955,
      "learning_rate": 0.26111101684890864,
      "loss": 0.3573,
      "num_input_tokens_seen": 17793040,
      "step": 9380
    },
    {
      "epoch": 1.5310384207521004,
      "grad_norm": 0.0010086748516187072,
      "learning_rate": 0.26107143648511205,
      "loss": 0.3502,
      "num_input_tokens_seen": 17802144,
      "step": 9385
    },
    {
      "epoch": 1.531854147972918,
      "grad_norm": 0.0012781850527971983,
      "learning_rate": 0.2610318389927042,
      "loss": 0.3464,
      "num_input_tokens_seen": 17812272,
      "step": 9390
    },
    {
      "epoch": 1.532669875193735,
      "grad_norm": 0.0018034453969448805,
      "learning_rate": 0.26099222437779146,
      "loss": 0.378,
      "num_input_tokens_seen": 17821488,
      "step": 9395
    },
    {
      "epoch": 1.5334856024145527,
      "grad_norm": 0.0014987074537202716,
      "learning_rate": 0.26095259264648285,
      "loss": 0.3424,
      "num_input_tokens_seen": 17831104,
      "step": 9400
    },
    {
      "epoch": 1.5334856024145527,
      "eval_loss": 0.35052230954170227,
      "eval_runtime": 152.4989,
      "eval_samples_per_second": 17.869,
      "eval_steps_per_second": 8.938,
      "num_input_tokens_seen": 17831104,
      "step": 9400
    },
    {
      "epoch": 1.5343013296353698,
      "grad_norm": 0.0019042114727199078,
      "learning_rate": 0.2609129438048902,
      "loss": 0.3406,
      "num_input_tokens_seen": 17841200,
      "step": 9405
    },
    {
      "epoch": 1.5351170568561874,
      "grad_norm": 0.001825696905143559,
      "learning_rate": 0.2608732778591278,
      "loss": 0.3524,
      "num_input_tokens_seen": 17849904,
      "step": 9410
    },
    {
      "epoch": 1.5359327840770045,
      "grad_norm": 0.002064695581793785,
      "learning_rate": 0.2608335948153126,
      "loss": 0.3456,
      "num_input_tokens_seen": 17859840,
      "step": 9415
    },
    {
      "epoch": 1.536748511297822,
      "grad_norm": 0.0009383309516124427,
      "learning_rate": 0.26079389467956426,
      "loss": 0.3259,
      "num_input_tokens_seen": 17870336,
      "step": 9420
    },
    {
      "epoch": 1.5375642385186392,
      "grad_norm": 0.0020615560933947563,
      "learning_rate": 0.26075417745800505,
      "loss": 0.3544,
      "num_input_tokens_seen": 17880304,
      "step": 9425
    },
    {
      "epoch": 1.5383799657394568,
      "grad_norm": 0.0015991092659533024,
      "learning_rate": 0.26071444315675985,
      "loss": 0.4106,
      "num_input_tokens_seen": 17889008,
      "step": 9430
    },
    {
      "epoch": 1.539195692960274,
      "grad_norm": 0.0007079422939568758,
      "learning_rate": 0.2606746917819562,
      "loss": 0.3394,
      "num_input_tokens_seen": 17898992,
      "step": 9435
    },
    {
      "epoch": 1.5400114201810915,
      "grad_norm": 0.0010872316779568791,
      "learning_rate": 0.2606349233397242,
      "loss": 0.3157,
      "num_input_tokens_seen": 17909168,
      "step": 9440
    },
    {
      "epoch": 1.5408271474019088,
      "grad_norm": 0.001314854365773499,
      "learning_rate": 0.26059513783619676,
      "loss": 0.3309,
      "num_input_tokens_seen": 17918720,
      "step": 9445
    },
    {
      "epoch": 1.5416428746227262,
      "grad_norm": 0.001880867755971849,
      "learning_rate": 0.26055533527750924,
      "loss": 0.4001,
      "num_input_tokens_seen": 17929136,
      "step": 9450
    },
    {
      "epoch": 1.5424586018435436,
      "grad_norm": 0.0006421580328606069,
      "learning_rate": 0.26051551566979964,
      "loss": 0.3411,
      "num_input_tokens_seen": 17938032,
      "step": 9455
    },
    {
      "epoch": 1.543274329064361,
      "grad_norm": 0.0018753822660073638,
      "learning_rate": 0.26047567901920876,
      "loss": 0.339,
      "num_input_tokens_seen": 17947840,
      "step": 9460
    },
    {
      "epoch": 1.5440900562851783,
      "grad_norm": 0.0010702981380745769,
      "learning_rate": 0.2604358253318798,
      "loss": 0.3323,
      "num_input_tokens_seen": 17958224,
      "step": 9465
    },
    {
      "epoch": 1.5449057835059956,
      "grad_norm": 0.0013132794992998242,
      "learning_rate": 0.26039595461395876,
      "loss": 0.3501,
      "num_input_tokens_seen": 17968144,
      "step": 9470
    },
    {
      "epoch": 1.545721510726813,
      "grad_norm": 0.001350431819446385,
      "learning_rate": 0.26035606687159424,
      "loss": 0.3275,
      "num_input_tokens_seen": 17978176,
      "step": 9475
    },
    {
      "epoch": 1.5465372379476303,
      "grad_norm": 0.0010288336779922247,
      "learning_rate": 0.26031616211093733,
      "loss": 0.3285,
      "num_input_tokens_seen": 17987040,
      "step": 9480
    },
    {
      "epoch": 1.5473529651684477,
      "grad_norm": 0.001788539346307516,
      "learning_rate": 0.26027624033814195,
      "loss": 0.3842,
      "num_input_tokens_seen": 17995904,
      "step": 9485
    },
    {
      "epoch": 1.548168692389265,
      "grad_norm": 0.0030202351044863462,
      "learning_rate": 0.2602363015593645,
      "loss": 0.3913,
      "num_input_tokens_seen": 18006832,
      "step": 9490
    },
    {
      "epoch": 1.5489844196100824,
      "grad_norm": 0.0008297240710817277,
      "learning_rate": 0.26019634578076395,
      "loss": 0.3484,
      "num_input_tokens_seen": 18016400,
      "step": 9495
    },
    {
      "epoch": 1.5498001468308997,
      "grad_norm": 0.001404733513481915,
      "learning_rate": 0.26015637300850214,
      "loss": 0.3759,
      "num_input_tokens_seen": 18026976,
      "step": 9500
    },
    {
      "epoch": 1.550615874051717,
      "grad_norm": 0.0006640206556767225,
      "learning_rate": 0.26011638324874325,
      "loss": 0.3243,
      "num_input_tokens_seen": 18037520,
      "step": 9505
    },
    {
      "epoch": 1.5514316012725344,
      "grad_norm": 0.001518457313068211,
      "learning_rate": 0.2600763765076543,
      "loss": 0.36,
      "num_input_tokens_seen": 18047264,
      "step": 9510
    },
    {
      "epoch": 1.5522473284933518,
      "grad_norm": 0.0010111377341672778,
      "learning_rate": 0.2600363527914048,
      "loss": 0.3408,
      "num_input_tokens_seen": 18056320,
      "step": 9515
    },
    {
      "epoch": 1.5530630557141691,
      "grad_norm": 0.0015420077834278345,
      "learning_rate": 0.25999631210616686,
      "loss": 0.3648,
      "num_input_tokens_seen": 18065968,
      "step": 9520
    },
    {
      "epoch": 1.5538787829349865,
      "grad_norm": 0.0009646931430324912,
      "learning_rate": 0.25995625445811527,
      "loss": 0.3695,
      "num_input_tokens_seen": 18074704,
      "step": 9525
    },
    {
      "epoch": 1.5546945101558038,
      "grad_norm": 0.0033785542473196983,
      "learning_rate": 0.2599161798534275,
      "loss": 0.3488,
      "num_input_tokens_seen": 18083808,
      "step": 9530
    },
    {
      "epoch": 1.5555102373766214,
      "grad_norm": 0.0028076160233467817,
      "learning_rate": 0.25987608829828346,
      "loss": 0.3278,
      "num_input_tokens_seen": 18093920,
      "step": 9535
    },
    {
      "epoch": 1.5563259645974385,
      "grad_norm": 0.001011007814668119,
      "learning_rate": 0.25983597979886586,
      "loss": 0.2876,
      "num_input_tokens_seen": 18104016,
      "step": 9540
    },
    {
      "epoch": 1.5571416918182561,
      "grad_norm": 0.0016925066011026502,
      "learning_rate": 0.2597958543613599,
      "loss": 0.4286,
      "num_input_tokens_seen": 18113264,
      "step": 9545
    },
    {
      "epoch": 1.5579574190390733,
      "grad_norm": 0.0016098087653517723,
      "learning_rate": 0.25975571199195335,
      "loss": 0.3736,
      "num_input_tokens_seen": 18120912,
      "step": 9550
    },
    {
      "epoch": 1.5587731462598908,
      "grad_norm": 0.0010269100312143564,
      "learning_rate": 0.25971555269683677,
      "loss": 0.366,
      "num_input_tokens_seen": 18130816,
      "step": 9555
    },
    {
      "epoch": 1.559588873480708,
      "grad_norm": 0.0012342424597591162,
      "learning_rate": 0.25967537648220324,
      "loss": 0.4042,
      "num_input_tokens_seen": 18140128,
      "step": 9560
    },
    {
      "epoch": 1.5604046007015255,
      "grad_norm": 0.0011217218125239015,
      "learning_rate": 0.2596351833542483,
      "loss": 0.3459,
      "num_input_tokens_seen": 18150000,
      "step": 9565
    },
    {
      "epoch": 1.5612203279223427,
      "grad_norm": 0.0010851913830265403,
      "learning_rate": 0.25959497331917036,
      "loss": 0.3427,
      "num_input_tokens_seen": 18159712,
      "step": 9570
    },
    {
      "epoch": 1.5620360551431602,
      "grad_norm": 0.0010470845736563206,
      "learning_rate": 0.2595547463831703,
      "loss": 0.3461,
      "num_input_tokens_seen": 18169936,
      "step": 9575
    },
    {
      "epoch": 1.5628517823639774,
      "grad_norm": 0.001405229326337576,
      "learning_rate": 0.25951450255245156,
      "loss": 0.3017,
      "num_input_tokens_seen": 18179120,
      "step": 9580
    },
    {
      "epoch": 1.563667509584795,
      "grad_norm": 0.0009584985091350973,
      "learning_rate": 0.2594742418332203,
      "loss": 0.4356,
      "num_input_tokens_seen": 18189504,
      "step": 9585
    },
    {
      "epoch": 1.564483236805612,
      "grad_norm": 0.00040123669896274805,
      "learning_rate": 0.2594339642316852,
      "loss": 0.3656,
      "num_input_tokens_seen": 18197088,
      "step": 9590
    },
    {
      "epoch": 1.5652989640264297,
      "grad_norm": 0.00069256970891729,
      "learning_rate": 0.2593936697540576,
      "loss": 0.3431,
      "num_input_tokens_seen": 18206048,
      "step": 9595
    },
    {
      "epoch": 1.5661146912472468,
      "grad_norm": 0.001949790632352233,
      "learning_rate": 0.2593533584065514,
      "loss": 0.3343,
      "num_input_tokens_seen": 18215168,
      "step": 9600
    },
    {
      "epoch": 1.5661146912472468,
      "eval_loss": 0.33461013436317444,
      "eval_runtime": 152.6782,
      "eval_samples_per_second": 17.848,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 18215168,
      "step": 9600
    },
    {
      "epoch": 1.5669304184680644,
      "grad_norm": 0.0027775561902672052,
      "learning_rate": 0.2593130301953831,
      "loss": 0.343,
      "num_input_tokens_seen": 18223648,
      "step": 9605
    },
    {
      "epoch": 1.5677461456888815,
      "grad_norm": 0.0025886562652885914,
      "learning_rate": 0.2592726851267718,
      "loss": 0.3705,
      "num_input_tokens_seen": 18233632,
      "step": 9610
    },
    {
      "epoch": 1.568561872909699,
      "grad_norm": 0.0011278624879196286,
      "learning_rate": 0.2592323232069393,
      "loss": 0.3826,
      "num_input_tokens_seen": 18242288,
      "step": 9615
    },
    {
      "epoch": 1.5693776001305162,
      "grad_norm": 0.0021767469588667154,
      "learning_rate": 0.25919194444210986,
      "loss": 0.3366,
      "num_input_tokens_seen": 18251456,
      "step": 9620
    },
    {
      "epoch": 1.5701933273513338,
      "grad_norm": 0.0006998165627010167,
      "learning_rate": 0.2591515488385103,
      "loss": 0.3344,
      "num_input_tokens_seen": 18262208,
      "step": 9625
    },
    {
      "epoch": 1.5710090545721511,
      "grad_norm": 0.0010645149741321802,
      "learning_rate": 0.2591111364023704,
      "loss": 0.3368,
      "num_input_tokens_seen": 18270976,
      "step": 9630
    },
    {
      "epoch": 1.5718247817929685,
      "grad_norm": 0.0018931747181341052,
      "learning_rate": 0.259070707139922,
      "loss": 0.3408,
      "num_input_tokens_seen": 18281152,
      "step": 9635
    },
    {
      "epoch": 1.5726405090137858,
      "grad_norm": 0.0013342495076358318,
      "learning_rate": 0.25903026105739985,
      "loss": 0.3214,
      "num_input_tokens_seen": 18290544,
      "step": 9640
    },
    {
      "epoch": 1.5734562362346032,
      "grad_norm": 0.004496992565691471,
      "learning_rate": 0.2589897981610413,
      "loss": 0.3944,
      "num_input_tokens_seen": 18300000,
      "step": 9645
    },
    {
      "epoch": 1.5742719634554205,
      "grad_norm": 0.000790402467828244,
      "learning_rate": 0.2589493184570863,
      "loss": 0.362,
      "num_input_tokens_seen": 18310240,
      "step": 9650
    },
    {
      "epoch": 1.5750876906762379,
      "grad_norm": 0.0019459432223811746,
      "learning_rate": 0.25890882195177717,
      "loss": 0.3319,
      "num_input_tokens_seen": 18318576,
      "step": 9655
    },
    {
      "epoch": 1.5759034178970552,
      "grad_norm": 0.0015542682958766818,
      "learning_rate": 0.25886830865135907,
      "loss": 0.3472,
      "num_input_tokens_seen": 18326592,
      "step": 9660
    },
    {
      "epoch": 1.5767191451178726,
      "grad_norm": 0.0012087911600247025,
      "learning_rate": 0.25882777856207967,
      "loss": 0.3406,
      "num_input_tokens_seen": 18336784,
      "step": 9665
    },
    {
      "epoch": 1.57753487233869,
      "grad_norm": 0.0016545143444091082,
      "learning_rate": 0.2587872316901892,
      "loss": 0.3467,
      "num_input_tokens_seen": 18345040,
      "step": 9670
    },
    {
      "epoch": 1.5783505995595073,
      "grad_norm": 0.0020186190959066153,
      "learning_rate": 0.25874666804194046,
      "loss": 0.3389,
      "num_input_tokens_seen": 18354256,
      "step": 9675
    },
    {
      "epoch": 1.5791663267803246,
      "grad_norm": 0.0019422279438003898,
      "learning_rate": 0.258706087623589,
      "loss": 0.3603,
      "num_input_tokens_seen": 18363856,
      "step": 9680
    },
    {
      "epoch": 1.579982054001142,
      "grad_norm": 0.0013522685039788485,
      "learning_rate": 0.25866549044139264,
      "loss": 0.3386,
      "num_input_tokens_seen": 18374288,
      "step": 9685
    },
    {
      "epoch": 1.5807977812219594,
      "grad_norm": 0.0023232598323374987,
      "learning_rate": 0.25862487650161214,
      "loss": 0.3561,
      "num_input_tokens_seen": 18383680,
      "step": 9690
    },
    {
      "epoch": 1.5816135084427767,
      "grad_norm": 0.0014203520258888602,
      "learning_rate": 0.2585842458105106,
      "loss": 0.3791,
      "num_input_tokens_seen": 18393360,
      "step": 9695
    },
    {
      "epoch": 1.582429235663594,
      "grad_norm": 0.006130042485892773,
      "learning_rate": 0.2585435983743538,
      "loss": 0.3702,
      "num_input_tokens_seen": 18403744,
      "step": 9700
    },
    {
      "epoch": 1.5832449628844114,
      "grad_norm": 0.0020275902934372425,
      "learning_rate": 0.2585029341994101,
      "loss": 0.344,
      "num_input_tokens_seen": 18413632,
      "step": 9705
    },
    {
      "epoch": 1.5840606901052288,
      "grad_norm": 0.0017953551141545177,
      "learning_rate": 0.2584622532919504,
      "loss": 0.3502,
      "num_input_tokens_seen": 18423120,
      "step": 9710
    },
    {
      "epoch": 1.5848764173260461,
      "grad_norm": 0.0016144238179549575,
      "learning_rate": 0.2584215556582482,
      "loss": 0.3637,
      "num_input_tokens_seen": 18432096,
      "step": 9715
    },
    {
      "epoch": 1.5856921445468637,
      "grad_norm": 0.0010383316548541188,
      "learning_rate": 0.25838084130457967,
      "loss": 0.3337,
      "num_input_tokens_seen": 18442160,
      "step": 9720
    },
    {
      "epoch": 1.5865078717676808,
      "grad_norm": 0.0007922188960947096,
      "learning_rate": 0.2583401102372234,
      "loss": 0.3358,
      "num_input_tokens_seen": 18450672,
      "step": 9725
    },
    {
      "epoch": 1.5873235989884984,
      "grad_norm": 0.0011821332154795527,
      "learning_rate": 0.2582993624624606,
      "loss": 0.3295,
      "num_input_tokens_seen": 18459312,
      "step": 9730
    },
    {
      "epoch": 1.5881393262093155,
      "grad_norm": 0.0012048491043969989,
      "learning_rate": 0.25825859798657513,
      "loss": 0.3881,
      "num_input_tokens_seen": 18468192,
      "step": 9735
    },
    {
      "epoch": 1.588955053430133,
      "grad_norm": 0.0008437793585471809,
      "learning_rate": 0.25821781681585343,
      "loss": 0.3494,
      "num_input_tokens_seen": 18478992,
      "step": 9740
    },
    {
      "epoch": 1.5897707806509502,
      "grad_norm": 0.0013338583521544933,
      "learning_rate": 0.2581770189565844,
      "loss": 0.3679,
      "num_input_tokens_seen": 18489344,
      "step": 9745
    },
    {
      "epoch": 1.5905865078717678,
      "grad_norm": 0.0007454873993992805,
      "learning_rate": 0.25813620441505963,
      "loss": 0.3696,
      "num_input_tokens_seen": 18497952,
      "step": 9750
    },
    {
      "epoch": 1.591402235092585,
      "grad_norm": 0.0023769319523125887,
      "learning_rate": 0.2580953731975732,
      "loss": 0.3358,
      "num_input_tokens_seen": 18507264,
      "step": 9755
    },
    {
      "epoch": 1.5922179623134025,
      "grad_norm": 0.0017434301553294063,
      "learning_rate": 0.2580545253104218,
      "loss": 0.3453,
      "num_input_tokens_seen": 18517648,
      "step": 9760
    },
    {
      "epoch": 1.5930336895342196,
      "grad_norm": 0.0009764518472366035,
      "learning_rate": 0.2580136607599047,
      "loss": 0.3779,
      "num_input_tokens_seen": 18526704,
      "step": 9765
    },
    {
      "epoch": 1.5938494167550372,
      "grad_norm": 0.0007353375549428165,
      "learning_rate": 0.2579727795523238,
      "loss": 0.3668,
      "num_input_tokens_seen": 18535888,
      "step": 9770
    },
    {
      "epoch": 1.5946651439758543,
      "grad_norm": 0.0007072296575643122,
      "learning_rate": 0.25793188169398334,
      "loss": 0.3569,
      "num_input_tokens_seen": 18545280,
      "step": 9775
    },
    {
      "epoch": 1.595480871196672,
      "grad_norm": 0.0010342407040297985,
      "learning_rate": 0.25789096719119037,
      "loss": 0.3575,
      "num_input_tokens_seen": 18555056,
      "step": 9780
    },
    {
      "epoch": 1.596296598417489,
      "grad_norm": 0.001783470856025815,
      "learning_rate": 0.2578500360502544,
      "loss": 0.3264,
      "num_input_tokens_seen": 18564368,
      "step": 9785
    },
    {
      "epoch": 1.5971123256383066,
      "grad_norm": 0.000646111264359206,
      "learning_rate": 0.2578090882774876,
      "loss": 0.3525,
      "num_input_tokens_seen": 18575008,
      "step": 9790
    },
    {
      "epoch": 1.5979280528591238,
      "grad_norm": 0.0011131232604384422,
      "learning_rate": 0.25776812387920456,
      "loss": 0.3264,
      "num_input_tokens_seen": 18583328,
      "step": 9795
    },
    {
      "epoch": 1.5987437800799413,
      "grad_norm": 0.0009791508782655,
      "learning_rate": 0.2577271428617225,
      "loss": 0.3255,
      "num_input_tokens_seen": 18592816,
      "step": 9800
    },
    {
      "epoch": 1.5987437800799413,
      "eval_loss": 0.3486329913139343,
      "eval_runtime": 152.6291,
      "eval_samples_per_second": 17.854,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 18592816,
      "step": 9800
    },
    {
      "epoch": 1.5995595073007585,
      "grad_norm": 0.0005781327490694821,
      "learning_rate": 0.25768614523136124,
      "loss": 0.3181,
      "num_input_tokens_seen": 18603504,
      "step": 9805
    },
    {
      "epoch": 1.600375234521576,
      "grad_norm": 0.0013657616218551993,
      "learning_rate": 0.25764513099444314,
      "loss": 0.3492,
      "num_input_tokens_seen": 18612992,
      "step": 9810
    },
    {
      "epoch": 1.6011909617423934,
      "grad_norm": 0.0019382814643904567,
      "learning_rate": 0.25760410015729307,
      "loss": 0.3418,
      "num_input_tokens_seen": 18623152,
      "step": 9815
    },
    {
      "epoch": 1.6020066889632107,
      "grad_norm": 0.0022751898504793644,
      "learning_rate": 0.2575630527262385,
      "loss": 0.3469,
      "num_input_tokens_seen": 18631696,
      "step": 9820
    },
    {
      "epoch": 1.602822416184028,
      "grad_norm": 0.0014714637072756886,
      "learning_rate": 0.25752198870760945,
      "loss": 0.4046,
      "num_input_tokens_seen": 18641376,
      "step": 9825
    },
    {
      "epoch": 1.6036381434048455,
      "grad_norm": 0.0018879863200709224,
      "learning_rate": 0.2574809081077386,
      "loss": 0.3609,
      "num_input_tokens_seen": 18651344,
      "step": 9830
    },
    {
      "epoch": 1.6044538706256628,
      "grad_norm": 0.001151335658505559,
      "learning_rate": 0.257439810932961,
      "loss": 0.3456,
      "num_input_tokens_seen": 18660880,
      "step": 9835
    },
    {
      "epoch": 1.6052695978464802,
      "grad_norm": 0.0011593118542805314,
      "learning_rate": 0.2573986971896144,
      "loss": 0.3453,
      "num_input_tokens_seen": 18669744,
      "step": 9840
    },
    {
      "epoch": 1.6060853250672975,
      "grad_norm": 0.002605149522423744,
      "learning_rate": 0.257357566884039,
      "loss": 0.3242,
      "num_input_tokens_seen": 18679776,
      "step": 9845
    },
    {
      "epoch": 1.6069010522881149,
      "grad_norm": 0.002560513559728861,
      "learning_rate": 0.25731642002257765,
      "loss": 0.3771,
      "num_input_tokens_seen": 18690128,
      "step": 9850
    },
    {
      "epoch": 1.6077167795089322,
      "grad_norm": 0.0015195385785773396,
      "learning_rate": 0.25727525661157574,
      "loss": 0.2741,
      "num_input_tokens_seen": 18699568,
      "step": 9855
    },
    {
      "epoch": 1.6085325067297496,
      "grad_norm": 0.0018480097642168403,
      "learning_rate": 0.2572340766573811,
      "loss": 0.3125,
      "num_input_tokens_seen": 18708832,
      "step": 9860
    },
    {
      "epoch": 1.609348233950567,
      "grad_norm": 0.002730721840634942,
      "learning_rate": 0.25719288016634434,
      "loss": 0.3098,
      "num_input_tokens_seen": 18717376,
      "step": 9865
    },
    {
      "epoch": 1.6101639611713843,
      "grad_norm": 0.0017399609787389636,
      "learning_rate": 0.25715166714481835,
      "loss": 0.2934,
      "num_input_tokens_seen": 18727168,
      "step": 9870
    },
    {
      "epoch": 1.6109796883922016,
      "grad_norm": 0.00258208648301661,
      "learning_rate": 0.2571104375991587,
      "loss": 0.305,
      "num_input_tokens_seen": 18737056,
      "step": 9875
    },
    {
      "epoch": 1.611795415613019,
      "grad_norm": 0.0011246050707995892,
      "learning_rate": 0.2570691915357236,
      "loss": 0.2371,
      "num_input_tokens_seen": 18745264,
      "step": 9880
    },
    {
      "epoch": 1.6126111428338363,
      "grad_norm": 0.005805058404803276,
      "learning_rate": 0.2570279289608736,
      "loss": 0.5208,
      "num_input_tokens_seen": 18754576,
      "step": 9885
    },
    {
      "epoch": 1.6134268700546537,
      "grad_norm": 0.000771152030210942,
      "learning_rate": 0.256986649880972,
      "loss": 0.3149,
      "num_input_tokens_seen": 18765440,
      "step": 9890
    },
    {
      "epoch": 1.614242597275471,
      "grad_norm": 0.001102885347791016,
      "learning_rate": 0.25694535430238447,
      "loss": 0.3315,
      "num_input_tokens_seen": 18775200,
      "step": 9895
    },
    {
      "epoch": 1.6150583244962884,
      "grad_norm": 0.0018569750245660543,
      "learning_rate": 0.25690404223147933,
      "loss": 0.3889,
      "num_input_tokens_seen": 18784496,
      "step": 9900
    },
    {
      "epoch": 1.615874051717106,
      "grad_norm": 0.0038089239969849586,
      "learning_rate": 0.2568627136746275,
      "loss": 0.3331,
      "num_input_tokens_seen": 18793344,
      "step": 9905
    },
    {
      "epoch": 1.616689778937923,
      "grad_norm": 0.0008534340304322541,
      "learning_rate": 0.25682136863820226,
      "loss": 0.3462,
      "num_input_tokens_seen": 18801952,
      "step": 9910
    },
    {
      "epoch": 1.6175055061587407,
      "grad_norm": 0.0007792014512233436,
      "learning_rate": 0.25678000712857957,
      "loss": 0.3641,
      "num_input_tokens_seen": 18812064,
      "step": 9915
    },
    {
      "epoch": 1.6183212333795578,
      "grad_norm": 0.0014902300899848342,
      "learning_rate": 0.2567386291521379,
      "loss": 0.3493,
      "num_input_tokens_seen": 18820704,
      "step": 9920
    },
    {
      "epoch": 1.6191369606003754,
      "grad_norm": 0.0007205671863630414,
      "learning_rate": 0.2566972347152583,
      "loss": 0.3601,
      "num_input_tokens_seen": 18830272,
      "step": 9925
    },
    {
      "epoch": 1.6199526878211925,
      "grad_norm": 0.0009256329503841698,
      "learning_rate": 0.2566558238243242,
      "loss": 0.4171,
      "num_input_tokens_seen": 18839872,
      "step": 9930
    },
    {
      "epoch": 1.62076841504201,
      "grad_norm": 0.0004612957709468901,
      "learning_rate": 0.25661439648572176,
      "loss": 0.3334,
      "num_input_tokens_seen": 18850144,
      "step": 9935
    },
    {
      "epoch": 1.6215841422628272,
      "grad_norm": 0.001106973853893578,
      "learning_rate": 0.25657295270583963,
      "loss": 0.3543,
      "num_input_tokens_seen": 18859424,
      "step": 9940
    },
    {
      "epoch": 1.6223998694836448,
      "grad_norm": 0.0019582442473620176,
      "learning_rate": 0.25653149249106894,
      "loss": 0.3664,
      "num_input_tokens_seen": 18869088,
      "step": 9945
    },
    {
      "epoch": 1.623215596704462,
      "grad_norm": 0.0015727546997368336,
      "learning_rate": 0.25649001584780323,
      "loss": 0.3602,
      "num_input_tokens_seen": 18878144,
      "step": 9950
    },
    {
      "epoch": 1.6240313239252795,
      "grad_norm": 0.0014534448273479939,
      "learning_rate": 0.2564485227824389,
      "loss": 0.3512,
      "num_input_tokens_seen": 18887328,
      "step": 9955
    },
    {
      "epoch": 1.6248470511460966,
      "grad_norm": 0.001147491275332868,
      "learning_rate": 0.25640701330137466,
      "loss": 0.323,
      "num_input_tokens_seen": 18895968,
      "step": 9960
    },
    {
      "epoch": 1.6256627783669142,
      "grad_norm": 0.0018099518492817879,
      "learning_rate": 0.2563654874110117,
      "loss": 0.3383,
      "num_input_tokens_seen": 18904672,
      "step": 9965
    },
    {
      "epoch": 1.6264785055877313,
      "grad_norm": 0.001372889499180019,
      "learning_rate": 0.256323945117754,
      "loss": 0.3887,
      "num_input_tokens_seen": 18914480,
      "step": 9970
    },
    {
      "epoch": 1.627294232808549,
      "grad_norm": 0.0027212672866880894,
      "learning_rate": 0.2562823864280078,
      "loss": 0.3213,
      "num_input_tokens_seen": 18924448,
      "step": 9975
    },
    {
      "epoch": 1.628109960029366,
      "grad_norm": 0.001132135046645999,
      "learning_rate": 0.25624081134818194,
      "loss": 0.3459,
      "num_input_tokens_seen": 18933888,
      "step": 9980
    },
    {
      "epoch": 1.6289256872501836,
      "grad_norm": 0.002126250183209777,
      "learning_rate": 0.2561992198846879,
      "loss": 0.3072,
      "num_input_tokens_seen": 18943472,
      "step": 9985
    },
    {
      "epoch": 1.6297414144710007,
      "grad_norm": 0.0016068648546934128,
      "learning_rate": 0.25615761204393955,
      "loss": 0.4276,
      "num_input_tokens_seen": 18952624,
      "step": 9990
    },
    {
      "epoch": 1.6305571416918183,
      "grad_norm": 0.0016465181251987815,
      "learning_rate": 0.2561159878323534,
      "loss": 0.3947,
      "num_input_tokens_seen": 18963456,
      "step": 9995
    },
    {
      "epoch": 1.6313728689126357,
      "grad_norm": 0.0008085024892352521,
      "learning_rate": 0.2560743472563483,
      "loss": 0.3551,
      "num_input_tokens_seen": 18972864,
      "step": 10000
    },
    {
      "epoch": 1.6313728689126357,
      "eval_loss": 0.3563860356807709,
      "eval_runtime": 152.6552,
      "eval_samples_per_second": 17.851,
      "eval_steps_per_second": 8.929,
      "num_input_tokens_seen": 18972864,
      "step": 10000
    },
    {
      "epoch": 1.632188596133453,
      "grad_norm": 0.0012687687994912267,
      "learning_rate": 0.25603269032234593,
      "loss": 0.3265,
      "num_input_tokens_seen": 18982160,
      "step": 10005
    },
    {
      "epoch": 1.6330043233542704,
      "grad_norm": 0.00232927268370986,
      "learning_rate": 0.2559910170367702,
      "loss": 0.3663,
      "num_input_tokens_seen": 18992384,
      "step": 10010
    },
    {
      "epoch": 1.6338200505750877,
      "grad_norm": 0.0011618505232036114,
      "learning_rate": 0.2559493274060477,
      "loss": 0.3494,
      "num_input_tokens_seen": 19002448,
      "step": 10015
    },
    {
      "epoch": 1.634635777795905,
      "grad_norm": 0.0018417385872453451,
      "learning_rate": 0.2559076214366074,
      "loss": 0.3438,
      "num_input_tokens_seen": 19012416,
      "step": 10020
    },
    {
      "epoch": 1.6354515050167224,
      "grad_norm": 0.0005660930182784796,
      "learning_rate": 0.25586589913488106,
      "loss": 0.3224,
      "num_input_tokens_seen": 19021472,
      "step": 10025
    },
    {
      "epoch": 1.6362672322375398,
      "grad_norm": 0.0014998591504991055,
      "learning_rate": 0.2558241605073026,
      "loss": 0.3482,
      "num_input_tokens_seen": 19030720,
      "step": 10030
    },
    {
      "epoch": 1.6370829594583571,
      "grad_norm": 0.001308815903030336,
      "learning_rate": 0.25578240556030873,
      "loss": 0.3227,
      "num_input_tokens_seen": 19041136,
      "step": 10035
    },
    {
      "epoch": 1.6378986866791745,
      "grad_norm": 0.0010830140672624111,
      "learning_rate": 0.2557406343003386,
      "loss": 0.3545,
      "num_input_tokens_seen": 19049696,
      "step": 10040
    },
    {
      "epoch": 1.6387144138999918,
      "grad_norm": 0.0013390925014391541,
      "learning_rate": 0.25569884673383375,
      "loss": 0.3394,
      "num_input_tokens_seen": 19059504,
      "step": 10045
    },
    {
      "epoch": 1.6395301411208092,
      "grad_norm": 0.0008630426018498838,
      "learning_rate": 0.25565704286723856,
      "loss": 0.3019,
      "num_input_tokens_seen": 19068480,
      "step": 10050
    },
    {
      "epoch": 1.6403458683416265,
      "grad_norm": 0.001178442151285708,
      "learning_rate": 0.25561522270699955,
      "loss": 0.3406,
      "num_input_tokens_seen": 19078560,
      "step": 10055
    },
    {
      "epoch": 1.641161595562444,
      "grad_norm": 0.00046957129961811006,
      "learning_rate": 0.25557338625956594,
      "loss": 0.3581,
      "num_input_tokens_seen": 19088080,
      "step": 10060
    },
    {
      "epoch": 1.6419773227832613,
      "grad_norm": 0.0008783351513557136,
      "learning_rate": 0.25553153353138947,
      "loss": 0.3362,
      "num_input_tokens_seen": 19097136,
      "step": 10065
    },
    {
      "epoch": 1.6427930500040786,
      "grad_norm": 0.001203106832690537,
      "learning_rate": 0.2554896645289243,
      "loss": 0.3332,
      "num_input_tokens_seen": 19105680,
      "step": 10070
    },
    {
      "epoch": 1.643608777224896,
      "grad_norm": 0.0011641817400231957,
      "learning_rate": 0.2554477792586272,
      "loss": 0.2932,
      "num_input_tokens_seen": 19114240,
      "step": 10075
    },
    {
      "epoch": 1.6444245044457133,
      "grad_norm": 0.001467230380512774,
      "learning_rate": 0.25540587772695744,
      "loss": 0.3898,
      "num_input_tokens_seen": 19124464,
      "step": 10080
    },
    {
      "epoch": 1.6452402316665307,
      "grad_norm": 0.0006956282304599881,
      "learning_rate": 0.2553639599403767,
      "loss": 0.391,
      "num_input_tokens_seen": 19134112,
      "step": 10085
    },
    {
      "epoch": 1.6460559588873482,
      "grad_norm": 0.001179989892989397,
      "learning_rate": 0.2553220259053493,
      "loss": 0.3322,
      "num_input_tokens_seen": 19143952,
      "step": 10090
    },
    {
      "epoch": 1.6468716861081654,
      "grad_norm": 0.0005524778971448541,
      "learning_rate": 0.2552800756283419,
      "loss": 0.3363,
      "num_input_tokens_seen": 19153632,
      "step": 10095
    },
    {
      "epoch": 1.647687413328983,
      "grad_norm": 0.0015575053403154016,
      "learning_rate": 0.25523810911582373,
      "loss": 0.3462,
      "num_input_tokens_seen": 19163568,
      "step": 10100
    },
    {
      "epoch": 1.6485031405498,
      "grad_norm": 0.0026931804604828358,
      "learning_rate": 0.25519612637426675,
      "loss": 0.3102,
      "num_input_tokens_seen": 19174128,
      "step": 10105
    },
    {
      "epoch": 1.6493188677706176,
      "grad_norm": 0.0017979848198592663,
      "learning_rate": 0.25515412741014504,
      "loss": 0.3325,
      "num_input_tokens_seen": 19183680,
      "step": 10110
    },
    {
      "epoch": 1.6501345949914348,
      "grad_norm": 0.0023531820625066757,
      "learning_rate": 0.2551121122299355,
      "loss": 0.3191,
      "num_input_tokens_seen": 19194272,
      "step": 10115
    },
    {
      "epoch": 1.6509503222122524,
      "grad_norm": 0.00193055032286793,
      "learning_rate": 0.2550700808401173,
      "loss": 0.3535,
      "num_input_tokens_seen": 19203024,
      "step": 10120
    },
    {
      "epoch": 1.6517660494330695,
      "grad_norm": 0.005255008582025766,
      "learning_rate": 0.2550280332471722,
      "loss": 0.3658,
      "num_input_tokens_seen": 19213536,
      "step": 10125
    },
    {
      "epoch": 1.652581776653887,
      "grad_norm": 0.0013082253281027079,
      "learning_rate": 0.2549859694575845,
      "loss": 0.3619,
      "num_input_tokens_seen": 19222752,
      "step": 10130
    },
    {
      "epoch": 1.6533975038747042,
      "grad_norm": 0.0008838636567816138,
      "learning_rate": 0.254943889477841,
      "loss": 0.3272,
      "num_input_tokens_seen": 19232336,
      "step": 10135
    },
    {
      "epoch": 1.6542132310955218,
      "grad_norm": 0.0016898653702810407,
      "learning_rate": 0.25490179331443097,
      "loss": 0.3725,
      "num_input_tokens_seen": 19241936,
      "step": 10140
    },
    {
      "epoch": 1.655028958316339,
      "grad_norm": 0.0025630472227931023,
      "learning_rate": 0.25485968097384615,
      "loss": 0.3421,
      "num_input_tokens_seen": 19251888,
      "step": 10145
    },
    {
      "epoch": 1.6558446855371565,
      "grad_norm": 0.0010777278803288937,
      "learning_rate": 0.25481755246258075,
      "loss": 0.311,
      "num_input_tokens_seen": 19261120,
      "step": 10150
    },
    {
      "epoch": 1.6566604127579736,
      "grad_norm": 0.001514066243544221,
      "learning_rate": 0.2547754077871315,
      "loss": 0.3899,
      "num_input_tokens_seen": 19270176,
      "step": 10155
    },
    {
      "epoch": 1.6574761399787912,
      "grad_norm": 0.0022943129297345877,
      "learning_rate": 0.25473324695399774,
      "loss": 0.322,
      "num_input_tokens_seen": 19279584,
      "step": 10160
    },
    {
      "epoch": 1.6582918671996083,
      "grad_norm": 0.0009528041700832546,
      "learning_rate": 0.25469106996968105,
      "loss": 0.3553,
      "num_input_tokens_seen": 19288496,
      "step": 10165
    },
    {
      "epoch": 1.6591075944204259,
      "grad_norm": 0.0008121443097479641,
      "learning_rate": 0.2546488768406858,
      "loss": 0.3542,
      "num_input_tokens_seen": 19297088,
      "step": 10170
    },
    {
      "epoch": 1.659923321641243,
      "grad_norm": 0.0018763300031423569,
      "learning_rate": 0.25460666757351863,
      "loss": 0.3684,
      "num_input_tokens_seen": 19306688,
      "step": 10175
    },
    {
      "epoch": 1.6607390488620606,
      "grad_norm": 0.0010827442165464163,
      "learning_rate": 0.25456444217468877,
      "loss": 0.314,
      "num_input_tokens_seen": 19315184,
      "step": 10180
    },
    {
      "epoch": 1.661554776082878,
      "grad_norm": 0.0034177375491708517,
      "learning_rate": 0.25452220065070785,
      "loss": 0.3885,
      "num_input_tokens_seen": 19323648,
      "step": 10185
    },
    {
      "epoch": 1.6623705033036953,
      "grad_norm": 0.002107700565829873,
      "learning_rate": 0.2544799430080901,
      "loss": 0.331,
      "num_input_tokens_seen": 19331616,
      "step": 10190
    },
    {
      "epoch": 1.6631862305245126,
      "grad_norm": 0.002891644835472107,
      "learning_rate": 0.2544376692533522,
      "loss": 0.341,
      "num_input_tokens_seen": 19340960,
      "step": 10195
    },
    {
      "epoch": 1.66400195774533,
      "grad_norm": 0.0015337757067754865,
      "learning_rate": 0.2543953793930132,
      "loss": 0.3223,
      "num_input_tokens_seen": 19350160,
      "step": 10200
    },
    {
      "epoch": 1.66400195774533,
      "eval_loss": 0.33092600107192993,
      "eval_runtime": 152.5982,
      "eval_samples_per_second": 17.857,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 19350160,
      "step": 10200
    },
    {
      "epoch": 1.6648176849661473,
      "grad_norm": 0.002200504532083869,
      "learning_rate": 0.2543530734335948,
      "loss": 0.2797,
      "num_input_tokens_seen": 19360160,
      "step": 10205
    },
    {
      "epoch": 1.6656334121869647,
      "grad_norm": 0.0035833835136145353,
      "learning_rate": 0.2543107513816211,
      "loss": 0.4834,
      "num_input_tokens_seen": 19369696,
      "step": 10210
    },
    {
      "epoch": 1.666449139407782,
      "grad_norm": 0.001313549466431141,
      "learning_rate": 0.25426841324361865,
      "loss": 0.33,
      "num_input_tokens_seen": 19379152,
      "step": 10215
    },
    {
      "epoch": 1.6672648666285994,
      "grad_norm": 0.0032093680929392576,
      "learning_rate": 0.2542260590261166,
      "loss": 0.3432,
      "num_input_tokens_seen": 19389616,
      "step": 10220
    },
    {
      "epoch": 1.6680805938494168,
      "grad_norm": 0.0010783199686557055,
      "learning_rate": 0.2541836887356465,
      "loss": 0.3681,
      "num_input_tokens_seen": 19398848,
      "step": 10225
    },
    {
      "epoch": 1.6688963210702341,
      "grad_norm": 0.001805326552130282,
      "learning_rate": 0.2541413023787423,
      "loss": 0.3724,
      "num_input_tokens_seen": 19408576,
      "step": 10230
    },
    {
      "epoch": 1.6697120482910515,
      "grad_norm": 0.0017785239033401012,
      "learning_rate": 0.2540988999619405,
      "loss": 0.359,
      "num_input_tokens_seen": 19417824,
      "step": 10235
    },
    {
      "epoch": 1.6705277755118688,
      "grad_norm": 0.0007933134911581874,
      "learning_rate": 0.25405648149178023,
      "loss": 0.3129,
      "num_input_tokens_seen": 19427776,
      "step": 10240
    },
    {
      "epoch": 1.6713435027326862,
      "grad_norm": 0.0018825592705979943,
      "learning_rate": 0.2540140469748028,
      "loss": 0.3228,
      "num_input_tokens_seen": 19438448,
      "step": 10245
    },
    {
      "epoch": 1.6721592299535035,
      "grad_norm": 0.0012346669100224972,
      "learning_rate": 0.25397159641755224,
      "loss": 0.3481,
      "num_input_tokens_seen": 19446592,
      "step": 10250
    },
    {
      "epoch": 1.6729749571743209,
      "grad_norm": 0.0017453690525144339,
      "learning_rate": 0.2539291298265749,
      "loss": 0.3569,
      "num_input_tokens_seen": 19456672,
      "step": 10255
    },
    {
      "epoch": 1.6737906843951382,
      "grad_norm": 0.0013370091328397393,
      "learning_rate": 0.2538866472084197,
      "loss": 0.3139,
      "num_input_tokens_seen": 19466960,
      "step": 10260
    },
    {
      "epoch": 1.6746064116159556,
      "grad_norm": 0.0013620206154882908,
      "learning_rate": 0.25384414856963794,
      "loss": 0.3967,
      "num_input_tokens_seen": 19476688,
      "step": 10265
    },
    {
      "epoch": 1.675422138836773,
      "grad_norm": 0.0010741313453763723,
      "learning_rate": 0.25380163391678356,
      "loss": 0.4348,
      "num_input_tokens_seen": 19486592,
      "step": 10270
    },
    {
      "epoch": 1.6762378660575903,
      "grad_norm": 0.0026609802152961493,
      "learning_rate": 0.2537591032564127,
      "loss": 0.3737,
      "num_input_tokens_seen": 19496128,
      "step": 10275
    },
    {
      "epoch": 1.6770535932784076,
      "grad_norm": 0.0021290213335305452,
      "learning_rate": 0.25371655659508424,
      "loss": 0.3869,
      "num_input_tokens_seen": 19506640,
      "step": 10280
    },
    {
      "epoch": 1.6778693204992252,
      "grad_norm": 0.0014691922115162015,
      "learning_rate": 0.25367399393935935,
      "loss": 0.3708,
      "num_input_tokens_seen": 19515648,
      "step": 10285
    },
    {
      "epoch": 1.6786850477200423,
      "grad_norm": 0.001254695700481534,
      "learning_rate": 0.25363141529580174,
      "loss": 0.3774,
      "num_input_tokens_seen": 19524400,
      "step": 10290
    },
    {
      "epoch": 1.67950077494086,
      "grad_norm": 0.0015642349608242512,
      "learning_rate": 0.2535888206709776,
      "loss": 0.3472,
      "num_input_tokens_seen": 19534608,
      "step": 10295
    },
    {
      "epoch": 1.680316502161677,
      "grad_norm": 0.0014373527374118567,
      "learning_rate": 0.2535462100714555,
      "loss": 0.3367,
      "num_input_tokens_seen": 19543936,
      "step": 10300
    },
    {
      "epoch": 1.6811322293824946,
      "grad_norm": 0.0011492334306240082,
      "learning_rate": 0.2535035835038066,
      "loss": 0.37,
      "num_input_tokens_seen": 19553200,
      "step": 10305
    },
    {
      "epoch": 1.6819479566033118,
      "grad_norm": 0.0017187432385981083,
      "learning_rate": 0.2534609409746044,
      "loss": 0.3687,
      "num_input_tokens_seen": 19562736,
      "step": 10310
    },
    {
      "epoch": 1.6827636838241293,
      "grad_norm": 0.001300085335969925,
      "learning_rate": 0.253418282490425,
      "loss": 0.3298,
      "num_input_tokens_seen": 19572016,
      "step": 10315
    },
    {
      "epoch": 1.6835794110449465,
      "grad_norm": 0.0034979660995304585,
      "learning_rate": 0.2533756080578467,
      "loss": 0.3387,
      "num_input_tokens_seen": 19581888,
      "step": 10320
    },
    {
      "epoch": 1.684395138265764,
      "grad_norm": 0.0019336057594045997,
      "learning_rate": 0.25333291768345056,
      "loss": 0.4103,
      "num_input_tokens_seen": 19591264,
      "step": 10325
    },
    {
      "epoch": 1.6852108654865812,
      "grad_norm": 0.0007693648803979158,
      "learning_rate": 0.25329021137381996,
      "loss": 0.3481,
      "num_input_tokens_seen": 19599568,
      "step": 10330
    },
    {
      "epoch": 1.6860265927073987,
      "grad_norm": 0.0015199016779661179,
      "learning_rate": 0.25324748913554074,
      "loss": 0.351,
      "num_input_tokens_seen": 19609504,
      "step": 10335
    },
    {
      "epoch": 1.6868423199282159,
      "grad_norm": 0.0007652914500795305,
      "learning_rate": 0.2532047509752013,
      "loss": 0.3442,
      "num_input_tokens_seen": 19617536,
      "step": 10340
    },
    {
      "epoch": 1.6876580471490334,
      "grad_norm": 0.0007692622020840645,
      "learning_rate": 0.25316199689939217,
      "loss": 0.3504,
      "num_input_tokens_seen": 19627936,
      "step": 10345
    },
    {
      "epoch": 1.6884737743698506,
      "grad_norm": 0.0014389563584700227,
      "learning_rate": 0.2531192269147068,
      "loss": 0.347,
      "num_input_tokens_seen": 19637552,
      "step": 10350
    },
    {
      "epoch": 1.6892895015906682,
      "grad_norm": 0.002770890248939395,
      "learning_rate": 0.2530764410277407,
      "loss": 0.3305,
      "num_input_tokens_seen": 19647424,
      "step": 10355
    },
    {
      "epoch": 1.6901052288114853,
      "grad_norm": 0.0025579375214874744,
      "learning_rate": 0.25303363924509203,
      "loss": 0.3203,
      "num_input_tokens_seen": 19657664,
      "step": 10360
    },
    {
      "epoch": 1.6909209560323029,
      "grad_norm": 0.0028479970060288906,
      "learning_rate": 0.25299082157336145,
      "loss": 0.3768,
      "num_input_tokens_seen": 19666432,
      "step": 10365
    },
    {
      "epoch": 1.6917366832531202,
      "grad_norm": 0.0016595585038885474,
      "learning_rate": 0.2529479880191519,
      "loss": 0.3457,
      "num_input_tokens_seen": 19675600,
      "step": 10370
    },
    {
      "epoch": 1.6925524104739376,
      "grad_norm": 0.0023920938838273287,
      "learning_rate": 0.2529051385890689,
      "loss": 0.3581,
      "num_input_tokens_seen": 19685216,
      "step": 10375
    },
    {
      "epoch": 1.693368137694755,
      "grad_norm": 0.004526253789663315,
      "learning_rate": 0.2528622732897203,
      "loss": 0.3499,
      "num_input_tokens_seen": 19694544,
      "step": 10380
    },
    {
      "epoch": 1.6941838649155723,
      "grad_norm": 0.002634921343997121,
      "learning_rate": 0.25281939212771654,
      "loss": 0.3227,
      "num_input_tokens_seen": 19704864,
      "step": 10385
    },
    {
      "epoch": 1.6949995921363896,
      "grad_norm": 0.0014804811216890812,
      "learning_rate": 0.2527764951096704,
      "loss": 0.39,
      "num_input_tokens_seen": 19714992,
      "step": 10390
    },
    {
      "epoch": 1.695815319357207,
      "grad_norm": 0.001020815339870751,
      "learning_rate": 0.2527335822421971,
      "loss": 0.3515,
      "num_input_tokens_seen": 19724784,
      "step": 10395
    },
    {
      "epoch": 1.6966310465780243,
      "grad_norm": 0.00273367902263999,
      "learning_rate": 0.25269065353191444,
      "loss": 0.3489,
      "num_input_tokens_seen": 19735024,
      "step": 10400
    },
    {
      "epoch": 1.6966310465780243,
      "eval_loss": 0.3512588143348694,
      "eval_runtime": 152.5947,
      "eval_samples_per_second": 17.858,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 19735024,
      "step": 10400
    },
    {
      "epoch": 1.6974467737988417,
      "grad_norm": 0.002295771846547723,
      "learning_rate": 0.2526477089854425,
      "loss": 0.3371,
      "num_input_tokens_seen": 19743360,
      "step": 10405
    },
    {
      "epoch": 1.698262501019659,
      "grad_norm": 0.0005873687914572656,
      "learning_rate": 0.25260474860940385,
      "loss": 0.3402,
      "num_input_tokens_seen": 19751760,
      "step": 10410
    },
    {
      "epoch": 1.6990782282404764,
      "grad_norm": 0.0017323531210422516,
      "learning_rate": 0.2525617724104236,
      "loss": 0.3086,
      "num_input_tokens_seen": 19760416,
      "step": 10415
    },
    {
      "epoch": 1.6998939554612937,
      "grad_norm": 0.002359993290156126,
      "learning_rate": 0.25251878039512915,
      "loss": 0.3688,
      "num_input_tokens_seen": 19769808,
      "step": 10420
    },
    {
      "epoch": 1.700709682682111,
      "grad_norm": 0.003161567496135831,
      "learning_rate": 0.25247577257015047,
      "loss": 0.3696,
      "num_input_tokens_seen": 19779200,
      "step": 10425
    },
    {
      "epoch": 1.7015254099029284,
      "grad_norm": 0.0013405421050265431,
      "learning_rate": 0.2524327489421198,
      "loss": 0.307,
      "num_input_tokens_seen": 19789712,
      "step": 10430
    },
    {
      "epoch": 1.7023411371237458,
      "grad_norm": 0.00393724488094449,
      "learning_rate": 0.25238970951767203,
      "loss": 0.3572,
      "num_input_tokens_seen": 19799808,
      "step": 10435
    },
    {
      "epoch": 1.7031568643445631,
      "grad_norm": 0.0015944726765155792,
      "learning_rate": 0.25234665430344433,
      "loss": 0.348,
      "num_input_tokens_seen": 19809760,
      "step": 10440
    },
    {
      "epoch": 1.7039725915653805,
      "grad_norm": 0.0013926830142736435,
      "learning_rate": 0.2523035833060764,
      "loss": 0.3324,
      "num_input_tokens_seen": 19819200,
      "step": 10445
    },
    {
      "epoch": 1.7047883187861979,
      "grad_norm": 0.0014034499181434512,
      "learning_rate": 0.2522604965322103,
      "loss": 0.3228,
      "num_input_tokens_seen": 19829744,
      "step": 10450
    },
    {
      "epoch": 1.7056040460070152,
      "grad_norm": 0.0017318383324891329,
      "learning_rate": 0.25221739398849047,
      "loss": 0.3286,
      "num_input_tokens_seen": 19839504,
      "step": 10455
    },
    {
      "epoch": 1.7064197732278326,
      "grad_norm": 0.0011984824668616056,
      "learning_rate": 0.252174275681564,
      "loss": 0.3834,
      "num_input_tokens_seen": 19848176,
      "step": 10460
    },
    {
      "epoch": 1.70723550044865,
      "grad_norm": 0.0023799885530024767,
      "learning_rate": 0.2521311416180802,
      "loss": 0.3285,
      "num_input_tokens_seen": 19856288,
      "step": 10465
    },
    {
      "epoch": 1.7080512276694675,
      "grad_norm": 0.0009999092435464263,
      "learning_rate": 0.25208799180469094,
      "loss": 0.3475,
      "num_input_tokens_seen": 19865808,
      "step": 10470
    },
    {
      "epoch": 1.7088669548902846,
      "grad_norm": 0.0007246713503263891,
      "learning_rate": 0.2520448262480504,
      "loss": 0.3565,
      "num_input_tokens_seen": 19874128,
      "step": 10475
    },
    {
      "epoch": 1.7096826821111022,
      "grad_norm": 0.0014207314234226942,
      "learning_rate": 0.25200164495481525,
      "loss": 0.3068,
      "num_input_tokens_seen": 19882864,
      "step": 10480
    },
    {
      "epoch": 1.7104984093319193,
      "grad_norm": 0.0013071062276139855,
      "learning_rate": 0.25195844793164474,
      "loss": 0.4091,
      "num_input_tokens_seen": 19892208,
      "step": 10485
    },
    {
      "epoch": 1.711314136552737,
      "grad_norm": 0.0010155192576348782,
      "learning_rate": 0.2519152351852001,
      "loss": 0.389,
      "num_input_tokens_seen": 19901856,
      "step": 10490
    },
    {
      "epoch": 1.712129863773554,
      "grad_norm": 0.000998414703644812,
      "learning_rate": 0.25187200672214555,
      "loss": 0.3664,
      "num_input_tokens_seen": 19911776,
      "step": 10495
    },
    {
      "epoch": 1.7129455909943716,
      "grad_norm": 0.0013968617422506213,
      "learning_rate": 0.2518287625491473,
      "loss": 0.3307,
      "num_input_tokens_seen": 19921088,
      "step": 10500
    },
    {
      "epoch": 1.7137613182151887,
      "grad_norm": 0.0008657986181788146,
      "learning_rate": 0.25178550267287425,
      "loss": 0.3373,
      "num_input_tokens_seen": 19929840,
      "step": 10505
    },
    {
      "epoch": 1.7145770454360063,
      "grad_norm": 0.0014805721584707499,
      "learning_rate": 0.2517422270999976,
      "loss": 0.3438,
      "num_input_tokens_seen": 19939168,
      "step": 10510
    },
    {
      "epoch": 1.7153927726568234,
      "grad_norm": 0.0023891364689916372,
      "learning_rate": 0.2516989358371909,
      "loss": 0.3214,
      "num_input_tokens_seen": 19948768,
      "step": 10515
    },
    {
      "epoch": 1.716208499877641,
      "grad_norm": 0.0010122864041477442,
      "learning_rate": 0.25165562889113025,
      "loss": 0.3576,
      "num_input_tokens_seen": 19957616,
      "step": 10520
    },
    {
      "epoch": 1.7170242270984581,
      "grad_norm": 0.0009224131936207414,
      "learning_rate": 0.2516123062684942,
      "loss": 0.3348,
      "num_input_tokens_seen": 19967216,
      "step": 10525
    },
    {
      "epoch": 1.7178399543192757,
      "grad_norm": 0.0029067504219710827,
      "learning_rate": 0.25156896797596356,
      "loss": 0.3268,
      "num_input_tokens_seen": 19977472,
      "step": 10530
    },
    {
      "epoch": 1.7186556815400928,
      "grad_norm": 0.0012433918891474605,
      "learning_rate": 0.2515256140202216,
      "loss": 0.3072,
      "num_input_tokens_seen": 19987904,
      "step": 10535
    },
    {
      "epoch": 1.7194714087609104,
      "grad_norm": 0.0012104081688448787,
      "learning_rate": 0.25148224440795425,
      "loss": 0.3813,
      "num_input_tokens_seen": 19997632,
      "step": 10540
    },
    {
      "epoch": 1.7202871359817276,
      "grad_norm": 0.001514956820756197,
      "learning_rate": 0.2514388591458494,
      "loss": 0.3755,
      "num_input_tokens_seen": 20008112,
      "step": 10545
    },
    {
      "epoch": 1.7211028632025451,
      "grad_norm": 0.002151055494323373,
      "learning_rate": 0.2513954582405977,
      "loss": 0.3498,
      "num_input_tokens_seen": 20017504,
      "step": 10550
    },
    {
      "epoch": 1.7219185904233625,
      "grad_norm": 0.0009850008646026254,
      "learning_rate": 0.2513520416988922,
      "loss": 0.339,
      "num_input_tokens_seen": 20026288,
      "step": 10555
    },
    {
      "epoch": 1.7227343176441798,
      "grad_norm": 0.0018650735728442669,
      "learning_rate": 0.2513086095274281,
      "loss": 0.3389,
      "num_input_tokens_seen": 20035968,
      "step": 10560
    },
    {
      "epoch": 1.7235500448649972,
      "grad_norm": 0.0009447772754356265,
      "learning_rate": 0.25126516173290336,
      "loss": 0.3485,
      "num_input_tokens_seen": 20044848,
      "step": 10565
    },
    {
      "epoch": 1.7243657720858145,
      "grad_norm": 0.0016464248765259981,
      "learning_rate": 0.2512216983220181,
      "loss": 0.3807,
      "num_input_tokens_seen": 20052288,
      "step": 10570
    },
    {
      "epoch": 1.725181499306632,
      "grad_norm": 0.003034319495782256,
      "learning_rate": 0.25117821930147494,
      "loss": 0.3707,
      "num_input_tokens_seen": 20061728,
      "step": 10575
    },
    {
      "epoch": 1.7259972265274492,
      "grad_norm": 0.0020909367594867945,
      "learning_rate": 0.2511347246779788,
      "loss": 0.3542,
      "num_input_tokens_seen": 20071680,
      "step": 10580
    },
    {
      "epoch": 1.7268129537482666,
      "grad_norm": 0.0016322002047672868,
      "learning_rate": 0.25109121445823723,
      "loss": 0.3193,
      "num_input_tokens_seen": 20081456,
      "step": 10585
    },
    {
      "epoch": 1.727628680969084,
      "grad_norm": 0.00266910414211452,
      "learning_rate": 0.25104768864896004,
      "loss": 0.3325,
      "num_input_tokens_seen": 20090512,
      "step": 10590
    },
    {
      "epoch": 1.7284444081899013,
      "grad_norm": 0.003031631466001272,
      "learning_rate": 0.2510041472568594,
      "loss": 0.3617,
      "num_input_tokens_seen": 20099552,
      "step": 10595
    },
    {
      "epoch": 1.7292601354107187,
      "grad_norm": 0.0016152323223650455,
      "learning_rate": 0.25096059028864987,
      "loss": 0.2748,
      "num_input_tokens_seen": 20108768,
      "step": 10600
    },
    {
      "epoch": 1.7292601354107187,
      "eval_loss": 0.3289847671985626,
      "eval_runtime": 152.6898,
      "eval_samples_per_second": 17.847,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 20108768,
      "step": 10600
    },
    {
      "epoch": 1.730075862631536,
      "grad_norm": 0.001691941055469215,
      "learning_rate": 0.25091701775104863,
      "loss": 0.3839,
      "num_input_tokens_seen": 20117600,
      "step": 10605
    },
    {
      "epoch": 1.7308915898523534,
      "grad_norm": 0.002781361574307084,
      "learning_rate": 0.250873429650775,
      "loss": 0.2914,
      "num_input_tokens_seen": 20127632,
      "step": 10610
    },
    {
      "epoch": 1.7317073170731707,
      "grad_norm": 0.001319479662925005,
      "learning_rate": 0.25082982599455095,
      "loss": 0.3916,
      "num_input_tokens_seen": 20135632,
      "step": 10615
    },
    {
      "epoch": 1.732523044293988,
      "grad_norm": 0.0012583803618326783,
      "learning_rate": 0.2507862067891006,
      "loss": 0.3301,
      "num_input_tokens_seen": 20144096,
      "step": 10620
    },
    {
      "epoch": 1.7333387715148054,
      "grad_norm": 0.0022410922683775425,
      "learning_rate": 0.25074257204115064,
      "loss": 0.3581,
      "num_input_tokens_seen": 20154288,
      "step": 10625
    },
    {
      "epoch": 1.7341544987356228,
      "grad_norm": 0.002037464175373316,
      "learning_rate": 0.25069892175742997,
      "loss": 0.3598,
      "num_input_tokens_seen": 20165824,
      "step": 10630
    },
    {
      "epoch": 1.7349702259564401,
      "grad_norm": 0.0005356563488021493,
      "learning_rate": 0.25065525594467014,
      "loss": 0.3381,
      "num_input_tokens_seen": 20176384,
      "step": 10635
    },
    {
      "epoch": 1.7357859531772575,
      "grad_norm": 0.0027210682164877653,
      "learning_rate": 0.2506115746096049,
      "loss": 0.3396,
      "num_input_tokens_seen": 20185664,
      "step": 10640
    },
    {
      "epoch": 1.7366016803980748,
      "grad_norm": 0.0023341418709605932,
      "learning_rate": 0.25056787775897055,
      "loss": 0.3473,
      "num_input_tokens_seen": 20195344,
      "step": 10645
    },
    {
      "epoch": 1.7374174076188922,
      "grad_norm": 0.0004957548226229846,
      "learning_rate": 0.2505241653995056,
      "loss": 0.3377,
      "num_input_tokens_seen": 20204112,
      "step": 10650
    },
    {
      "epoch": 1.7382331348397098,
      "grad_norm": 0.001637584762647748,
      "learning_rate": 0.25048043753795113,
      "loss": 0.3702,
      "num_input_tokens_seen": 20213472,
      "step": 10655
    },
    {
      "epoch": 1.739048862060527,
      "grad_norm": 0.0013136430643498898,
      "learning_rate": 0.2504366941810504,
      "loss": 0.343,
      "num_input_tokens_seen": 20223280,
      "step": 10660
    },
    {
      "epoch": 1.7398645892813445,
      "grad_norm": 0.0010611391626298428,
      "learning_rate": 0.2503929353355493,
      "loss": 0.377,
      "num_input_tokens_seen": 20231696,
      "step": 10665
    },
    {
      "epoch": 1.7406803165021616,
      "grad_norm": 0.0014534477377310395,
      "learning_rate": 0.250349161008196,
      "loss": 0.3055,
      "num_input_tokens_seen": 20240704,
      "step": 10670
    },
    {
      "epoch": 1.7414960437229792,
      "grad_norm": 0.0030051919165998697,
      "learning_rate": 0.2503053712057409,
      "loss": 0.3627,
      "num_input_tokens_seen": 20249792,
      "step": 10675
    },
    {
      "epoch": 1.7423117709437963,
      "grad_norm": 0.001810617046430707,
      "learning_rate": 0.25026156593493715,
      "loss": 0.3327,
      "num_input_tokens_seen": 20259120,
      "step": 10680
    },
    {
      "epoch": 1.7431274981646139,
      "grad_norm": 0.0008397328201681376,
      "learning_rate": 0.2502177452025399,
      "loss": 0.3104,
      "num_input_tokens_seen": 20268288,
      "step": 10685
    },
    {
      "epoch": 1.743943225385431,
      "grad_norm": 0.0032719294540584087,
      "learning_rate": 0.25017390901530695,
      "loss": 0.3594,
      "num_input_tokens_seen": 20277024,
      "step": 10690
    },
    {
      "epoch": 1.7447589526062486,
      "grad_norm": 0.0020510954782366753,
      "learning_rate": 0.2501300573799984,
      "loss": 0.3453,
      "num_input_tokens_seen": 20286560,
      "step": 10695
    },
    {
      "epoch": 1.7455746798270657,
      "grad_norm": 0.001893661799840629,
      "learning_rate": 0.2500861903033766,
      "loss": 0.3347,
      "num_input_tokens_seen": 20296688,
      "step": 10700
    },
    {
      "epoch": 1.7463904070478833,
      "grad_norm": 0.0012105208588764071,
      "learning_rate": 0.25004230779220654,
      "loss": 0.2809,
      "num_input_tokens_seen": 20307024,
      "step": 10705
    },
    {
      "epoch": 1.7472061342687004,
      "grad_norm": 0.0009635318419896066,
      "learning_rate": 0.24999840985325542,
      "loss": 0.4032,
      "num_input_tokens_seen": 20316336,
      "step": 10710
    },
    {
      "epoch": 1.748021861489518,
      "grad_norm": 0.0017622385639697313,
      "learning_rate": 0.24995449649329285,
      "loss": 0.34,
      "num_input_tokens_seen": 20325280,
      "step": 10715
    },
    {
      "epoch": 1.7488375887103351,
      "grad_norm": 0.0016354245599359274,
      "learning_rate": 0.2499105677190908,
      "loss": 0.3532,
      "num_input_tokens_seen": 20335392,
      "step": 10720
    },
    {
      "epoch": 1.7496533159311527,
      "grad_norm": 0.002882373984903097,
      "learning_rate": 0.24986662353742364,
      "loss": 0.3361,
      "num_input_tokens_seen": 20343936,
      "step": 10725
    },
    {
      "epoch": 1.7504690431519698,
      "grad_norm": 0.0024759217631071806,
      "learning_rate": 0.24982266395506814,
      "loss": 0.2856,
      "num_input_tokens_seen": 20352704,
      "step": 10730
    },
    {
      "epoch": 1.7512847703727874,
      "grad_norm": 0.0027678650803864002,
      "learning_rate": 0.2497786889788034,
      "loss": 0.3412,
      "num_input_tokens_seen": 20361696,
      "step": 10735
    },
    {
      "epoch": 1.7521004975936045,
      "grad_norm": 0.0017237914726138115,
      "learning_rate": 0.24973469861541095,
      "loss": 0.3057,
      "num_input_tokens_seen": 20371248,
      "step": 10740
    },
    {
      "epoch": 1.752916224814422,
      "grad_norm": 0.003829790046438575,
      "learning_rate": 0.24969069287167456,
      "loss": 0.3084,
      "num_input_tokens_seen": 20381376,
      "step": 10745
    },
    {
      "epoch": 1.7537319520352395,
      "grad_norm": 0.0009760879329405725,
      "learning_rate": 0.2496466717543806,
      "loss": 0.3227,
      "num_input_tokens_seen": 20390688,
      "step": 10750
    },
    {
      "epoch": 1.7545476792560568,
      "grad_norm": 0.0014791880967095494,
      "learning_rate": 0.24960263527031762,
      "loss": 0.3873,
      "num_input_tokens_seen": 20399504,
      "step": 10755
    },
    {
      "epoch": 1.7553634064768742,
      "grad_norm": 0.001167726586572826,
      "learning_rate": 0.24955858342627657,
      "loss": 0.3573,
      "num_input_tokens_seen": 20410240,
      "step": 10760
    },
    {
      "epoch": 1.7561791336976915,
      "grad_norm": 0.0032307938672602177,
      "learning_rate": 0.24951451622905083,
      "loss": 0.3446,
      "num_input_tokens_seen": 20420688,
      "step": 10765
    },
    {
      "epoch": 1.7569948609185089,
      "grad_norm": 0.0017921068938449025,
      "learning_rate": 0.24947043368543612,
      "loss": 0.3703,
      "num_input_tokens_seen": 20431856,
      "step": 10770
    },
    {
      "epoch": 1.7578105881393262,
      "grad_norm": 0.0012880238937214017,
      "learning_rate": 0.2494263358022305,
      "loss": 0.3369,
      "num_input_tokens_seen": 20441168,
      "step": 10775
    },
    {
      "epoch": 1.7586263153601436,
      "grad_norm": 0.0016118005150929093,
      "learning_rate": 0.24938222258623444,
      "loss": 0.3283,
      "num_input_tokens_seen": 20450336,
      "step": 10780
    },
    {
      "epoch": 1.759442042580961,
      "grad_norm": 0.001114063197746873,
      "learning_rate": 0.24933809404425075,
      "loss": 0.2881,
      "num_input_tokens_seen": 20460496,
      "step": 10785
    },
    {
      "epoch": 1.7602577698017783,
      "grad_norm": 0.001983939902856946,
      "learning_rate": 0.24929395018308453,
      "loss": 0.4212,
      "num_input_tokens_seen": 20470032,
      "step": 10790
    },
    {
      "epoch": 1.7610734970225956,
      "grad_norm": 0.0010364582994952798,
      "learning_rate": 0.24924979100954348,
      "loss": 0.3516,
      "num_input_tokens_seen": 20480208,
      "step": 10795
    },
    {
      "epoch": 1.761889224243413,
      "grad_norm": 0.0006127640372142196,
      "learning_rate": 0.24920561653043735,
      "loss": 0.3823,
      "num_input_tokens_seen": 20489424,
      "step": 10800
    },
    {
      "epoch": 1.761889224243413,
      "eval_loss": 0.337137907743454,
      "eval_runtime": 152.613,
      "eval_samples_per_second": 17.856,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 20489424,
      "step": 10800
    },
    {
      "epoch": 1.7627049514642303,
      "grad_norm": 0.0008739382028579712,
      "learning_rate": 0.24916142675257846,
      "loss": 0.2993,
      "num_input_tokens_seen": 20498464,
      "step": 10805
    },
    {
      "epoch": 1.7635206786850477,
      "grad_norm": 0.0011169322533532977,
      "learning_rate": 0.24911722168278144,
      "loss": 0.4687,
      "num_input_tokens_seen": 20507328,
      "step": 10810
    },
    {
      "epoch": 1.764336405905865,
      "grad_norm": 0.0011682065669447184,
      "learning_rate": 0.24907300132786328,
      "loss": 0.352,
      "num_input_tokens_seen": 20516704,
      "step": 10815
    },
    {
      "epoch": 1.7651521331266824,
      "grad_norm": 0.002152572851628065,
      "learning_rate": 0.24902876569464322,
      "loss": 0.3169,
      "num_input_tokens_seen": 20527296,
      "step": 10820
    },
    {
      "epoch": 1.7659678603474998,
      "grad_norm": 0.001757334335707128,
      "learning_rate": 0.24898451478994305,
      "loss": 0.3295,
      "num_input_tokens_seen": 20536384,
      "step": 10825
    },
    {
      "epoch": 1.766783587568317,
      "grad_norm": 0.0009271187009289861,
      "learning_rate": 0.2489402486205868,
      "loss": 0.2947,
      "num_input_tokens_seen": 20547536,
      "step": 10830
    },
    {
      "epoch": 1.7675993147891345,
      "grad_norm": 0.002120593562722206,
      "learning_rate": 0.24889596719340085,
      "loss": 0.3772,
      "num_input_tokens_seen": 20555680,
      "step": 10835
    },
    {
      "epoch": 1.768415042009952,
      "grad_norm": 0.0020933314226567745,
      "learning_rate": 0.24885167051521392,
      "loss": 0.3686,
      "num_input_tokens_seen": 20563824,
      "step": 10840
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.0008089946350082755,
      "learning_rate": 0.24880735859285716,
      "loss": 0.344,
      "num_input_tokens_seen": 20573840,
      "step": 10845
    },
    {
      "epoch": 1.7700464964515867,
      "grad_norm": 0.0016704053850844502,
      "learning_rate": 0.24876303143316406,
      "loss": 0.3612,
      "num_input_tokens_seen": 20582480,
      "step": 10850
    },
    {
      "epoch": 1.7708622236724039,
      "grad_norm": 0.0010663510765880346,
      "learning_rate": 0.24871868904297031,
      "loss": 0.3262,
      "num_input_tokens_seen": 20592640,
      "step": 10855
    },
    {
      "epoch": 1.7716779508932214,
      "grad_norm": 0.0006108595989644527,
      "learning_rate": 0.24867433142911416,
      "loss": 0.2945,
      "num_input_tokens_seen": 20602384,
      "step": 10860
    },
    {
      "epoch": 1.7724936781140386,
      "grad_norm": 0.002614156110212207,
      "learning_rate": 0.24862995859843612,
      "loss": 0.4113,
      "num_input_tokens_seen": 20610976,
      "step": 10865
    },
    {
      "epoch": 1.7733094053348561,
      "grad_norm": 0.0016540414653718472,
      "learning_rate": 0.24858557055777897,
      "loss": 0.2854,
      "num_input_tokens_seen": 20620752,
      "step": 10870
    },
    {
      "epoch": 1.7741251325556733,
      "grad_norm": 0.0008876384817995131,
      "learning_rate": 0.24854116731398793,
      "loss": 0.3655,
      "num_input_tokens_seen": 20630016,
      "step": 10875
    },
    {
      "epoch": 1.7749408597764909,
      "grad_norm": 0.0016695831436663866,
      "learning_rate": 0.24849674887391052,
      "loss": 0.3295,
      "num_input_tokens_seen": 20640896,
      "step": 10880
    },
    {
      "epoch": 1.775756586997308,
      "grad_norm": 0.002111308043822646,
      "learning_rate": 0.2484523152443967,
      "loss": 0.4021,
      "num_input_tokens_seen": 20651040,
      "step": 10885
    },
    {
      "epoch": 1.7765723142181256,
      "grad_norm": 0.002906143432483077,
      "learning_rate": 0.24840786643229862,
      "loss": 0.3553,
      "num_input_tokens_seen": 20660192,
      "step": 10890
    },
    {
      "epoch": 1.7773880414389427,
      "grad_norm": 0.0011998900445178151,
      "learning_rate": 0.2483634024444709,
      "loss": 0.3414,
      "num_input_tokens_seen": 20669728,
      "step": 10895
    },
    {
      "epoch": 1.7782037686597603,
      "grad_norm": 0.0016766376793384552,
      "learning_rate": 0.24831892328777033,
      "loss": 0.3575,
      "num_input_tokens_seen": 20679920,
      "step": 10900
    },
    {
      "epoch": 1.7790194958805774,
      "grad_norm": 0.0008294403669424355,
      "learning_rate": 0.2482744289690563,
      "loss": 0.3341,
      "num_input_tokens_seen": 20689040,
      "step": 10905
    },
    {
      "epoch": 1.779835223101395,
      "grad_norm": 0.001781053957529366,
      "learning_rate": 0.2482299194951903,
      "loss": 0.3248,
      "num_input_tokens_seen": 20698496,
      "step": 10910
    },
    {
      "epoch": 1.780650950322212,
      "grad_norm": 0.0009371531778015196,
      "learning_rate": 0.2481853948730363,
      "loss": 0.3734,
      "num_input_tokens_seen": 20707440,
      "step": 10915
    },
    {
      "epoch": 1.7814666775430297,
      "grad_norm": 0.0011428437428548932,
      "learning_rate": 0.24814085510946052,
      "loss": 0.3291,
      "num_input_tokens_seen": 20717008,
      "step": 10920
    },
    {
      "epoch": 1.7822824047638468,
      "grad_norm": 0.0012332332553341985,
      "learning_rate": 0.24809630021133158,
      "loss": 0.3526,
      "num_input_tokens_seen": 20726480,
      "step": 10925
    },
    {
      "epoch": 1.7830981319846644,
      "grad_norm": 0.0029296078719198704,
      "learning_rate": 0.24805173018552037,
      "loss": 0.3157,
      "num_input_tokens_seen": 20735472,
      "step": 10930
    },
    {
      "epoch": 1.7839138592054817,
      "grad_norm": 0.00125355354975909,
      "learning_rate": 0.2480071450389002,
      "loss": 0.3177,
      "num_input_tokens_seen": 20744720,
      "step": 10935
    },
    {
      "epoch": 1.784729586426299,
      "grad_norm": 0.002291103359311819,
      "learning_rate": 0.24796254477834662,
      "loss": 0.3782,
      "num_input_tokens_seen": 20753600,
      "step": 10940
    },
    {
      "epoch": 1.7855453136471164,
      "grad_norm": 0.0006885069306008518,
      "learning_rate": 0.24791792941073754,
      "loss": 0.3139,
      "num_input_tokens_seen": 20763024,
      "step": 10945
    },
    {
      "epoch": 1.7863610408679338,
      "grad_norm": 0.0012774320784956217,
      "learning_rate": 0.2478732989429533,
      "loss": 0.3617,
      "num_input_tokens_seen": 20773120,
      "step": 10950
    },
    {
      "epoch": 1.7871767680887511,
      "grad_norm": 0.001001552096568048,
      "learning_rate": 0.24782865338187632,
      "loss": 0.3472,
      "num_input_tokens_seen": 20781984,
      "step": 10955
    },
    {
      "epoch": 1.7879924953095685,
      "grad_norm": 0.0019728897605091333,
      "learning_rate": 0.2477839927343916,
      "loss": 0.3829,
      "num_input_tokens_seen": 20790688,
      "step": 10960
    },
    {
      "epoch": 1.7888082225303858,
      "grad_norm": 0.0008870852761901915,
      "learning_rate": 0.2477393170073864,
      "loss": 0.314,
      "num_input_tokens_seen": 20801008,
      "step": 10965
    },
    {
      "epoch": 1.7896239497512032,
      "grad_norm": 0.0025414787232875824,
      "learning_rate": 0.2476946262077503,
      "loss": 0.3473,
      "num_input_tokens_seen": 20811328,
      "step": 10970
    },
    {
      "epoch": 1.7904396769720206,
      "grad_norm": 0.0030013341456651688,
      "learning_rate": 0.24764992034237507,
      "loss": 0.3384,
      "num_input_tokens_seen": 20821376,
      "step": 10975
    },
    {
      "epoch": 1.791255404192838,
      "grad_norm": 0.0012863631127402186,
      "learning_rate": 0.24760519941815498,
      "loss": 0.3184,
      "num_input_tokens_seen": 20831936,
      "step": 10980
    },
    {
      "epoch": 1.7920711314136553,
      "grad_norm": 0.0010182020487263799,
      "learning_rate": 0.2475604634419866,
      "loss": 0.289,
      "num_input_tokens_seen": 20842208,
      "step": 10985
    },
    {
      "epoch": 1.7928868586344726,
      "grad_norm": 0.0014567767502740026,
      "learning_rate": 0.24751571242076872,
      "loss": 0.3241,
      "num_input_tokens_seen": 20852288,
      "step": 10990
    },
    {
      "epoch": 1.79370258585529,
      "grad_norm": 0.002131383866071701,
      "learning_rate": 0.2474709463614025,
      "loss": 0.3649,
      "num_input_tokens_seen": 20861536,
      "step": 10995
    },
    {
      "epoch": 1.7945183130761073,
      "grad_norm": 0.002628362737596035,
      "learning_rate": 0.24742616527079145,
      "loss": 0.3643,
      "num_input_tokens_seen": 20870832,
      "step": 11000
    },
    {
      "epoch": 1.7945183130761073,
      "eval_loss": 0.3393862545490265,
      "eval_runtime": 152.6449,
      "eval_samples_per_second": 17.852,
      "eval_steps_per_second": 8.929,
      "num_input_tokens_seen": 20870832,
      "step": 11000
    },
    {
      "epoch": 1.7953340402969247,
      "grad_norm": 0.0010687147732824087,
      "learning_rate": 0.24738136915584139,
      "loss": 0.3193,
      "num_input_tokens_seen": 20879216,
      "step": 11005
    },
    {
      "epoch": 1.796149767517742,
      "grad_norm": 0.0006344877765513957,
      "learning_rate": 0.24733655802346047,
      "loss": 0.3562,
      "num_input_tokens_seen": 20888320,
      "step": 11010
    },
    {
      "epoch": 1.7969654947385594,
      "grad_norm": 0.0015486010815948248,
      "learning_rate": 0.24729173188055906,
      "loss": 0.3623,
      "num_input_tokens_seen": 20898064,
      "step": 11015
    },
    {
      "epoch": 1.7977812219593767,
      "grad_norm": 0.0018083370523527265,
      "learning_rate": 0.24724689073404996,
      "loss": 0.3206,
      "num_input_tokens_seen": 20906784,
      "step": 11020
    },
    {
      "epoch": 1.7985969491801943,
      "grad_norm": 0.002205795608460903,
      "learning_rate": 0.24720203459084822,
      "loss": 0.3494,
      "num_input_tokens_seen": 20916736,
      "step": 11025
    },
    {
      "epoch": 1.7994126764010114,
      "grad_norm": 0.0031758523546159267,
      "learning_rate": 0.24715716345787123,
      "loss": 0.4153,
      "num_input_tokens_seen": 20926112,
      "step": 11030
    },
    {
      "epoch": 1.800228403621829,
      "grad_norm": 0.003970603924244642,
      "learning_rate": 0.2471122773420387,
      "loss": 0.3568,
      "num_input_tokens_seen": 20937104,
      "step": 11035
    },
    {
      "epoch": 1.8010441308426461,
      "grad_norm": 0.002833325881510973,
      "learning_rate": 0.24706737625027259,
      "loss": 0.3389,
      "num_input_tokens_seen": 20947040,
      "step": 11040
    },
    {
      "epoch": 1.8018598580634637,
      "grad_norm": 0.0017678934382274747,
      "learning_rate": 0.24702246018949725,
      "loss": 0.3547,
      "num_input_tokens_seen": 20956096,
      "step": 11045
    },
    {
      "epoch": 1.8026755852842808,
      "grad_norm": 0.0015625820960849524,
      "learning_rate": 0.2469775291666393,
      "loss": 0.3822,
      "num_input_tokens_seen": 20965456,
      "step": 11050
    },
    {
      "epoch": 1.8034913125050984,
      "grad_norm": 0.0016759526915848255,
      "learning_rate": 0.24693258318862765,
      "loss": 0.2813,
      "num_input_tokens_seen": 20974688,
      "step": 11055
    },
    {
      "epoch": 1.8043070397259156,
      "grad_norm": 0.0018425638554617763,
      "learning_rate": 0.2468876222623935,
      "loss": 0.4668,
      "num_input_tokens_seen": 20984464,
      "step": 11060
    },
    {
      "epoch": 1.8051227669467331,
      "grad_norm": 0.000515251187607646,
      "learning_rate": 0.2468426463948705,
      "loss": 0.3509,
      "num_input_tokens_seen": 20992928,
      "step": 11065
    },
    {
      "epoch": 1.8059384941675503,
      "grad_norm": 0.0015966705977916718,
      "learning_rate": 0.24679765559299438,
      "loss": 0.3327,
      "num_input_tokens_seen": 21002208,
      "step": 11070
    },
    {
      "epoch": 1.8067542213883678,
      "grad_norm": 0.001540074939839542,
      "learning_rate": 0.24675264986370332,
      "loss": 0.356,
      "num_input_tokens_seen": 21011808,
      "step": 11075
    },
    {
      "epoch": 1.807569948609185,
      "grad_norm": 0.0010276687098667026,
      "learning_rate": 0.2467076292139378,
      "loss": 0.2981,
      "num_input_tokens_seen": 21021360,
      "step": 11080
    },
    {
      "epoch": 1.8083856758300025,
      "grad_norm": 0.0009188342373818159,
      "learning_rate": 0.24666259365064055,
      "loss": 0.3531,
      "num_input_tokens_seen": 21030032,
      "step": 11085
    },
    {
      "epoch": 1.8092014030508197,
      "grad_norm": 0.0007054214947856963,
      "learning_rate": 0.24661754318075663,
      "loss": 0.33,
      "num_input_tokens_seen": 21039520,
      "step": 11090
    },
    {
      "epoch": 1.8100171302716372,
      "grad_norm": 0.0008906634175218642,
      "learning_rate": 0.2465724778112334,
      "loss": 0.3933,
      "num_input_tokens_seen": 21048128,
      "step": 11095
    },
    {
      "epoch": 1.8108328574924544,
      "grad_norm": 0.0014245039783418179,
      "learning_rate": 0.24652739754902042,
      "loss": 0.3077,
      "num_input_tokens_seen": 21058096,
      "step": 11100
    },
    {
      "epoch": 1.811648584713272,
      "grad_norm": 0.001448179711587727,
      "learning_rate": 0.24648230240106975,
      "loss": 0.3374,
      "num_input_tokens_seen": 21065968,
      "step": 11105
    },
    {
      "epoch": 1.812464311934089,
      "grad_norm": 0.0032603100407868624,
      "learning_rate": 0.2464371923743356,
      "loss": 0.3737,
      "num_input_tokens_seen": 21074448,
      "step": 11110
    },
    {
      "epoch": 1.8132800391549067,
      "grad_norm": 0.0012270765146240592,
      "learning_rate": 0.24639206747577444,
      "loss": 0.3343,
      "num_input_tokens_seen": 21083312,
      "step": 11115
    },
    {
      "epoch": 1.814095766375724,
      "grad_norm": 0.0005591180524788797,
      "learning_rate": 0.24634692771234515,
      "loss": 0.3444,
      "num_input_tokens_seen": 21091376,
      "step": 11120
    },
    {
      "epoch": 1.8149114935965414,
      "grad_norm": 0.0008276433800347149,
      "learning_rate": 0.2463017730910088,
      "loss": 0.3477,
      "num_input_tokens_seen": 21100192,
      "step": 11125
    },
    {
      "epoch": 1.8157272208173587,
      "grad_norm": 0.0018344840500503778,
      "learning_rate": 0.2462566036187289,
      "loss": 0.3729,
      "num_input_tokens_seen": 21109056,
      "step": 11130
    },
    {
      "epoch": 1.816542948038176,
      "grad_norm": 0.0023385626263916492,
      "learning_rate": 0.24621141930247106,
      "loss": 0.3848,
      "num_input_tokens_seen": 21117392,
      "step": 11135
    },
    {
      "epoch": 1.8173586752589934,
      "grad_norm": 0.0026142550632357597,
      "learning_rate": 0.2461662201492033,
      "loss": 0.3799,
      "num_input_tokens_seen": 21126112,
      "step": 11140
    },
    {
      "epoch": 1.8181744024798108,
      "grad_norm": 0.0017819356871768832,
      "learning_rate": 0.24612100616589586,
      "loss": 0.2883,
      "num_input_tokens_seen": 21135008,
      "step": 11145
    },
    {
      "epoch": 1.8189901297006281,
      "grad_norm": 0.0009772825287654996,
      "learning_rate": 0.24607577735952135,
      "loss": 0.2922,
      "num_input_tokens_seen": 21143216,
      "step": 11150
    },
    {
      "epoch": 1.8198058569214455,
      "grad_norm": 0.0020451124291867018,
      "learning_rate": 0.24603053373705464,
      "loss": 0.3112,
      "num_input_tokens_seen": 21152464,
      "step": 11155
    },
    {
      "epoch": 1.8206215841422628,
      "grad_norm": 0.005673659034073353,
      "learning_rate": 0.2459852753054728,
      "loss": 0.354,
      "num_input_tokens_seen": 21162352,
      "step": 11160
    },
    {
      "epoch": 1.8214373113630802,
      "grad_norm": 0.0015107851941138506,
      "learning_rate": 0.24594000207175526,
      "loss": 0.309,
      "num_input_tokens_seen": 21172864,
      "step": 11165
    },
    {
      "epoch": 1.8222530385838975,
      "grad_norm": 0.007650075014680624,
      "learning_rate": 0.2458947140428838,
      "loss": 0.3359,
      "num_input_tokens_seen": 21182208,
      "step": 11170
    },
    {
      "epoch": 1.8230687658047149,
      "grad_norm": 0.002202556701377034,
      "learning_rate": 0.24584941122584233,
      "loss": 0.3777,
      "num_input_tokens_seen": 21192128,
      "step": 11175
    },
    {
      "epoch": 1.8238844930255322,
      "grad_norm": 0.001555692870169878,
      "learning_rate": 0.24580409362761713,
      "loss": 0.409,
      "num_input_tokens_seen": 21201248,
      "step": 11180
    },
    {
      "epoch": 1.8247002202463496,
      "grad_norm": 0.0022137584164738655,
      "learning_rate": 0.2457587612551967,
      "loss": 0.3672,
      "num_input_tokens_seen": 21211216,
      "step": 11185
    },
    {
      "epoch": 1.825515947467167,
      "grad_norm": 0.0024084204342216253,
      "learning_rate": 0.24571341411557193,
      "loss": 0.3173,
      "num_input_tokens_seen": 21221168,
      "step": 11190
    },
    {
      "epoch": 1.8263316746879843,
      "grad_norm": 0.0003256124327890575,
      "learning_rate": 0.2456680522157359,
      "loss": 0.3164,
      "num_input_tokens_seen": 21231472,
      "step": 11195
    },
    {
      "epoch": 1.8271474019088016,
      "grad_norm": 0.0010201133554801345,
      "learning_rate": 0.245622675562684,
      "loss": 0.3802,
      "num_input_tokens_seen": 21240960,
      "step": 11200
    },
    {
      "epoch": 1.8271474019088016,
      "eval_loss": 0.3317229449748993,
      "eval_runtime": 152.6284,
      "eval_samples_per_second": 17.854,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 21240960,
      "step": 11200
    },
    {
      "epoch": 1.827963129129619,
      "grad_norm": 0.001755463657900691,
      "learning_rate": 0.24557728416341384,
      "loss": 0.3441,
      "num_input_tokens_seen": 21250784,
      "step": 11205
    },
    {
      "epoch": 1.8287788563504366,
      "grad_norm": 0.0017400692449882627,
      "learning_rate": 0.24553187802492538,
      "loss": 0.3754,
      "num_input_tokens_seen": 21260432,
      "step": 11210
    },
    {
      "epoch": 1.8295945835712537,
      "grad_norm": 0.001273578149266541,
      "learning_rate": 0.24548645715422074,
      "loss": 0.3488,
      "num_input_tokens_seen": 21268688,
      "step": 11215
    },
    {
      "epoch": 1.8304103107920713,
      "grad_norm": 0.0014002294046804309,
      "learning_rate": 0.2454410215583045,
      "loss": 0.2969,
      "num_input_tokens_seen": 21277792,
      "step": 11220
    },
    {
      "epoch": 1.8312260380128884,
      "grad_norm": 0.0025232795160263777,
      "learning_rate": 0.24539557124418332,
      "loss": 0.3973,
      "num_input_tokens_seen": 21287312,
      "step": 11225
    },
    {
      "epoch": 1.832041765233706,
      "grad_norm": 0.0012750624446198344,
      "learning_rate": 0.24535010621886624,
      "loss": 0.3724,
      "num_input_tokens_seen": 21295792,
      "step": 11230
    },
    {
      "epoch": 1.8328574924545231,
      "grad_norm": 0.0018980810418725014,
      "learning_rate": 0.2453046264893646,
      "loss": 0.3432,
      "num_input_tokens_seen": 21306064,
      "step": 11235
    },
    {
      "epoch": 1.8336732196753407,
      "grad_norm": 0.0013140903320163488,
      "learning_rate": 0.24525913206269184,
      "loss": 0.3424,
      "num_input_tokens_seen": 21316080,
      "step": 11240
    },
    {
      "epoch": 1.8344889468961578,
      "grad_norm": 0.0012429046910256147,
      "learning_rate": 0.2452136229458638,
      "loss": 0.3546,
      "num_input_tokens_seen": 21325568,
      "step": 11245
    },
    {
      "epoch": 1.8353046741169754,
      "grad_norm": 0.0015622887294739485,
      "learning_rate": 0.24516809914589857,
      "loss": 0.3104,
      "num_input_tokens_seen": 21335296,
      "step": 11250
    },
    {
      "epoch": 1.8361204013377925,
      "grad_norm": 0.0018598996102809906,
      "learning_rate": 0.2451225606698165,
      "loss": 0.38,
      "num_input_tokens_seen": 21344720,
      "step": 11255
    },
    {
      "epoch": 1.83693612855861,
      "grad_norm": 0.0008818313363008201,
      "learning_rate": 0.2450770075246402,
      "loss": 0.3368,
      "num_input_tokens_seen": 21354320,
      "step": 11260
    },
    {
      "epoch": 1.8377518557794272,
      "grad_norm": 0.0018084030598402023,
      "learning_rate": 0.24503143971739455,
      "loss": 0.3226,
      "num_input_tokens_seen": 21363312,
      "step": 11265
    },
    {
      "epoch": 1.8385675830002448,
      "grad_norm": 0.0010593731421977282,
      "learning_rate": 0.24498585725510663,
      "loss": 0.3414,
      "num_input_tokens_seen": 21373024,
      "step": 11270
    },
    {
      "epoch": 1.839383310221062,
      "grad_norm": 0.0009067380451597273,
      "learning_rate": 0.24494026014480583,
      "loss": 0.3707,
      "num_input_tokens_seen": 21382192,
      "step": 11275
    },
    {
      "epoch": 1.8401990374418795,
      "grad_norm": 0.0006837673136033118,
      "learning_rate": 0.24489464839352387,
      "loss": 0.3527,
      "num_input_tokens_seen": 21392416,
      "step": 11280
    },
    {
      "epoch": 1.8410147646626966,
      "grad_norm": 0.000783375115133822,
      "learning_rate": 0.2448490220082946,
      "loss": 0.3509,
      "num_input_tokens_seen": 21400576,
      "step": 11285
    },
    {
      "epoch": 1.8418304918835142,
      "grad_norm": 0.0017105760052800179,
      "learning_rate": 0.24480338099615415,
      "loss": 0.3651,
      "num_input_tokens_seen": 21410096,
      "step": 11290
    },
    {
      "epoch": 1.8426462191043314,
      "grad_norm": 0.0020197504200041294,
      "learning_rate": 0.244757725364141,
      "loss": 0.345,
      "num_input_tokens_seen": 21418576,
      "step": 11295
    },
    {
      "epoch": 1.843461946325149,
      "grad_norm": 0.0004691259819082916,
      "learning_rate": 0.24471205511929583,
      "loss": 0.3096,
      "num_input_tokens_seen": 21428032,
      "step": 11300
    },
    {
      "epoch": 1.8442776735459663,
      "grad_norm": 0.0008487607119604945,
      "learning_rate": 0.24466637026866145,
      "loss": 0.3697,
      "num_input_tokens_seen": 21436528,
      "step": 11305
    },
    {
      "epoch": 1.8450934007667836,
      "grad_norm": 0.0014929197495803237,
      "learning_rate": 0.2446206708192832,
      "loss": 0.3251,
      "num_input_tokens_seen": 21447072,
      "step": 11310
    },
    {
      "epoch": 1.845909127987601,
      "grad_norm": 0.000677551026456058,
      "learning_rate": 0.2445749567782084,
      "loss": 0.3359,
      "num_input_tokens_seen": 21456000,
      "step": 11315
    },
    {
      "epoch": 1.8467248552084183,
      "grad_norm": 0.0015565772773697972,
      "learning_rate": 0.2445292281524868,
      "loss": 0.3256,
      "num_input_tokens_seen": 21465888,
      "step": 11320
    },
    {
      "epoch": 1.8475405824292357,
      "grad_norm": 0.0009499933803454041,
      "learning_rate": 0.24448348494917022,
      "loss": 0.3102,
      "num_input_tokens_seen": 21475616,
      "step": 11325
    },
    {
      "epoch": 1.848356309650053,
      "grad_norm": 0.0010900149354711175,
      "learning_rate": 0.24443772717531295,
      "loss": 0.3195,
      "num_input_tokens_seen": 21485216,
      "step": 11330
    },
    {
      "epoch": 1.8491720368708704,
      "grad_norm": 0.0032317168079316616,
      "learning_rate": 0.24439195483797138,
      "loss": 0.383,
      "num_input_tokens_seen": 21493824,
      "step": 11335
    },
    {
      "epoch": 1.8499877640916877,
      "grad_norm": 0.001051778206601739,
      "learning_rate": 0.24434616794420416,
      "loss": 0.3883,
      "num_input_tokens_seen": 21503552,
      "step": 11340
    },
    {
      "epoch": 1.850803491312505,
      "grad_norm": 0.002236738335341215,
      "learning_rate": 0.24430036650107223,
      "loss": 0.3381,
      "num_input_tokens_seen": 21512544,
      "step": 11345
    },
    {
      "epoch": 1.8516192185333225,
      "grad_norm": 0.0033347641583532095,
      "learning_rate": 0.2442545505156387,
      "loss": 0.3862,
      "num_input_tokens_seen": 21522128,
      "step": 11350
    },
    {
      "epoch": 1.8524349457541398,
      "grad_norm": 0.0005919477553106844,
      "learning_rate": 0.24420871999496904,
      "loss": 0.3361,
      "num_input_tokens_seen": 21531920,
      "step": 11355
    },
    {
      "epoch": 1.8532506729749572,
      "grad_norm": 0.001672523096203804,
      "learning_rate": 0.24416287494613084,
      "loss": 0.3277,
      "num_input_tokens_seen": 21542016,
      "step": 11360
    },
    {
      "epoch": 1.8540664001957745,
      "grad_norm": 0.002261393703520298,
      "learning_rate": 0.24411701537619399,
      "loss": 0.2752,
      "num_input_tokens_seen": 21551808,
      "step": 11365
    },
    {
      "epoch": 1.8548821274165919,
      "grad_norm": 0.003068141872063279,
      "learning_rate": 0.24407114129223062,
      "loss": 0.3051,
      "num_input_tokens_seen": 21560944,
      "step": 11370
    },
    {
      "epoch": 1.8556978546374092,
      "grad_norm": 0.001474719843827188,
      "learning_rate": 0.2440252527013151,
      "loss": 0.3829,
      "num_input_tokens_seen": 21569664,
      "step": 11375
    },
    {
      "epoch": 1.8565135818582266,
      "grad_norm": 0.0022808206267654896,
      "learning_rate": 0.24397934961052403,
      "loss": 0.3806,
      "num_input_tokens_seen": 21579552,
      "step": 11380
    },
    {
      "epoch": 1.857329309079044,
      "grad_norm": 0.0015099215088412166,
      "learning_rate": 0.24393343202693618,
      "loss": 0.3282,
      "num_input_tokens_seen": 21587424,
      "step": 11385
    },
    {
      "epoch": 1.8581450362998613,
      "grad_norm": 0.0027260237839072943,
      "learning_rate": 0.2438874999576327,
      "loss": 0.3762,
      "num_input_tokens_seen": 21597296,
      "step": 11390
    },
    {
      "epoch": 1.8589607635206788,
      "grad_norm": 0.0025510319974273443,
      "learning_rate": 0.24384155340969688,
      "loss": 0.3765,
      "num_input_tokens_seen": 21605904,
      "step": 11395
    },
    {
      "epoch": 1.859776490741496,
      "grad_norm": 0.002008997369557619,
      "learning_rate": 0.24379559239021423,
      "loss": 0.3923,
      "num_input_tokens_seen": 21615744,
      "step": 11400
    },
    {
      "epoch": 1.859776490741496,
      "eval_loss": 0.34030041098594666,
      "eval_runtime": 152.5976,
      "eval_samples_per_second": 17.857,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 21615744,
      "step": 11400
    },
    {
      "epoch": 1.8605922179623136,
      "grad_norm": 0.0020516044460237026,
      "learning_rate": 0.2437496169062725,
      "loss": 0.3715,
      "num_input_tokens_seen": 21625008,
      "step": 11405
    },
    {
      "epoch": 1.8614079451831307,
      "grad_norm": 0.002411672379821539,
      "learning_rate": 0.24370362696496176,
      "loss": 0.3634,
      "num_input_tokens_seen": 21634192,
      "step": 11410
    },
    {
      "epoch": 1.8622236724039483,
      "grad_norm": 0.002078381832689047,
      "learning_rate": 0.24365762257337417,
      "loss": 0.3735,
      "num_input_tokens_seen": 21642256,
      "step": 11415
    },
    {
      "epoch": 1.8630393996247654,
      "grad_norm": 0.001540415221825242,
      "learning_rate": 0.2436116037386042,
      "loss": 0.3307,
      "num_input_tokens_seen": 21650128,
      "step": 11420
    },
    {
      "epoch": 1.863855126845583,
      "grad_norm": 0.0017382123041898012,
      "learning_rate": 0.24356557046774852,
      "loss": 0.3611,
      "num_input_tokens_seen": 21660080,
      "step": 11425
    },
    {
      "epoch": 1.8646708540664,
      "grad_norm": 0.001306416466832161,
      "learning_rate": 0.24351952276790606,
      "loss": 0.3328,
      "num_input_tokens_seen": 21670480,
      "step": 11430
    },
    {
      "epoch": 1.8654865812872177,
      "grad_norm": 0.002118972595781088,
      "learning_rate": 0.24347346064617797,
      "loss": 0.3696,
      "num_input_tokens_seen": 21679680,
      "step": 11435
    },
    {
      "epoch": 1.8663023085080348,
      "grad_norm": 0.0008206577622331679,
      "learning_rate": 0.24342738410966758,
      "loss": 0.3365,
      "num_input_tokens_seen": 21688528,
      "step": 11440
    },
    {
      "epoch": 1.8671180357288524,
      "grad_norm": 0.0019398524891585112,
      "learning_rate": 0.24338129316548046,
      "loss": 0.342,
      "num_input_tokens_seen": 21697504,
      "step": 11445
    },
    {
      "epoch": 1.8679337629496695,
      "grad_norm": 0.0016794500406831503,
      "learning_rate": 0.24333518782072444,
      "loss": 0.3216,
      "num_input_tokens_seen": 21707296,
      "step": 11450
    },
    {
      "epoch": 1.868749490170487,
      "grad_norm": 0.0009783896384760737,
      "learning_rate": 0.24328906808250952,
      "loss": 0.3089,
      "num_input_tokens_seen": 21716672,
      "step": 11455
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 0.0007495476165786386,
      "learning_rate": 0.243242933957948,
      "loss": 0.3325,
      "num_input_tokens_seen": 21725616,
      "step": 11460
    },
    {
      "epoch": 1.8703809446121218,
      "grad_norm": 0.00245479354634881,
      "learning_rate": 0.24319678545415427,
      "loss": 0.3359,
      "num_input_tokens_seen": 21735296,
      "step": 11465
    },
    {
      "epoch": 1.871196671832939,
      "grad_norm": 0.002362926257774234,
      "learning_rate": 0.24315062257824507,
      "loss": 0.3212,
      "num_input_tokens_seen": 21744480,
      "step": 11470
    },
    {
      "epoch": 1.8720123990537565,
      "grad_norm": 0.0015663010999560356,
      "learning_rate": 0.24310444533733921,
      "loss": 0.2858,
      "num_input_tokens_seen": 21753376,
      "step": 11475
    },
    {
      "epoch": 1.8728281262745736,
      "grad_norm": 0.0018837767420336604,
      "learning_rate": 0.2430582537385579,
      "loss": 0.333,
      "num_input_tokens_seen": 21761312,
      "step": 11480
    },
    {
      "epoch": 1.8736438534953912,
      "grad_norm": 0.0016417012084275484,
      "learning_rate": 0.2430120477890244,
      "loss": 0.3141,
      "num_input_tokens_seen": 21770304,
      "step": 11485
    },
    {
      "epoch": 1.8744595807162086,
      "grad_norm": 0.0017996624810621142,
      "learning_rate": 0.24296582749586426,
      "loss": 0.2933,
      "num_input_tokens_seen": 21778496,
      "step": 11490
    },
    {
      "epoch": 1.875275307937026,
      "grad_norm": 0.0028192815370857716,
      "learning_rate": 0.24291959286620526,
      "loss": 0.3802,
      "num_input_tokens_seen": 21789056,
      "step": 11495
    },
    {
      "epoch": 1.8760910351578433,
      "grad_norm": 0.000998629373498261,
      "learning_rate": 0.24287334390717738,
      "loss": 0.3349,
      "num_input_tokens_seen": 21799664,
      "step": 11500
    },
    {
      "epoch": 1.8769067623786606,
      "grad_norm": 0.0018390710465610027,
      "learning_rate": 0.24282708062591268,
      "loss": 0.3083,
      "num_input_tokens_seen": 21809376,
      "step": 11505
    },
    {
      "epoch": 1.877722489599478,
      "grad_norm": 0.001795196789316833,
      "learning_rate": 0.24278080302954563,
      "loss": 0.3595,
      "num_input_tokens_seen": 21819456,
      "step": 11510
    },
    {
      "epoch": 1.8785382168202953,
      "grad_norm": 0.0033840415999293327,
      "learning_rate": 0.24273451112521283,
      "loss": 0.3129,
      "num_input_tokens_seen": 21828976,
      "step": 11515
    },
    {
      "epoch": 1.8793539440411127,
      "grad_norm": 0.0024046481121331453,
      "learning_rate": 0.242688204920053,
      "loss": 0.4114,
      "num_input_tokens_seen": 21837632,
      "step": 11520
    },
    {
      "epoch": 1.88016967126193,
      "grad_norm": 0.0013600117526948452,
      "learning_rate": 0.24264188442120715,
      "loss": 0.3251,
      "num_input_tokens_seen": 21846944,
      "step": 11525
    },
    {
      "epoch": 1.8809853984827474,
      "grad_norm": 0.0010941976215690374,
      "learning_rate": 0.24259554963581853,
      "loss": 0.3518,
      "num_input_tokens_seen": 21855728,
      "step": 11530
    },
    {
      "epoch": 1.8818011257035647,
      "grad_norm": 0.001413918798789382,
      "learning_rate": 0.24254920057103257,
      "loss": 0.2999,
      "num_input_tokens_seen": 21866512,
      "step": 11535
    },
    {
      "epoch": 1.882616852924382,
      "grad_norm": 0.0016234806971624494,
      "learning_rate": 0.24250283723399685,
      "loss": 0.3395,
      "num_input_tokens_seen": 21875616,
      "step": 11540
    },
    {
      "epoch": 1.8834325801451994,
      "grad_norm": 0.003019687719643116,
      "learning_rate": 0.24245645963186108,
      "loss": 0.3066,
      "num_input_tokens_seen": 21883632,
      "step": 11545
    },
    {
      "epoch": 1.8842483073660168,
      "grad_norm": 0.0030794376507401466,
      "learning_rate": 0.2424100677717774,
      "loss": 0.3225,
      "num_input_tokens_seen": 21892544,
      "step": 11550
    },
    {
      "epoch": 1.8850640345868341,
      "grad_norm": 0.000933828589040786,
      "learning_rate": 0.24236366166090004,
      "loss": 0.1679,
      "num_input_tokens_seen": 21903408,
      "step": 11555
    },
    {
      "epoch": 1.8858797618076515,
      "grad_norm": 0.0032636343967169523,
      "learning_rate": 0.24231724130638527,
      "loss": 0.4918,
      "num_input_tokens_seen": 21911744,
      "step": 11560
    },
    {
      "epoch": 1.8866954890284688,
      "grad_norm": 0.0010951666627079248,
      "learning_rate": 0.2422708067153917,
      "loss": 0.3699,
      "num_input_tokens_seen": 21922016,
      "step": 11565
    },
    {
      "epoch": 1.8875112162492862,
      "grad_norm": 0.0008557241526432335,
      "learning_rate": 0.24222435789508026,
      "loss": 0.3513,
      "num_input_tokens_seen": 21932912,
      "step": 11570
    },
    {
      "epoch": 1.8883269434701035,
      "grad_norm": 0.0017476676730439067,
      "learning_rate": 0.24217789485261387,
      "loss": 0.3593,
      "num_input_tokens_seen": 21942864,
      "step": 11575
    },
    {
      "epoch": 1.8891426706909211,
      "grad_norm": 0.0012502750614657998,
      "learning_rate": 0.2421314175951577,
      "loss": 0.3321,
      "num_input_tokens_seen": 21953296,
      "step": 11580
    },
    {
      "epoch": 1.8899583979117383,
      "grad_norm": 0.0027794106863439083,
      "learning_rate": 0.2420849261298791,
      "loss": 0.2934,
      "num_input_tokens_seen": 21962480,
      "step": 11585
    },
    {
      "epoch": 1.8907741251325558,
      "grad_norm": 0.0045115333050489426,
      "learning_rate": 0.24203842046394775,
      "loss": 0.3388,
      "num_input_tokens_seen": 21972368,
      "step": 11590
    },
    {
      "epoch": 1.891589852353373,
      "grad_norm": 0.0025819200091063976,
      "learning_rate": 0.24199190060453535,
      "loss": 0.3147,
      "num_input_tokens_seen": 21981632,
      "step": 11595
    },
    {
      "epoch": 1.8924055795741905,
      "grad_norm": 0.0032012625597417355,
      "learning_rate": 0.2419453665588158,
      "loss": 0.3803,
      "num_input_tokens_seen": 21991984,
      "step": 11600
    },
    {
      "epoch": 1.8924055795741905,
      "eval_loss": 0.32890668511390686,
      "eval_runtime": 152.6384,
      "eval_samples_per_second": 17.853,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 21991984,
      "step": 11600
    },
    {
      "epoch": 1.8932213067950077,
      "grad_norm": 0.0020705382339656353,
      "learning_rate": 0.24189881833396523,
      "loss": 0.3637,
      "num_input_tokens_seen": 22001280,
      "step": 11605
    },
    {
      "epoch": 1.8940370340158252,
      "grad_norm": 0.0023454115726053715,
      "learning_rate": 0.24185225593716203,
      "loss": 0.3663,
      "num_input_tokens_seen": 22010336,
      "step": 11610
    },
    {
      "epoch": 1.8948527612366424,
      "grad_norm": 0.0027065437752753496,
      "learning_rate": 0.2418056793755867,
      "loss": 0.3251,
      "num_input_tokens_seen": 22020720,
      "step": 11615
    },
    {
      "epoch": 1.89566848845746,
      "grad_norm": 0.002446261001750827,
      "learning_rate": 0.24175908865642187,
      "loss": 0.3546,
      "num_input_tokens_seen": 22028400,
      "step": 11620
    },
    {
      "epoch": 1.896484215678277,
      "grad_norm": 0.001056081149727106,
      "learning_rate": 0.24171248378685248,
      "loss": 0.3551,
      "num_input_tokens_seen": 22038448,
      "step": 11625
    },
    {
      "epoch": 1.8972999428990946,
      "grad_norm": 0.003108436008915305,
      "learning_rate": 0.24166586477406554,
      "loss": 0.3692,
      "num_input_tokens_seen": 22048512,
      "step": 11630
    },
    {
      "epoch": 1.8981156701199118,
      "grad_norm": 0.0022683364804834127,
      "learning_rate": 0.24161923162525034,
      "loss": 0.3405,
      "num_input_tokens_seen": 22057440,
      "step": 11635
    },
    {
      "epoch": 1.8989313973407294,
      "grad_norm": 0.0011116544483229518,
      "learning_rate": 0.2415725843475982,
      "loss": 0.3598,
      "num_input_tokens_seen": 22066208,
      "step": 11640
    },
    {
      "epoch": 1.8997471245615465,
      "grad_norm": 0.0012627732940018177,
      "learning_rate": 0.24152592294830286,
      "loss": 0.2772,
      "num_input_tokens_seen": 22075792,
      "step": 11645
    },
    {
      "epoch": 1.900562851782364,
      "grad_norm": 0.0005611422238871455,
      "learning_rate": 0.24147924743455995,
      "loss": 0.3445,
      "num_input_tokens_seen": 22084752,
      "step": 11650
    },
    {
      "epoch": 1.9013785790031812,
      "grad_norm": 0.0032596909441053867,
      "learning_rate": 0.24143255781356754,
      "loss": 0.3189,
      "num_input_tokens_seen": 22093568,
      "step": 11655
    },
    {
      "epoch": 1.9021943062239988,
      "grad_norm": 0.0017137369140982628,
      "learning_rate": 0.24138585409252566,
      "loss": 0.4146,
      "num_input_tokens_seen": 22103296,
      "step": 11660
    },
    {
      "epoch": 1.903010033444816,
      "grad_norm": 0.0017617834964767098,
      "learning_rate": 0.24133913627863662,
      "loss": 0.3867,
      "num_input_tokens_seen": 22113792,
      "step": 11665
    },
    {
      "epoch": 1.9038257606656335,
      "grad_norm": 0.0017486201832070947,
      "learning_rate": 0.241292404379105,
      "loss": 0.3853,
      "num_input_tokens_seen": 22123648,
      "step": 11670
    },
    {
      "epoch": 1.9046414878864508,
      "grad_norm": 0.0016949379350990057,
      "learning_rate": 0.24124565840113735,
      "loss": 0.3628,
      "num_input_tokens_seen": 22133568,
      "step": 11675
    },
    {
      "epoch": 1.9054572151072682,
      "grad_norm": 0.00036950159119442105,
      "learning_rate": 0.2411988983519425,
      "loss": 0.3358,
      "num_input_tokens_seen": 22143152,
      "step": 11680
    },
    {
      "epoch": 1.9062729423280855,
      "grad_norm": 0.0015390811022371054,
      "learning_rate": 0.24115212423873145,
      "loss": 0.3906,
      "num_input_tokens_seen": 22152144,
      "step": 11685
    },
    {
      "epoch": 1.9070886695489029,
      "grad_norm": 0.001093006692826748,
      "learning_rate": 0.24110533606871737,
      "loss": 0.3121,
      "num_input_tokens_seen": 22161888,
      "step": 11690
    },
    {
      "epoch": 1.9079043967697202,
      "grad_norm": 0.0017842260422185063,
      "learning_rate": 0.24105853384911552,
      "loss": 0.3566,
      "num_input_tokens_seen": 22171072,
      "step": 11695
    },
    {
      "epoch": 1.9087201239905376,
      "grad_norm": 0.001931467093527317,
      "learning_rate": 0.24101171758714346,
      "loss": 0.3668,
      "num_input_tokens_seen": 22179632,
      "step": 11700
    },
    {
      "epoch": 1.909535851211355,
      "grad_norm": 0.0013806623173877597,
      "learning_rate": 0.24096488729002086,
      "loss": 0.3161,
      "num_input_tokens_seen": 22189072,
      "step": 11705
    },
    {
      "epoch": 1.9103515784321723,
      "grad_norm": 0.001138112274929881,
      "learning_rate": 0.24091804296496946,
      "loss": 0.3471,
      "num_input_tokens_seen": 22197632,
      "step": 11710
    },
    {
      "epoch": 1.9111673056529896,
      "grad_norm": 0.0014703955966979265,
      "learning_rate": 0.2408711846192133,
      "loss": 0.3584,
      "num_input_tokens_seen": 22207888,
      "step": 11715
    },
    {
      "epoch": 1.911983032873807,
      "grad_norm": 0.001067585777491331,
      "learning_rate": 0.24082431225997855,
      "loss": 0.247,
      "num_input_tokens_seen": 22216288,
      "step": 11720
    },
    {
      "epoch": 1.9127987600946244,
      "grad_norm": 0.0020015526097267866,
      "learning_rate": 0.24077742589449344,
      "loss": 0.3248,
      "num_input_tokens_seen": 22225664,
      "step": 11725
    },
    {
      "epoch": 1.9136144873154417,
      "grad_norm": 0.002042268170043826,
      "learning_rate": 0.24073052552998844,
      "loss": 0.3679,
      "num_input_tokens_seen": 22235200,
      "step": 11730
    },
    {
      "epoch": 1.914430214536259,
      "grad_norm": 0.001550660701468587,
      "learning_rate": 0.2406836111736963,
      "loss": 0.328,
      "num_input_tokens_seen": 22243216,
      "step": 11735
    },
    {
      "epoch": 1.9152459417570764,
      "grad_norm": 0.00156681751832366,
      "learning_rate": 0.2406366828328517,
      "loss": 0.3331,
      "num_input_tokens_seen": 22251968,
      "step": 11740
    },
    {
      "epoch": 1.9160616689778938,
      "grad_norm": 0.0017219537403434515,
      "learning_rate": 0.2405897405146915,
      "loss": 0.3281,
      "num_input_tokens_seen": 22261632,
      "step": 11745
    },
    {
      "epoch": 1.9168773961987111,
      "grad_norm": 0.002662660088390112,
      "learning_rate": 0.240542784226455,
      "loss": 0.3658,
      "num_input_tokens_seen": 22270448,
      "step": 11750
    },
    {
      "epoch": 1.9176931234195285,
      "grad_norm": 0.0029620365239679813,
      "learning_rate": 0.24049581397538328,
      "loss": 0.3492,
      "num_input_tokens_seen": 22279792,
      "step": 11755
    },
    {
      "epoch": 1.9185088506403458,
      "grad_norm": 0.0020717941224575043,
      "learning_rate": 0.24044882976871984,
      "loss": 0.3444,
      "num_input_tokens_seen": 22290160,
      "step": 11760
    },
    {
      "epoch": 1.9193245778611632,
      "grad_norm": 0.0018596602603793144,
      "learning_rate": 0.2404018316137102,
      "loss": 0.3584,
      "num_input_tokens_seen": 22300848,
      "step": 11765
    },
    {
      "epoch": 1.9201403050819805,
      "grad_norm": 0.004353513941168785,
      "learning_rate": 0.24035481951760204,
      "loss": 0.346,
      "num_input_tokens_seen": 22309184,
      "step": 11770
    },
    {
      "epoch": 1.920956032302798,
      "grad_norm": 0.0024045081809163094,
      "learning_rate": 0.2403077934876452,
      "loss": 0.3234,
      "num_input_tokens_seen": 22319632,
      "step": 11775
    },
    {
      "epoch": 1.9217717595236152,
      "grad_norm": 0.0008537769899703562,
      "learning_rate": 0.2402607535310918,
      "loss": 0.3441,
      "num_input_tokens_seen": 22327552,
      "step": 11780
    },
    {
      "epoch": 1.9225874867444328,
      "grad_norm": 0.0013143247924745083,
      "learning_rate": 0.2402136996551959,
      "loss": 0.3189,
      "num_input_tokens_seen": 22338112,
      "step": 11785
    },
    {
      "epoch": 1.92340321396525,
      "grad_norm": 0.0018237852491438389,
      "learning_rate": 0.24016663186721376,
      "loss": 0.3317,
      "num_input_tokens_seen": 22346576,
      "step": 11790
    },
    {
      "epoch": 1.9242189411860675,
      "grad_norm": 0.0022539410274475813,
      "learning_rate": 0.24011955017440395,
      "loss": 0.3489,
      "num_input_tokens_seen": 22355984,
      "step": 11795
    },
    {
      "epoch": 1.9250346684068846,
      "grad_norm": 0.0034274866338819265,
      "learning_rate": 0.24007245458402696,
      "loss": 0.3738,
      "num_input_tokens_seen": 22366624,
      "step": 11800
    },
    {
      "epoch": 1.9250346684068846,
      "eval_loss": 0.33501869440078735,
      "eval_runtime": 152.607,
      "eval_samples_per_second": 17.856,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 22366624,
      "step": 11800
    },
    {
      "epoch": 1.9258503956277022,
      "grad_norm": 0.001433099154382944,
      "learning_rate": 0.2400253451033456,
      "loss": 0.3392,
      "num_input_tokens_seen": 22376304,
      "step": 11805
    },
    {
      "epoch": 1.9266661228485193,
      "grad_norm": 0.0019253462087363005,
      "learning_rate": 0.23997822173962463,
      "loss": 0.3686,
      "num_input_tokens_seen": 22386448,
      "step": 11810
    },
    {
      "epoch": 1.927481850069337,
      "grad_norm": 0.0014029436279088259,
      "learning_rate": 0.23993108450013118,
      "loss": 0.3363,
      "num_input_tokens_seen": 22396912,
      "step": 11815
    },
    {
      "epoch": 1.928297577290154,
      "grad_norm": 0.002389009576290846,
      "learning_rate": 0.2398839333921343,
      "loss": 0.313,
      "num_input_tokens_seen": 22404944,
      "step": 11820
    },
    {
      "epoch": 1.9291133045109716,
      "grad_norm": 0.0010898554464802146,
      "learning_rate": 0.23983676842290536,
      "loss": 0.3265,
      "num_input_tokens_seen": 22414848,
      "step": 11825
    },
    {
      "epoch": 1.9299290317317888,
      "grad_norm": 0.0016364576295018196,
      "learning_rate": 0.2397895895997178,
      "loss": 0.2381,
      "num_input_tokens_seen": 22424256,
      "step": 11830
    },
    {
      "epoch": 1.9307447589526063,
      "grad_norm": 0.004400459583848715,
      "learning_rate": 0.23974239692984714,
      "loss": 0.3523,
      "num_input_tokens_seen": 22433488,
      "step": 11835
    },
    {
      "epoch": 1.9315604861734235,
      "grad_norm": 0.002622745931148529,
      "learning_rate": 0.2396951904205711,
      "loss": 0.2922,
      "num_input_tokens_seen": 22442704,
      "step": 11840
    },
    {
      "epoch": 1.932376213394241,
      "grad_norm": 0.0033518250565975904,
      "learning_rate": 0.23964797007916952,
      "loss": 0.3824,
      "num_input_tokens_seen": 22451056,
      "step": 11845
    },
    {
      "epoch": 1.9331919406150582,
      "grad_norm": 0.0012870935024693608,
      "learning_rate": 0.23960073591292436,
      "loss": 0.316,
      "num_input_tokens_seen": 22462016,
      "step": 11850
    },
    {
      "epoch": 1.9340076678358757,
      "grad_norm": 0.14679501950740814,
      "learning_rate": 0.2395534879291197,
      "loss": 0.3986,
      "num_input_tokens_seen": 22470128,
      "step": 11855
    },
    {
      "epoch": 1.934823395056693,
      "grad_norm": 0.0020531509071588516,
      "learning_rate": 0.23950622613504186,
      "loss": 0.3662,
      "num_input_tokens_seen": 22478720,
      "step": 11860
    },
    {
      "epoch": 1.9356391222775104,
      "grad_norm": 0.0019798618741333485,
      "learning_rate": 0.2394589505379791,
      "loss": 0.4068,
      "num_input_tokens_seen": 22488416,
      "step": 11865
    },
    {
      "epoch": 1.9364548494983278,
      "grad_norm": 0.002801400376483798,
      "learning_rate": 0.23941166114522197,
      "loss": 0.3542,
      "num_input_tokens_seen": 22497328,
      "step": 11870
    },
    {
      "epoch": 1.9372705767191452,
      "grad_norm": 0.002135330578312278,
      "learning_rate": 0.23936435796406308,
      "loss": 0.3812,
      "num_input_tokens_seen": 22506064,
      "step": 11875
    },
    {
      "epoch": 1.9380863039399625,
      "grad_norm": 0.0022931494750082493,
      "learning_rate": 0.23931704100179715,
      "loss": 0.3567,
      "num_input_tokens_seen": 22515760,
      "step": 11880
    },
    {
      "epoch": 1.9389020311607799,
      "grad_norm": 0.0017269582021981478,
      "learning_rate": 0.2392697102657211,
      "loss": 0.3089,
      "num_input_tokens_seen": 22525888,
      "step": 11885
    },
    {
      "epoch": 1.9397177583815972,
      "grad_norm": 0.0012613777071237564,
      "learning_rate": 0.23922236576313388,
      "loss": 0.3942,
      "num_input_tokens_seen": 22535936,
      "step": 11890
    },
    {
      "epoch": 1.9405334856024146,
      "grad_norm": 0.0010162158869206905,
      "learning_rate": 0.2391750075013366,
      "loss": 0.3689,
      "num_input_tokens_seen": 22544800,
      "step": 11895
    },
    {
      "epoch": 1.941349212823232,
      "grad_norm": 0.0008387927082367241,
      "learning_rate": 0.2391276354876326,
      "loss": 0.3433,
      "num_input_tokens_seen": 22554368,
      "step": 11900
    },
    {
      "epoch": 1.9421649400440493,
      "grad_norm": 0.0009052188252098858,
      "learning_rate": 0.23908024972932707,
      "loss": 0.3298,
      "num_input_tokens_seen": 22562864,
      "step": 11905
    },
    {
      "epoch": 1.9429806672648666,
      "grad_norm": 0.0011065510334447026,
      "learning_rate": 0.2390328502337276,
      "loss": 0.3674,
      "num_input_tokens_seen": 22573152,
      "step": 11910
    },
    {
      "epoch": 1.943796394485684,
      "grad_norm": 0.0010187607258558273,
      "learning_rate": 0.23898543700814376,
      "loss": 0.3686,
      "num_input_tokens_seen": 22581728,
      "step": 11915
    },
    {
      "epoch": 1.9446121217065013,
      "grad_norm": 0.000798387045506388,
      "learning_rate": 0.2389380100598873,
      "loss": 0.3653,
      "num_input_tokens_seen": 22590688,
      "step": 11920
    },
    {
      "epoch": 1.9454278489273187,
      "grad_norm": 0.0025687275920063257,
      "learning_rate": 0.23889056939627207,
      "loss": 0.3705,
      "num_input_tokens_seen": 22599808,
      "step": 11925
    },
    {
      "epoch": 1.946243576148136,
      "grad_norm": 0.0023285136558115482,
      "learning_rate": 0.23884311502461386,
      "loss": 0.3699,
      "num_input_tokens_seen": 22610128,
      "step": 11930
    },
    {
      "epoch": 1.9470593033689534,
      "grad_norm": 0.0018742562970146537,
      "learning_rate": 0.23879564695223088,
      "loss": 0.4027,
      "num_input_tokens_seen": 22618944,
      "step": 11935
    },
    {
      "epoch": 1.9478750305897707,
      "grad_norm": 0.0015558860031887889,
      "learning_rate": 0.23874816518644332,
      "loss": 0.3749,
      "num_input_tokens_seen": 22628912,
      "step": 11940
    },
    {
      "epoch": 1.948690757810588,
      "grad_norm": 0.0011425073025748134,
      "learning_rate": 0.23870066973457335,
      "loss": 0.3576,
      "num_input_tokens_seen": 22638880,
      "step": 11945
    },
    {
      "epoch": 1.9495064850314054,
      "grad_norm": 0.0017198253190144897,
      "learning_rate": 0.23865316060394545,
      "loss": 0.3736,
      "num_input_tokens_seen": 22648576,
      "step": 11950
    },
    {
      "epoch": 1.9503222122522228,
      "grad_norm": 0.0014963459689170122,
      "learning_rate": 0.2386056378018861,
      "loss": 0.3572,
      "num_input_tokens_seen": 22658112,
      "step": 11955
    },
    {
      "epoch": 1.9511379394730404,
      "grad_norm": 0.0007514870958402753,
      "learning_rate": 0.2385581013357239,
      "loss": 0.3368,
      "num_input_tokens_seen": 22666288,
      "step": 11960
    },
    {
      "epoch": 1.9519536666938575,
      "grad_norm": 0.0030612086411565542,
      "learning_rate": 0.23851055121278958,
      "loss": 0.3415,
      "num_input_tokens_seen": 22676720,
      "step": 11965
    },
    {
      "epoch": 1.952769393914675,
      "grad_norm": 0.0014656081330031157,
      "learning_rate": 0.23846298744041594,
      "loss": 0.3514,
      "num_input_tokens_seen": 22687424,
      "step": 11970
    },
    {
      "epoch": 1.9535851211354922,
      "grad_norm": 0.0029661189764738083,
      "learning_rate": 0.23841541002593802,
      "loss": 0.3392,
      "num_input_tokens_seen": 22697616,
      "step": 11975
    },
    {
      "epoch": 1.9544008483563098,
      "grad_norm": 0.0016323919408023357,
      "learning_rate": 0.23836781897669276,
      "loss": 0.3413,
      "num_input_tokens_seen": 22706688,
      "step": 11980
    },
    {
      "epoch": 1.955216575577127,
      "grad_norm": 0.0012871641665697098,
      "learning_rate": 0.23832021430001926,
      "loss": 0.3745,
      "num_input_tokens_seen": 22716016,
      "step": 11985
    },
    {
      "epoch": 1.9560323027979445,
      "grad_norm": 0.00108522386290133,
      "learning_rate": 0.2382725960032588,
      "loss": 0.3416,
      "num_input_tokens_seen": 22725936,
      "step": 11990
    },
    {
      "epoch": 1.9568480300187616,
      "grad_norm": 0.0031401009764522314,
      "learning_rate": 0.23822496409375482,
      "loss": 0.3031,
      "num_input_tokens_seen": 22735760,
      "step": 11995
    },
    {
      "epoch": 1.9576637572395792,
      "grad_norm": 0.0017925945576280355,
      "learning_rate": 0.2381773185788526,
      "loss": 0.3144,
      "num_input_tokens_seen": 22746000,
      "step": 12000
    },
    {
      "epoch": 1.9576637572395792,
      "eval_loss": 0.3475615084171295,
      "eval_runtime": 152.3559,
      "eval_samples_per_second": 17.886,
      "eval_steps_per_second": 8.946,
      "num_input_tokens_seen": 22746000,
      "step": 12000
    },
    {
      "epoch": 1.9584794844603963,
      "grad_norm": 0.001319893985055387,
      "learning_rate": 0.2381296594658998,
      "loss": 0.2306,
      "num_input_tokens_seen": 22754928,
      "step": 12005
    },
    {
      "epoch": 1.959295211681214,
      "grad_norm": 0.0027959428261965513,
      "learning_rate": 0.238081986762246,
      "loss": 0.5391,
      "num_input_tokens_seen": 22763280,
      "step": 12010
    },
    {
      "epoch": 1.960110938902031,
      "grad_norm": 0.0012892434606328607,
      "learning_rate": 0.23803430047524293,
      "loss": 0.3088,
      "num_input_tokens_seen": 22772368,
      "step": 12015
    },
    {
      "epoch": 1.9609266661228486,
      "grad_norm": 0.0015497779240831733,
      "learning_rate": 0.23798660061224441,
      "loss": 0.3682,
      "num_input_tokens_seen": 22781296,
      "step": 12020
    },
    {
      "epoch": 1.9617423933436657,
      "grad_norm": 0.0021416593808680773,
      "learning_rate": 0.23793888718060632,
      "loss": 0.3947,
      "num_input_tokens_seen": 22791264,
      "step": 12025
    },
    {
      "epoch": 1.9625581205644833,
      "grad_norm": 0.000873857585247606,
      "learning_rate": 0.23789116018768675,
      "loss": 0.3305,
      "num_input_tokens_seen": 22801136,
      "step": 12030
    },
    {
      "epoch": 1.9633738477853004,
      "grad_norm": 0.001280890661291778,
      "learning_rate": 0.2378434196408458,
      "loss": 0.3196,
      "num_input_tokens_seen": 22811792,
      "step": 12035
    },
    {
      "epoch": 1.964189575006118,
      "grad_norm": 0.0018694524187594652,
      "learning_rate": 0.23779566554744563,
      "loss": 0.3414,
      "num_input_tokens_seen": 22821616,
      "step": 12040
    },
    {
      "epoch": 1.9650053022269351,
      "grad_norm": 0.0013460420304909348,
      "learning_rate": 0.23774789791485051,
      "loss": 0.351,
      "num_input_tokens_seen": 22831968,
      "step": 12045
    },
    {
      "epoch": 1.9658210294477527,
      "grad_norm": 0.0016671480843797326,
      "learning_rate": 0.2377001167504268,
      "loss": 0.3549,
      "num_input_tokens_seen": 22841040,
      "step": 12050
    },
    {
      "epoch": 1.96663675666857,
      "grad_norm": 0.0007745596813037992,
      "learning_rate": 0.23765232206154302,
      "loss": 0.3396,
      "num_input_tokens_seen": 22849232,
      "step": 12055
    },
    {
      "epoch": 1.9674524838893874,
      "grad_norm": 0.002055006567388773,
      "learning_rate": 0.23760451385556966,
      "loss": 0.3418,
      "num_input_tokens_seen": 22859632,
      "step": 12060
    },
    {
      "epoch": 1.9682682111102048,
      "grad_norm": 0.0012625331291928887,
      "learning_rate": 0.23755669213987932,
      "loss": 0.3278,
      "num_input_tokens_seen": 22868944,
      "step": 12065
    },
    {
      "epoch": 1.9690839383310221,
      "grad_norm": 0.0016391054959967732,
      "learning_rate": 0.23750885692184676,
      "loss": 0.2957,
      "num_input_tokens_seen": 22878784,
      "step": 12070
    },
    {
      "epoch": 1.9698996655518395,
      "grad_norm": 0.001391268684528768,
      "learning_rate": 0.23746100820884875,
      "loss": 0.3314,
      "num_input_tokens_seen": 22888544,
      "step": 12075
    },
    {
      "epoch": 1.9707153927726568,
      "grad_norm": 0.0013348957290872931,
      "learning_rate": 0.23741314600826421,
      "loss": 0.359,
      "num_input_tokens_seen": 22897840,
      "step": 12080
    },
    {
      "epoch": 1.9715311199934742,
      "grad_norm": 0.0010373718105256557,
      "learning_rate": 0.23736527032747406,
      "loss": 0.3043,
      "num_input_tokens_seen": 22907472,
      "step": 12085
    },
    {
      "epoch": 1.9723468472142915,
      "grad_norm": 0.0009713033214211464,
      "learning_rate": 0.23731738117386128,
      "loss": 0.3261,
      "num_input_tokens_seen": 22916800,
      "step": 12090
    },
    {
      "epoch": 1.973162574435109,
      "grad_norm": 0.0024477029219269753,
      "learning_rate": 0.237269478554811,
      "loss": 0.3694,
      "num_input_tokens_seen": 22925840,
      "step": 12095
    },
    {
      "epoch": 1.9739783016559262,
      "grad_norm": 0.0012843416770920157,
      "learning_rate": 0.23722156247771053,
      "loss": 0.3644,
      "num_input_tokens_seen": 22935680,
      "step": 12100
    },
    {
      "epoch": 1.9747940288767436,
      "grad_norm": 0.0017443338874727488,
      "learning_rate": 0.23717363294994895,
      "loss": 0.3495,
      "num_input_tokens_seen": 22944144,
      "step": 12105
    },
    {
      "epoch": 1.975609756097561,
      "grad_norm": 0.004917157348245382,
      "learning_rate": 0.2371256899789177,
      "loss": 0.3972,
      "num_input_tokens_seen": 22954400,
      "step": 12110
    },
    {
      "epoch": 1.9764254833183783,
      "grad_norm": 0.004327651113271713,
      "learning_rate": 0.23707773357201017,
      "loss": 0.4198,
      "num_input_tokens_seen": 22963984,
      "step": 12115
    },
    {
      "epoch": 1.9772412105391957,
      "grad_norm": 0.0010898510226979852,
      "learning_rate": 0.2370297637366218,
      "loss": 0.289,
      "num_input_tokens_seen": 22972672,
      "step": 12120
    },
    {
      "epoch": 1.978056937760013,
      "grad_norm": 0.004517805762588978,
      "learning_rate": 0.23698178048015026,
      "loss": 0.4915,
      "num_input_tokens_seen": 22982464,
      "step": 12125
    },
    {
      "epoch": 1.9788726649808304,
      "grad_norm": 0.0006289648590609431,
      "learning_rate": 0.236933783809995,
      "loss": 0.3447,
      "num_input_tokens_seen": 22991936,
      "step": 12130
    },
    {
      "epoch": 1.9796883922016477,
      "grad_norm": 0.0009801106061786413,
      "learning_rate": 0.23688577373355785,
      "loss": 0.3368,
      "num_input_tokens_seen": 23002128,
      "step": 12135
    },
    {
      "epoch": 1.980504119422465,
      "grad_norm": 0.0013267174363136292,
      "learning_rate": 0.23683775025824247,
      "loss": 0.3377,
      "num_input_tokens_seen": 23011920,
      "step": 12140
    },
    {
      "epoch": 1.9813198466432826,
      "grad_norm": 0.003554630558937788,
      "learning_rate": 0.2367897133914548,
      "loss": 0.3633,
      "num_input_tokens_seen": 23021216,
      "step": 12145
    },
    {
      "epoch": 1.9821355738640998,
      "grad_norm": 0.0010422006016597152,
      "learning_rate": 0.2367416631406026,
      "loss": 0.3324,
      "num_input_tokens_seen": 23029616,
      "step": 12150
    },
    {
      "epoch": 1.9829513010849174,
      "grad_norm": 0.0023517711088061333,
      "learning_rate": 0.23669359951309588,
      "loss": 0.3309,
      "num_input_tokens_seen": 23038304,
      "step": 12155
    },
    {
      "epoch": 1.9837670283057345,
      "grad_norm": 0.003264110768213868,
      "learning_rate": 0.23664552251634666,
      "loss": 0.337,
      "num_input_tokens_seen": 23047536,
      "step": 12160
    },
    {
      "epoch": 1.984582755526552,
      "grad_norm": 0.003935564309358597,
      "learning_rate": 0.23659743215776907,
      "loss": 0.3646,
      "num_input_tokens_seen": 23057456,
      "step": 12165
    },
    {
      "epoch": 1.9853984827473692,
      "grad_norm": 0.0018364995485171676,
      "learning_rate": 0.23654932844477908,
      "loss": 0.3441,
      "num_input_tokens_seen": 23066496,
      "step": 12170
    },
    {
      "epoch": 1.9862142099681868,
      "grad_norm": 0.001951990183442831,
      "learning_rate": 0.23650121138479507,
      "loss": 0.3035,
      "num_input_tokens_seen": 23076256,
      "step": 12175
    },
    {
      "epoch": 1.987029937189004,
      "grad_norm": 0.0015532418619841337,
      "learning_rate": 0.23645308098523724,
      "loss": 0.3954,
      "num_input_tokens_seen": 23084976,
      "step": 12180
    },
    {
      "epoch": 1.9878456644098215,
      "grad_norm": 0.0023117074742913246,
      "learning_rate": 0.23640493725352785,
      "loss": 0.3464,
      "num_input_tokens_seen": 23094816,
      "step": 12185
    },
    {
      "epoch": 1.9886613916306386,
      "grad_norm": 0.0010050158016383648,
      "learning_rate": 0.2363567801970913,
      "loss": 0.3364,
      "num_input_tokens_seen": 23103552,
      "step": 12190
    },
    {
      "epoch": 1.9894771188514562,
      "grad_norm": 0.0017211806261911988,
      "learning_rate": 0.236308609823354,
      "loss": 0.3637,
      "num_input_tokens_seen": 23113184,
      "step": 12195
    },
    {
      "epoch": 1.9902928460722733,
      "grad_norm": 0.0011832349700853229,
      "learning_rate": 0.23626042613974452,
      "loss": 0.3205,
      "num_input_tokens_seen": 23122688,
      "step": 12200
    },
    {
      "epoch": 1.9902928460722733,
      "eval_loss": 0.3513506352901459,
      "eval_runtime": 152.4182,
      "eval_samples_per_second": 17.878,
      "eval_steps_per_second": 8.943,
      "num_input_tokens_seen": 23122688,
      "step": 12200
    },
    {
      "epoch": 1.9911085732930909,
      "grad_norm": 0.0018990852404385805,
      "learning_rate": 0.23621222915369325,
      "loss": 0.3228,
      "num_input_tokens_seen": 23131520,
      "step": 12205
    },
    {
      "epoch": 1.991924300513908,
      "grad_norm": 0.0029491668101400137,
      "learning_rate": 0.23616401887263283,
      "loss": 0.3495,
      "num_input_tokens_seen": 23141248,
      "step": 12210
    },
    {
      "epoch": 1.9927400277347256,
      "grad_norm": 0.0028946036472916603,
      "learning_rate": 0.23611579530399793,
      "loss": 0.386,
      "num_input_tokens_seen": 23150080,
      "step": 12215
    },
    {
      "epoch": 1.9935557549555427,
      "grad_norm": 0.003068783087655902,
      "learning_rate": 0.23606755845522517,
      "loss": 0.3594,
      "num_input_tokens_seen": 23159936,
      "step": 12220
    },
    {
      "epoch": 1.9943714821763603,
      "grad_norm": 0.0024236489553004503,
      "learning_rate": 0.23601930833375329,
      "loss": 0.3401,
      "num_input_tokens_seen": 23168464,
      "step": 12225
    },
    {
      "epoch": 1.9951872093971774,
      "grad_norm": 0.0017310632392764091,
      "learning_rate": 0.23597104494702312,
      "loss": 0.2921,
      "num_input_tokens_seen": 23177600,
      "step": 12230
    },
    {
      "epoch": 1.996002936617995,
      "grad_norm": 0.0029869454447180033,
      "learning_rate": 0.23592276830247744,
      "loss": 0.3876,
      "num_input_tokens_seen": 23185872,
      "step": 12235
    },
    {
      "epoch": 1.9968186638388123,
      "grad_norm": 0.0023795003071427345,
      "learning_rate": 0.2358744784075611,
      "loss": 0.3407,
      "num_input_tokens_seen": 23194848,
      "step": 12240
    },
    {
      "epoch": 1.9976343910596297,
      "grad_norm": 0.0013272813521325588,
      "learning_rate": 0.235826175269721,
      "loss": 0.3015,
      "num_input_tokens_seen": 23203952,
      "step": 12245
    },
    {
      "epoch": 1.998450118280447,
      "grad_norm": 0.0019670468755066395,
      "learning_rate": 0.23577785889640612,
      "loss": 0.3495,
      "num_input_tokens_seen": 23213760,
      "step": 12250
    },
    {
      "epoch": 1.9992658455012644,
      "grad_norm": 0.001661204732954502,
      "learning_rate": 0.23572952929506744,
      "loss": 0.3247,
      "num_input_tokens_seen": 23224064,
      "step": 12255
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.002564790192991495,
      "learning_rate": 0.23568118647315803,
      "loss": 0.3926,
      "num_input_tokens_seen": 23232704,
      "step": 12260
    },
    {
      "epoch": 2.0008157272208176,
      "grad_norm": 0.0008393545867875218,
      "learning_rate": 0.23563283043813296,
      "loss": 0.3568,
      "num_input_tokens_seen": 23241536,
      "step": 12265
    },
    {
      "epoch": 2.0016314544416347,
      "grad_norm": 0.0015428097685799003,
      "learning_rate": 0.23558446119744922,
      "loss": 0.3975,
      "num_input_tokens_seen": 23249984,
      "step": 12270
    },
    {
      "epoch": 2.0024471816624523,
      "grad_norm": 0.0015803177375346422,
      "learning_rate": 0.23553607875856608,
      "loss": 0.3527,
      "num_input_tokens_seen": 23260144,
      "step": 12275
    },
    {
      "epoch": 2.0032629088832694,
      "grad_norm": 0.0024982155300676823,
      "learning_rate": 0.2354876831289447,
      "loss": 0.3833,
      "num_input_tokens_seen": 23268880,
      "step": 12280
    },
    {
      "epoch": 2.004078636104087,
      "grad_norm": 0.002970847999677062,
      "learning_rate": 0.23543927431604827,
      "loss": 0.3835,
      "num_input_tokens_seen": 23277056,
      "step": 12285
    },
    {
      "epoch": 2.004894363324904,
      "grad_norm": 0.002108647720888257,
      "learning_rate": 0.23539085232734203,
      "loss": 0.3769,
      "num_input_tokens_seen": 23286144,
      "step": 12290
    },
    {
      "epoch": 2.0057100905457217,
      "grad_norm": 0.0011940093245357275,
      "learning_rate": 0.2353424171702933,
      "loss": 0.3242,
      "num_input_tokens_seen": 23294688,
      "step": 12295
    },
    {
      "epoch": 2.006525817766539,
      "grad_norm": 0.0030293543823063374,
      "learning_rate": 0.23529396885237133,
      "loss": 0.3777,
      "num_input_tokens_seen": 23304672,
      "step": 12300
    },
    {
      "epoch": 2.0073415449873564,
      "grad_norm": 0.002384280087426305,
      "learning_rate": 0.2352455073810475,
      "loss": 0.3643,
      "num_input_tokens_seen": 23314544,
      "step": 12305
    },
    {
      "epoch": 2.0081572722081735,
      "grad_norm": 0.000993159832432866,
      "learning_rate": 0.23519703276379517,
      "loss": 0.3318,
      "num_input_tokens_seen": 23324848,
      "step": 12310
    },
    {
      "epoch": 2.008972999428991,
      "grad_norm": 0.0014423042302951217,
      "learning_rate": 0.2351485450080897,
      "loss": 0.3355,
      "num_input_tokens_seen": 23333216,
      "step": 12315
    },
    {
      "epoch": 2.0097887266498082,
      "grad_norm": 0.0016775570111349225,
      "learning_rate": 0.2351000441214086,
      "loss": 0.3177,
      "num_input_tokens_seen": 23343200,
      "step": 12320
    },
    {
      "epoch": 2.010604453870626,
      "grad_norm": 0.0014153398806229234,
      "learning_rate": 0.23505153011123125,
      "loss": 0.3285,
      "num_input_tokens_seen": 23352736,
      "step": 12325
    },
    {
      "epoch": 2.011420181091443,
      "grad_norm": 0.0011734738945960999,
      "learning_rate": 0.23500300298503912,
      "loss": 0.302,
      "num_input_tokens_seen": 23362416,
      "step": 12330
    },
    {
      "epoch": 2.0122359083122605,
      "grad_norm": 0.004068582318723202,
      "learning_rate": 0.23495446275031576,
      "loss": 0.3728,
      "num_input_tokens_seen": 23370656,
      "step": 12335
    },
    {
      "epoch": 2.0130516355330776,
      "grad_norm": 0.0013550483854487538,
      "learning_rate": 0.2349059094145466,
      "loss": 0.3246,
      "num_input_tokens_seen": 23380288,
      "step": 12340
    },
    {
      "epoch": 2.013867362753895,
      "grad_norm": 0.0018603153293952346,
      "learning_rate": 0.2348573429852192,
      "loss": 0.3214,
      "num_input_tokens_seen": 23390272,
      "step": 12345
    },
    {
      "epoch": 2.0146830899747123,
      "grad_norm": 0.0022521705832332373,
      "learning_rate": 0.23480876346982313,
      "loss": 0.2982,
      "num_input_tokens_seen": 23399744,
      "step": 12350
    },
    {
      "epoch": 2.01549881719553,
      "grad_norm": 0.00228908215649426,
      "learning_rate": 0.23476017087585,
      "loss": 0.3008,
      "num_input_tokens_seen": 23409616,
      "step": 12355
    },
    {
      "epoch": 2.016314544416347,
      "grad_norm": 0.0019239324610680342,
      "learning_rate": 0.23471156521079334,
      "loss": 0.3313,
      "num_input_tokens_seen": 23418064,
      "step": 12360
    },
    {
      "epoch": 2.0171302716371646,
      "grad_norm": 0.0023530046455562115,
      "learning_rate": 0.23466294648214875,
      "loss": 0.3356,
      "num_input_tokens_seen": 23428160,
      "step": 12365
    },
    {
      "epoch": 2.0179459988579818,
      "grad_norm": 0.0018849793123081326,
      "learning_rate": 0.2346143146974139,
      "loss": 0.3392,
      "num_input_tokens_seen": 23437760,
      "step": 12370
    },
    {
      "epoch": 2.0187617260787993,
      "grad_norm": 0.0015339229721575975,
      "learning_rate": 0.23456566986408836,
      "loss": 0.3367,
      "num_input_tokens_seen": 23445584,
      "step": 12375
    },
    {
      "epoch": 2.0195774532996165,
      "grad_norm": 0.0026114440988749266,
      "learning_rate": 0.23451701198967384,
      "loss": 0.3796,
      "num_input_tokens_seen": 23454704,
      "step": 12380
    },
    {
      "epoch": 2.020393180520434,
      "grad_norm": 0.0014122298453003168,
      "learning_rate": 0.23446834108167397,
      "loss": 0.3878,
      "num_input_tokens_seen": 23464720,
      "step": 12385
    },
    {
      "epoch": 2.021208907741251,
      "grad_norm": 0.0015638089971616864,
      "learning_rate": 0.23441965714759438,
      "loss": 0.3406,
      "num_input_tokens_seen": 23473888,
      "step": 12390
    },
    {
      "epoch": 2.0220246349620687,
      "grad_norm": 0.0026458457577973604,
      "learning_rate": 0.23437096019494277,
      "loss": 0.4629,
      "num_input_tokens_seen": 23483536,
      "step": 12395
    },
    {
      "epoch": 2.022840362182886,
      "grad_norm": 0.0019662796985358,
      "learning_rate": 0.23432225023122885,
      "loss": 0.3868,
      "num_input_tokens_seen": 23494112,
      "step": 12400
    },
    {
      "epoch": 2.022840362182886,
      "eval_loss": 0.33898940682411194,
      "eval_runtime": 152.437,
      "eval_samples_per_second": 17.876,
      "eval_steps_per_second": 8.941,
      "num_input_tokens_seen": 23494112,
      "step": 12400
    },
    {
      "epoch": 2.0236560894037035,
      "grad_norm": 0.0032534771598875523,
      "learning_rate": 0.23427352726396428,
      "loss": 0.3795,
      "num_input_tokens_seen": 23503232,
      "step": 12405
    },
    {
      "epoch": 2.0244718166245206,
      "grad_norm": 0.0027075426187366247,
      "learning_rate": 0.2342247913006628,
      "loss": 0.4186,
      "num_input_tokens_seen": 23513472,
      "step": 12410
    },
    {
      "epoch": 2.025287543845338,
      "grad_norm": 0.002185160294175148,
      "learning_rate": 0.23417604234883999,
      "loss": 0.333,
      "num_input_tokens_seen": 23522640,
      "step": 12415
    },
    {
      "epoch": 2.0261032710661553,
      "grad_norm": 0.0008639141451567411,
      "learning_rate": 0.23412728041601363,
      "loss": 0.3169,
      "num_input_tokens_seen": 23532784,
      "step": 12420
    },
    {
      "epoch": 2.026918998286973,
      "grad_norm": 0.0008510379120707512,
      "learning_rate": 0.23407850550970347,
      "loss": 0.4295,
      "num_input_tokens_seen": 23541408,
      "step": 12425
    },
    {
      "epoch": 2.02773472550779,
      "grad_norm": 0.0008258064044639468,
      "learning_rate": 0.23402971763743116,
      "loss": 0.3117,
      "num_input_tokens_seen": 23551872,
      "step": 12430
    },
    {
      "epoch": 2.0285504527286076,
      "grad_norm": 0.0008321022614836693,
      "learning_rate": 0.23398091680672037,
      "loss": 0.3155,
      "num_input_tokens_seen": 23559936,
      "step": 12435
    },
    {
      "epoch": 2.0293661799494247,
      "grad_norm": 0.0032397648319602013,
      "learning_rate": 0.23393210302509687,
      "loss": 0.3716,
      "num_input_tokens_seen": 23570208,
      "step": 12440
    },
    {
      "epoch": 2.0301819071702423,
      "grad_norm": 0.0034830914810299873,
      "learning_rate": 0.23388327630008832,
      "loss": 0.3261,
      "num_input_tokens_seen": 23580000,
      "step": 12445
    },
    {
      "epoch": 2.03099763439106,
      "grad_norm": 0.0021577407605946064,
      "learning_rate": 0.23383443663922443,
      "loss": 0.2877,
      "num_input_tokens_seen": 23589824,
      "step": 12450
    },
    {
      "epoch": 2.031813361611877,
      "grad_norm": 0.0012497339630499482,
      "learning_rate": 0.23378558405003685,
      "loss": 0.432,
      "num_input_tokens_seen": 23599408,
      "step": 12455
    },
    {
      "epoch": 2.0326290888326946,
      "grad_norm": 0.0014503187267109752,
      "learning_rate": 0.2337367185400593,
      "loss": 0.3106,
      "num_input_tokens_seen": 23608720,
      "step": 12460
    },
    {
      "epoch": 2.0334448160535117,
      "grad_norm": 0.0028789055068045855,
      "learning_rate": 0.23368784011682747,
      "loss": 0.3251,
      "num_input_tokens_seen": 23619232,
      "step": 12465
    },
    {
      "epoch": 2.0342605432743293,
      "grad_norm": 0.002217155648395419,
      "learning_rate": 0.23363894878787902,
      "loss": 0.3552,
      "num_input_tokens_seen": 23628864,
      "step": 12470
    },
    {
      "epoch": 2.0350762704951464,
      "grad_norm": 0.0017143961740657687,
      "learning_rate": 0.23359004456075352,
      "loss": 0.3493,
      "num_input_tokens_seen": 23638768,
      "step": 12475
    },
    {
      "epoch": 2.035891997715964,
      "grad_norm": 0.0008811427396722138,
      "learning_rate": 0.23354112744299277,
      "loss": 0.3232,
      "num_input_tokens_seen": 23649056,
      "step": 12480
    },
    {
      "epoch": 2.036707724936781,
      "grad_norm": 0.0013579843798652291,
      "learning_rate": 0.2334921974421403,
      "loss": 0.3287,
      "num_input_tokens_seen": 23658096,
      "step": 12485
    },
    {
      "epoch": 2.0375234521575987,
      "grad_norm": 0.0009923195466399193,
      "learning_rate": 0.23344325456574178,
      "loss": 0.3915,
      "num_input_tokens_seen": 23668096,
      "step": 12490
    },
    {
      "epoch": 2.038339179378416,
      "grad_norm": 0.0013736215187236667,
      "learning_rate": 0.23339429882134477,
      "loss": 0.372,
      "num_input_tokens_seen": 23676832,
      "step": 12495
    },
    {
      "epoch": 2.0391549065992334,
      "grad_norm": 0.0008896479266695678,
      "learning_rate": 0.23334533021649884,
      "loss": 0.3224,
      "num_input_tokens_seen": 23685328,
      "step": 12500
    },
    {
      "epoch": 2.0399706338200505,
      "grad_norm": 0.0007154257618822157,
      "learning_rate": 0.23329634875875566,
      "loss": 0.3248,
      "num_input_tokens_seen": 23695936,
      "step": 12505
    },
    {
      "epoch": 2.040786361040868,
      "grad_norm": 0.0014746417291462421,
      "learning_rate": 0.23324735445566874,
      "loss": 0.3409,
      "num_input_tokens_seen": 23706288,
      "step": 12510
    },
    {
      "epoch": 2.041602088261685,
      "grad_norm": 0.0011726216180250049,
      "learning_rate": 0.2331983473147936,
      "loss": 0.3475,
      "num_input_tokens_seen": 23715744,
      "step": 12515
    },
    {
      "epoch": 2.042417815482503,
      "grad_norm": 0.0011965828016400337,
      "learning_rate": 0.23314932734368776,
      "loss": 0.3361,
      "num_input_tokens_seen": 23725968,
      "step": 12520
    },
    {
      "epoch": 2.04323354270332,
      "grad_norm": 0.0017847036942839622,
      "learning_rate": 0.2331002945499107,
      "loss": 0.3165,
      "num_input_tokens_seen": 23735664,
      "step": 12525
    },
    {
      "epoch": 2.0440492699241375,
      "grad_norm": 0.0015287469141185284,
      "learning_rate": 0.23305124894102397,
      "loss": 0.3631,
      "num_input_tokens_seen": 23745520,
      "step": 12530
    },
    {
      "epoch": 2.0448649971449546,
      "grad_norm": 0.0012942691100761294,
      "learning_rate": 0.23300219052459092,
      "loss": 0.3624,
      "num_input_tokens_seen": 23754720,
      "step": 12535
    },
    {
      "epoch": 2.045680724365772,
      "grad_norm": 0.002329540438950062,
      "learning_rate": 0.23295311930817708,
      "loss": 0.3505,
      "num_input_tokens_seen": 23764432,
      "step": 12540
    },
    {
      "epoch": 2.0464964515865893,
      "grad_norm": 0.0014497085940092802,
      "learning_rate": 0.23290403529934972,
      "loss": 0.3605,
      "num_input_tokens_seen": 23773728,
      "step": 12545
    },
    {
      "epoch": 2.047312178807407,
      "grad_norm": 0.0012156631564721465,
      "learning_rate": 0.23285493850567832,
      "loss": 0.3491,
      "num_input_tokens_seen": 23782736,
      "step": 12550
    },
    {
      "epoch": 2.048127906028224,
      "grad_norm": 0.0014586709439754486,
      "learning_rate": 0.23280582893473414,
      "loss": 0.3599,
      "num_input_tokens_seen": 23791520,
      "step": 12555
    },
    {
      "epoch": 2.0489436332490416,
      "grad_norm": 0.001527779153548181,
      "learning_rate": 0.2327567065940906,
      "loss": 0.3181,
      "num_input_tokens_seen": 23800608,
      "step": 12560
    },
    {
      "epoch": 2.0497593604698587,
      "grad_norm": 0.0014585291501134634,
      "learning_rate": 0.23270757149132285,
      "loss": 0.3381,
      "num_input_tokens_seen": 23810384,
      "step": 12565
    },
    {
      "epoch": 2.0505750876906763,
      "grad_norm": 0.0019747100304812193,
      "learning_rate": 0.23265842363400827,
      "loss": 0.402,
      "num_input_tokens_seen": 23819856,
      "step": 12570
    },
    {
      "epoch": 2.0513908149114934,
      "grad_norm": 0.0019897243473678827,
      "learning_rate": 0.23260926302972595,
      "loss": 0.3087,
      "num_input_tokens_seen": 23828912,
      "step": 12575
    },
    {
      "epoch": 2.052206542132311,
      "grad_norm": 0.001944493385963142,
      "learning_rate": 0.2325600896860572,
      "loss": 0.3734,
      "num_input_tokens_seen": 23839168,
      "step": 12580
    },
    {
      "epoch": 2.053022269353128,
      "grad_norm": 0.001936294138431549,
      "learning_rate": 0.23251090361058505,
      "loss": 0.336,
      "num_input_tokens_seen": 23848752,
      "step": 12585
    },
    {
      "epoch": 2.0538379965739457,
      "grad_norm": 0.002362916711717844,
      "learning_rate": 0.23246170481089476,
      "loss": 0.3653,
      "num_input_tokens_seen": 23857120,
      "step": 12590
    },
    {
      "epoch": 2.054653723794763,
      "grad_norm": 0.0004920873907394707,
      "learning_rate": 0.23241249329457317,
      "loss": 0.3153,
      "num_input_tokens_seen": 23866592,
      "step": 12595
    },
    {
      "epoch": 2.0554694510155804,
      "grad_norm": 0.0010641812114045024,
      "learning_rate": 0.23236326906920957,
      "loss": 0.3732,
      "num_input_tokens_seen": 23876160,
      "step": 12600
    },
    {
      "epoch": 2.0554694510155804,
      "eval_loss": 0.33186185359954834,
      "eval_runtime": 152.4069,
      "eval_samples_per_second": 17.88,
      "eval_steps_per_second": 8.943,
      "num_input_tokens_seen": 23876160,
      "step": 12600
    },
    {
      "epoch": 2.0562851782363976,
      "grad_norm": 0.0012404299341142178,
      "learning_rate": 0.2323140321423948,
      "loss": 0.3432,
      "num_input_tokens_seen": 23885344,
      "step": 12605
    },
    {
      "epoch": 2.057100905457215,
      "grad_norm": 0.0025474324356764555,
      "learning_rate": 0.23226478252172184,
      "loss": 0.3555,
      "num_input_tokens_seen": 23895024,
      "step": 12610
    },
    {
      "epoch": 2.0579166326780323,
      "grad_norm": 0.0008479031384922564,
      "learning_rate": 0.23221552021478561,
      "loss": 0.3302,
      "num_input_tokens_seen": 23905376,
      "step": 12615
    },
    {
      "epoch": 2.05873235989885,
      "grad_norm": 0.0003696276980917901,
      "learning_rate": 0.232166245229183,
      "loss": 0.3468,
      "num_input_tokens_seen": 23914432,
      "step": 12620
    },
    {
      "epoch": 2.059548087119667,
      "grad_norm": 0.0007972944877110422,
      "learning_rate": 0.2321169575725128,
      "loss": 0.3646,
      "num_input_tokens_seen": 23922560,
      "step": 12625
    },
    {
      "epoch": 2.0603638143404845,
      "grad_norm": 0.0016905618831515312,
      "learning_rate": 0.23206765725237577,
      "loss": 0.2996,
      "num_input_tokens_seen": 23933008,
      "step": 12630
    },
    {
      "epoch": 2.0611795415613017,
      "grad_norm": 0.0018836238887161016,
      "learning_rate": 0.2320183442763747,
      "loss": 0.3493,
      "num_input_tokens_seen": 23942080,
      "step": 12635
    },
    {
      "epoch": 2.0619952687821193,
      "grad_norm": 0.0011434301268309355,
      "learning_rate": 0.23196901865211422,
      "loss": 0.3062,
      "num_input_tokens_seen": 23952608,
      "step": 12640
    },
    {
      "epoch": 2.062810996002937,
      "grad_norm": 0.0012325126444920897,
      "learning_rate": 0.231919680387201,
      "loss": 0.3363,
      "num_input_tokens_seen": 23962848,
      "step": 12645
    },
    {
      "epoch": 2.063626723223754,
      "grad_norm": 0.0020462102256715298,
      "learning_rate": 0.23187032948924358,
      "loss": 0.304,
      "num_input_tokens_seen": 23972592,
      "step": 12650
    },
    {
      "epoch": 2.0644424504445715,
      "grad_norm": 0.0016621487447991967,
      "learning_rate": 0.23182096596585247,
      "loss": 0.3127,
      "num_input_tokens_seen": 23981776,
      "step": 12655
    },
    {
      "epoch": 2.0652581776653887,
      "grad_norm": 0.0010760148288682103,
      "learning_rate": 0.23177158982464025,
      "loss": 0.3382,
      "num_input_tokens_seen": 23991904,
      "step": 12660
    },
    {
      "epoch": 2.0660739048862062,
      "grad_norm": 0.000987932668067515,
      "learning_rate": 0.23172220107322122,
      "loss": 0.383,
      "num_input_tokens_seen": 24002896,
      "step": 12665
    },
    {
      "epoch": 2.0668896321070234,
      "grad_norm": 0.0018181346822530031,
      "learning_rate": 0.23167279971921184,
      "loss": 0.3004,
      "num_input_tokens_seen": 24012224,
      "step": 12670
    },
    {
      "epoch": 2.067705359327841,
      "grad_norm": 0.001699586515314877,
      "learning_rate": 0.23162338577023034,
      "loss": 0.3449,
      "num_input_tokens_seen": 24020672,
      "step": 12675
    },
    {
      "epoch": 2.068521086548658,
      "grad_norm": 0.002949852729216218,
      "learning_rate": 0.23157395923389704,
      "loss": 0.3146,
      "num_input_tokens_seen": 24029024,
      "step": 12680
    },
    {
      "epoch": 2.0693368137694756,
      "grad_norm": 0.0011900492245331407,
      "learning_rate": 0.2315245201178341,
      "loss": 0.3187,
      "num_input_tokens_seen": 24039568,
      "step": 12685
    },
    {
      "epoch": 2.0701525409902928,
      "grad_norm": 0.0021671925205737352,
      "learning_rate": 0.23147506842966564,
      "loss": 0.3267,
      "num_input_tokens_seen": 24049088,
      "step": 12690
    },
    {
      "epoch": 2.0709682682111104,
      "grad_norm": 0.0010725357569754124,
      "learning_rate": 0.23142560417701774,
      "loss": 0.3662,
      "num_input_tokens_seen": 24058928,
      "step": 12695
    },
    {
      "epoch": 2.0717839954319275,
      "grad_norm": 0.0020288412924855947,
      "learning_rate": 0.23137612736751845,
      "loss": 0.2979,
      "num_input_tokens_seen": 24068896,
      "step": 12700
    },
    {
      "epoch": 2.072599722652745,
      "grad_norm": 0.0030300235375761986,
      "learning_rate": 0.23132663800879766,
      "loss": 0.3498,
      "num_input_tokens_seen": 24078992,
      "step": 12705
    },
    {
      "epoch": 2.073415449873562,
      "grad_norm": 0.0016016741283237934,
      "learning_rate": 0.2312771361084873,
      "loss": 0.3073,
      "num_input_tokens_seen": 24088864,
      "step": 12710
    },
    {
      "epoch": 2.0742311770943798,
      "grad_norm": 0.0016069182893261313,
      "learning_rate": 0.23122762167422112,
      "loss": 0.3305,
      "num_input_tokens_seen": 24099056,
      "step": 12715
    },
    {
      "epoch": 2.075046904315197,
      "grad_norm": 0.0013898536562919617,
      "learning_rate": 0.23117809471363493,
      "loss": 0.2785,
      "num_input_tokens_seen": 24108608,
      "step": 12720
    },
    {
      "epoch": 2.0758626315360145,
      "grad_norm": 0.00037363756564445794,
      "learning_rate": 0.23112855523436637,
      "loss": 0.329,
      "num_input_tokens_seen": 24117696,
      "step": 12725
    },
    {
      "epoch": 2.0766783587568316,
      "grad_norm": 0.005499252118170261,
      "learning_rate": 0.23107900324405511,
      "loss": 0.3811,
      "num_input_tokens_seen": 24128272,
      "step": 12730
    },
    {
      "epoch": 2.077494085977649,
      "grad_norm": 0.0015117103466764092,
      "learning_rate": 0.2310294387503426,
      "loss": 0.3012,
      "num_input_tokens_seen": 24137264,
      "step": 12735
    },
    {
      "epoch": 2.0783098131984663,
      "grad_norm": 0.0007153202896006405,
      "learning_rate": 0.23097986176087237,
      "loss": 0.3518,
      "num_input_tokens_seen": 24145808,
      "step": 12740
    },
    {
      "epoch": 2.079125540419284,
      "grad_norm": 0.0016263973666355014,
      "learning_rate": 0.23093027228328986,
      "loss": 0.3603,
      "num_input_tokens_seen": 24155888,
      "step": 12745
    },
    {
      "epoch": 2.079941267640101,
      "grad_norm": 0.0022584160324186087,
      "learning_rate": 0.23088067032524226,
      "loss": 0.3557,
      "num_input_tokens_seen": 24166784,
      "step": 12750
    },
    {
      "epoch": 2.0807569948609186,
      "grad_norm": 0.0014272031839936972,
      "learning_rate": 0.23083105589437888,
      "loss": 0.2985,
      "num_input_tokens_seen": 24176176,
      "step": 12755
    },
    {
      "epoch": 2.0815727220817357,
      "grad_norm": 0.0033126019407063723,
      "learning_rate": 0.23078142899835094,
      "loss": 0.3256,
      "num_input_tokens_seen": 24186112,
      "step": 12760
    },
    {
      "epoch": 2.0823884493025533,
      "grad_norm": 0.0017154073575511575,
      "learning_rate": 0.23073178964481147,
      "loss": 0.3193,
      "num_input_tokens_seen": 24195344,
      "step": 12765
    },
    {
      "epoch": 2.0832041765233704,
      "grad_norm": 0.0035797960590571165,
      "learning_rate": 0.2306821378414155,
      "loss": 0.3251,
      "num_input_tokens_seen": 24204528,
      "step": 12770
    },
    {
      "epoch": 2.084019903744188,
      "grad_norm": 0.0018135823775082827,
      "learning_rate": 0.2306324735958199,
      "loss": 0.4328,
      "num_input_tokens_seen": 24214192,
      "step": 12775
    },
    {
      "epoch": 2.084835630965005,
      "grad_norm": 0.0027640284970402718,
      "learning_rate": 0.23058279691568362,
      "loss": 0.3481,
      "num_input_tokens_seen": 24224000,
      "step": 12780
    },
    {
      "epoch": 2.0856513581858227,
      "grad_norm": 0.0016015457222238183,
      "learning_rate": 0.23053310780866745,
      "loss": 0.323,
      "num_input_tokens_seen": 24233680,
      "step": 12785
    },
    {
      "epoch": 2.08646708540664,
      "grad_norm": 0.0013023581122979522,
      "learning_rate": 0.23048340628243397,
      "loss": 0.3646,
      "num_input_tokens_seen": 24243680,
      "step": 12790
    },
    {
      "epoch": 2.0872828126274574,
      "grad_norm": 0.001435036538168788,
      "learning_rate": 0.23043369234464783,
      "loss": 0.2658,
      "num_input_tokens_seen": 24253344,
      "step": 12795
    },
    {
      "epoch": 2.0880985398482745,
      "grad_norm": 0.0023292223922908306,
      "learning_rate": 0.2303839660029755,
      "loss": 0.3627,
      "num_input_tokens_seen": 24261904,
      "step": 12800
    },
    {
      "epoch": 2.0880985398482745,
      "eval_loss": 0.32856374979019165,
      "eval_runtime": 152.4602,
      "eval_samples_per_second": 17.874,
      "eval_steps_per_second": 8.94,
      "num_input_tokens_seen": 24261904,
      "step": 12800
    },
    {
      "epoch": 2.088914267069092,
      "grad_norm": 0.002724752761423588,
      "learning_rate": 0.23033422726508548,
      "loss": 0.3278,
      "num_input_tokens_seen": 24270480,
      "step": 12805
    },
    {
      "epoch": 2.0897299942899092,
      "grad_norm": 0.002071324037387967,
      "learning_rate": 0.23028447613864808,
      "loss": 0.356,
      "num_input_tokens_seen": 24280288,
      "step": 12810
    },
    {
      "epoch": 2.090545721510727,
      "grad_norm": 0.0018076522974297404,
      "learning_rate": 0.2302347126313355,
      "loss": 0.345,
      "num_input_tokens_seen": 24289568,
      "step": 12815
    },
    {
      "epoch": 2.0913614487315444,
      "grad_norm": 0.002618051366880536,
      "learning_rate": 0.23018493675082197,
      "loss": 0.3473,
      "num_input_tokens_seen": 24297888,
      "step": 12820
    },
    {
      "epoch": 2.0921771759523615,
      "grad_norm": 0.0010224785655736923,
      "learning_rate": 0.2301351485047835,
      "loss": 0.2972,
      "num_input_tokens_seen": 24307168,
      "step": 12825
    },
    {
      "epoch": 2.092992903173179,
      "grad_norm": 0.0026276051066815853,
      "learning_rate": 0.23008534790089813,
      "loss": 0.2872,
      "num_input_tokens_seen": 24317920,
      "step": 12830
    },
    {
      "epoch": 2.0938086303939962,
      "grad_norm": 0.002164266537874937,
      "learning_rate": 0.2300355349468457,
      "loss": 0.3465,
      "num_input_tokens_seen": 24327056,
      "step": 12835
    },
    {
      "epoch": 2.094624357614814,
      "grad_norm": 0.001755847712047398,
      "learning_rate": 0.22998570965030793,
      "loss": 0.3383,
      "num_input_tokens_seen": 24336480,
      "step": 12840
    },
    {
      "epoch": 2.095440084835631,
      "grad_norm": 0.0036077946424484253,
      "learning_rate": 0.22993587201896862,
      "loss": 0.2997,
      "num_input_tokens_seen": 24344560,
      "step": 12845
    },
    {
      "epoch": 2.0962558120564485,
      "grad_norm": 0.0020760009065270424,
      "learning_rate": 0.2298860220605133,
      "loss": 0.3949,
      "num_input_tokens_seen": 24354144,
      "step": 12850
    },
    {
      "epoch": 2.0970715392772656,
      "grad_norm": 0.0017707080114632845,
      "learning_rate": 0.22983615978262942,
      "loss": 0.2977,
      "num_input_tokens_seen": 24363904,
      "step": 12855
    },
    {
      "epoch": 2.097887266498083,
      "grad_norm": 0.0015798417152836919,
      "learning_rate": 0.22978628519300648,
      "loss": 0.3117,
      "num_input_tokens_seen": 24373680,
      "step": 12860
    },
    {
      "epoch": 2.0987029937189003,
      "grad_norm": 0.003844177583232522,
      "learning_rate": 0.22973639829933568,
      "loss": 0.3901,
      "num_input_tokens_seen": 24382048,
      "step": 12865
    },
    {
      "epoch": 2.099518720939718,
      "grad_norm": 0.0023066592402756214,
      "learning_rate": 0.22968649910931027,
      "loss": 0.349,
      "num_input_tokens_seen": 24392448,
      "step": 12870
    },
    {
      "epoch": 2.100334448160535,
      "grad_norm": 0.0015421869466081262,
      "learning_rate": 0.22963658763062528,
      "loss": 0.327,
      "num_input_tokens_seen": 24402336,
      "step": 12875
    },
    {
      "epoch": 2.1011501753813526,
      "grad_norm": 0.0009004730964079499,
      "learning_rate": 0.22958666387097765,
      "loss": 0.3225,
      "num_input_tokens_seen": 24412064,
      "step": 12880
    },
    {
      "epoch": 2.1019659026021698,
      "grad_norm": 0.0015632358845323324,
      "learning_rate": 0.22953672783806633,
      "loss": 0.3686,
      "num_input_tokens_seen": 24422528,
      "step": 12885
    },
    {
      "epoch": 2.1027816298229873,
      "grad_norm": 0.0010366067290306091,
      "learning_rate": 0.22948677953959207,
      "loss": 0.3215,
      "num_input_tokens_seen": 24432192,
      "step": 12890
    },
    {
      "epoch": 2.1035973570438045,
      "grad_norm": 0.004283059388399124,
      "learning_rate": 0.2294368189832575,
      "loss": 0.328,
      "num_input_tokens_seen": 24441936,
      "step": 12895
    },
    {
      "epoch": 2.104413084264622,
      "grad_norm": 0.001940351678058505,
      "learning_rate": 0.2293868461767672,
      "loss": 0.2837,
      "num_input_tokens_seen": 24451504,
      "step": 12900
    },
    {
      "epoch": 2.105228811485439,
      "grad_norm": 0.001583268167451024,
      "learning_rate": 0.22933686112782758,
      "loss": 0.423,
      "num_input_tokens_seen": 24461392,
      "step": 12905
    },
    {
      "epoch": 2.1060445387062567,
      "grad_norm": 0.0018477117409929633,
      "learning_rate": 0.22928686384414698,
      "loss": 0.3013,
      "num_input_tokens_seen": 24470496,
      "step": 12910
    },
    {
      "epoch": 2.106860265927074,
      "grad_norm": 0.0016297688707709312,
      "learning_rate": 0.22923685433343552,
      "loss": 0.3606,
      "num_input_tokens_seen": 24481024,
      "step": 12915
    },
    {
      "epoch": 2.1076759931478914,
      "grad_norm": 0.002187862526625395,
      "learning_rate": 0.22918683260340542,
      "loss": 0.3662,
      "num_input_tokens_seen": 24491424,
      "step": 12920
    },
    {
      "epoch": 2.1084917203687086,
      "grad_norm": 0.0020019086077809334,
      "learning_rate": 0.2291367986617706,
      "loss": 0.3095,
      "num_input_tokens_seen": 24499936,
      "step": 12925
    },
    {
      "epoch": 2.109307447589526,
      "grad_norm": 0.009416231885552406,
      "learning_rate": 0.22908675251624697,
      "loss": 0.4275,
      "num_input_tokens_seen": 24510368,
      "step": 12930
    },
    {
      "epoch": 2.1101231748103433,
      "grad_norm": 0.0017207273049280047,
      "learning_rate": 0.22903669417455216,
      "loss": 0.3586,
      "num_input_tokens_seen": 24520144,
      "step": 12935
    },
    {
      "epoch": 2.110938902031161,
      "grad_norm": 0.0007368624792434275,
      "learning_rate": 0.22898662364440592,
      "loss": 0.2655,
      "num_input_tokens_seen": 24529392,
      "step": 12940
    },
    {
      "epoch": 2.111754629251978,
      "grad_norm": 0.0002983103913720697,
      "learning_rate": 0.2289365409335297,
      "loss": 0.358,
      "num_input_tokens_seen": 24538976,
      "step": 12945
    },
    {
      "epoch": 2.1125703564727956,
      "grad_norm": 0.00040784396696835756,
      "learning_rate": 0.2288864460496469,
      "loss": 0.3425,
      "num_input_tokens_seen": 24548240,
      "step": 12950
    },
    {
      "epoch": 2.1133860836936127,
      "grad_norm": 0.0007234673830680549,
      "learning_rate": 0.22883633900048272,
      "loss": 0.3196,
      "num_input_tokens_seen": 24557584,
      "step": 12955
    },
    {
      "epoch": 2.1142018109144303,
      "grad_norm": 0.0011278094025328755,
      "learning_rate": 0.2287862197937644,
      "loss": 0.3644,
      "num_input_tokens_seen": 24565808,
      "step": 12960
    },
    {
      "epoch": 2.1150175381352474,
      "grad_norm": 0.0015343423001468182,
      "learning_rate": 0.2287360884372209,
      "loss": 0.3522,
      "num_input_tokens_seen": 24575968,
      "step": 12965
    },
    {
      "epoch": 2.115833265356065,
      "grad_norm": 0.0021357128862291574,
      "learning_rate": 0.22868594493858307,
      "loss": 0.3396,
      "num_input_tokens_seen": 24584800,
      "step": 12970
    },
    {
      "epoch": 2.116648992576882,
      "grad_norm": 0.0017755338922142982,
      "learning_rate": 0.2286357893055837,
      "loss": 0.3845,
      "num_input_tokens_seen": 24595968,
      "step": 12975
    },
    {
      "epoch": 2.1174647197976997,
      "grad_norm": 0.0008813436143100262,
      "learning_rate": 0.22858562154595746,
      "loss": 0.3366,
      "num_input_tokens_seen": 24605168,
      "step": 12980
    },
    {
      "epoch": 2.118280447018517,
      "grad_norm": 0.000583598215598613,
      "learning_rate": 0.22853544166744078,
      "loss": 0.3475,
      "num_input_tokens_seen": 24615216,
      "step": 12985
    },
    {
      "epoch": 2.1190961742393344,
      "grad_norm": 0.0009084368939511478,
      "learning_rate": 0.22848524967777206,
      "loss": 0.3619,
      "num_input_tokens_seen": 24624848,
      "step": 12990
    },
    {
      "epoch": 2.1199119014601515,
      "grad_norm": 0.0003685315605252981,
      "learning_rate": 0.22843504558469152,
      "loss": 0.3393,
      "num_input_tokens_seen": 24634128,
      "step": 12995
    },
    {
      "epoch": 2.120727628680969,
      "grad_norm": 0.0013705672463402152,
      "learning_rate": 0.2283848293959413,
      "loss": 0.3366,
      "num_input_tokens_seen": 24643776,
      "step": 13000
    },
    {
      "epoch": 2.120727628680969,
      "eval_loss": 0.3426823019981384,
      "eval_runtime": 152.4972,
      "eval_samples_per_second": 17.869,
      "eval_steps_per_second": 8.938,
      "num_input_tokens_seen": 24643776,
      "step": 13000
    },
    {
      "epoch": 2.121543355901786,
      "grad_norm": 0.0013918853364884853,
      "learning_rate": 0.22833460111926532,
      "loss": 0.3106,
      "num_input_tokens_seen": 24654064,
      "step": 13005
    },
    {
      "epoch": 2.122359083122604,
      "grad_norm": 0.0010816779686138034,
      "learning_rate": 0.22828436076240946,
      "loss": 0.3473,
      "num_input_tokens_seen": 24662112,
      "step": 13010
    },
    {
      "epoch": 2.1231748103434214,
      "grad_norm": 0.0013356487033888698,
      "learning_rate": 0.22823410833312135,
      "loss": 0.3699,
      "num_input_tokens_seen": 24670944,
      "step": 13015
    },
    {
      "epoch": 2.1239905375642385,
      "grad_norm": 0.0012259449576959014,
      "learning_rate": 0.2281838438391506,
      "loss": 0.3087,
      "num_input_tokens_seen": 24680608,
      "step": 13020
    },
    {
      "epoch": 2.124806264785056,
      "grad_norm": 0.00218389043584466,
      "learning_rate": 0.22813356728824863,
      "loss": 0.3307,
      "num_input_tokens_seen": 24690848,
      "step": 13025
    },
    {
      "epoch": 2.125621992005873,
      "grad_norm": 0.001441996544599533,
      "learning_rate": 0.2280832786881687,
      "loss": 0.3218,
      "num_input_tokens_seen": 24700656,
      "step": 13030
    },
    {
      "epoch": 2.126437719226691,
      "grad_norm": 0.002228668425232172,
      "learning_rate": 0.22803297804666592,
      "loss": 0.3607,
      "num_input_tokens_seen": 24710576,
      "step": 13035
    },
    {
      "epoch": 2.127253446447508,
      "grad_norm": 0.0011974043445661664,
      "learning_rate": 0.22798266537149728,
      "loss": 0.321,
      "num_input_tokens_seen": 24719168,
      "step": 13040
    },
    {
      "epoch": 2.1280691736683255,
      "grad_norm": 0.0003142386849503964,
      "learning_rate": 0.22793234067042167,
      "loss": 0.3638,
      "num_input_tokens_seen": 24728064,
      "step": 13045
    },
    {
      "epoch": 2.1288849008891426,
      "grad_norm": 0.0011830904986709356,
      "learning_rate": 0.22788200395119979,
      "loss": 0.3109,
      "num_input_tokens_seen": 24738480,
      "step": 13050
    },
    {
      "epoch": 2.12970062810996,
      "grad_norm": 0.000787861121352762,
      "learning_rate": 0.2278316552215942,
      "loss": 0.3683,
      "num_input_tokens_seen": 24747840,
      "step": 13055
    },
    {
      "epoch": 2.1305163553307773,
      "grad_norm": 0.001105556613765657,
      "learning_rate": 0.22778129448936918,
      "loss": 0.3611,
      "num_input_tokens_seen": 24756752,
      "step": 13060
    },
    {
      "epoch": 2.131332082551595,
      "grad_norm": 0.0010241797426715493,
      "learning_rate": 0.22773092176229118,
      "loss": 0.3397,
      "num_input_tokens_seen": 24766000,
      "step": 13065
    },
    {
      "epoch": 2.132147809772412,
      "grad_norm": 0.0020111287012696266,
      "learning_rate": 0.22768053704812816,
      "loss": 0.3218,
      "num_input_tokens_seen": 24775728,
      "step": 13070
    },
    {
      "epoch": 2.1329635369932296,
      "grad_norm": 0.0009419272537343204,
      "learning_rate": 0.22763014035465018,
      "loss": 0.3216,
      "num_input_tokens_seen": 24786176,
      "step": 13075
    },
    {
      "epoch": 2.1337792642140467,
      "grad_norm": 0.0015554673736914992,
      "learning_rate": 0.22757973168962892,
      "loss": 0.3452,
      "num_input_tokens_seen": 24794384,
      "step": 13080
    },
    {
      "epoch": 2.1345949914348643,
      "grad_norm": 0.0026720718014985323,
      "learning_rate": 0.22752931106083818,
      "loss": 0.3434,
      "num_input_tokens_seen": 24803008,
      "step": 13085
    },
    {
      "epoch": 2.1354107186556814,
      "grad_norm": 0.0011945216683670878,
      "learning_rate": 0.22747887847605341,
      "loss": 0.3305,
      "num_input_tokens_seen": 24813168,
      "step": 13090
    },
    {
      "epoch": 2.136226445876499,
      "grad_norm": 0.0017432828899472952,
      "learning_rate": 0.22742843394305184,
      "loss": 0.301,
      "num_input_tokens_seen": 24821792,
      "step": 13095
    },
    {
      "epoch": 2.137042173097316,
      "grad_norm": 0.003927092533558607,
      "learning_rate": 0.22737797746961272,
      "loss": 0.3959,
      "num_input_tokens_seen": 24831984,
      "step": 13100
    },
    {
      "epoch": 2.1378579003181337,
      "grad_norm": 0.0022620304953306913,
      "learning_rate": 0.22732750906351712,
      "loss": 0.3396,
      "num_input_tokens_seen": 24841296,
      "step": 13105
    },
    {
      "epoch": 2.138673627538951,
      "grad_norm": 0.0032655152026563883,
      "learning_rate": 0.22727702873254785,
      "loss": 0.3502,
      "num_input_tokens_seen": 24852208,
      "step": 13110
    },
    {
      "epoch": 2.1394893547597684,
      "grad_norm": 0.0021408535540103912,
      "learning_rate": 0.22722653648448968,
      "loss": 0.3061,
      "num_input_tokens_seen": 24861264,
      "step": 13115
    },
    {
      "epoch": 2.1403050819805856,
      "grad_norm": 0.0022965860553085804,
      "learning_rate": 0.22717603232712902,
      "loss": 0.3232,
      "num_input_tokens_seen": 24869856,
      "step": 13120
    },
    {
      "epoch": 2.141120809201403,
      "grad_norm": 0.0015865149907767773,
      "learning_rate": 0.22712551626825436,
      "loss": 0.309,
      "num_input_tokens_seen": 24878880,
      "step": 13125
    },
    {
      "epoch": 2.1419365364222203,
      "grad_norm": 0.001002489821985364,
      "learning_rate": 0.2270749883156559,
      "loss": 0.3148,
      "num_input_tokens_seen": 24889568,
      "step": 13130
    },
    {
      "epoch": 2.142752263643038,
      "grad_norm": 0.0031518912874162197,
      "learning_rate": 0.22702444847712563,
      "loss": 0.3235,
      "num_input_tokens_seen": 24898224,
      "step": 13135
    },
    {
      "epoch": 2.143567990863855,
      "grad_norm": 0.001686863019131124,
      "learning_rate": 0.22697389676045743,
      "loss": 0.2908,
      "num_input_tokens_seen": 24907024,
      "step": 13140
    },
    {
      "epoch": 2.1443837180846725,
      "grad_norm": 0.005229317583143711,
      "learning_rate": 0.22692333317344704,
      "loss": 0.3514,
      "num_input_tokens_seen": 24916480,
      "step": 13145
    },
    {
      "epoch": 2.1451994453054897,
      "grad_norm": 0.002307638991624117,
      "learning_rate": 0.22687275772389198,
      "loss": 0.2867,
      "num_input_tokens_seen": 24925328,
      "step": 13150
    },
    {
      "epoch": 2.1460151725263072,
      "grad_norm": 0.005630282685160637,
      "learning_rate": 0.22682217041959168,
      "loss": 0.4536,
      "num_input_tokens_seen": 24935248,
      "step": 13155
    },
    {
      "epoch": 2.1468308997471244,
      "grad_norm": 0.0007497041951864958,
      "learning_rate": 0.2267715712683473,
      "loss": 0.3515,
      "num_input_tokens_seen": 24944272,
      "step": 13160
    },
    {
      "epoch": 2.147646626967942,
      "grad_norm": 0.0022558877244591713,
      "learning_rate": 0.22672096027796182,
      "loss": 0.3603,
      "num_input_tokens_seen": 24953616,
      "step": 13165
    },
    {
      "epoch": 2.148462354188759,
      "grad_norm": 0.0014658046420663595,
      "learning_rate": 0.22667033745624016,
      "loss": 0.2789,
      "num_input_tokens_seen": 24962352,
      "step": 13170
    },
    {
      "epoch": 2.1492780814095767,
      "grad_norm": 0.004885718692094088,
      "learning_rate": 0.22661970281098895,
      "loss": 0.369,
      "num_input_tokens_seen": 24971504,
      "step": 13175
    },
    {
      "epoch": 2.150093808630394,
      "grad_norm": 0.003944504074752331,
      "learning_rate": 0.22656905635001667,
      "loss": 0.3623,
      "num_input_tokens_seen": 24980976,
      "step": 13180
    },
    {
      "epoch": 2.1509095358512114,
      "grad_norm": 0.002011008094996214,
      "learning_rate": 0.2265183980811337,
      "loss": 0.3906,
      "num_input_tokens_seen": 24992096,
      "step": 13185
    },
    {
      "epoch": 2.151725263072029,
      "grad_norm": 0.0007691297796554863,
      "learning_rate": 0.22646772801215218,
      "loss": 0.333,
      "num_input_tokens_seen": 25001408,
      "step": 13190
    },
    {
      "epoch": 2.152540990292846,
      "grad_norm": 0.0007077573682181537,
      "learning_rate": 0.22641704615088598,
      "loss": 0.3278,
      "num_input_tokens_seen": 25011264,
      "step": 13195
    },
    {
      "epoch": 2.153356717513663,
      "grad_norm": 0.0017220606096088886,
      "learning_rate": 0.22636635250515103,
      "loss": 0.3541,
      "num_input_tokens_seen": 25020496,
      "step": 13200
    },
    {
      "epoch": 2.153356717513663,
      "eval_loss": 0.3606691062450409,
      "eval_runtime": 152.4816,
      "eval_samples_per_second": 17.871,
      "eval_steps_per_second": 8.939,
      "num_input_tokens_seen": 25020496,
      "step": 13200
    },
    {
      "epoch": 2.1541724447344808,
      "grad_norm": 0.0021737380884587765,
      "learning_rate": 0.2263156470827648,
      "loss": 0.4272,
      "num_input_tokens_seen": 25031200,
      "step": 13205
    },
    {
      "epoch": 2.1549881719552983,
      "grad_norm": 0.0008907775627449155,
      "learning_rate": 0.22626492989154678,
      "loss": 0.3569,
      "num_input_tokens_seen": 25039680,
      "step": 13210
    },
    {
      "epoch": 2.1558038991761155,
      "grad_norm": 0.001790808979421854,
      "learning_rate": 0.22621420093931813,
      "loss": 0.3163,
      "num_input_tokens_seen": 25048544,
      "step": 13215
    },
    {
      "epoch": 2.156619626396933,
      "grad_norm": 0.003166577545925975,
      "learning_rate": 0.22616346023390194,
      "loss": 0.4015,
      "num_input_tokens_seen": 25057408,
      "step": 13220
    },
    {
      "epoch": 2.15743535361775,
      "grad_norm": 0.0010300737340003252,
      "learning_rate": 0.22611270778312306,
      "loss": 0.3266,
      "num_input_tokens_seen": 25068272,
      "step": 13225
    },
    {
      "epoch": 2.1582510808385678,
      "grad_norm": 0.0015169765101745725,
      "learning_rate": 0.2260619435948081,
      "loss": 0.3576,
      "num_input_tokens_seen": 25078032,
      "step": 13230
    },
    {
      "epoch": 2.159066808059385,
      "grad_norm": 0.0017021058592945337,
      "learning_rate": 0.22601116767678567,
      "loss": 0.3021,
      "num_input_tokens_seen": 25087440,
      "step": 13235
    },
    {
      "epoch": 2.1598825352802025,
      "grad_norm": 0.0004916352336294949,
      "learning_rate": 0.2259603800368859,
      "loss": 0.3311,
      "num_input_tokens_seen": 25096208,
      "step": 13240
    },
    {
      "epoch": 2.1606982625010196,
      "grad_norm": 0.0012812811182811856,
      "learning_rate": 0.22590958068294098,
      "loss": 0.3805,
      "num_input_tokens_seen": 25106544,
      "step": 13245
    },
    {
      "epoch": 2.161513989721837,
      "grad_norm": 0.0008832095772959292,
      "learning_rate": 0.22585876962278478,
      "loss": 0.247,
      "num_input_tokens_seen": 25116816,
      "step": 13250
    },
    {
      "epoch": 2.1623297169426543,
      "grad_norm": 0.001811258145608008,
      "learning_rate": 0.22580794686425298,
      "loss": 0.3813,
      "num_input_tokens_seen": 25125888,
      "step": 13255
    },
    {
      "epoch": 2.163145444163472,
      "grad_norm": 0.0007548151770606637,
      "learning_rate": 0.22575711241518312,
      "loss": 0.329,
      "num_input_tokens_seen": 25135520,
      "step": 13260
    },
    {
      "epoch": 2.163961171384289,
      "grad_norm": 0.0011504130670800805,
      "learning_rate": 0.22570626628341453,
      "loss": 0.3361,
      "num_input_tokens_seen": 25144416,
      "step": 13265
    },
    {
      "epoch": 2.1647768986051066,
      "grad_norm": 0.0017025198321789503,
      "learning_rate": 0.22565540847678828,
      "loss": 0.333,
      "num_input_tokens_seen": 25154032,
      "step": 13270
    },
    {
      "epoch": 2.1655926258259237,
      "grad_norm": 0.0011376782786101103,
      "learning_rate": 0.2256045390031473,
      "loss": 0.3629,
      "num_input_tokens_seen": 25163024,
      "step": 13275
    },
    {
      "epoch": 2.1664083530467413,
      "grad_norm": 0.0014055612264201045,
      "learning_rate": 0.22555365787033627,
      "loss": 0.3444,
      "num_input_tokens_seen": 25171664,
      "step": 13280
    },
    {
      "epoch": 2.1672240802675584,
      "grad_norm": 0.0012313753832131624,
      "learning_rate": 0.22550276508620173,
      "loss": 0.328,
      "num_input_tokens_seen": 25180960,
      "step": 13285
    },
    {
      "epoch": 2.168039807488376,
      "grad_norm": 0.000899986014701426,
      "learning_rate": 0.22545186065859202,
      "loss": 0.3143,
      "num_input_tokens_seen": 25190064,
      "step": 13290
    },
    {
      "epoch": 2.168855534709193,
      "grad_norm": 0.0013816652353852987,
      "learning_rate": 0.2254009445953572,
      "loss": 0.2569,
      "num_input_tokens_seen": 25199872,
      "step": 13295
    },
    {
      "epoch": 2.1696712619300107,
      "grad_norm": 0.0018692264566197991,
      "learning_rate": 0.22535001690434917,
      "loss": 0.3916,
      "num_input_tokens_seen": 25209680,
      "step": 13300
    },
    {
      "epoch": 2.170486989150828,
      "grad_norm": 0.0018605801742523909,
      "learning_rate": 0.22529907759342163,
      "loss": 0.198,
      "num_input_tokens_seen": 25219024,
      "step": 13305
    },
    {
      "epoch": 2.1713027163716454,
      "grad_norm": 0.0030039618723094463,
      "learning_rate": 0.22524812667043007,
      "loss": 0.4183,
      "num_input_tokens_seen": 25229136,
      "step": 13310
    },
    {
      "epoch": 2.1721184435924625,
      "grad_norm": 0.0033020437695086002,
      "learning_rate": 0.22519716414323177,
      "loss": 0.4002,
      "num_input_tokens_seen": 25238576,
      "step": 13315
    },
    {
      "epoch": 2.17293417081328,
      "grad_norm": 0.004547236952930689,
      "learning_rate": 0.22514619001968567,
      "loss": 0.339,
      "num_input_tokens_seen": 25247568,
      "step": 13320
    },
    {
      "epoch": 2.1737498980340972,
      "grad_norm": 0.000986401573754847,
      "learning_rate": 0.2250952043076528,
      "loss": 0.3382,
      "num_input_tokens_seen": 25254896,
      "step": 13325
    },
    {
      "epoch": 2.174565625254915,
      "grad_norm": 0.001689652563072741,
      "learning_rate": 0.2250442070149957,
      "loss": 0.3942,
      "num_input_tokens_seen": 25263216,
      "step": 13330
    },
    {
      "epoch": 2.175381352475732,
      "grad_norm": 0.0010197118390351534,
      "learning_rate": 0.22499319814957885,
      "loss": 0.3129,
      "num_input_tokens_seen": 25273072,
      "step": 13335
    },
    {
      "epoch": 2.1761970796965495,
      "grad_norm": 0.0011969665065407753,
      "learning_rate": 0.2249421777192684,
      "loss": 0.3007,
      "num_input_tokens_seen": 25282544,
      "step": 13340
    },
    {
      "epoch": 2.1770128069173666,
      "grad_norm": 0.0022412531543523073,
      "learning_rate": 0.22489114573193236,
      "loss": 0.4141,
      "num_input_tokens_seen": 25291136,
      "step": 13345
    },
    {
      "epoch": 2.1778285341381842,
      "grad_norm": 0.0012162828352302313,
      "learning_rate": 0.2248401021954405,
      "loss": 0.3449,
      "num_input_tokens_seen": 25299984,
      "step": 13350
    },
    {
      "epoch": 2.1786442613590014,
      "grad_norm": 0.0022393246181309223,
      "learning_rate": 0.22478904711766443,
      "loss": 0.2919,
      "num_input_tokens_seen": 25308896,
      "step": 13355
    },
    {
      "epoch": 2.179459988579819,
      "grad_norm": 0.002959558041766286,
      "learning_rate": 0.22473798050647734,
      "loss": 0.3647,
      "num_input_tokens_seen": 25317872,
      "step": 13360
    },
    {
      "epoch": 2.180275715800636,
      "grad_norm": 0.0023242789320647717,
      "learning_rate": 0.22468690236975453,
      "loss": 0.3223,
      "num_input_tokens_seen": 25326112,
      "step": 13365
    },
    {
      "epoch": 2.1810914430214536,
      "grad_norm": 0.0009106943034566939,
      "learning_rate": 0.22463581271537272,
      "loss": 0.3262,
      "num_input_tokens_seen": 25336624,
      "step": 13370
    },
    {
      "epoch": 2.1819071702422708,
      "grad_norm": 0.0019756697583943605,
      "learning_rate": 0.22458471155121076,
      "loss": 0.3479,
      "num_input_tokens_seen": 25346720,
      "step": 13375
    },
    {
      "epoch": 2.1827228974630883,
      "grad_norm": 0.0012503910111263394,
      "learning_rate": 0.2245335988851489,
      "loss": 0.33,
      "num_input_tokens_seen": 25355392,
      "step": 13380
    },
    {
      "epoch": 2.183538624683906,
      "grad_norm": 0.0014385093236342072,
      "learning_rate": 0.2244824747250695,
      "loss": 0.3418,
      "num_input_tokens_seen": 25364736,
      "step": 13385
    },
    {
      "epoch": 2.184354351904723,
      "grad_norm": 0.0016354414401575923,
      "learning_rate": 0.22443133907885646,
      "loss": 0.3167,
      "num_input_tokens_seen": 25372192,
      "step": 13390
    },
    {
      "epoch": 2.1851700791255406,
      "grad_norm": 0.004151939880102873,
      "learning_rate": 0.22438019195439557,
      "loss": 0.3497,
      "num_input_tokens_seen": 25382064,
      "step": 13395
    },
    {
      "epoch": 2.1859858063463578,
      "grad_norm": 0.0016248570755124092,
      "learning_rate": 0.22432903335957435,
      "loss": 0.3233,
      "num_input_tokens_seen": 25391072,
      "step": 13400
    },
    {
      "epoch": 2.1859858063463578,
      "eval_loss": 0.3271985948085785,
      "eval_runtime": 152.5108,
      "eval_samples_per_second": 17.868,
      "eval_steps_per_second": 8.937,
      "num_input_tokens_seen": 25391072,
      "step": 13400
    },
    {
      "epoch": 2.1868015335671753,
      "grad_norm": 0.0013315363321453333,
      "learning_rate": 0.22427786330228214,
      "loss": 0.3421,
      "num_input_tokens_seen": 25400688,
      "step": 13405
    },
    {
      "epoch": 2.1876172607879925,
      "grad_norm": 0.0017781687201932073,
      "learning_rate": 0.22422668179040997,
      "loss": 0.3422,
      "num_input_tokens_seen": 25410768,
      "step": 13410
    },
    {
      "epoch": 2.18843298800881,
      "grad_norm": 0.0017742872005328536,
      "learning_rate": 0.2241754888318507,
      "loss": 0.3277,
      "num_input_tokens_seen": 25420576,
      "step": 13415
    },
    {
      "epoch": 2.189248715229627,
      "grad_norm": 0.0022312544751912355,
      "learning_rate": 0.22412428443449886,
      "loss": 0.327,
      "num_input_tokens_seen": 25428688,
      "step": 13420
    },
    {
      "epoch": 2.1900644424504447,
      "grad_norm": 0.0030459254048764706,
      "learning_rate": 0.22407306860625087,
      "loss": 0.3294,
      "num_input_tokens_seen": 25438112,
      "step": 13425
    },
    {
      "epoch": 2.190880169671262,
      "grad_norm": 0.0016891133273020387,
      "learning_rate": 0.22402184135500483,
      "loss": 0.3655,
      "num_input_tokens_seen": 25447024,
      "step": 13430
    },
    {
      "epoch": 2.1916958968920794,
      "grad_norm": 0.0039040714036673307,
      "learning_rate": 0.22397060268866067,
      "loss": 0.3867,
      "num_input_tokens_seen": 25455824,
      "step": 13435
    },
    {
      "epoch": 2.1925116241128966,
      "grad_norm": 0.0009121721959672868,
      "learning_rate": 0.22391935261511994,
      "loss": 0.2976,
      "num_input_tokens_seen": 25464688,
      "step": 13440
    },
    {
      "epoch": 2.193327351333714,
      "grad_norm": 0.00093493377789855,
      "learning_rate": 0.22386809114228615,
      "loss": 0.3559,
      "num_input_tokens_seen": 25474480,
      "step": 13445
    },
    {
      "epoch": 2.1941430785545313,
      "grad_norm": 0.0027852579951286316,
      "learning_rate": 0.22381681827806446,
      "loss": 0.363,
      "num_input_tokens_seen": 25484384,
      "step": 13450
    },
    {
      "epoch": 2.194958805775349,
      "grad_norm": 0.0012129121460020542,
      "learning_rate": 0.22376553403036173,
      "loss": 0.3167,
      "num_input_tokens_seen": 25495104,
      "step": 13455
    },
    {
      "epoch": 2.195774532996166,
      "grad_norm": 0.002524586860090494,
      "learning_rate": 0.22371423840708662,
      "loss": 0.3084,
      "num_input_tokens_seen": 25504800,
      "step": 13460
    },
    {
      "epoch": 2.1965902602169836,
      "grad_norm": 0.0015865231398493052,
      "learning_rate": 0.22366293141614962,
      "loss": 0.3512,
      "num_input_tokens_seen": 25513760,
      "step": 13465
    },
    {
      "epoch": 2.1974059874378007,
      "grad_norm": 0.0011117610847577453,
      "learning_rate": 0.22361161306546287,
      "loss": 0.3521,
      "num_input_tokens_seen": 25523808,
      "step": 13470
    },
    {
      "epoch": 2.1982217146586183,
      "grad_norm": 0.0008518864051438868,
      "learning_rate": 0.22356028336294037,
      "loss": 0.3288,
      "num_input_tokens_seen": 25531920,
      "step": 13475
    },
    {
      "epoch": 2.1990374418794354,
      "grad_norm": 0.0028832091484218836,
      "learning_rate": 0.2235089423164977,
      "loss": 0.3235,
      "num_input_tokens_seen": 25541696,
      "step": 13480
    },
    {
      "epoch": 2.199853169100253,
      "grad_norm": 0.0029438775964081287,
      "learning_rate": 0.22345758993405243,
      "loss": 0.3692,
      "num_input_tokens_seen": 25551184,
      "step": 13485
    },
    {
      "epoch": 2.20066889632107,
      "grad_norm": 0.002437043935060501,
      "learning_rate": 0.2234062262235236,
      "loss": 0.3859,
      "num_input_tokens_seen": 25559840,
      "step": 13490
    },
    {
      "epoch": 2.2014846235418877,
      "grad_norm": 0.0013567208079621196,
      "learning_rate": 0.22335485119283222,
      "loss": 0.3417,
      "num_input_tokens_seen": 25568448,
      "step": 13495
    },
    {
      "epoch": 2.202300350762705,
      "grad_norm": 0.00452877813950181,
      "learning_rate": 0.22330346484990093,
      "loss": 0.3672,
      "num_input_tokens_seen": 25578352,
      "step": 13500
    },
    {
      "epoch": 2.2031160779835224,
      "grad_norm": 0.0013698127586394548,
      "learning_rate": 0.22325206720265425,
      "loss": 0.3642,
      "num_input_tokens_seen": 25587360,
      "step": 13505
    },
    {
      "epoch": 2.2039318052043395,
      "grad_norm": 0.0015414603985846043,
      "learning_rate": 0.2232006582590182,
      "loss": 0.3539,
      "num_input_tokens_seen": 25596112,
      "step": 13510
    },
    {
      "epoch": 2.204747532425157,
      "grad_norm": 0.003790806280449033,
      "learning_rate": 0.22314923802692077,
      "loss": 0.3515,
      "num_input_tokens_seen": 25606336,
      "step": 13515
    },
    {
      "epoch": 2.205563259645974,
      "grad_norm": 0.0019030609400942922,
      "learning_rate": 0.22309780651429156,
      "loss": 0.3009,
      "num_input_tokens_seen": 25615248,
      "step": 13520
    },
    {
      "epoch": 2.206378986866792,
      "grad_norm": 0.001600555726327002,
      "learning_rate": 0.22304636372906203,
      "loss": 0.3769,
      "num_input_tokens_seen": 25624064,
      "step": 13525
    },
    {
      "epoch": 2.207194714087609,
      "grad_norm": 0.0019759174901992083,
      "learning_rate": 0.22299490967916522,
      "loss": 0.3191,
      "num_input_tokens_seen": 25633472,
      "step": 13530
    },
    {
      "epoch": 2.2080104413084265,
      "grad_norm": 0.001544798957183957,
      "learning_rate": 0.22294344437253602,
      "loss": 0.3092,
      "num_input_tokens_seen": 25642240,
      "step": 13535
    },
    {
      "epoch": 2.2088261685292436,
      "grad_norm": 0.0015578164020553231,
      "learning_rate": 0.22289196781711101,
      "loss": 0.3482,
      "num_input_tokens_seen": 25651680,
      "step": 13540
    },
    {
      "epoch": 2.209641895750061,
      "grad_norm": 0.0031178980134427547,
      "learning_rate": 0.2228404800208286,
      "loss": 0.3797,
      "num_input_tokens_seen": 25660512,
      "step": 13545
    },
    {
      "epoch": 2.2104576229708783,
      "grad_norm": 0.00392446992918849,
      "learning_rate": 0.22278898099162875,
      "loss": 0.3801,
      "num_input_tokens_seen": 25670144,
      "step": 13550
    },
    {
      "epoch": 2.211273350191696,
      "grad_norm": 0.0025143513921648264,
      "learning_rate": 0.22273747073745337,
      "loss": 0.3573,
      "num_input_tokens_seen": 25678352,
      "step": 13555
    },
    {
      "epoch": 2.2120890774125135,
      "grad_norm": 0.00038176486850716174,
      "learning_rate": 0.22268594926624588,
      "loss": 0.3516,
      "num_input_tokens_seen": 25688112,
      "step": 13560
    },
    {
      "epoch": 2.2129048046333306,
      "grad_norm": 0.0015880652936175466,
      "learning_rate": 0.22263441658595162,
      "loss": 0.3449,
      "num_input_tokens_seen": 25697680,
      "step": 13565
    },
    {
      "epoch": 2.2137205318541477,
      "grad_norm": 0.0006372809875756502,
      "learning_rate": 0.2225828727045175,
      "loss": 0.3611,
      "num_input_tokens_seen": 25706496,
      "step": 13570
    },
    {
      "epoch": 2.2145362590749653,
      "grad_norm": 0.001706136274151504,
      "learning_rate": 0.22253131762989228,
      "loss": 0.3248,
      "num_input_tokens_seen": 25715888,
      "step": 13575
    },
    {
      "epoch": 2.215351986295783,
      "grad_norm": 0.0014354151207953691,
      "learning_rate": 0.2224797513700264,
      "loss": 0.3494,
      "num_input_tokens_seen": 25724448,
      "step": 13580
    },
    {
      "epoch": 2.2161677135166,
      "grad_norm": 0.0015809581382200122,
      "learning_rate": 0.22242817393287204,
      "loss": 0.3357,
      "num_input_tokens_seen": 25733952,
      "step": 13585
    },
    {
      "epoch": 2.2169834407374176,
      "grad_norm": 0.0008599716238677502,
      "learning_rate": 0.22237658532638305,
      "loss": 0.3342,
      "num_input_tokens_seen": 25743152,
      "step": 13590
    },
    {
      "epoch": 2.2177991679582347,
      "grad_norm": 0.0015101744793355465,
      "learning_rate": 0.22232498555851513,
      "loss": 0.3344,
      "num_input_tokens_seen": 25752784,
      "step": 13595
    },
    {
      "epoch": 2.2186148951790523,
      "grad_norm": 0.0014726646477356553,
      "learning_rate": 0.22227337463722546,
      "loss": 0.3341,
      "num_input_tokens_seen": 25762416,
      "step": 13600
    },
    {
      "epoch": 2.2186148951790523,
      "eval_loss": 0.34018057584762573,
      "eval_runtime": 152.5049,
      "eval_samples_per_second": 17.868,
      "eval_steps_per_second": 8.937,
      "num_input_tokens_seen": 25762416,
      "step": 13600
    },
    {
      "epoch": 2.2194306223998694,
      "grad_norm": 0.002495590830221772,
      "learning_rate": 0.2222217525704732,
      "loss": 0.3282,
      "num_input_tokens_seen": 25770752,
      "step": 13605
    },
    {
      "epoch": 2.220246349620687,
      "grad_norm": 0.002142902696505189,
      "learning_rate": 0.22217011936621908,
      "loss": 0.3456,
      "num_input_tokens_seen": 25779728,
      "step": 13610
    },
    {
      "epoch": 2.221062076841504,
      "grad_norm": 0.0015826656017452478,
      "learning_rate": 0.22211847503242566,
      "loss": 0.2655,
      "num_input_tokens_seen": 25788496,
      "step": 13615
    },
    {
      "epoch": 2.2218778040623217,
      "grad_norm": 0.0014027145225554705,
      "learning_rate": 0.22206681957705704,
      "loss": 0.4316,
      "num_input_tokens_seen": 25799280,
      "step": 13620
    },
    {
      "epoch": 2.222693531283139,
      "grad_norm": 0.0018620726186782122,
      "learning_rate": 0.2220151530080792,
      "loss": 0.4076,
      "num_input_tokens_seen": 25808208,
      "step": 13625
    },
    {
      "epoch": 2.2235092585039564,
      "grad_norm": 0.001895873574540019,
      "learning_rate": 0.2219634753334598,
      "loss": 0.3371,
      "num_input_tokens_seen": 25818000,
      "step": 13630
    },
    {
      "epoch": 2.2243249857247736,
      "grad_norm": 0.002917452249675989,
      "learning_rate": 0.22191178656116817,
      "loss": 0.3585,
      "num_input_tokens_seen": 25827200,
      "step": 13635
    },
    {
      "epoch": 2.225140712945591,
      "grad_norm": 0.001954446779564023,
      "learning_rate": 0.2218600866991753,
      "loss": 0.3434,
      "num_input_tokens_seen": 25836144,
      "step": 13640
    },
    {
      "epoch": 2.2259564401664083,
      "grad_norm": 0.001206322805956006,
      "learning_rate": 0.221808375755454,
      "loss": 0.3477,
      "num_input_tokens_seen": 25845696,
      "step": 13645
    },
    {
      "epoch": 2.226772167387226,
      "grad_norm": 0.0010401264298707247,
      "learning_rate": 0.22175665373797881,
      "loss": 0.3448,
      "num_input_tokens_seen": 25855408,
      "step": 13650
    },
    {
      "epoch": 2.227587894608043,
      "grad_norm": 0.0011307407403364778,
      "learning_rate": 0.22170492065472583,
      "loss": 0.3906,
      "num_input_tokens_seen": 25865456,
      "step": 13655
    },
    {
      "epoch": 2.2284036218288605,
      "grad_norm": 0.0009460892179049551,
      "learning_rate": 0.221653176513673,
      "loss": 0.3158,
      "num_input_tokens_seen": 25875088,
      "step": 13660
    },
    {
      "epoch": 2.2292193490496777,
      "grad_norm": 0.0009036572882905602,
      "learning_rate": 0.2216014213227999,
      "loss": 0.3353,
      "num_input_tokens_seen": 25884288,
      "step": 13665
    },
    {
      "epoch": 2.2300350762704952,
      "grad_norm": 0.002574351616203785,
      "learning_rate": 0.22154965509008784,
      "loss": 0.3602,
      "num_input_tokens_seen": 25895072,
      "step": 13670
    },
    {
      "epoch": 2.2308508034913124,
      "grad_norm": 0.0011269109090790153,
      "learning_rate": 0.2214978778235198,
      "loss": 0.3306,
      "num_input_tokens_seen": 25904480,
      "step": 13675
    },
    {
      "epoch": 2.23166653071213,
      "grad_norm": 0.002114287344738841,
      "learning_rate": 0.2214460895310805,
      "loss": 0.3752,
      "num_input_tokens_seen": 25913648,
      "step": 13680
    },
    {
      "epoch": 2.232482257932947,
      "grad_norm": 0.0005371870938688517,
      "learning_rate": 0.22139429022075635,
      "loss": 0.3186,
      "num_input_tokens_seen": 25923968,
      "step": 13685
    },
    {
      "epoch": 2.2332979851537647,
      "grad_norm": 0.0014654129045084119,
      "learning_rate": 0.22134247990053546,
      "loss": 0.3205,
      "num_input_tokens_seen": 25934256,
      "step": 13690
    },
    {
      "epoch": 2.234113712374582,
      "grad_norm": 0.0009063169127330184,
      "learning_rate": 0.2212906585784076,
      "loss": 0.349,
      "num_input_tokens_seen": 25943888,
      "step": 13695
    },
    {
      "epoch": 2.2349294395953994,
      "grad_norm": 0.002516753040254116,
      "learning_rate": 0.22123882626236432,
      "loss": 0.3352,
      "num_input_tokens_seen": 25953808,
      "step": 13700
    },
    {
      "epoch": 2.2357451668162165,
      "grad_norm": 0.0011390689760446548,
      "learning_rate": 0.2211869829603988,
      "loss": 0.3358,
      "num_input_tokens_seen": 25962752,
      "step": 13705
    },
    {
      "epoch": 2.236560894037034,
      "grad_norm": 0.0018355563515797257,
      "learning_rate": 0.22113512868050592,
      "loss": 0.3725,
      "num_input_tokens_seen": 25972288,
      "step": 13710
    },
    {
      "epoch": 2.237376621257851,
      "grad_norm": 0.0022504699882119894,
      "learning_rate": 0.2210832634306822,
      "loss": 0.3317,
      "num_input_tokens_seen": 25981504,
      "step": 13715
    },
    {
      "epoch": 2.2381923484786688,
      "grad_norm": 0.0017151916399598122,
      "learning_rate": 0.22103138721892598,
      "loss": 0.3099,
      "num_input_tokens_seen": 25989984,
      "step": 13720
    },
    {
      "epoch": 2.239008075699486,
      "grad_norm": 0.0018114999402314425,
      "learning_rate": 0.22097950005323724,
      "loss": 0.2988,
      "num_input_tokens_seen": 25999920,
      "step": 13725
    },
    {
      "epoch": 2.2398238029203035,
      "grad_norm": 0.001259637763723731,
      "learning_rate": 0.22092760194161762,
      "loss": 0.3661,
      "num_input_tokens_seen": 26008528,
      "step": 13730
    },
    {
      "epoch": 2.2406395301411206,
      "grad_norm": 0.0018667116528376937,
      "learning_rate": 0.2208756928920704,
      "loss": 0.3079,
      "num_input_tokens_seen": 26018800,
      "step": 13735
    },
    {
      "epoch": 2.241455257361938,
      "grad_norm": 0.001568939653225243,
      "learning_rate": 0.22082377291260072,
      "loss": 0.3215,
      "num_input_tokens_seen": 26027072,
      "step": 13740
    },
    {
      "epoch": 2.2422709845827553,
      "grad_norm": 0.001932755927555263,
      "learning_rate": 0.2207718420112152,
      "loss": 0.3388,
      "num_input_tokens_seen": 26037584,
      "step": 13745
    },
    {
      "epoch": 2.243086711803573,
      "grad_norm": 0.0012222561053931713,
      "learning_rate": 0.22071990019592228,
      "loss": 0.3169,
      "num_input_tokens_seen": 26047440,
      "step": 13750
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.001321048242971301,
      "learning_rate": 0.22066794747473198,
      "loss": 0.4063,
      "num_input_tokens_seen": 26056896,
      "step": 13755
    },
    {
      "epoch": 2.2447181662452076,
      "grad_norm": 0.0012537961592897773,
      "learning_rate": 0.2206159838556562,
      "loss": 0.2802,
      "num_input_tokens_seen": 26067184,
      "step": 13760
    },
    {
      "epoch": 2.245533893466025,
      "grad_norm": 0.0029519773088395596,
      "learning_rate": 0.2205640093467082,
      "loss": 0.4093,
      "num_input_tokens_seen": 26075952,
      "step": 13765
    },
    {
      "epoch": 2.2463496206868423,
      "grad_norm": 0.0010728705674409866,
      "learning_rate": 0.22051202395590322,
      "loss": 0.3672,
      "num_input_tokens_seen": 26084656,
      "step": 13770
    },
    {
      "epoch": 2.24716534790766,
      "grad_norm": 0.0015664579113945365,
      "learning_rate": 0.22046002769125808,
      "loss": 0.3447,
      "num_input_tokens_seen": 26094496,
      "step": 13775
    },
    {
      "epoch": 2.247981075128477,
      "grad_norm": 0.0014824571553617716,
      "learning_rate": 0.2204080205607912,
      "loss": 0.3169,
      "num_input_tokens_seen": 26104112,
      "step": 13780
    },
    {
      "epoch": 2.2487968023492946,
      "grad_norm": 0.0007931285072118044,
      "learning_rate": 0.22035600257252272,
      "loss": 0.3518,
      "num_input_tokens_seen": 26112032,
      "step": 13785
    },
    {
      "epoch": 2.2496125295701117,
      "grad_norm": 0.0010672274511307478,
      "learning_rate": 0.2203039737344745,
      "loss": 0.3101,
      "num_input_tokens_seen": 26120720,
      "step": 13790
    },
    {
      "epoch": 2.2504282567909293,
      "grad_norm": 0.000726126367226243,
      "learning_rate": 0.22025193405467003,
      "loss": 0.3482,
      "num_input_tokens_seen": 26129328,
      "step": 13795
    },
    {
      "epoch": 2.2512439840117464,
      "grad_norm": 0.002684695413336158,
      "learning_rate": 0.2201998835411345,
      "loss": 0.3562,
      "num_input_tokens_seen": 26139456,
      "step": 13800
    },
    {
      "epoch": 2.2512439840117464,
      "eval_loss": 0.3295580744743347,
      "eval_runtime": 152.5005,
      "eval_samples_per_second": 17.869,
      "eval_steps_per_second": 8.938,
      "num_input_tokens_seen": 26139456,
      "step": 13800
    },
    {
      "epoch": 2.252059711232564,
      "grad_norm": 0.0016390461241826415,
      "learning_rate": 0.22014782220189474,
      "loss": 0.3511,
      "num_input_tokens_seen": 26147936,
      "step": 13805
    },
    {
      "epoch": 2.252875438453381,
      "grad_norm": 0.003143212990835309,
      "learning_rate": 0.2200957500449793,
      "loss": 0.3128,
      "num_input_tokens_seen": 26156432,
      "step": 13810
    },
    {
      "epoch": 2.2536911656741987,
      "grad_norm": 0.0012048924108967185,
      "learning_rate": 0.22004366707841827,
      "loss": 0.3729,
      "num_input_tokens_seen": 26165632,
      "step": 13815
    },
    {
      "epoch": 2.254506892895016,
      "grad_norm": 0.0036668211687356234,
      "learning_rate": 0.21999157331024358,
      "loss": 0.3788,
      "num_input_tokens_seen": 26175168,
      "step": 13820
    },
    {
      "epoch": 2.2553226201158334,
      "grad_norm": 0.0010996183846145868,
      "learning_rate": 0.21993946874848871,
      "loss": 0.3425,
      "num_input_tokens_seen": 26184832,
      "step": 13825
    },
    {
      "epoch": 2.2561383473366505,
      "grad_norm": 0.0020646313205361366,
      "learning_rate": 0.2198873534011888,
      "loss": 0.3353,
      "num_input_tokens_seen": 26193616,
      "step": 13830
    },
    {
      "epoch": 2.256954074557468,
      "grad_norm": 0.0022442929912358522,
      "learning_rate": 0.2198352272763808,
      "loss": 0.3386,
      "num_input_tokens_seen": 26204208,
      "step": 13835
    },
    {
      "epoch": 2.2577698017782852,
      "grad_norm": 0.0012103734770789742,
      "learning_rate": 0.2197830903821031,
      "loss": 0.3034,
      "num_input_tokens_seen": 26212176,
      "step": 13840
    },
    {
      "epoch": 2.258585528999103,
      "grad_norm": 0.0015312869800254703,
      "learning_rate": 0.21973094272639598,
      "loss": 0.2991,
      "num_input_tokens_seen": 26222064,
      "step": 13845
    },
    {
      "epoch": 2.25940125621992,
      "grad_norm": 0.004489441867917776,
      "learning_rate": 0.21967878431730117,
      "loss": 0.3556,
      "num_input_tokens_seen": 26231392,
      "step": 13850
    },
    {
      "epoch": 2.2602169834407375,
      "grad_norm": 0.002584145637229085,
      "learning_rate": 0.21962661516286217,
      "loss": 0.3496,
      "num_input_tokens_seen": 26240224,
      "step": 13855
    },
    {
      "epoch": 2.2610327106615546,
      "grad_norm": 0.0026385465171188116,
      "learning_rate": 0.21957443527112414,
      "loss": 0.3691,
      "num_input_tokens_seen": 26251248,
      "step": 13860
    },
    {
      "epoch": 2.261848437882372,
      "grad_norm": 0.00119578477460891,
      "learning_rate": 0.21952224465013384,
      "loss": 0.3829,
      "num_input_tokens_seen": 26259344,
      "step": 13865
    },
    {
      "epoch": 2.2626641651031894,
      "grad_norm": 0.0019738462287932634,
      "learning_rate": 0.21947004330793976,
      "loss": 0.3409,
      "num_input_tokens_seen": 26268288,
      "step": 13870
    },
    {
      "epoch": 2.263479892324007,
      "grad_norm": 0.0024613807909190655,
      "learning_rate": 0.21941783125259198,
      "loss": 0.3347,
      "num_input_tokens_seen": 26277488,
      "step": 13875
    },
    {
      "epoch": 2.264295619544824,
      "grad_norm": 0.0011524790897965431,
      "learning_rate": 0.21936560849214226,
      "loss": 0.3488,
      "num_input_tokens_seen": 26286976,
      "step": 13880
    },
    {
      "epoch": 2.2651113467656416,
      "grad_norm": 0.002474455628544092,
      "learning_rate": 0.21931337503464404,
      "loss": 0.3307,
      "num_input_tokens_seen": 26295984,
      "step": 13885
    },
    {
      "epoch": 2.2659270739864588,
      "grad_norm": 0.0025373490061610937,
      "learning_rate": 0.21926113088815233,
      "loss": 0.3695,
      "num_input_tokens_seen": 26304832,
      "step": 13890
    },
    {
      "epoch": 2.2667428012072763,
      "grad_norm": 0.002767360769212246,
      "learning_rate": 0.2192088760607238,
      "loss": 0.3085,
      "num_input_tokens_seen": 26313808,
      "step": 13895
    },
    {
      "epoch": 2.2675585284280935,
      "grad_norm": 0.0005969550693407655,
      "learning_rate": 0.2191566105604169,
      "loss": 0.3709,
      "num_input_tokens_seen": 26322816,
      "step": 13900
    },
    {
      "epoch": 2.268374255648911,
      "grad_norm": 0.0012610313715413213,
      "learning_rate": 0.21910433439529153,
      "loss": 0.3359,
      "num_input_tokens_seen": 26331792,
      "step": 13905
    },
    {
      "epoch": 2.269189982869728,
      "grad_norm": 0.001478724181652069,
      "learning_rate": 0.2190520475734094,
      "loss": 0.3385,
      "num_input_tokens_seen": 26340944,
      "step": 13910
    },
    {
      "epoch": 2.2700057100905457,
      "grad_norm": 0.0012971573742106557,
      "learning_rate": 0.2189997501028338,
      "loss": 0.3418,
      "num_input_tokens_seen": 26349216,
      "step": 13915
    },
    {
      "epoch": 2.270821437311363,
      "grad_norm": 0.0012475443072617054,
      "learning_rate": 0.2189474419916296,
      "loss": 0.3914,
      "num_input_tokens_seen": 26358624,
      "step": 13920
    },
    {
      "epoch": 2.2716371645321805,
      "grad_norm": 0.0025332150980830193,
      "learning_rate": 0.21889512324786342,
      "loss": 0.3415,
      "num_input_tokens_seen": 26366928,
      "step": 13925
    },
    {
      "epoch": 2.272452891752998,
      "grad_norm": 0.0013281519059091806,
      "learning_rate": 0.21884279387960345,
      "loss": 0.3526,
      "num_input_tokens_seen": 26376192,
      "step": 13930
    },
    {
      "epoch": 2.273268618973815,
      "grad_norm": 0.0016081358771771193,
      "learning_rate": 0.2187904538949195,
      "loss": 0.3366,
      "num_input_tokens_seen": 26384768,
      "step": 13935
    },
    {
      "epoch": 2.2740843461946323,
      "grad_norm": 0.0016436397563666105,
      "learning_rate": 0.2187381033018831,
      "loss": 0.3685,
      "num_input_tokens_seen": 26394848,
      "step": 13940
    },
    {
      "epoch": 2.27490007341545,
      "grad_norm": 0.000869942654389888,
      "learning_rate": 0.2186857421085673,
      "loss": 0.3636,
      "num_input_tokens_seen": 26403936,
      "step": 13945
    },
    {
      "epoch": 2.2757158006362674,
      "grad_norm": 0.0018420652486383915,
      "learning_rate": 0.21863337032304697,
      "loss": 0.3327,
      "num_input_tokens_seen": 26414064,
      "step": 13950
    },
    {
      "epoch": 2.2765315278570846,
      "grad_norm": 0.0009475345141254365,
      "learning_rate": 0.21858098795339845,
      "loss": 0.3077,
      "num_input_tokens_seen": 26424224,
      "step": 13955
    },
    {
      "epoch": 2.2773472550779017,
      "grad_norm": 0.0015826028538867831,
      "learning_rate": 0.21852859500769975,
      "loss": 0.3098,
      "num_input_tokens_seen": 26434256,
      "step": 13960
    },
    {
      "epoch": 2.2781629822987193,
      "grad_norm": 0.0016816556453704834,
      "learning_rate": 0.21847619149403044,
      "loss": 0.3879,
      "num_input_tokens_seen": 26444336,
      "step": 13965
    },
    {
      "epoch": 2.278978709519537,
      "grad_norm": 0.001630181330256164,
      "learning_rate": 0.21842377742047195,
      "loss": 0.3281,
      "num_input_tokens_seen": 26453952,
      "step": 13970
    },
    {
      "epoch": 2.279794436740354,
      "grad_norm": 0.0007778381113894284,
      "learning_rate": 0.21837135279510705,
      "loss": 0.3443,
      "num_input_tokens_seen": 26463024,
      "step": 13975
    },
    {
      "epoch": 2.2806101639611716,
      "grad_norm": 0.0014414828037843108,
      "learning_rate": 0.21831891762602038,
      "loss": 0.3668,
      "num_input_tokens_seen": 26472128,
      "step": 13980
    },
    {
      "epoch": 2.2814258911819887,
      "grad_norm": 0.0023533001076430082,
      "learning_rate": 0.21826647192129806,
      "loss": 0.3233,
      "num_input_tokens_seen": 26481312,
      "step": 13985
    },
    {
      "epoch": 2.2822416184028063,
      "grad_norm": 0.0005879959207959473,
      "learning_rate": 0.21821401568902787,
      "loss": 0.3612,
      "num_input_tokens_seen": 26491104,
      "step": 13990
    },
    {
      "epoch": 2.2830573456236234,
      "grad_norm": 0.0012186148669570684,
      "learning_rate": 0.21816154893729925,
      "loss": 0.3758,
      "num_input_tokens_seen": 26501584,
      "step": 13995
    },
    {
      "epoch": 2.283873072844441,
      "grad_norm": 0.0006256807246245444,
      "learning_rate": 0.2181090716742032,
      "loss": 0.3395,
      "num_input_tokens_seen": 26511344,
      "step": 14000
    },
    {
      "epoch": 2.283873072844441,
      "eval_loss": 0.33109185099601746,
      "eval_runtime": 152.5758,
      "eval_samples_per_second": 17.86,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 26511344,
      "step": 14000
    },
    {
      "epoch": 2.284688800065258,
      "grad_norm": 0.0017748120008036494,
      "learning_rate": 0.21805658390783236,
      "loss": 0.3568,
      "num_input_tokens_seen": 26521392,
      "step": 14005
    },
    {
      "epoch": 2.2855045272860757,
      "grad_norm": 0.0006158183678053319,
      "learning_rate": 0.21800408564628107,
      "loss": 0.3326,
      "num_input_tokens_seen": 26530896,
      "step": 14010
    },
    {
      "epoch": 2.286320254506893,
      "grad_norm": 0.0009689107537269592,
      "learning_rate": 0.21795157689764516,
      "loss": 0.3677,
      "num_input_tokens_seen": 26539744,
      "step": 14015
    },
    {
      "epoch": 2.2871359817277104,
      "grad_norm": 0.0010427890811115503,
      "learning_rate": 0.21789905767002216,
      "loss": 0.3378,
      "num_input_tokens_seen": 26549824,
      "step": 14020
    },
    {
      "epoch": 2.2879517089485275,
      "grad_norm": 0.0018109644297510386,
      "learning_rate": 0.2178465279715112,
      "loss": 0.3458,
      "num_input_tokens_seen": 26558320,
      "step": 14025
    },
    {
      "epoch": 2.288767436169345,
      "grad_norm": 0.0012183316284790635,
      "learning_rate": 0.21779398781021303,
      "loss": 0.3421,
      "num_input_tokens_seen": 26567568,
      "step": 14030
    },
    {
      "epoch": 2.289583163390162,
      "grad_norm": 0.001492315554060042,
      "learning_rate": 0.21774143719422998,
      "loss": 0.3407,
      "num_input_tokens_seen": 26577360,
      "step": 14035
    },
    {
      "epoch": 2.29039889061098,
      "grad_norm": 0.0009976859437301755,
      "learning_rate": 0.21768887613166601,
      "loss": 0.3295,
      "num_input_tokens_seen": 26586592,
      "step": 14040
    },
    {
      "epoch": 2.291214617831797,
      "grad_norm": 0.0008899665554054081,
      "learning_rate": 0.2176363046306267,
      "loss": 0.3189,
      "num_input_tokens_seen": 26595024,
      "step": 14045
    },
    {
      "epoch": 2.2920303450526145,
      "grad_norm": 0.001002711826004088,
      "learning_rate": 0.21758372269921925,
      "loss": 0.3043,
      "num_input_tokens_seen": 26604864,
      "step": 14050
    },
    {
      "epoch": 2.2928460722734316,
      "grad_norm": 0.0018757075304165483,
      "learning_rate": 0.21753113034555244,
      "loss": 0.3271,
      "num_input_tokens_seen": 26613984,
      "step": 14055
    },
    {
      "epoch": 2.293661799494249,
      "grad_norm": 0.0020848033018410206,
      "learning_rate": 0.2174785275777367,
      "loss": 0.3135,
      "num_input_tokens_seen": 26624128,
      "step": 14060
    },
    {
      "epoch": 2.2944775267150663,
      "grad_norm": 0.0014210538938641548,
      "learning_rate": 0.21742591440388404,
      "loss": 0.3501,
      "num_input_tokens_seen": 26633312,
      "step": 14065
    },
    {
      "epoch": 2.295293253935884,
      "grad_norm": 0.0009468105272389948,
      "learning_rate": 0.21737329083210802,
      "loss": 0.3447,
      "num_input_tokens_seen": 26642416,
      "step": 14070
    },
    {
      "epoch": 2.296108981156701,
      "grad_norm": 0.0016924783121794462,
      "learning_rate": 0.2173206568705239,
      "loss": 0.3419,
      "num_input_tokens_seen": 26652048,
      "step": 14075
    },
    {
      "epoch": 2.2969247083775186,
      "grad_norm": 0.0023372219875454903,
      "learning_rate": 0.2172680125272485,
      "loss": 0.3687,
      "num_input_tokens_seen": 26661904,
      "step": 14080
    },
    {
      "epoch": 2.2977404355983357,
      "grad_norm": 0.0017544180154800415,
      "learning_rate": 0.2172153578104002,
      "loss": 0.3272,
      "num_input_tokens_seen": 26671392,
      "step": 14085
    },
    {
      "epoch": 2.2985561628191533,
      "grad_norm": 0.0019683099817484617,
      "learning_rate": 0.21716269272809902,
      "loss": 0.3427,
      "num_input_tokens_seen": 26681248,
      "step": 14090
    },
    {
      "epoch": 2.2993718900399704,
      "grad_norm": 0.0014838306233286858,
      "learning_rate": 0.21711001728846666,
      "loss": 0.3466,
      "num_input_tokens_seen": 26691344,
      "step": 14095
    },
    {
      "epoch": 2.300187617260788,
      "grad_norm": 0.002614712342619896,
      "learning_rate": 0.21705733149962628,
      "loss": 0.34,
      "num_input_tokens_seen": 26700432,
      "step": 14100
    },
    {
      "epoch": 2.3010033444816056,
      "grad_norm": 0.0029457961209118366,
      "learning_rate": 0.21700463536970263,
      "loss": 0.3382,
      "num_input_tokens_seen": 26710112,
      "step": 14105
    },
    {
      "epoch": 2.3018190717024227,
      "grad_norm": 0.0010566513519734144,
      "learning_rate": 0.21695192890682222,
      "loss": 0.3328,
      "num_input_tokens_seen": 26719040,
      "step": 14110
    },
    {
      "epoch": 2.30263479892324,
      "grad_norm": 0.001810796675272286,
      "learning_rate": 0.21689921211911298,
      "loss": 0.2765,
      "num_input_tokens_seen": 26728640,
      "step": 14115
    },
    {
      "epoch": 2.3034505261440574,
      "grad_norm": 0.002226949669420719,
      "learning_rate": 0.21684648501470452,
      "loss": 0.3102,
      "num_input_tokens_seen": 26738720,
      "step": 14120
    },
    {
      "epoch": 2.304266253364875,
      "grad_norm": 0.0025203616824001074,
      "learning_rate": 0.216793747601728,
      "loss": 0.4167,
      "num_input_tokens_seen": 26747296,
      "step": 14125
    },
    {
      "epoch": 2.305081980585692,
      "grad_norm": 0.0014798222109675407,
      "learning_rate": 0.21674099988831627,
      "loss": 0.3249,
      "num_input_tokens_seen": 26756368,
      "step": 14130
    },
    {
      "epoch": 2.3058977078065093,
      "grad_norm": 0.004187475424259901,
      "learning_rate": 0.21668824188260363,
      "loss": 0.3789,
      "num_input_tokens_seen": 26765984,
      "step": 14135
    },
    {
      "epoch": 2.306713435027327,
      "grad_norm": 0.0012064077891409397,
      "learning_rate": 0.21663547359272606,
      "loss": 0.3397,
      "num_input_tokens_seen": 26774768,
      "step": 14140
    },
    {
      "epoch": 2.3075291622481444,
      "grad_norm": 0.0013632908230647445,
      "learning_rate": 0.216582695026821,
      "loss": 0.3803,
      "num_input_tokens_seen": 26784464,
      "step": 14145
    },
    {
      "epoch": 2.3083448894689615,
      "grad_norm": 0.002585518406704068,
      "learning_rate": 0.21652990619302767,
      "loss": 0.2898,
      "num_input_tokens_seen": 26794848,
      "step": 14150
    },
    {
      "epoch": 2.309160616689779,
      "grad_norm": 0.001794485840946436,
      "learning_rate": 0.21647710709948673,
      "loss": 0.379,
      "num_input_tokens_seen": 26803872,
      "step": 14155
    },
    {
      "epoch": 2.3099763439105963,
      "grad_norm": 0.0020029633305966854,
      "learning_rate": 0.2164242977543405,
      "loss": 0.3192,
      "num_input_tokens_seen": 26812544,
      "step": 14160
    },
    {
      "epoch": 2.310792071131414,
      "grad_norm": 0.001852586050517857,
      "learning_rate": 0.21637147816573277,
      "loss": 0.3164,
      "num_input_tokens_seen": 26822608,
      "step": 14165
    },
    {
      "epoch": 2.311607798352231,
      "grad_norm": 0.0004114567709621042,
      "learning_rate": 0.21631864834180908,
      "loss": 0.3156,
      "num_input_tokens_seen": 26831680,
      "step": 14170
    },
    {
      "epoch": 2.3124235255730485,
      "grad_norm": 0.004010517615824938,
      "learning_rate": 0.21626580829071637,
      "loss": 0.2943,
      "num_input_tokens_seen": 26841264,
      "step": 14175
    },
    {
      "epoch": 2.3132392527938657,
      "grad_norm": 0.0015718096401542425,
      "learning_rate": 0.21621295802060328,
      "loss": 0.3691,
      "num_input_tokens_seen": 26852464,
      "step": 14180
    },
    {
      "epoch": 2.3140549800146832,
      "grad_norm": 0.003983223345130682,
      "learning_rate": 0.21616009753961996,
      "loss": 0.3699,
      "num_input_tokens_seen": 26861536,
      "step": 14185
    },
    {
      "epoch": 2.3148707072355004,
      "grad_norm": 0.0018683323869481683,
      "learning_rate": 0.2161072268559182,
      "loss": 0.3413,
      "num_input_tokens_seen": 26871712,
      "step": 14190
    },
    {
      "epoch": 2.315686434456318,
      "grad_norm": 0.002001747488975525,
      "learning_rate": 0.21605434597765133,
      "loss": 0.3424,
      "num_input_tokens_seen": 26882208,
      "step": 14195
    },
    {
      "epoch": 2.316502161677135,
      "grad_norm": 0.0013009440153837204,
      "learning_rate": 0.21600145491297418,
      "loss": 0.3766,
      "num_input_tokens_seen": 26891616,
      "step": 14200
    },
    {
      "epoch": 2.316502161677135,
      "eval_loss": 0.3432576060295105,
      "eval_runtime": 152.4944,
      "eval_samples_per_second": 17.87,
      "eval_steps_per_second": 8.938,
      "num_input_tokens_seen": 26891616,
      "step": 14200
    },
    {
      "epoch": 2.3173178888979526,
      "grad_norm": 0.002400047145783901,
      "learning_rate": 0.21594855367004326,
      "loss": 0.3324,
      "num_input_tokens_seen": 26902032,
      "step": 14205
    },
    {
      "epoch": 2.31813361611877,
      "grad_norm": 0.0033239643089473248,
      "learning_rate": 0.21589564225701663,
      "loss": 0.3207,
      "num_input_tokens_seen": 26911664,
      "step": 14210
    },
    {
      "epoch": 2.3189493433395874,
      "grad_norm": 0.0036479244008660316,
      "learning_rate": 0.21584272068205385,
      "loss": 0.3321,
      "num_input_tokens_seen": 26922320,
      "step": 14215
    },
    {
      "epoch": 2.3197650705604045,
      "grad_norm": 0.0023771494161337614,
      "learning_rate": 0.2157897889533161,
      "loss": 0.3166,
      "num_input_tokens_seen": 26932640,
      "step": 14220
    },
    {
      "epoch": 2.320580797781222,
      "grad_norm": 0.002305226633325219,
      "learning_rate": 0.21573684707896612,
      "loss": 0.3218,
      "num_input_tokens_seen": 26942608,
      "step": 14225
    },
    {
      "epoch": 2.321396525002039,
      "grad_norm": 0.001592840300872922,
      "learning_rate": 0.21568389506716826,
      "loss": 0.3456,
      "num_input_tokens_seen": 26951264,
      "step": 14230
    },
    {
      "epoch": 2.3222122522228568,
      "grad_norm": 0.0031065994407981634,
      "learning_rate": 0.21563093292608831,
      "loss": 0.2991,
      "num_input_tokens_seen": 26960944,
      "step": 14235
    },
    {
      "epoch": 2.323027979443674,
      "grad_norm": 0.005020554643124342,
      "learning_rate": 0.21557796066389376,
      "loss": 0.4039,
      "num_input_tokens_seen": 26969968,
      "step": 14240
    },
    {
      "epoch": 2.3238437066644915,
      "grad_norm": 0.002498710062354803,
      "learning_rate": 0.21552497828875353,
      "loss": 0.3231,
      "num_input_tokens_seen": 26979168,
      "step": 14245
    },
    {
      "epoch": 2.3246594338853086,
      "grad_norm": 0.0034353816881775856,
      "learning_rate": 0.21547198580883828,
      "loss": 0.3273,
      "num_input_tokens_seen": 26989680,
      "step": 14250
    },
    {
      "epoch": 2.325475161106126,
      "grad_norm": 0.001486252062022686,
      "learning_rate": 0.21541898323232,
      "loss": 0.3308,
      "num_input_tokens_seen": 26999824,
      "step": 14255
    },
    {
      "epoch": 2.3262908883269433,
      "grad_norm": 0.002202654490247369,
      "learning_rate": 0.2153659705673724,
      "loss": 0.3889,
      "num_input_tokens_seen": 27009216,
      "step": 14260
    },
    {
      "epoch": 2.327106615547761,
      "grad_norm": 0.0028224026318639517,
      "learning_rate": 0.2153129478221707,
      "loss": 0.388,
      "num_input_tokens_seen": 27016384,
      "step": 14265
    },
    {
      "epoch": 2.327922342768578,
      "grad_norm": 0.001628249534405768,
      "learning_rate": 0.21525991500489164,
      "loss": 0.3223,
      "num_input_tokens_seen": 27026384,
      "step": 14270
    },
    {
      "epoch": 2.3287380699893956,
      "grad_norm": 0.003890537191182375,
      "learning_rate": 0.21520687212371362,
      "loss": 0.3998,
      "num_input_tokens_seen": 27035504,
      "step": 14275
    },
    {
      "epoch": 2.3295537972102127,
      "grad_norm": 0.0013169508893042803,
      "learning_rate": 0.21515381918681648,
      "loss": 0.3265,
      "num_input_tokens_seen": 27045472,
      "step": 14280
    },
    {
      "epoch": 2.3303695244310303,
      "grad_norm": 0.0009804564760997891,
      "learning_rate": 0.21510075620238167,
      "loss": 0.2733,
      "num_input_tokens_seen": 27054992,
      "step": 14285
    },
    {
      "epoch": 2.3311852516518474,
      "grad_norm": 0.003936870489269495,
      "learning_rate": 0.21504768317859208,
      "loss": 0.3935,
      "num_input_tokens_seen": 27065344,
      "step": 14290
    },
    {
      "epoch": 2.332000978872665,
      "grad_norm": 0.00142448500264436,
      "learning_rate": 0.2149946001236323,
      "loss": 0.3255,
      "num_input_tokens_seen": 27076160,
      "step": 14295
    },
    {
      "epoch": 2.3328167060934826,
      "grad_norm": 0.0005374245229177177,
      "learning_rate": 0.21494150704568848,
      "loss": 0.3509,
      "num_input_tokens_seen": 27084464,
      "step": 14300
    },
    {
      "epoch": 2.3336324333142997,
      "grad_norm": 0.0007674898952245712,
      "learning_rate": 0.21488840395294811,
      "loss": 0.3407,
      "num_input_tokens_seen": 27093264,
      "step": 14305
    },
    {
      "epoch": 2.334448160535117,
      "grad_norm": 0.0019995090551674366,
      "learning_rate": 0.21483529085360042,
      "loss": 0.3294,
      "num_input_tokens_seen": 27102208,
      "step": 14310
    },
    {
      "epoch": 2.3352638877559344,
      "grad_norm": 0.0016328126657754183,
      "learning_rate": 0.2147821677558361,
      "loss": 0.2966,
      "num_input_tokens_seen": 27110592,
      "step": 14315
    },
    {
      "epoch": 2.336079614976752,
      "grad_norm": 0.0032766612712293863,
      "learning_rate": 0.2147290346678475,
      "loss": 0.3166,
      "num_input_tokens_seen": 27120256,
      "step": 14320
    },
    {
      "epoch": 2.336895342197569,
      "grad_norm": 0.0016809084918349981,
      "learning_rate": 0.21467589159782827,
      "loss": 0.4591,
      "num_input_tokens_seen": 27130080,
      "step": 14325
    },
    {
      "epoch": 2.3377110694183862,
      "grad_norm": 0.0018815239891409874,
      "learning_rate": 0.21462273855397374,
      "loss": 0.3527,
      "num_input_tokens_seen": 27138816,
      "step": 14330
    },
    {
      "epoch": 2.338526796639204,
      "grad_norm": 0.002365630120038986,
      "learning_rate": 0.21456957554448083,
      "loss": 0.341,
      "num_input_tokens_seen": 27148832,
      "step": 14335
    },
    {
      "epoch": 2.3393425238600214,
      "grad_norm": 0.0023956364020705223,
      "learning_rate": 0.21451640257754795,
      "loss": 0.3488,
      "num_input_tokens_seen": 27158304,
      "step": 14340
    },
    {
      "epoch": 2.3401582510808385,
      "grad_norm": 0.002361949533224106,
      "learning_rate": 0.21446321966137508,
      "loss": 0.3028,
      "num_input_tokens_seen": 27167712,
      "step": 14345
    },
    {
      "epoch": 2.340973978301656,
      "grad_norm": 0.0014973778743296862,
      "learning_rate": 0.21441002680416354,
      "loss": 0.3487,
      "num_input_tokens_seen": 27176672,
      "step": 14350
    },
    {
      "epoch": 2.3417897055224732,
      "grad_norm": 0.0022175048943609,
      "learning_rate": 0.21435682401411654,
      "loss": 0.3478,
      "num_input_tokens_seen": 27185648,
      "step": 14355
    },
    {
      "epoch": 2.342605432743291,
      "grad_norm": 0.0017329455586150289,
      "learning_rate": 0.2143036112994385,
      "loss": 0.3248,
      "num_input_tokens_seen": 27196192,
      "step": 14360
    },
    {
      "epoch": 2.343421159964108,
      "grad_norm": 0.0017858303617686033,
      "learning_rate": 0.21425038866833548,
      "loss": 0.3074,
      "num_input_tokens_seen": 27207152,
      "step": 14365
    },
    {
      "epoch": 2.3442368871849255,
      "grad_norm": 0.0028581772930920124,
      "learning_rate": 0.21419715612901508,
      "loss": 0.3429,
      "num_input_tokens_seen": 27216752,
      "step": 14370
    },
    {
      "epoch": 2.3450526144057426,
      "grad_norm": 0.0027592326514422894,
      "learning_rate": 0.21414391368968652,
      "loss": 0.3481,
      "num_input_tokens_seen": 27226752,
      "step": 14375
    },
    {
      "epoch": 2.34586834162656,
      "grad_norm": 0.0013954290188848972,
      "learning_rate": 0.21409066135856034,
      "loss": 0.3069,
      "num_input_tokens_seen": 27237488,
      "step": 14380
    },
    {
      "epoch": 2.3466840688473773,
      "grad_norm": 0.00111006002407521,
      "learning_rate": 0.21403739914384878,
      "loss": 0.3391,
      "num_input_tokens_seen": 27247008,
      "step": 14385
    },
    {
      "epoch": 2.347499796068195,
      "grad_norm": 0.0022246739827096462,
      "learning_rate": 0.21398412705376554,
      "loss": 0.3483,
      "num_input_tokens_seen": 27257408,
      "step": 14390
    },
    {
      "epoch": 2.348315523289012,
      "grad_norm": 0.002569869626313448,
      "learning_rate": 0.2139308450965258,
      "loss": 0.3637,
      "num_input_tokens_seen": 27266480,
      "step": 14395
    },
    {
      "epoch": 2.3491312505098296,
      "grad_norm": 0.0008341901702806354,
      "learning_rate": 0.21387755328034638,
      "loss": 0.3558,
      "num_input_tokens_seen": 27274960,
      "step": 14400
    },
    {
      "epoch": 2.3491312505098296,
      "eval_loss": 0.3296234905719757,
      "eval_runtime": 152.573,
      "eval_samples_per_second": 17.86,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 27274960,
      "step": 14400
    },
    {
      "epoch": 2.3499469777306468,
      "grad_norm": 0.0003416312683839351,
      "learning_rate": 0.2138242516134455,
      "loss": 0.343,
      "num_input_tokens_seen": 27284032,
      "step": 14405
    },
    {
      "epoch": 2.3507627049514643,
      "grad_norm": 0.0009828415932133794,
      "learning_rate": 0.2137709401040429,
      "loss": 0.3645,
      "num_input_tokens_seen": 27292848,
      "step": 14410
    },
    {
      "epoch": 2.3515784321722815,
      "grad_norm": 0.003207904053851962,
      "learning_rate": 0.21371761876036,
      "loss": 0.3484,
      "num_input_tokens_seen": 27302336,
      "step": 14415
    },
    {
      "epoch": 2.352394159393099,
      "grad_norm": 0.0011287882225587964,
      "learning_rate": 0.21366428759061956,
      "loss": 0.3313,
      "num_input_tokens_seen": 27312384,
      "step": 14420
    },
    {
      "epoch": 2.353209886613916,
      "grad_norm": 0.000759411312174052,
      "learning_rate": 0.2136109466030459,
      "loss": 0.3089,
      "num_input_tokens_seen": 27321360,
      "step": 14425
    },
    {
      "epoch": 2.3540256138347337,
      "grad_norm": 0.0016763278981670737,
      "learning_rate": 0.2135575958058649,
      "loss": 0.3322,
      "num_input_tokens_seen": 27331584,
      "step": 14430
    },
    {
      "epoch": 2.354841341055551,
      "grad_norm": 0.0009603751823306084,
      "learning_rate": 0.2135042352073039,
      "loss": 0.4006,
      "num_input_tokens_seen": 27340208,
      "step": 14435
    },
    {
      "epoch": 2.3556570682763684,
      "grad_norm": 0.0007661922718398273,
      "learning_rate": 0.2134508648155918,
      "loss": 0.3337,
      "num_input_tokens_seen": 27350544,
      "step": 14440
    },
    {
      "epoch": 2.3564727954971856,
      "grad_norm": 0.0024269134737551212,
      "learning_rate": 0.213397484638959,
      "loss": 0.3336,
      "num_input_tokens_seen": 27359680,
      "step": 14445
    },
    {
      "epoch": 2.357288522718003,
      "grad_norm": 0.0008884232374839485,
      "learning_rate": 0.21334409468563728,
      "loss": 0.2766,
      "num_input_tokens_seen": 27367760,
      "step": 14450
    },
    {
      "epoch": 2.3581042499388203,
      "grad_norm": 0.00114395038690418,
      "learning_rate": 0.2132906949638602,
      "loss": 0.3805,
      "num_input_tokens_seen": 27378272,
      "step": 14455
    },
    {
      "epoch": 2.358919977159638,
      "grad_norm": 0.0015859566628932953,
      "learning_rate": 0.21323728548186255,
      "loss": 0.2966,
      "num_input_tokens_seen": 27388880,
      "step": 14460
    },
    {
      "epoch": 2.359735704380455,
      "grad_norm": 0.002162608318030834,
      "learning_rate": 0.21318386624788088,
      "loss": 0.3162,
      "num_input_tokens_seen": 27398656,
      "step": 14465
    },
    {
      "epoch": 2.3605514316012726,
      "grad_norm": 0.0013710286002606153,
      "learning_rate": 0.21313043727015288,
      "loss": 0.3134,
      "num_input_tokens_seen": 27407968,
      "step": 14470
    },
    {
      "epoch": 2.36136715882209,
      "grad_norm": 0.0016392486868426204,
      "learning_rate": 0.2130769985569182,
      "loss": 0.3457,
      "num_input_tokens_seen": 27416544,
      "step": 14475
    },
    {
      "epoch": 2.3621828860429073,
      "grad_norm": 0.002524743089452386,
      "learning_rate": 0.21302355011641766,
      "loss": 0.2981,
      "num_input_tokens_seen": 27426608,
      "step": 14480
    },
    {
      "epoch": 2.3629986132637244,
      "grad_norm": 0.00414297403767705,
      "learning_rate": 0.21297009195689365,
      "loss": 0.309,
      "num_input_tokens_seen": 27436768,
      "step": 14485
    },
    {
      "epoch": 2.363814340484542,
      "grad_norm": 0.0023647809866815805,
      "learning_rate": 0.21291662408659015,
      "loss": 0.3039,
      "num_input_tokens_seen": 27444896,
      "step": 14490
    },
    {
      "epoch": 2.3646300677053596,
      "grad_norm": 0.0021361634135246277,
      "learning_rate": 0.21286314651375254,
      "loss": 0.3625,
      "num_input_tokens_seen": 27454768,
      "step": 14495
    },
    {
      "epoch": 2.3654457949261767,
      "grad_norm": 0.0016555963084101677,
      "learning_rate": 0.2128096592466278,
      "loss": 0.3786,
      "num_input_tokens_seen": 27463584,
      "step": 14500
    },
    {
      "epoch": 2.366261522146994,
      "grad_norm": 0.002625335706397891,
      "learning_rate": 0.21275616229346428,
      "loss": 0.3199,
      "num_input_tokens_seen": 27472304,
      "step": 14505
    },
    {
      "epoch": 2.3670772493678114,
      "grad_norm": 0.0007508857524953783,
      "learning_rate": 0.21270265566251184,
      "loss": 0.3425,
      "num_input_tokens_seen": 27481120,
      "step": 14510
    },
    {
      "epoch": 2.367892976588629,
      "grad_norm": 0.0021209269762039185,
      "learning_rate": 0.21264913936202193,
      "loss": 0.281,
      "num_input_tokens_seen": 27491008,
      "step": 14515
    },
    {
      "epoch": 2.368708703809446,
      "grad_norm": 0.0008503316785208881,
      "learning_rate": 0.2125956134002475,
      "loss": 0.3267,
      "num_input_tokens_seen": 27500752,
      "step": 14520
    },
    {
      "epoch": 2.3695244310302637,
      "grad_norm": 0.0037107167299836874,
      "learning_rate": 0.2125420777854428,
      "loss": 0.3866,
      "num_input_tokens_seen": 27509504,
      "step": 14525
    },
    {
      "epoch": 2.370340158251081,
      "grad_norm": 0.0005474525387398899,
      "learning_rate": 0.21248853252586372,
      "loss": 0.3177,
      "num_input_tokens_seen": 27518032,
      "step": 14530
    },
    {
      "epoch": 2.3711558854718984,
      "grad_norm": 0.001657177577726543,
      "learning_rate": 0.21243497762976774,
      "loss": 0.3419,
      "num_input_tokens_seen": 27526112,
      "step": 14535
    },
    {
      "epoch": 2.3719716126927155,
      "grad_norm": 0.0027666001114994287,
      "learning_rate": 0.21238141310541356,
      "loss": 0.3365,
      "num_input_tokens_seen": 27535488,
      "step": 14540
    },
    {
      "epoch": 2.372787339913533,
      "grad_norm": 0.0009852309012785554,
      "learning_rate": 0.21232783896106153,
      "loss": 0.3248,
      "num_input_tokens_seen": 27544784,
      "step": 14545
    },
    {
      "epoch": 2.37360306713435,
      "grad_norm": 0.0018384108552709222,
      "learning_rate": 0.21227425520497345,
      "loss": 0.3968,
      "num_input_tokens_seen": 27554480,
      "step": 14550
    },
    {
      "epoch": 2.374418794355168,
      "grad_norm": 0.001971052261069417,
      "learning_rate": 0.2122206618454127,
      "loss": 0.3704,
      "num_input_tokens_seen": 27563728,
      "step": 14555
    },
    {
      "epoch": 2.375234521575985,
      "grad_norm": 0.0023673069663345814,
      "learning_rate": 0.2121670588906439,
      "loss": 0.3015,
      "num_input_tokens_seen": 27573152,
      "step": 14560
    },
    {
      "epoch": 2.3760502487968025,
      "grad_norm": 0.0021279132924973965,
      "learning_rate": 0.21211344634893345,
      "loss": 0.3425,
      "num_input_tokens_seen": 27583072,
      "step": 14565
    },
    {
      "epoch": 2.3768659760176196,
      "grad_norm": 0.00396018847823143,
      "learning_rate": 0.21205982422854897,
      "loss": 0.2889,
      "num_input_tokens_seen": 27591984,
      "step": 14570
    },
    {
      "epoch": 2.377681703238437,
      "grad_norm": 0.0005375200998969376,
      "learning_rate": 0.21200619253775974,
      "loss": 0.305,
      "num_input_tokens_seen": 27602096,
      "step": 14575
    },
    {
      "epoch": 2.3784974304592543,
      "grad_norm": 0.002539324341341853,
      "learning_rate": 0.21195255128483637,
      "loss": 0.3328,
      "num_input_tokens_seen": 27611744,
      "step": 14580
    },
    {
      "epoch": 2.379313157680072,
      "grad_norm": 0.0023041781969368458,
      "learning_rate": 0.21189890047805102,
      "loss": 0.4041,
      "num_input_tokens_seen": 27622000,
      "step": 14585
    },
    {
      "epoch": 2.380128884900889,
      "grad_norm": 0.0022947886027395725,
      "learning_rate": 0.21184524012567735,
      "loss": 0.3514,
      "num_input_tokens_seen": 27632112,
      "step": 14590
    },
    {
      "epoch": 2.3809446121217066,
      "grad_norm": 0.0010775203118100762,
      "learning_rate": 0.2117915702359905,
      "loss": 0.3261,
      "num_input_tokens_seen": 27642256,
      "step": 14595
    },
    {
      "epoch": 2.3817603393425237,
      "grad_norm": 0.0012053977698087692,
      "learning_rate": 0.211737890817267,
      "loss": 0.3056,
      "num_input_tokens_seen": 27652224,
      "step": 14600
    },
    {
      "epoch": 2.3817603393425237,
      "eval_loss": 0.47137317061424255,
      "eval_runtime": 152.6182,
      "eval_samples_per_second": 17.855,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 27652224,
      "step": 14600
    },
    {
      "epoch": 2.3825760665633413,
      "grad_norm": 0.006162581965327263,
      "learning_rate": 0.21168420187778483,
      "loss": 0.4482,
      "num_input_tokens_seen": 27661616,
      "step": 14605
    },
    {
      "epoch": 2.3833917937841584,
      "grad_norm": 0.001278150943107903,
      "learning_rate": 0.21163050342582362,
      "loss": 0.3693,
      "num_input_tokens_seen": 27670304,
      "step": 14610
    },
    {
      "epoch": 2.384207521004976,
      "grad_norm": 0.0008740540943108499,
      "learning_rate": 0.21157679546966426,
      "loss": 0.3102,
      "num_input_tokens_seen": 27678784,
      "step": 14615
    },
    {
      "epoch": 2.385023248225793,
      "grad_norm": 0.0021433669608086348,
      "learning_rate": 0.2115230780175892,
      "loss": 0.3909,
      "num_input_tokens_seen": 27688064,
      "step": 14620
    },
    {
      "epoch": 2.3858389754466107,
      "grad_norm": 0.002391066635027528,
      "learning_rate": 0.21146935107788237,
      "loss": 0.3801,
      "num_input_tokens_seen": 27697904,
      "step": 14625
    },
    {
      "epoch": 2.386654702667428,
      "grad_norm": 0.0013249865733087063,
      "learning_rate": 0.21141561465882916,
      "loss": 0.3093,
      "num_input_tokens_seen": 27708528,
      "step": 14630
    },
    {
      "epoch": 2.3874704298882454,
      "grad_norm": 0.0005857261130586267,
      "learning_rate": 0.21136186876871635,
      "loss": 0.3209,
      "num_input_tokens_seen": 27717552,
      "step": 14635
    },
    {
      "epoch": 2.3882861571090626,
      "grad_norm": 0.0028857679571956396,
      "learning_rate": 0.21130811341583225,
      "loss": 0.3142,
      "num_input_tokens_seen": 27727088,
      "step": 14640
    },
    {
      "epoch": 2.38910188432988,
      "grad_norm": 0.001161823864094913,
      "learning_rate": 0.21125434860846667,
      "loss": 0.3181,
      "num_input_tokens_seen": 27736800,
      "step": 14645
    },
    {
      "epoch": 2.3899176115506973,
      "grad_norm": 0.0019855787977576256,
      "learning_rate": 0.2112005743549107,
      "loss": 0.3529,
      "num_input_tokens_seen": 27746608,
      "step": 14650
    },
    {
      "epoch": 2.390733338771515,
      "grad_norm": 0.0024147904478013515,
      "learning_rate": 0.21114679066345707,
      "loss": 0.3532,
      "num_input_tokens_seen": 27756464,
      "step": 14655
    },
    {
      "epoch": 2.391549065992332,
      "grad_norm": 0.0018712853780016303,
      "learning_rate": 0.21109299754239993,
      "loss": 0.37,
      "num_input_tokens_seen": 27766688,
      "step": 14660
    },
    {
      "epoch": 2.3923647932131495,
      "grad_norm": 0.0012790741166099906,
      "learning_rate": 0.21103919500003482,
      "loss": 0.3666,
      "num_input_tokens_seen": 27775792,
      "step": 14665
    },
    {
      "epoch": 2.393180520433967,
      "grad_norm": 0.0020701149478554726,
      "learning_rate": 0.21098538304465872,
      "loss": 0.3596,
      "num_input_tokens_seen": 27786720,
      "step": 14670
    },
    {
      "epoch": 2.3939962476547842,
      "grad_norm": 0.002462870441377163,
      "learning_rate": 0.2109315616845702,
      "loss": 0.2669,
      "num_input_tokens_seen": 27796832,
      "step": 14675
    },
    {
      "epoch": 2.3948119748756014,
      "grad_norm": 0.0015910458751022816,
      "learning_rate": 0.21087773092806925,
      "loss": 0.3373,
      "num_input_tokens_seen": 27805584,
      "step": 14680
    },
    {
      "epoch": 2.395627702096419,
      "grad_norm": 0.0017236822750419378,
      "learning_rate": 0.21082389078345704,
      "loss": 0.3206,
      "num_input_tokens_seen": 27816208,
      "step": 14685
    },
    {
      "epoch": 2.3964434293172365,
      "grad_norm": 0.0026476162020117044,
      "learning_rate": 0.2107700412590365,
      "loss": 0.3552,
      "num_input_tokens_seen": 27825536,
      "step": 14690
    },
    {
      "epoch": 2.3972591565380537,
      "grad_norm": 0.001762735890224576,
      "learning_rate": 0.210716182363112,
      "loss": 0.3421,
      "num_input_tokens_seen": 27835488,
      "step": 14695
    },
    {
      "epoch": 2.398074883758871,
      "grad_norm": 0.0019079078920185566,
      "learning_rate": 0.2106623141039891,
      "loss": 0.2824,
      "num_input_tokens_seen": 27844368,
      "step": 14700
    },
    {
      "epoch": 2.3988906109796884,
      "grad_norm": 0.0013004351640120149,
      "learning_rate": 0.21060843648997507,
      "loss": 0.3942,
      "num_input_tokens_seen": 27853104,
      "step": 14705
    },
    {
      "epoch": 2.399706338200506,
      "grad_norm": 0.0021707199048250914,
      "learning_rate": 0.21055454952937844,
      "loss": 0.2916,
      "num_input_tokens_seen": 27863120,
      "step": 14710
    },
    {
      "epoch": 2.400522065421323,
      "grad_norm": 0.0012960155727341771,
      "learning_rate": 0.21050065323050937,
      "loss": 0.3163,
      "num_input_tokens_seen": 27872560,
      "step": 14715
    },
    {
      "epoch": 2.4013377926421406,
      "grad_norm": 0.0014347108080983162,
      "learning_rate": 0.21044674760167928,
      "loss": 0.3421,
      "num_input_tokens_seen": 27881728,
      "step": 14720
    },
    {
      "epoch": 2.4021535198629578,
      "grad_norm": 0.0012458863202482462,
      "learning_rate": 0.210392832651201,
      "loss": 0.3557,
      "num_input_tokens_seen": 27891664,
      "step": 14725
    },
    {
      "epoch": 2.4029692470837754,
      "grad_norm": 0.001040878938511014,
      "learning_rate": 0.210338908387389,
      "loss": 0.339,
      "num_input_tokens_seen": 27901504,
      "step": 14730
    },
    {
      "epoch": 2.4037849743045925,
      "grad_norm": 0.0012436440447345376,
      "learning_rate": 0.21028497481855912,
      "loss": 0.3289,
      "num_input_tokens_seen": 27910720,
      "step": 14735
    },
    {
      "epoch": 2.40460070152541,
      "grad_norm": 0.0011375658214092255,
      "learning_rate": 0.21023103195302847,
      "loss": 0.307,
      "num_input_tokens_seen": 27920288,
      "step": 14740
    },
    {
      "epoch": 2.405416428746227,
      "grad_norm": 0.0024592040572315454,
      "learning_rate": 0.21017707979911582,
      "loss": 0.2765,
      "num_input_tokens_seen": 27929312,
      "step": 14745
    },
    {
      "epoch": 2.4062321559670448,
      "grad_norm": 0.0009883624734357,
      "learning_rate": 0.21012311836514122,
      "loss": 0.3359,
      "num_input_tokens_seen": 27938656,
      "step": 14750
    },
    {
      "epoch": 2.407047883187862,
      "grad_norm": 0.002984774997457862,
      "learning_rate": 0.21006914765942622,
      "loss": 0.3262,
      "num_input_tokens_seen": 27949184,
      "step": 14755
    },
    {
      "epoch": 2.4078636104086795,
      "grad_norm": 0.002753032138571143,
      "learning_rate": 0.2100151676902938,
      "loss": 0.2874,
      "num_input_tokens_seen": 27957632,
      "step": 14760
    },
    {
      "epoch": 2.4086793376294966,
      "grad_norm": 0.0012681358493864536,
      "learning_rate": 0.2099611784660683,
      "loss": 0.3368,
      "num_input_tokens_seen": 27966976,
      "step": 14765
    },
    {
      "epoch": 2.409495064850314,
      "grad_norm": 0.0017391863511875272,
      "learning_rate": 0.20990717999507552,
      "loss": 0.3536,
      "num_input_tokens_seen": 27975856,
      "step": 14770
    },
    {
      "epoch": 2.4103107920711313,
      "grad_norm": 0.0017764839576557279,
      "learning_rate": 0.20985317228564276,
      "loss": 0.3032,
      "num_input_tokens_seen": 27985600,
      "step": 14775
    },
    {
      "epoch": 2.411126519291949,
      "grad_norm": 0.003503483021631837,
      "learning_rate": 0.20979915534609872,
      "loss": 0.3148,
      "num_input_tokens_seen": 27996160,
      "step": 14780
    },
    {
      "epoch": 2.411942246512766,
      "grad_norm": 0.004776437766849995,
      "learning_rate": 0.20974512918477342,
      "loss": 0.4393,
      "num_input_tokens_seen": 28006240,
      "step": 14785
    },
    {
      "epoch": 2.4127579737335836,
      "grad_norm": 0.0007430670084431767,
      "learning_rate": 0.2096910938099984,
      "loss": 0.3175,
      "num_input_tokens_seen": 28014912,
      "step": 14790
    },
    {
      "epoch": 2.4135737009544007,
      "grad_norm": 0.0014439105289056897,
      "learning_rate": 0.2096370492301066,
      "loss": 0.3381,
      "num_input_tokens_seen": 28023472,
      "step": 14795
    },
    {
      "epoch": 2.4143894281752183,
      "grad_norm": 0.004545621573925018,
      "learning_rate": 0.2095829954534323,
      "loss": 0.3646,
      "num_input_tokens_seen": 28033168,
      "step": 14800
    },
    {
      "epoch": 2.4143894281752183,
      "eval_loss": 0.3234986960887909,
      "eval_runtime": 152.7566,
      "eval_samples_per_second": 17.839,
      "eval_steps_per_second": 8.923,
      "num_input_tokens_seen": 28033168,
      "step": 14800
    },
    {
      "epoch": 2.4152051553960354,
      "grad_norm": 0.002642726991325617,
      "learning_rate": 0.2095289324883114,
      "loss": 0.3112,
      "num_input_tokens_seen": 28042560,
      "step": 14805
    },
    {
      "epoch": 2.416020882616853,
      "grad_norm": 0.0016760589787736535,
      "learning_rate": 0.20947486034308097,
      "loss": 0.3467,
      "num_input_tokens_seen": 28051488,
      "step": 14810
    },
    {
      "epoch": 2.41683660983767,
      "grad_norm": 0.00271931872703135,
      "learning_rate": 0.2094207790260797,
      "loss": 0.3184,
      "num_input_tokens_seen": 28061072,
      "step": 14815
    },
    {
      "epoch": 2.4176523370584877,
      "grad_norm": 0.0007102055824361742,
      "learning_rate": 0.20936668854564758,
      "loss": 0.3697,
      "num_input_tokens_seen": 28070256,
      "step": 14820
    },
    {
      "epoch": 2.418468064279305,
      "grad_norm": 0.0018611564300954342,
      "learning_rate": 0.20931258891012602,
      "loss": 0.2951,
      "num_input_tokens_seen": 28080064,
      "step": 14825
    },
    {
      "epoch": 2.4192837915001224,
      "grad_norm": 0.0016471542185172439,
      "learning_rate": 0.20925848012785792,
      "loss": 0.3575,
      "num_input_tokens_seen": 28087712,
      "step": 14830
    },
    {
      "epoch": 2.4200995187209395,
      "grad_norm": 0.000956164498347789,
      "learning_rate": 0.20920436220718747,
      "loss": 0.3426,
      "num_input_tokens_seen": 28098160,
      "step": 14835
    },
    {
      "epoch": 2.420915245941757,
      "grad_norm": 0.0009957501897588372,
      "learning_rate": 0.20915023515646033,
      "loss": 0.302,
      "num_input_tokens_seen": 28107920,
      "step": 14840
    },
    {
      "epoch": 2.4217309731625742,
      "grad_norm": 0.0018259426578879356,
      "learning_rate": 0.20909609898402368,
      "loss": 0.3572,
      "num_input_tokens_seen": 28117312,
      "step": 14845
    },
    {
      "epoch": 2.422546700383392,
      "grad_norm": 0.0018216677708551288,
      "learning_rate": 0.2090419536982258,
      "loss": 0.3259,
      "num_input_tokens_seen": 28126192,
      "step": 14850
    },
    {
      "epoch": 2.423362427604209,
      "grad_norm": 0.0010387536603957415,
      "learning_rate": 0.2089877993074168,
      "loss": 0.3573,
      "num_input_tokens_seen": 28135040,
      "step": 14855
    },
    {
      "epoch": 2.4241781548250265,
      "grad_norm": 0.003302802797406912,
      "learning_rate": 0.20893363581994784,
      "loss": 0.3633,
      "num_input_tokens_seen": 28143488,
      "step": 14860
    },
    {
      "epoch": 2.424993882045844,
      "grad_norm": 0.0011961685959249735,
      "learning_rate": 0.2088794632441716,
      "loss": 0.3235,
      "num_input_tokens_seen": 28154304,
      "step": 14865
    },
    {
      "epoch": 2.4258096092666612,
      "grad_norm": 0.0009732794715091586,
      "learning_rate": 0.20882528158844219,
      "loss": 0.3028,
      "num_input_tokens_seen": 28163632,
      "step": 14870
    },
    {
      "epoch": 2.4266253364874784,
      "grad_norm": 0.0012057594722136855,
      "learning_rate": 0.20877109086111514,
      "loss": 0.3531,
      "num_input_tokens_seen": 28173248,
      "step": 14875
    },
    {
      "epoch": 2.427441063708296,
      "grad_norm": 0.0005974527448415756,
      "learning_rate": 0.2087168910705473,
      "loss": 0.4097,
      "num_input_tokens_seen": 28183600,
      "step": 14880
    },
    {
      "epoch": 2.4282567909291135,
      "grad_norm": 0.0013897273456677794,
      "learning_rate": 0.208662682225097,
      "loss": 0.2681,
      "num_input_tokens_seen": 28194672,
      "step": 14885
    },
    {
      "epoch": 2.4290725181499306,
      "grad_norm": 0.001900488045066595,
      "learning_rate": 0.2086084643331239,
      "loss": 0.3808,
      "num_input_tokens_seen": 28204288,
      "step": 14890
    },
    {
      "epoch": 2.429888245370748,
      "grad_norm": 0.0013938164338469505,
      "learning_rate": 0.20855423740298906,
      "loss": 0.3117,
      "num_input_tokens_seen": 28213392,
      "step": 14895
    },
    {
      "epoch": 2.4307039725915653,
      "grad_norm": 0.0036484356969594955,
      "learning_rate": 0.208500001443055,
      "loss": 0.3114,
      "num_input_tokens_seen": 28222464,
      "step": 14900
    },
    {
      "epoch": 2.431519699812383,
      "grad_norm": 0.0027041565626859665,
      "learning_rate": 0.20844575646168553,
      "loss": 0.3093,
      "num_input_tokens_seen": 28231440,
      "step": 14905
    },
    {
      "epoch": 2.4323354270332,
      "grad_norm": 0.0011433549225330353,
      "learning_rate": 0.20839150246724594,
      "loss": 0.3347,
      "num_input_tokens_seen": 28242736,
      "step": 14910
    },
    {
      "epoch": 2.4331511542540176,
      "grad_norm": 0.001840839977376163,
      "learning_rate": 0.20833723946810287,
      "loss": 0.3131,
      "num_input_tokens_seen": 28252880,
      "step": 14915
    },
    {
      "epoch": 2.4339668814748348,
      "grad_norm": 0.000975103466771543,
      "learning_rate": 0.20828296747262437,
      "loss": 0.3175,
      "num_input_tokens_seen": 28262464,
      "step": 14920
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 0.0014068346936255693,
      "learning_rate": 0.20822868648917986,
      "loss": 0.4006,
      "num_input_tokens_seen": 28271424,
      "step": 14925
    },
    {
      "epoch": 2.4355983359164695,
      "grad_norm": 0.001011966960504651,
      "learning_rate": 0.20817439652614017,
      "loss": 0.3296,
      "num_input_tokens_seen": 28281456,
      "step": 14930
    },
    {
      "epoch": 2.436414063137287,
      "grad_norm": 0.0023311874829232693,
      "learning_rate": 0.20812009759187744,
      "loss": 0.3236,
      "num_input_tokens_seen": 28291984,
      "step": 14935
    },
    {
      "epoch": 2.437229790358104,
      "grad_norm": 0.002317806938663125,
      "learning_rate": 0.2080657896947653,
      "loss": 0.2927,
      "num_input_tokens_seen": 28300592,
      "step": 14940
    },
    {
      "epoch": 2.4380455175789217,
      "grad_norm": 0.0012628146214410663,
      "learning_rate": 0.2080114728431787,
      "loss": 0.3402,
      "num_input_tokens_seen": 28309248,
      "step": 14945
    },
    {
      "epoch": 2.438861244799739,
      "grad_norm": 0.0014831300359219313,
      "learning_rate": 0.20795714704549392,
      "loss": 0.2506,
      "num_input_tokens_seen": 28318704,
      "step": 14950
    },
    {
      "epoch": 2.4396769720205564,
      "grad_norm": 0.0013525622198358178,
      "learning_rate": 0.20790281231008875,
      "loss": 0.2959,
      "num_input_tokens_seen": 28328704,
      "step": 14955
    },
    {
      "epoch": 2.4404926992413736,
      "grad_norm": 0.001979571534320712,
      "learning_rate": 0.20784846864534226,
      "loss": 0.3329,
      "num_input_tokens_seen": 28337248,
      "step": 14960
    },
    {
      "epoch": 2.441308426462191,
      "grad_norm": 0.0008529658662155271,
      "learning_rate": 0.20779411605963496,
      "loss": 0.4067,
      "num_input_tokens_seen": 28347040,
      "step": 14965
    },
    {
      "epoch": 2.4421241536830083,
      "grad_norm": 0.002056591212749481,
      "learning_rate": 0.2077397545613487,
      "loss": 0.2617,
      "num_input_tokens_seen": 28356960,
      "step": 14970
    },
    {
      "epoch": 2.442939880903826,
      "grad_norm": 0.0025060216430574656,
      "learning_rate": 0.20768538415886661,
      "loss": 0.342,
      "num_input_tokens_seen": 28367008,
      "step": 14975
    },
    {
      "epoch": 2.443755608124643,
      "grad_norm": 0.002071838127449155,
      "learning_rate": 0.20763100486057343,
      "loss": 0.3205,
      "num_input_tokens_seen": 28376592,
      "step": 14980
    },
    {
      "epoch": 2.4445713353454606,
      "grad_norm": 0.0024582778569310904,
      "learning_rate": 0.20757661667485502,
      "loss": 0.3879,
      "num_input_tokens_seen": 28386688,
      "step": 14985
    },
    {
      "epoch": 2.4453870625662777,
      "grad_norm": 0.0013953095767647028,
      "learning_rate": 0.2075222196100988,
      "loss": 0.2661,
      "num_input_tokens_seen": 28396304,
      "step": 14990
    },
    {
      "epoch": 2.4462027897870953,
      "grad_norm": 0.001333547756075859,
      "learning_rate": 0.20746781367469344,
      "loss": 0.2843,
      "num_input_tokens_seen": 28405712,
      "step": 14995
    },
    {
      "epoch": 2.4470185170079124,
      "grad_norm": 0.0007852283888496459,
      "learning_rate": 0.207413398877029,
      "loss": 0.3433,
      "num_input_tokens_seen": 28414784,
      "step": 15000
    },
    {
      "epoch": 2.4470185170079124,
      "eval_loss": 0.3337409198284149,
      "eval_runtime": 152.5912,
      "eval_samples_per_second": 17.858,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 28414784,
      "step": 15000
    },
    {
      "epoch": 2.44783424422873,
      "grad_norm": 0.0013283414300531149,
      "learning_rate": 0.20735897522549698,
      "loss": 0.3027,
      "num_input_tokens_seen": 28423024,
      "step": 15005
    },
    {
      "epoch": 2.448649971449547,
      "grad_norm": 0.000512058730237186,
      "learning_rate": 0.2073045427284902,
      "loss": 0.2719,
      "num_input_tokens_seen": 28430992,
      "step": 15010
    },
    {
      "epoch": 2.4494656986703647,
      "grad_norm": 0.0031955649610608816,
      "learning_rate": 0.2072501013944027,
      "loss": 0.4127,
      "num_input_tokens_seen": 28440432,
      "step": 15015
    },
    {
      "epoch": 2.450281425891182,
      "grad_norm": 0.0016472704010084271,
      "learning_rate": 0.20719565123163017,
      "loss": 0.332,
      "num_input_tokens_seen": 28449248,
      "step": 15020
    },
    {
      "epoch": 2.4510971531119994,
      "grad_norm": 0.0017492335755378008,
      "learning_rate": 0.20714119224856944,
      "loss": 0.3375,
      "num_input_tokens_seen": 28458240,
      "step": 15025
    },
    {
      "epoch": 2.4519128803328165,
      "grad_norm": 0.001842833706177771,
      "learning_rate": 0.2070867244536188,
      "loss": 0.3128,
      "num_input_tokens_seen": 28467200,
      "step": 15030
    },
    {
      "epoch": 2.452728607553634,
      "grad_norm": 0.0025777220726013184,
      "learning_rate": 0.20703224785517785,
      "loss": 0.2954,
      "num_input_tokens_seen": 28476432,
      "step": 15035
    },
    {
      "epoch": 2.4535443347744517,
      "grad_norm": 0.0013025650987401605,
      "learning_rate": 0.20697776246164754,
      "loss": 0.3106,
      "num_input_tokens_seen": 28485840,
      "step": 15040
    },
    {
      "epoch": 2.454360061995269,
      "grad_norm": 0.0003779028484132141,
      "learning_rate": 0.2069232682814303,
      "loss": 0.3799,
      "num_input_tokens_seen": 28494576,
      "step": 15045
    },
    {
      "epoch": 2.455175789216086,
      "grad_norm": 0.0027550379745662212,
      "learning_rate": 0.20686876532292972,
      "loss": 0.3523,
      "num_input_tokens_seen": 28504960,
      "step": 15050
    },
    {
      "epoch": 2.4559915164369035,
      "grad_norm": 0.0006874407408758998,
      "learning_rate": 0.20681425359455083,
      "loss": 0.3087,
      "num_input_tokens_seen": 28515072,
      "step": 15055
    },
    {
      "epoch": 2.456807243657721,
      "grad_norm": 0.0015553113771602511,
      "learning_rate": 0.20675973310470008,
      "loss": 0.3743,
      "num_input_tokens_seen": 28524096,
      "step": 15060
    },
    {
      "epoch": 2.457622970878538,
      "grad_norm": 0.001990947872400284,
      "learning_rate": 0.2067052038617852,
      "loss": 0.3688,
      "num_input_tokens_seen": 28533056,
      "step": 15065
    },
    {
      "epoch": 2.4584386980993553,
      "grad_norm": 0.001170413801446557,
      "learning_rate": 0.2066506658742153,
      "loss": 0.297,
      "num_input_tokens_seen": 28543424,
      "step": 15070
    },
    {
      "epoch": 2.459254425320173,
      "grad_norm": 0.0012414626544341445,
      "learning_rate": 0.20659611915040077,
      "loss": 0.3452,
      "num_input_tokens_seen": 28553632,
      "step": 15075
    },
    {
      "epoch": 2.4600701525409905,
      "grad_norm": 0.0008168646018020809,
      "learning_rate": 0.20654156369875348,
      "loss": 0.2863,
      "num_input_tokens_seen": 28563536,
      "step": 15080
    },
    {
      "epoch": 2.4608858797618076,
      "grad_norm": 0.003916288260370493,
      "learning_rate": 0.20648699952768648,
      "loss": 0.3596,
      "num_input_tokens_seen": 28572784,
      "step": 15085
    },
    {
      "epoch": 2.461701606982625,
      "grad_norm": 0.0029722375329583883,
      "learning_rate": 0.20643242664561437,
      "loss": 0.3015,
      "num_input_tokens_seen": 28581008,
      "step": 15090
    },
    {
      "epoch": 2.4625173342034423,
      "grad_norm": 0.0008164079044945538,
      "learning_rate": 0.20637784506095277,
      "loss": 0.3428,
      "num_input_tokens_seen": 28590192,
      "step": 15095
    },
    {
      "epoch": 2.46333306142426,
      "grad_norm": 0.0014761657221242785,
      "learning_rate": 0.20632325478211908,
      "loss": 0.3725,
      "num_input_tokens_seen": 28600016,
      "step": 15100
    },
    {
      "epoch": 2.464148788645077,
      "grad_norm": 0.0022573296446353197,
      "learning_rate": 0.20626865581753165,
      "loss": 0.3209,
      "num_input_tokens_seen": 28609664,
      "step": 15105
    },
    {
      "epoch": 2.4649645158658946,
      "grad_norm": 0.0036071946378797293,
      "learning_rate": 0.2062140481756104,
      "loss": 0.3537,
      "num_input_tokens_seen": 28617648,
      "step": 15110
    },
    {
      "epoch": 2.4657802430867117,
      "grad_norm": 0.002540511079132557,
      "learning_rate": 0.20615943186477648,
      "loss": 0.3329,
      "num_input_tokens_seen": 28625952,
      "step": 15115
    },
    {
      "epoch": 2.4665959703075293,
      "grad_norm": 0.002902921987697482,
      "learning_rate": 0.20610480689345242,
      "loss": 0.303,
      "num_input_tokens_seen": 28636256,
      "step": 15120
    },
    {
      "epoch": 2.4674116975283464,
      "grad_norm": 0.0036472436040639877,
      "learning_rate": 0.2060501732700621,
      "loss": 0.3502,
      "num_input_tokens_seen": 28645280,
      "step": 15125
    },
    {
      "epoch": 2.468227424749164,
      "grad_norm": 0.0019888021051883698,
      "learning_rate": 0.20599553100303067,
      "loss": 0.3509,
      "num_input_tokens_seen": 28655312,
      "step": 15130
    },
    {
      "epoch": 2.469043151969981,
      "grad_norm": 0.0012347534066066146,
      "learning_rate": 0.20594088010078465,
      "loss": 0.3288,
      "num_input_tokens_seen": 28663904,
      "step": 15135
    },
    {
      "epoch": 2.4698588791907987,
      "grad_norm": 0.002477551344782114,
      "learning_rate": 0.20588622057175196,
      "loss": 0.3337,
      "num_input_tokens_seen": 28673920,
      "step": 15140
    },
    {
      "epoch": 2.470674606411616,
      "grad_norm": 0.003274915274232626,
      "learning_rate": 0.20583155242436177,
      "loss": 0.3881,
      "num_input_tokens_seen": 28683024,
      "step": 15145
    },
    {
      "epoch": 2.4714903336324334,
      "grad_norm": 0.0017890670569613576,
      "learning_rate": 0.20577687566704453,
      "loss": 0.3098,
      "num_input_tokens_seen": 28691840,
      "step": 15150
    },
    {
      "epoch": 2.4723060608532506,
      "grad_norm": 0.002184398239478469,
      "learning_rate": 0.20572219030823213,
      "loss": 0.3713,
      "num_input_tokens_seen": 28700000,
      "step": 15155
    },
    {
      "epoch": 2.473121788074068,
      "grad_norm": 0.002954936819151044,
      "learning_rate": 0.20566749635635775,
      "loss": 0.3419,
      "num_input_tokens_seen": 28710336,
      "step": 15160
    },
    {
      "epoch": 2.4739375152948853,
      "grad_norm": 0.001364743453450501,
      "learning_rate": 0.20561279381985587,
      "loss": 0.3029,
      "num_input_tokens_seen": 28719632,
      "step": 15165
    },
    {
      "epoch": 2.474753242515703,
      "grad_norm": 0.0008822826202958822,
      "learning_rate": 0.2055580827071623,
      "loss": 0.2908,
      "num_input_tokens_seen": 28729008,
      "step": 15170
    },
    {
      "epoch": 2.47556896973652,
      "grad_norm": 0.0027039279229938984,
      "learning_rate": 0.20550336302671418,
      "loss": 0.3907,
      "num_input_tokens_seen": 28738256,
      "step": 15175
    },
    {
      "epoch": 2.4763846969573375,
      "grad_norm": 0.004082357045263052,
      "learning_rate": 0.20544863478695,
      "loss": 0.3976,
      "num_input_tokens_seen": 28748544,
      "step": 15180
    },
    {
      "epoch": 2.4772004241781547,
      "grad_norm": 0.002700315322726965,
      "learning_rate": 0.20539389799630953,
      "loss": 0.3832,
      "num_input_tokens_seen": 28758448,
      "step": 15185
    },
    {
      "epoch": 2.4780161513989722,
      "grad_norm": 0.0013972775777801871,
      "learning_rate": 0.20533915266323388,
      "loss": 0.3256,
      "num_input_tokens_seen": 28768704,
      "step": 15190
    },
    {
      "epoch": 2.4788318786197894,
      "grad_norm": 0.0022448962554335594,
      "learning_rate": 0.20528439879616542,
      "loss": 0.309,
      "num_input_tokens_seen": 28778640,
      "step": 15195
    },
    {
      "epoch": 2.479647605840607,
      "grad_norm": 0.0013719069538637996,
      "learning_rate": 0.20522963640354794,
      "loss": 0.3656,
      "num_input_tokens_seen": 28787168,
      "step": 15200
    },
    {
      "epoch": 2.479647605840607,
      "eval_loss": 0.3254866898059845,
      "eval_runtime": 152.6749,
      "eval_samples_per_second": 17.848,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 28787168,
      "step": 15200
    },
    {
      "epoch": 2.480463333061424,
      "grad_norm": 0.002504071919247508,
      "learning_rate": 0.20517486549382644,
      "loss": 0.3614,
      "num_input_tokens_seen": 28796512,
      "step": 15205
    },
    {
      "epoch": 2.4812790602822417,
      "grad_norm": 0.0006108260713517666,
      "learning_rate": 0.20512008607544735,
      "loss": 0.3284,
      "num_input_tokens_seen": 28805392,
      "step": 15210
    },
    {
      "epoch": 2.482094787503059,
      "grad_norm": 0.0012022322043776512,
      "learning_rate": 0.20506529815685826,
      "loss": 0.3755,
      "num_input_tokens_seen": 28815024,
      "step": 15215
    },
    {
      "epoch": 2.4829105147238764,
      "grad_norm": 0.0018008820479735732,
      "learning_rate": 0.2050105017465082,
      "loss": 0.3718,
      "num_input_tokens_seen": 28824464,
      "step": 15220
    },
    {
      "epoch": 2.4837262419446935,
      "grad_norm": 0.0028105913661420345,
      "learning_rate": 0.20495569685284754,
      "loss": 0.3467,
      "num_input_tokens_seen": 28833968,
      "step": 15225
    },
    {
      "epoch": 2.484541969165511,
      "grad_norm": 0.002196379005908966,
      "learning_rate": 0.20490088348432778,
      "loss": 0.3231,
      "num_input_tokens_seen": 28843232,
      "step": 15230
    },
    {
      "epoch": 2.4853576963863286,
      "grad_norm": 0.0034478018060326576,
      "learning_rate": 0.2048460616494018,
      "loss": 0.3241,
      "num_input_tokens_seen": 28851040,
      "step": 15235
    },
    {
      "epoch": 2.4861734236071458,
      "grad_norm": 0.0006609843112528324,
      "learning_rate": 0.2047912313565239,
      "loss": 0.3137,
      "num_input_tokens_seen": 28860896,
      "step": 15240
    },
    {
      "epoch": 2.486989150827963,
      "grad_norm": 0.0010036317398771644,
      "learning_rate": 0.20473639261414958,
      "loss": 0.335,
      "num_input_tokens_seen": 28869904,
      "step": 15245
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 0.0004238925175741315,
      "learning_rate": 0.2046815454307357,
      "loss": 0.2875,
      "num_input_tokens_seen": 28879600,
      "step": 15250
    },
    {
      "epoch": 2.488620605269598,
      "grad_norm": 0.002554841572418809,
      "learning_rate": 0.20462668981474028,
      "loss": 0.4036,
      "num_input_tokens_seen": 28888944,
      "step": 15255
    },
    {
      "epoch": 2.489436332490415,
      "grad_norm": 0.0012081974418833852,
      "learning_rate": 0.20457182577462288,
      "loss": 0.2918,
      "num_input_tokens_seen": 28898016,
      "step": 15260
    },
    {
      "epoch": 2.4902520597112328,
      "grad_norm": 0.002024971880018711,
      "learning_rate": 0.2045169533188441,
      "loss": 0.2957,
      "num_input_tokens_seen": 28907232,
      "step": 15265
    },
    {
      "epoch": 2.49106778693205,
      "grad_norm": 0.0012975212885066867,
      "learning_rate": 0.20446207245586603,
      "loss": 0.3442,
      "num_input_tokens_seen": 28916240,
      "step": 15270
    },
    {
      "epoch": 2.4918835141528675,
      "grad_norm": 0.0022170771844685078,
      "learning_rate": 0.20440718319415196,
      "loss": 0.3521,
      "num_input_tokens_seen": 28925936,
      "step": 15275
    },
    {
      "epoch": 2.4926992413736846,
      "grad_norm": 0.0017375507159158587,
      "learning_rate": 0.20435228554216653,
      "loss": 0.3085,
      "num_input_tokens_seen": 28935440,
      "step": 15280
    },
    {
      "epoch": 2.493514968594502,
      "grad_norm": 0.0026815393939614296,
      "learning_rate": 0.20429737950837565,
      "loss": 0.3289,
      "num_input_tokens_seen": 28946048,
      "step": 15285
    },
    {
      "epoch": 2.4943306958153193,
      "grad_norm": 0.000610642833635211,
      "learning_rate": 0.20424246510124647,
      "loss": 0.3233,
      "num_input_tokens_seen": 28955664,
      "step": 15290
    },
    {
      "epoch": 2.495146423036137,
      "grad_norm": 0.001191953313536942,
      "learning_rate": 0.20418754232924755,
      "loss": 0.3569,
      "num_input_tokens_seen": 28964704,
      "step": 15295
    },
    {
      "epoch": 2.495962150256954,
      "grad_norm": 0.0020104674622416496,
      "learning_rate": 0.20413261120084863,
      "loss": 0.3445,
      "num_input_tokens_seen": 28974176,
      "step": 15300
    },
    {
      "epoch": 2.4967778774777716,
      "grad_norm": 0.002735377522185445,
      "learning_rate": 0.2040776717245208,
      "loss": 0.3927,
      "num_input_tokens_seen": 28983184,
      "step": 15305
    },
    {
      "epoch": 2.4975936046985887,
      "grad_norm": 0.0012269228463992476,
      "learning_rate": 0.2040227239087364,
      "loss": 0.3201,
      "num_input_tokens_seen": 28992080,
      "step": 15310
    },
    {
      "epoch": 2.4984093319194063,
      "grad_norm": 0.0037244008854031563,
      "learning_rate": 0.20396776776196904,
      "loss": 0.4146,
      "num_input_tokens_seen": 29002368,
      "step": 15315
    },
    {
      "epoch": 2.4992250591402234,
      "grad_norm": 0.002300404477864504,
      "learning_rate": 0.20391280329269373,
      "loss": 0.3634,
      "num_input_tokens_seen": 29011392,
      "step": 15320
    },
    {
      "epoch": 2.500040786361041,
      "grad_norm": 0.0009386052843183279,
      "learning_rate": 0.20385783050938663,
      "loss": 0.2773,
      "num_input_tokens_seen": 29020832,
      "step": 15325
    },
    {
      "epoch": 2.500856513581858,
      "grad_norm": 0.0012452146038413048,
      "learning_rate": 0.20380284942052526,
      "loss": 0.2692,
      "num_input_tokens_seen": 29030272,
      "step": 15330
    },
    {
      "epoch": 2.5016722408026757,
      "grad_norm": 0.002237240318208933,
      "learning_rate": 0.2037478600345884,
      "loss": 0.3909,
      "num_input_tokens_seen": 29040800,
      "step": 15335
    },
    {
      "epoch": 2.502487968023493,
      "grad_norm": 0.001276531140320003,
      "learning_rate": 0.20369286236005604,
      "loss": 0.3155,
      "num_input_tokens_seen": 29050512,
      "step": 15340
    },
    {
      "epoch": 2.5033036952443104,
      "grad_norm": 0.0021831877529621124,
      "learning_rate": 0.20363785640540957,
      "loss": 0.336,
      "num_input_tokens_seen": 29060528,
      "step": 15345
    },
    {
      "epoch": 2.5041194224651275,
      "grad_norm": 0.002345562446862459,
      "learning_rate": 0.2035828421791316,
      "loss": 0.307,
      "num_input_tokens_seen": 29070448,
      "step": 15350
    },
    {
      "epoch": 2.504935149685945,
      "grad_norm": 0.0025475926231592894,
      "learning_rate": 0.20352781968970599,
      "loss": 0.3073,
      "num_input_tokens_seen": 29079760,
      "step": 15355
    },
    {
      "epoch": 2.5057508769067622,
      "grad_norm": 0.0014591667568311095,
      "learning_rate": 0.2034727889456179,
      "loss": 0.3326,
      "num_input_tokens_seen": 29089008,
      "step": 15360
    },
    {
      "epoch": 2.50656660412758,
      "grad_norm": 0.003019869327545166,
      "learning_rate": 0.2034177499553538,
      "loss": 0.3575,
      "num_input_tokens_seen": 29099968,
      "step": 15365
    },
    {
      "epoch": 2.507382331348397,
      "grad_norm": 0.003252894850447774,
      "learning_rate": 0.2033627027274014,
      "loss": 0.2963,
      "num_input_tokens_seen": 29108912,
      "step": 15370
    },
    {
      "epoch": 2.5081980585692145,
      "grad_norm": 0.001847125357016921,
      "learning_rate": 0.20330764727024955,
      "loss": 0.4096,
      "num_input_tokens_seen": 29118656,
      "step": 15375
    },
    {
      "epoch": 2.5090137857900316,
      "grad_norm": 0.002125159837305546,
      "learning_rate": 0.20325258359238868,
      "loss": 0.3082,
      "num_input_tokens_seen": 29128560,
      "step": 15380
    },
    {
      "epoch": 2.5098295130108492,
      "grad_norm": 0.001576687442138791,
      "learning_rate": 0.20319751170231018,
      "loss": 0.3329,
      "num_input_tokens_seen": 29138032,
      "step": 15385
    },
    {
      "epoch": 2.510645240231667,
      "grad_norm": 0.002098450902849436,
      "learning_rate": 0.2031424316085068,
      "loss": 0.3205,
      "num_input_tokens_seen": 29147120,
      "step": 15390
    },
    {
      "epoch": 2.511460967452484,
      "grad_norm": 0.0018260091310366988,
      "learning_rate": 0.20308734331947265,
      "loss": 0.3269,
      "num_input_tokens_seen": 29155504,
      "step": 15395
    },
    {
      "epoch": 2.512276694673301,
      "grad_norm": 0.0031766307074576616,
      "learning_rate": 0.20303224684370305,
      "loss": 0.3642,
      "num_input_tokens_seen": 29164512,
      "step": 15400
    },
    {
      "epoch": 2.512276694673301,
      "eval_loss": 0.3283045291900635,
      "eval_runtime": 152.5969,
      "eval_samples_per_second": 17.858,
      "eval_steps_per_second": 8.932,
      "num_input_tokens_seen": 29164512,
      "step": 15400
    },
    {
      "epoch": 2.5130924218941186,
      "grad_norm": 0.001720725791528821,
      "learning_rate": 0.20297714218969456,
      "loss": 0.3121,
      "num_input_tokens_seen": 29172336,
      "step": 15405
    },
    {
      "epoch": 2.513908149114936,
      "grad_norm": 0.0016841371543705463,
      "learning_rate": 0.20292202936594497,
      "loss": 0.3333,
      "num_input_tokens_seen": 29182592,
      "step": 15410
    },
    {
      "epoch": 2.5147238763357533,
      "grad_norm": 0.0008540397393517196,
      "learning_rate": 0.2028669083809534,
      "loss": 0.2796,
      "num_input_tokens_seen": 29191968,
      "step": 15415
    },
    {
      "epoch": 2.5155396035565705,
      "grad_norm": 0.0031070783734321594,
      "learning_rate": 0.20281177924322016,
      "loss": 0.3453,
      "num_input_tokens_seen": 29202944,
      "step": 15420
    },
    {
      "epoch": 2.516355330777388,
      "grad_norm": 0.002333835233002901,
      "learning_rate": 0.2027566419612469,
      "loss": 0.3108,
      "num_input_tokens_seen": 29213328,
      "step": 15425
    },
    {
      "epoch": 2.5171710579982056,
      "grad_norm": 0.0016016429290175438,
      "learning_rate": 0.20270149654353647,
      "loss": 0.3415,
      "num_input_tokens_seen": 29223120,
      "step": 15430
    },
    {
      "epoch": 2.5179867852190227,
      "grad_norm": 0.0008877075742930174,
      "learning_rate": 0.202646342998593,
      "loss": 0.3474,
      "num_input_tokens_seen": 29232240,
      "step": 15435
    },
    {
      "epoch": 2.51880251243984,
      "grad_norm": 0.0020630776416510344,
      "learning_rate": 0.20259118133492185,
      "loss": 0.2995,
      "num_input_tokens_seen": 29241376,
      "step": 15440
    },
    {
      "epoch": 2.5196182396606575,
      "grad_norm": 0.0028656248468905687,
      "learning_rate": 0.20253601156102966,
      "loss": 0.3235,
      "num_input_tokens_seen": 29250704,
      "step": 15445
    },
    {
      "epoch": 2.520433966881475,
      "grad_norm": 0.0034606484696269035,
      "learning_rate": 0.20248083368542422,
      "loss": 0.3394,
      "num_input_tokens_seen": 29260976,
      "step": 15450
    },
    {
      "epoch": 2.521249694102292,
      "grad_norm": 0.003632253734394908,
      "learning_rate": 0.2024256477166147,
      "loss": 0.387,
      "num_input_tokens_seen": 29270576,
      "step": 15455
    },
    {
      "epoch": 2.5220654213231093,
      "grad_norm": 0.001569664105772972,
      "learning_rate": 0.2023704536631115,
      "loss": 0.3717,
      "num_input_tokens_seen": 29279888,
      "step": 15460
    },
    {
      "epoch": 2.522881148543927,
      "grad_norm": 0.0018398945685476065,
      "learning_rate": 0.20231525153342625,
      "loss": 0.3408,
      "num_input_tokens_seen": 29287936,
      "step": 15465
    },
    {
      "epoch": 2.5236968757647444,
      "grad_norm": 0.0013159903464838862,
      "learning_rate": 0.20226004133607173,
      "loss": 0.3353,
      "num_input_tokens_seen": 29297216,
      "step": 15470
    },
    {
      "epoch": 2.5245126029855616,
      "grad_norm": 0.0019699959084391594,
      "learning_rate": 0.20220482307956214,
      "loss": 0.3553,
      "num_input_tokens_seen": 29306848,
      "step": 15475
    },
    {
      "epoch": 2.525328330206379,
      "grad_norm": 0.000800458830781281,
      "learning_rate": 0.20214959677241276,
      "loss": 0.2994,
      "num_input_tokens_seen": 29316368,
      "step": 15480
    },
    {
      "epoch": 2.5261440574271963,
      "grad_norm": 0.0016022383933886886,
      "learning_rate": 0.20209436242314022,
      "loss": 0.3651,
      "num_input_tokens_seen": 29326032,
      "step": 15485
    },
    {
      "epoch": 2.526959784648014,
      "grad_norm": 0.0013657899107784033,
      "learning_rate": 0.2020391200402623,
      "loss": 0.3519,
      "num_input_tokens_seen": 29335328,
      "step": 15490
    },
    {
      "epoch": 2.527775511868831,
      "grad_norm": 0.0026466569397598505,
      "learning_rate": 0.2019838696322981,
      "loss": 0.3378,
      "num_input_tokens_seen": 29345392,
      "step": 15495
    },
    {
      "epoch": 2.5285912390896486,
      "grad_norm": 0.0011016944190487266,
      "learning_rate": 0.20192861120776798,
      "loss": 0.3476,
      "num_input_tokens_seen": 29354816,
      "step": 15500
    },
    {
      "epoch": 2.5294069663104657,
      "grad_norm": 0.0067222630605101585,
      "learning_rate": 0.20187334477519345,
      "loss": 0.3802,
      "num_input_tokens_seen": 29364928,
      "step": 15505
    },
    {
      "epoch": 2.5302226935312833,
      "grad_norm": 0.0018485645996406674,
      "learning_rate": 0.20181807034309726,
      "loss": 0.3444,
      "num_input_tokens_seen": 29374576,
      "step": 15510
    },
    {
      "epoch": 2.5310384207521004,
      "grad_norm": 0.0013808740768581629,
      "learning_rate": 0.2017627879200034,
      "loss": 0.3154,
      "num_input_tokens_seen": 29383760,
      "step": 15515
    },
    {
      "epoch": 2.531854147972918,
      "grad_norm": 0.002097177319228649,
      "learning_rate": 0.2017074975144372,
      "loss": 0.3432,
      "num_input_tokens_seen": 29393248,
      "step": 15520
    },
    {
      "epoch": 2.532669875193735,
      "grad_norm": 0.0016030378174036741,
      "learning_rate": 0.20165219913492508,
      "loss": 0.3242,
      "num_input_tokens_seen": 29402960,
      "step": 15525
    },
    {
      "epoch": 2.5334856024145527,
      "grad_norm": 0.002863076515495777,
      "learning_rate": 0.20159689278999468,
      "loss": 0.3179,
      "num_input_tokens_seen": 29413344,
      "step": 15530
    },
    {
      "epoch": 2.53430132963537,
      "grad_norm": 0.0019019471947103739,
      "learning_rate": 0.20154157848817508,
      "loss": 0.3525,
      "num_input_tokens_seen": 29422432,
      "step": 15535
    },
    {
      "epoch": 2.5351170568561874,
      "grad_norm": 0.0019255527295172215,
      "learning_rate": 0.20148625623799632,
      "loss": 0.2996,
      "num_input_tokens_seen": 29431920,
      "step": 15540
    },
    {
      "epoch": 2.5359327840770045,
      "grad_norm": 0.002347463509067893,
      "learning_rate": 0.20143092604798984,
      "loss": 0.364,
      "num_input_tokens_seen": 29441152,
      "step": 15545
    },
    {
      "epoch": 2.536748511297822,
      "grad_norm": 0.0012984528439119458,
      "learning_rate": 0.2013755879266883,
      "loss": 0.3417,
      "num_input_tokens_seen": 29451648,
      "step": 15550
    },
    {
      "epoch": 2.537564238518639,
      "grad_norm": 0.0008975945529527962,
      "learning_rate": 0.20132024188262543,
      "loss": 0.3553,
      "num_input_tokens_seen": 29460256,
      "step": 15555
    },
    {
      "epoch": 2.538379965739457,
      "grad_norm": 0.0016854420537129045,
      "learning_rate": 0.2012648879243363,
      "loss": 0.3354,
      "num_input_tokens_seen": 29468496,
      "step": 15560
    },
    {
      "epoch": 2.539195692960274,
      "grad_norm": 0.002544727176427841,
      "learning_rate": 0.20120952606035725,
      "loss": 0.3242,
      "num_input_tokens_seen": 29477888,
      "step": 15565
    },
    {
      "epoch": 2.5400114201810915,
      "grad_norm": 0.0019638012163341045,
      "learning_rate": 0.20115415629922576,
      "loss": 0.3197,
      "num_input_tokens_seen": 29488224,
      "step": 15570
    },
    {
      "epoch": 2.5408271474019086,
      "grad_norm": 0.0013704515295103192,
      "learning_rate": 0.20109877864948048,
      "loss": 0.3216,
      "num_input_tokens_seen": 29497632,
      "step": 15575
    },
    {
      "epoch": 2.541642874622726,
      "grad_norm": 0.0009191979770548642,
      "learning_rate": 0.20104339311966138,
      "loss": 0.3142,
      "num_input_tokens_seen": 29507936,
      "step": 15580
    },
    {
      "epoch": 2.5424586018435438,
      "grad_norm": 0.0018067419296130538,
      "learning_rate": 0.2009879997183097,
      "loss": 0.3378,
      "num_input_tokens_seen": 29518048,
      "step": 15585
    },
    {
      "epoch": 2.543274329064361,
      "grad_norm": 0.0019674068316817284,
      "learning_rate": 0.20093259845396763,
      "loss": 0.278,
      "num_input_tokens_seen": 29526848,
      "step": 15590
    },
    {
      "epoch": 2.544090056285178,
      "grad_norm": 0.0024911295622587204,
      "learning_rate": 0.20087718933517884,
      "loss": 0.2988,
      "num_input_tokens_seen": 29535024,
      "step": 15595
    },
    {
      "epoch": 2.5449057835059956,
      "grad_norm": 0.0025638516526669264,
      "learning_rate": 0.20082177237048807,
      "loss": 0.3603,
      "num_input_tokens_seen": 29545056,
      "step": 15600
    },
    {
      "epoch": 2.5449057835059956,
      "eval_loss": 0.3529355525970459,
      "eval_runtime": 152.639,
      "eval_samples_per_second": 17.853,
      "eval_steps_per_second": 8.93,
      "num_input_tokens_seen": 29545056,
      "step": 15600
    },
    {
      "epoch": 2.545721510726813,
      "grad_norm": 0.002613308373838663,
      "learning_rate": 0.20076634756844133,
      "loss": 0.3749,
      "num_input_tokens_seen": 29553664,
      "step": 15605
    },
    {
      "epoch": 2.5465372379476303,
      "grad_norm": 0.0027475589886307716,
      "learning_rate": 0.20071091493758586,
      "loss": 0.3258,
      "num_input_tokens_seen": 29563904,
      "step": 15610
    },
    {
      "epoch": 2.5473529651684474,
      "grad_norm": 0.003029825631529093,
      "learning_rate": 0.20065547448647003,
      "loss": 0.359,
      "num_input_tokens_seen": 29573968,
      "step": 15615
    },
    {
      "epoch": 2.548168692389265,
      "grad_norm": 0.0023546728771179914,
      "learning_rate": 0.20060002622364348,
      "loss": 0.311,
      "num_input_tokens_seen": 29582768,
      "step": 15620
    },
    {
      "epoch": 2.5489844196100826,
      "grad_norm": 0.0021060360595583916,
      "learning_rate": 0.20054457015765695,
      "loss": 0.3336,
      "num_input_tokens_seen": 29591744,
      "step": 15625
    },
    {
      "epoch": 2.5498001468308997,
      "grad_norm": 0.0016976840561255813,
      "learning_rate": 0.20048910629706254,
      "loss": 0.3458,
      "num_input_tokens_seen": 29599984,
      "step": 15630
    },
    {
      "epoch": 2.550615874051717,
      "grad_norm": 0.0016783735482022166,
      "learning_rate": 0.20043363465041347,
      "loss": 0.3487,
      "num_input_tokens_seen": 29607936,
      "step": 15635
    },
    {
      "epoch": 2.5514316012725344,
      "grad_norm": 0.0019969912245869637,
      "learning_rate": 0.2003781552262641,
      "loss": 0.3574,
      "num_input_tokens_seen": 29617680,
      "step": 15640
    },
    {
      "epoch": 2.552247328493352,
      "grad_norm": 0.0008200027514249086,
      "learning_rate": 0.20032266803317014,
      "loss": 0.3375,
      "num_input_tokens_seen": 29627216,
      "step": 15645
    },
    {
      "epoch": 2.553063055714169,
      "grad_norm": 0.002122915117070079,
      "learning_rate": 0.2002671730796884,
      "loss": 0.4022,
      "num_input_tokens_seen": 29636480,
      "step": 15650
    },
    {
      "epoch": 2.5538787829349863,
      "grad_norm": 0.0011917531955987215,
      "learning_rate": 0.20021167037437684,
      "loss": 0.3163,
      "num_input_tokens_seen": 29646400,
      "step": 15655
    },
    {
      "epoch": 2.554694510155804,
      "grad_norm": 0.002765474608168006,
      "learning_rate": 0.20015615992579472,
      "loss": 0.3398,
      "num_input_tokens_seen": 29655856,
      "step": 15660
    },
    {
      "epoch": 2.5555102373766214,
      "grad_norm": 0.0006127864471636713,
      "learning_rate": 0.20010064174250244,
      "loss": 0.3347,
      "num_input_tokens_seen": 29666336,
      "step": 15665
    },
    {
      "epoch": 2.5563259645974385,
      "grad_norm": 0.0014253059634938836,
      "learning_rate": 0.2000451158330616,
      "loss": 0.3564,
      "num_input_tokens_seen": 29675504,
      "step": 15670
    },
    {
      "epoch": 2.557141691818256,
      "grad_norm": 0.0009291842579841614,
      "learning_rate": 0.199989582206035,
      "loss": 0.3502,
      "num_input_tokens_seen": 29684192,
      "step": 15675
    },
    {
      "epoch": 2.5579574190390733,
      "grad_norm": 0.0011900614481419325,
      "learning_rate": 0.1999340408699866,
      "loss": 0.3651,
      "num_input_tokens_seen": 29693744,
      "step": 15680
    },
    {
      "epoch": 2.558773146259891,
      "grad_norm": 0.0014451330062001944,
      "learning_rate": 0.19987849183348155,
      "loss": 0.3354,
      "num_input_tokens_seen": 29702224,
      "step": 15685
    },
    {
      "epoch": 2.559588873480708,
      "grad_norm": 0.0009974505519494414,
      "learning_rate": 0.19982293510508628,
      "loss": 0.3144,
      "num_input_tokens_seen": 29712304,
      "step": 15690
    },
    {
      "epoch": 2.5604046007015255,
      "grad_norm": 0.0013990940060466528,
      "learning_rate": 0.19976737069336833,
      "loss": 0.3662,
      "num_input_tokens_seen": 29721680,
      "step": 15695
    },
    {
      "epoch": 2.5612203279223427,
      "grad_norm": 0.0015779259847477078,
      "learning_rate": 0.1997117986068964,
      "loss": 0.3361,
      "num_input_tokens_seen": 29730992,
      "step": 15700
    },
    {
      "epoch": 2.5620360551431602,
      "grad_norm": 0.0023326051887124777,
      "learning_rate": 0.19965621885424037,
      "loss": 0.2542,
      "num_input_tokens_seen": 29740576,
      "step": 15705
    },
    {
      "epoch": 2.5628517823639774,
      "grad_norm": 0.001525240601040423,
      "learning_rate": 0.19960063144397142,
      "loss": 0.3486,
      "num_input_tokens_seen": 29751776,
      "step": 15710
    },
    {
      "epoch": 2.563667509584795,
      "grad_norm": 0.0016375030390918255,
      "learning_rate": 0.19954503638466176,
      "loss": 0.3793,
      "num_input_tokens_seen": 29761840,
      "step": 15715
    },
    {
      "epoch": 2.564483236805612,
      "grad_norm": 0.0023006475530564785,
      "learning_rate": 0.1994894336848848,
      "loss": 0.3624,
      "num_input_tokens_seen": 29770784,
      "step": 15720
    },
    {
      "epoch": 2.5652989640264297,
      "grad_norm": 0.0019330340437591076,
      "learning_rate": 0.1994338233532153,
      "loss": 0.376,
      "num_input_tokens_seen": 29779984,
      "step": 15725
    },
    {
      "epoch": 2.566114691247247,
      "grad_norm": 0.001521644415333867,
      "learning_rate": 0.19937820539822904,
      "loss": 0.333,
      "num_input_tokens_seen": 29788800,
      "step": 15730
    },
    {
      "epoch": 2.5669304184680644,
      "grad_norm": 0.0026073826011270285,
      "learning_rate": 0.199322579828503,
      "loss": 0.3353,
      "num_input_tokens_seen": 29797760,
      "step": 15735
    },
    {
      "epoch": 2.5677461456888815,
      "grad_norm": 0.0014000770170241594,
      "learning_rate": 0.19926694665261527,
      "loss": 0.3892,
      "num_input_tokens_seen": 29805552,
      "step": 15740
    },
    {
      "epoch": 2.568561872909699,
      "grad_norm": 0.0025778727140277624,
      "learning_rate": 0.19921130587914526,
      "loss": 0.3401,
      "num_input_tokens_seen": 29814512,
      "step": 15745
    },
    {
      "epoch": 2.569377600130516,
      "grad_norm": 0.001884360215626657,
      "learning_rate": 0.19915565751667344,
      "loss": 0.33,
      "num_input_tokens_seen": 29823952,
      "step": 15750
    },
    {
      "epoch": 2.5701933273513338,
      "grad_norm": 0.0030353162437677383,
      "learning_rate": 0.19910000157378152,
      "loss": 0.2973,
      "num_input_tokens_seen": 29832592,
      "step": 15755
    },
    {
      "epoch": 2.5710090545721513,
      "grad_norm": 0.00213158642873168,
      "learning_rate": 0.1990443380590523,
      "loss": 0.3327,
      "num_input_tokens_seen": 29842288,
      "step": 15760
    },
    {
      "epoch": 2.5718247817929685,
      "grad_norm": 0.003396642627194524,
      "learning_rate": 0.19898866698106984,
      "loss": 0.3456,
      "num_input_tokens_seen": 29852224,
      "step": 15765
    },
    {
      "epoch": 2.5726405090137856,
      "grad_norm": 0.0018926153425127268,
      "learning_rate": 0.19893298834841933,
      "loss": 0.4124,
      "num_input_tokens_seen": 29861296,
      "step": 15770
    },
    {
      "epoch": 2.573456236234603,
      "grad_norm": 0.0009093980770558119,
      "learning_rate": 0.19887730216968705,
      "loss": 0.3096,
      "num_input_tokens_seen": 29870480,
      "step": 15775
    },
    {
      "epoch": 2.5742719634554208,
      "grad_norm": 0.0021727599669247866,
      "learning_rate": 0.19882160845346053,
      "loss": 0.3676,
      "num_input_tokens_seen": 29880880,
      "step": 15780
    },
    {
      "epoch": 2.575087690676238,
      "grad_norm": 0.0025181779637932777,
      "learning_rate": 0.1987659072083285,
      "loss": 0.3437,
      "num_input_tokens_seen": 29891952,
      "step": 15785
    },
    {
      "epoch": 2.575903417897055,
      "grad_norm": 0.0016589075094088912,
      "learning_rate": 0.1987101984428807,
      "loss": 0.3164,
      "num_input_tokens_seen": 29902528,
      "step": 15790
    },
    {
      "epoch": 2.5767191451178726,
      "grad_norm": 0.001508632441982627,
      "learning_rate": 0.19865448216570822,
      "loss": 0.3446,
      "num_input_tokens_seen": 29911616,
      "step": 15795
    },
    {
      "epoch": 2.57753487233869,
      "grad_norm": 0.0015406813472509384,
      "learning_rate": 0.19859875838540317,
      "loss": 0.3109,
      "num_input_tokens_seen": 29922176,
      "step": 15800
    },
    {
      "epoch": 2.57753487233869,
      "eval_loss": 0.3304975926876068,
      "eval_runtime": 152.5254,
      "eval_samples_per_second": 17.866,
      "eval_steps_per_second": 8.936,
      "num_input_tokens_seen": 29922176,
      "step": 15800
    },
    {
      "epoch": 2.5783505995595073,
      "grad_norm": 0.001814668532460928,
      "learning_rate": 0.1985430271105588,
      "loss": 0.3645,
      "num_input_tokens_seen": 29930912,
      "step": 15805
    },
    {
      "epoch": 2.5791663267803244,
      "grad_norm": 0.0011930580949410796,
      "learning_rate": 0.19848728834976961,
      "loss": 0.3931,
      "num_input_tokens_seen": 29940464,
      "step": 15810
    },
    {
      "epoch": 2.579982054001142,
      "grad_norm": 0.0009194129961542785,
      "learning_rate": 0.19843154211163128,
      "loss": 0.3422,
      "num_input_tokens_seen": 29949632,
      "step": 15815
    },
    {
      "epoch": 2.5807977812219596,
      "grad_norm": 0.000802507740445435,
      "learning_rate": 0.1983757884047405,
      "loss": 0.2881,
      "num_input_tokens_seen": 29959008,
      "step": 15820
    },
    {
      "epoch": 2.5816135084427767,
      "grad_norm": 0.002700173994526267,
      "learning_rate": 0.1983200272376952,
      "loss": 0.3519,
      "num_input_tokens_seen": 29968528,
      "step": 15825
    },
    {
      "epoch": 2.582429235663594,
      "grad_norm": 0.002499060006812215,
      "learning_rate": 0.1982642586190945,
      "loss": 0.3627,
      "num_input_tokens_seen": 29977488,
      "step": 15830
    },
    {
      "epoch": 2.5832449628844114,
      "grad_norm": 0.002694193972274661,
      "learning_rate": 0.1982084825575386,
      "loss": 0.3339,
      "num_input_tokens_seen": 29987280,
      "step": 15835
    },
    {
      "epoch": 2.584060690105229,
      "grad_norm": 0.0020966504234820604,
      "learning_rate": 0.19815269906162883,
      "loss": 0.3349,
      "num_input_tokens_seen": 29996896,
      "step": 15840
    },
    {
      "epoch": 2.584876417326046,
      "grad_norm": 0.0017303064232692122,
      "learning_rate": 0.19809690813996775,
      "loss": 0.2988,
      "num_input_tokens_seen": 30006496,
      "step": 15845
    },
    {
      "epoch": 2.5856921445468637,
      "grad_norm": 0.0024192703422158957,
      "learning_rate": 0.19804110980115905,
      "loss": 0.3098,
      "num_input_tokens_seen": 30016032,
      "step": 15850
    },
    {
      "epoch": 2.586507871767681,
      "grad_norm": 0.0013587572611868382,
      "learning_rate": 0.19798530405380746,
      "loss": 0.2997,
      "num_input_tokens_seen": 30025168,
      "step": 15855
    },
    {
      "epoch": 2.5873235989884984,
      "grad_norm": 0.0017412295565009117,
      "learning_rate": 0.19792949090651893,
      "loss": 0.3541,
      "num_input_tokens_seen": 30034128,
      "step": 15860
    },
    {
      "epoch": 2.5881393262093155,
      "grad_norm": 0.003244841704145074,
      "learning_rate": 0.19787367036790066,
      "loss": 0.3292,
      "num_input_tokens_seen": 30043504,
      "step": 15865
    },
    {
      "epoch": 2.588955053430133,
      "grad_norm": 0.00199018232524395,
      "learning_rate": 0.19781784244656075,
      "loss": 0.2966,
      "num_input_tokens_seen": 30052992,
      "step": 15870
    },
    {
      "epoch": 2.5897707806509502,
      "grad_norm": 0.0014034101041033864,
      "learning_rate": 0.19776200715110864,
      "loss": 0.2564,
      "num_input_tokens_seen": 30062528,
      "step": 15875
    },
    {
      "epoch": 2.590586507871768,
      "grad_norm": 0.004507888108491898,
      "learning_rate": 0.1977061644901548,
      "loss": 0.3099,
      "num_input_tokens_seen": 30072688,
      "step": 15880
    },
    {
      "epoch": 2.591402235092585,
      "grad_norm": 0.0026266095228493214,
      "learning_rate": 0.1976503144723109,
      "loss": 0.3313,
      "num_input_tokens_seen": 30080752,
      "step": 15885
    },
    {
      "epoch": 2.5922179623134025,
      "grad_norm": 0.002945743501186371,
      "learning_rate": 0.19759445710618967,
      "loss": 0.2931,
      "num_input_tokens_seen": 30089504,
      "step": 15890
    },
    {
      "epoch": 2.5930336895342196,
      "grad_norm": 0.002539140172302723,
      "learning_rate": 0.19753859240040508,
      "loss": 0.3607,
      "num_input_tokens_seen": 30098912,
      "step": 15895
    },
    {
      "epoch": 2.593849416755037,
      "grad_norm": 0.002050805604085326,
      "learning_rate": 0.1974827203635721,
      "loss": 0.3522,
      "num_input_tokens_seen": 30108208,
      "step": 15900
    },
    {
      "epoch": 2.5946651439758543,
      "grad_norm": 0.0011258028680458665,
      "learning_rate": 0.19742684100430694,
      "loss": 0.3277,
      "num_input_tokens_seen": 30118352,
      "step": 15905
    },
    {
      "epoch": 2.595480871196672,
      "grad_norm": 0.0022204008419066668,
      "learning_rate": 0.19737095433122692,
      "loss": 0.3397,
      "num_input_tokens_seen": 30126640,
      "step": 15910
    },
    {
      "epoch": 2.596296598417489,
      "grad_norm": 0.0018935841508209705,
      "learning_rate": 0.19731506035295046,
      "loss": 0.3215,
      "num_input_tokens_seen": 30136800,
      "step": 15915
    },
    {
      "epoch": 2.5971123256383066,
      "grad_norm": 0.0018194244476035237,
      "learning_rate": 0.19725915907809702,
      "loss": 0.3635,
      "num_input_tokens_seen": 30146480,
      "step": 15920
    },
    {
      "epoch": 2.5979280528591238,
      "grad_norm": 0.0003715509665198624,
      "learning_rate": 0.1972032505152874,
      "loss": 0.3405,
      "num_input_tokens_seen": 30156448,
      "step": 15925
    },
    {
      "epoch": 2.5987437800799413,
      "grad_norm": 0.0013440776383504272,
      "learning_rate": 0.19714733467314338,
      "loss": 0.2971,
      "num_input_tokens_seen": 30166048,
      "step": 15930
    },
    {
      "epoch": 2.5995595073007585,
      "grad_norm": 0.0015021502040326595,
      "learning_rate": 0.19709141156028784,
      "loss": 0.319,
      "num_input_tokens_seen": 30175856,
      "step": 15935
    },
    {
      "epoch": 2.600375234521576,
      "grad_norm": 0.0003745588182937354,
      "learning_rate": 0.1970354811853448,
      "loss": 0.3613,
      "num_input_tokens_seen": 30187008,
      "step": 15940
    },
    {
      "epoch": 2.601190961742393,
      "grad_norm": 0.0010769421933218837,
      "learning_rate": 0.19697954355693953,
      "loss": 0.3585,
      "num_input_tokens_seen": 30196384,
      "step": 15945
    },
    {
      "epoch": 2.6020066889632107,
      "grad_norm": 0.0020127047318965197,
      "learning_rate": 0.19692359868369827,
      "loss": 0.2997,
      "num_input_tokens_seen": 30206848,
      "step": 15950
    },
    {
      "epoch": 2.6028224161840283,
      "grad_norm": 0.0019698364194482565,
      "learning_rate": 0.1968676465742484,
      "loss": 0.3044,
      "num_input_tokens_seen": 30216176,
      "step": 15955
    },
    {
      "epoch": 2.6036381434048455,
      "grad_norm": 0.0015819232212379575,
      "learning_rate": 0.19681168723721845,
      "loss": 0.3218,
      "num_input_tokens_seen": 30226720,
      "step": 15960
    },
    {
      "epoch": 2.6044538706256626,
      "grad_norm": 0.0008177856216207147,
      "learning_rate": 0.19675572068123803,
      "loss": 0.399,
      "num_input_tokens_seen": 30236832,
      "step": 15965
    },
    {
      "epoch": 2.60526959784648,
      "grad_norm": 0.002196901012212038,
      "learning_rate": 0.19669974691493794,
      "loss": 0.3541,
      "num_input_tokens_seen": 30247424,
      "step": 15970
    },
    {
      "epoch": 2.6060853250672977,
      "grad_norm": 0.0016857233131304383,
      "learning_rate": 0.19664376594695002,
      "loss": 0.3568,
      "num_input_tokens_seen": 30256960,
      "step": 15975
    },
    {
      "epoch": 2.606901052288115,
      "grad_norm": 0.0014482919359579682,
      "learning_rate": 0.19658777778590722,
      "loss": 0.3545,
      "num_input_tokens_seen": 30266544,
      "step": 15980
    },
    {
      "epoch": 2.607716779508932,
      "grad_norm": 0.0014782603830099106,
      "learning_rate": 0.19653178244044364,
      "loss": 0.3234,
      "num_input_tokens_seen": 30276848,
      "step": 15985
    },
    {
      "epoch": 2.6085325067297496,
      "grad_norm": 0.00196535955183208,
      "learning_rate": 0.19647577991919443,
      "loss": 0.3628,
      "num_input_tokens_seen": 30285696,
      "step": 15990
    },
    {
      "epoch": 2.609348233950567,
      "grad_norm": 0.0016775918193161488,
      "learning_rate": 0.1964197702307959,
      "loss": 0.3464,
      "num_input_tokens_seen": 30295664,
      "step": 15995
    },
    {
      "epoch": 2.6101639611713843,
      "grad_norm": 0.0011220388114452362,
      "learning_rate": 0.19636375338388545,
      "loss": 0.3171,
      "num_input_tokens_seen": 30304336,
      "step": 16000
    },
    {
      "epoch": 2.6101639611713843,
      "eval_loss": 0.32561859488487244,
      "eval_runtime": 152.6975,
      "eval_samples_per_second": 17.846,
      "eval_steps_per_second": 8.926,
      "num_input_tokens_seen": 30304336,
      "step": 16000
    },
    {
      "epoch": 2.6109796883922014,
      "grad_norm": 0.0011026073480024934,
      "learning_rate": 0.1963077293871016,
      "loss": 0.3396,
      "num_input_tokens_seen": 30313696,
      "step": 16005
    },
    {
      "epoch": 2.611795415613019,
      "grad_norm": 0.0008073618519119918,
      "learning_rate": 0.19625169824908395,
      "loss": 0.3209,
      "num_input_tokens_seen": 30323168,
      "step": 16010
    },
    {
      "epoch": 2.6126111428338366,
      "grad_norm": 0.0011211058590561152,
      "learning_rate": 0.19619565997847319,
      "loss": 0.3606,
      "num_input_tokens_seen": 30333152,
      "step": 16015
    },
    {
      "epoch": 2.6134268700546537,
      "grad_norm": 0.000800504582002759,
      "learning_rate": 0.19613961458391113,
      "loss": 0.3215,
      "num_input_tokens_seen": 30344176,
      "step": 16020
    },
    {
      "epoch": 2.614242597275471,
      "grad_norm": 0.0022999891079962254,
      "learning_rate": 0.19608356207404065,
      "loss": 0.3482,
      "num_input_tokens_seen": 30354336,
      "step": 16025
    },
    {
      "epoch": 2.6150583244962884,
      "grad_norm": 0.001916227862238884,
      "learning_rate": 0.1960275024575058,
      "loss": 0.321,
      "num_input_tokens_seen": 30363552,
      "step": 16030
    },
    {
      "epoch": 2.615874051717106,
      "grad_norm": 0.0019358554854989052,
      "learning_rate": 0.19597143574295164,
      "loss": 0.4091,
      "num_input_tokens_seen": 30373392,
      "step": 16035
    },
    {
      "epoch": 2.616689778937923,
      "grad_norm": 0.0014986059395596385,
      "learning_rate": 0.1959153619390244,
      "loss": 0.312,
      "num_input_tokens_seen": 30383408,
      "step": 16040
    },
    {
      "epoch": 2.6175055061587407,
      "grad_norm": 0.0018835163209587336,
      "learning_rate": 0.1958592810543713,
      "loss": 0.3004,
      "num_input_tokens_seen": 30391760,
      "step": 16045
    },
    {
      "epoch": 2.618321233379558,
      "grad_norm": 0.0029494687914848328,
      "learning_rate": 0.19580319309764077,
      "loss": 0.3298,
      "num_input_tokens_seen": 30400400,
      "step": 16050
    },
    {
      "epoch": 2.6191369606003754,
      "grad_norm": 0.003132457844913006,
      "learning_rate": 0.1957470980774823,
      "loss": 0.323,
      "num_input_tokens_seen": 30409984,
      "step": 16055
    },
    {
      "epoch": 2.6199526878211925,
      "grad_norm": 0.0024705093819648027,
      "learning_rate": 0.19569099600254639,
      "loss": 0.2971,
      "num_input_tokens_seen": 30418224,
      "step": 16060
    },
    {
      "epoch": 2.62076841504201,
      "grad_norm": 0.0018151390831917524,
      "learning_rate": 0.1956348868814847,
      "loss": 0.3016,
      "num_input_tokens_seen": 30427040,
      "step": 16065
    },
    {
      "epoch": 2.621584142262827,
      "grad_norm": 0.002482388401404023,
      "learning_rate": 0.19557877072295,
      "loss": 0.3155,
      "num_input_tokens_seen": 30436640,
      "step": 16070
    },
    {
      "epoch": 2.622399869483645,
      "grad_norm": 0.0030982207972556353,
      "learning_rate": 0.19552264753559603,
      "loss": 0.2498,
      "num_input_tokens_seen": 30446288,
      "step": 16075
    },
    {
      "epoch": 2.623215596704462,
      "grad_norm": 0.003471283707767725,
      "learning_rate": 0.19546651732807774,
      "loss": 0.4274,
      "num_input_tokens_seen": 30456080,
      "step": 16080
    },
    {
      "epoch": 2.6240313239252795,
      "grad_norm": 0.0030396876391023397,
      "learning_rate": 0.19541038010905112,
      "loss": 0.3181,
      "num_input_tokens_seen": 30465552,
      "step": 16085
    },
    {
      "epoch": 2.6248470511460966,
      "grad_norm": 0.002991790184751153,
      "learning_rate": 0.19535423588717324,
      "loss": 0.3532,
      "num_input_tokens_seen": 30475488,
      "step": 16090
    },
    {
      "epoch": 2.625662778366914,
      "grad_norm": 0.0028255193028599024,
      "learning_rate": 0.19529808467110224,
      "loss": 0.3306,
      "num_input_tokens_seen": 30484656,
      "step": 16095
    },
    {
      "epoch": 2.6264785055877313,
      "grad_norm": 0.003378732595592737,
      "learning_rate": 0.19524192646949734,
      "loss": 0.3827,
      "num_input_tokens_seen": 30494848,
      "step": 16100
    },
    {
      "epoch": 2.627294232808549,
      "grad_norm": 0.0005124058807268739,
      "learning_rate": 0.19518576129101878,
      "loss": 0.3609,
      "num_input_tokens_seen": 30503712,
      "step": 16105
    },
    {
      "epoch": 2.628109960029366,
      "grad_norm": 0.0025022157933562994,
      "learning_rate": 0.19512958914432804,
      "loss": 0.3262,
      "num_input_tokens_seen": 30512240,
      "step": 16110
    },
    {
      "epoch": 2.6289256872501836,
      "grad_norm": 0.00042983933235518634,
      "learning_rate": 0.1950734100380875,
      "loss": 0.3152,
      "num_input_tokens_seen": 30520464,
      "step": 16115
    },
    {
      "epoch": 2.6297414144710007,
      "grad_norm": 0.0027579180896282196,
      "learning_rate": 0.19501722398096066,
      "loss": 0.348,
      "num_input_tokens_seen": 30530560,
      "step": 16120
    },
    {
      "epoch": 2.6305571416918183,
      "grad_norm": 0.0012671687873080373,
      "learning_rate": 0.1949610309816122,
      "loss": 0.3832,
      "num_input_tokens_seen": 30540448,
      "step": 16125
    },
    {
      "epoch": 2.631372868912636,
      "grad_norm": 0.0011714432621374726,
      "learning_rate": 0.1949048310487078,
      "loss": 0.3249,
      "num_input_tokens_seen": 30550944,
      "step": 16130
    },
    {
      "epoch": 2.632188596133453,
      "grad_norm": 0.0005722786299884319,
      "learning_rate": 0.19484862419091406,
      "loss": 0.3372,
      "num_input_tokens_seen": 30560368,
      "step": 16135
    },
    {
      "epoch": 2.63300432335427,
      "grad_norm": 0.00032207576441578567,
      "learning_rate": 0.19479241041689893,
      "loss": 0.3325,
      "num_input_tokens_seen": 30569024,
      "step": 16140
    },
    {
      "epoch": 2.6338200505750877,
      "grad_norm": 0.0008912334451451898,
      "learning_rate": 0.19473618973533116,
      "loss": 0.3356,
      "num_input_tokens_seen": 30580304,
      "step": 16145
    },
    {
      "epoch": 2.6346357777959053,
      "grad_norm": 0.002339682774618268,
      "learning_rate": 0.19467996215488076,
      "loss": 0.361,
      "num_input_tokens_seen": 30590096,
      "step": 16150
    },
    {
      "epoch": 2.6354515050167224,
      "grad_norm": 0.0023674287367612123,
      "learning_rate": 0.1946237276842187,
      "loss": 0.3628,
      "num_input_tokens_seen": 30600032,
      "step": 16155
    },
    {
      "epoch": 2.6362672322375396,
      "grad_norm": 0.0016564264660701156,
      "learning_rate": 0.19456748633201712,
      "loss": 0.351,
      "num_input_tokens_seen": 30609744,
      "step": 16160
    },
    {
      "epoch": 2.637082959458357,
      "grad_norm": 0.003267489606514573,
      "learning_rate": 0.194511238106949,
      "loss": 0.3205,
      "num_input_tokens_seen": 30619072,
      "step": 16165
    },
    {
      "epoch": 2.6378986866791747,
      "grad_norm": 0.002166968071833253,
      "learning_rate": 0.19445498301768863,
      "loss": 0.3187,
      "num_input_tokens_seen": 30628656,
      "step": 16170
    },
    {
      "epoch": 2.638714413899992,
      "grad_norm": 0.0010565845295786858,
      "learning_rate": 0.19439872107291126,
      "loss": 0.2823,
      "num_input_tokens_seen": 30638304,
      "step": 16175
    },
    {
      "epoch": 2.639530141120809,
      "grad_norm": 0.002423040335997939,
      "learning_rate": 0.1943424522812931,
      "loss": 0.331,
      "num_input_tokens_seen": 30648336,
      "step": 16180
    },
    {
      "epoch": 2.6403458683416265,
      "grad_norm": 0.00336930388584733,
      "learning_rate": 0.19428617665151157,
      "loss": 0.4096,
      "num_input_tokens_seen": 30658000,
      "step": 16185
    },
    {
      "epoch": 2.641161595562444,
      "grad_norm": 0.0021231838036328554,
      "learning_rate": 0.19422989419224507,
      "loss": 0.2984,
      "num_input_tokens_seen": 30667920,
      "step": 16190
    },
    {
      "epoch": 2.6419773227832613,
      "grad_norm": 0.0008286051452159882,
      "learning_rate": 0.19417360491217303,
      "loss": 0.3396,
      "num_input_tokens_seen": 30677808,
      "step": 16195
    },
    {
      "epoch": 2.6427930500040784,
      "grad_norm": 0.001740719540975988,
      "learning_rate": 0.19411730881997605,
      "loss": 0.3595,
      "num_input_tokens_seen": 30688608,
      "step": 16200
    },
    {
      "epoch": 2.6427930500040784,
      "eval_loss": 0.3383682072162628,
      "eval_runtime": 152.6837,
      "eval_samples_per_second": 17.847,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 30688608,
      "step": 16200
    },
    {
      "epoch": 2.643608777224896,
      "grad_norm": 0.0038273376412689686,
      "learning_rate": 0.1940610059243356,
      "loss": 0.3279,
      "num_input_tokens_seen": 30698352,
      "step": 16205
    },
    {
      "epoch": 2.6444245044457135,
      "grad_norm": 0.0013289761263877153,
      "learning_rate": 0.19400469623393435,
      "loss": 0.2962,
      "num_input_tokens_seen": 30707952,
      "step": 16210
    },
    {
      "epoch": 2.6452402316665307,
      "grad_norm": 0.0013935774331912398,
      "learning_rate": 0.1939483797574559,
      "loss": 0.2868,
      "num_input_tokens_seen": 30717424,
      "step": 16215
    },
    {
      "epoch": 2.6460559588873482,
      "grad_norm": 0.002985071623697877,
      "learning_rate": 0.19389205650358504,
      "loss": 0.3503,
      "num_input_tokens_seen": 30726720,
      "step": 16220
    },
    {
      "epoch": 2.6468716861081654,
      "grad_norm": 0.0017634272808209062,
      "learning_rate": 0.19383572648100747,
      "loss": 0.3631,
      "num_input_tokens_seen": 30736480,
      "step": 16225
    },
    {
      "epoch": 2.647687413328983,
      "grad_norm": 0.0007498295162804425,
      "learning_rate": 0.19377938969841,
      "loss": 0.318,
      "num_input_tokens_seen": 30745680,
      "step": 16230
    },
    {
      "epoch": 2.6485031405498,
      "grad_norm": 0.0013451193226501346,
      "learning_rate": 0.1937230461644805,
      "loss": 0.3799,
      "num_input_tokens_seen": 30754736,
      "step": 16235
    },
    {
      "epoch": 2.6493188677706176,
      "grad_norm": 0.001824690611101687,
      "learning_rate": 0.19366669588790777,
      "loss": 0.3628,
      "num_input_tokens_seen": 30763392,
      "step": 16240
    },
    {
      "epoch": 2.6501345949914348,
      "grad_norm": 0.0010753765236586332,
      "learning_rate": 0.19361033887738185,
      "loss": 0.3348,
      "num_input_tokens_seen": 30773408,
      "step": 16245
    },
    {
      "epoch": 2.6509503222122524,
      "grad_norm": 0.0022896702867001295,
      "learning_rate": 0.19355397514159361,
      "loss": 0.312,
      "num_input_tokens_seen": 30782656,
      "step": 16250
    },
    {
      "epoch": 2.6517660494330695,
      "grad_norm": 0.0010328018106520176,
      "learning_rate": 0.19349760468923508,
      "loss": 0.3634,
      "num_input_tokens_seen": 30793136,
      "step": 16255
    },
    {
      "epoch": 2.652581776653887,
      "grad_norm": 0.0020148332696408033,
      "learning_rate": 0.19344122752899925,
      "loss": 0.3275,
      "num_input_tokens_seen": 30801760,
      "step": 16260
    },
    {
      "epoch": 2.653397503874704,
      "grad_norm": 0.003029941115528345,
      "learning_rate": 0.1933848436695802,
      "loss": 0.3962,
      "num_input_tokens_seen": 30811728,
      "step": 16265
    },
    {
      "epoch": 2.6542132310955218,
      "grad_norm": 0.0012603894574567676,
      "learning_rate": 0.1933284531196731,
      "loss": 0.3541,
      "num_input_tokens_seen": 30821840,
      "step": 16270
    },
    {
      "epoch": 2.655028958316339,
      "grad_norm": 0.0008046120055951178,
      "learning_rate": 0.19327205588797403,
      "loss": 0.322,
      "num_input_tokens_seen": 30831632,
      "step": 16275
    },
    {
      "epoch": 2.6558446855371565,
      "grad_norm": 0.0017774823354557157,
      "learning_rate": 0.19321565198318014,
      "loss": 0.3459,
      "num_input_tokens_seen": 30840976,
      "step": 16280
    },
    {
      "epoch": 2.6566604127579736,
      "grad_norm": 0.0009059170843102038,
      "learning_rate": 0.1931592414139896,
      "loss": 0.3044,
      "num_input_tokens_seen": 30850576,
      "step": 16285
    },
    {
      "epoch": 2.657476139978791,
      "grad_norm": 0.0012172652641311288,
      "learning_rate": 0.19310282418910169,
      "loss": 0.3339,
      "num_input_tokens_seen": 30860032,
      "step": 16290
    },
    {
      "epoch": 2.6582918671996083,
      "grad_norm": 0.0016731087816879153,
      "learning_rate": 0.1930464003172166,
      "loss": 0.3146,
      "num_input_tokens_seen": 30870224,
      "step": 16295
    },
    {
      "epoch": 2.659107594420426,
      "grad_norm": 0.0014397691702470183,
      "learning_rate": 0.19298996980703567,
      "loss": 0.3549,
      "num_input_tokens_seen": 30879728,
      "step": 16300
    },
    {
      "epoch": 2.659923321641243,
      "grad_norm": 0.0017627788474783301,
      "learning_rate": 0.19293353266726113,
      "loss": 0.3659,
      "num_input_tokens_seen": 30887600,
      "step": 16305
    },
    {
      "epoch": 2.6607390488620606,
      "grad_norm": 0.002109035151079297,
      "learning_rate": 0.19287708890659633,
      "loss": 0.3733,
      "num_input_tokens_seen": 30896752,
      "step": 16310
    },
    {
      "epoch": 2.6615547760828777,
      "grad_norm": 0.0016208522720262408,
      "learning_rate": 0.19282063853374556,
      "loss": 0.3343,
      "num_input_tokens_seen": 30905664,
      "step": 16315
    },
    {
      "epoch": 2.6623705033036953,
      "grad_norm": 0.001063663512468338,
      "learning_rate": 0.19276418155741423,
      "loss": 0.2892,
      "num_input_tokens_seen": 30916240,
      "step": 16320
    },
    {
      "epoch": 2.663186230524513,
      "grad_norm": 0.0014839436626061797,
      "learning_rate": 0.19270771798630867,
      "loss": 0.3105,
      "num_input_tokens_seen": 30924656,
      "step": 16325
    },
    {
      "epoch": 2.66400195774533,
      "grad_norm": 0.0007470163982361555,
      "learning_rate": 0.1926512478291363,
      "loss": 0.3316,
      "num_input_tokens_seen": 30934832,
      "step": 16330
    },
    {
      "epoch": 2.664817684966147,
      "grad_norm": 0.0009544100612401962,
      "learning_rate": 0.19259477109460557,
      "loss": 0.4189,
      "num_input_tokens_seen": 30943248,
      "step": 16335
    },
    {
      "epoch": 2.6656334121869647,
      "grad_norm": 0.000939323625061661,
      "learning_rate": 0.19253828779142584,
      "loss": 0.3131,
      "num_input_tokens_seen": 30952720,
      "step": 16340
    },
    {
      "epoch": 2.6664491394077823,
      "grad_norm": 0.0013138545909896493,
      "learning_rate": 0.19248179792830755,
      "loss": 0.3412,
      "num_input_tokens_seen": 30963056,
      "step": 16345
    },
    {
      "epoch": 2.6672648666285994,
      "grad_norm": 0.0023173142690211535,
      "learning_rate": 0.19242530151396217,
      "loss": 0.3384,
      "num_input_tokens_seen": 30973520,
      "step": 16350
    },
    {
      "epoch": 2.6680805938494165,
      "grad_norm": 0.001964168855920434,
      "learning_rate": 0.19236879855710215,
      "loss": 0.3615,
      "num_input_tokens_seen": 30982064,
      "step": 16355
    },
    {
      "epoch": 2.668896321070234,
      "grad_norm": 0.0010950227733701468,
      "learning_rate": 0.19231228906644096,
      "loss": 0.3403,
      "num_input_tokens_seen": 30991488,
      "step": 16360
    },
    {
      "epoch": 2.6697120482910517,
      "grad_norm": 0.002174975350499153,
      "learning_rate": 0.19225577305069302,
      "loss": 0.3758,
      "num_input_tokens_seen": 31002192,
      "step": 16365
    },
    {
      "epoch": 2.670527775511869,
      "grad_norm": 0.0011300873011350632,
      "learning_rate": 0.1921992505185739,
      "loss": 0.3509,
      "num_input_tokens_seen": 31011488,
      "step": 16370
    },
    {
      "epoch": 2.671343502732686,
      "grad_norm": 0.0004093459283467382,
      "learning_rate": 0.19214272147880004,
      "loss": 0.3432,
      "num_input_tokens_seen": 31020352,
      "step": 16375
    },
    {
      "epoch": 2.6721592299535035,
      "grad_norm": 0.0007627535960637033,
      "learning_rate": 0.19208618594008892,
      "loss": 0.3272,
      "num_input_tokens_seen": 31029824,
      "step": 16380
    },
    {
      "epoch": 2.672974957174321,
      "grad_norm": 0.0010661885607987642,
      "learning_rate": 0.19202964391115904,
      "loss": 0.3636,
      "num_input_tokens_seen": 31039808,
      "step": 16385
    },
    {
      "epoch": 2.6737906843951382,
      "grad_norm": 0.002332880860194564,
      "learning_rate": 0.1919730954007299,
      "loss": 0.3874,
      "num_input_tokens_seen": 31049696,
      "step": 16390
    },
    {
      "epoch": 2.6746064116159554,
      "grad_norm": 0.0017514561768621206,
      "learning_rate": 0.19191654041752199,
      "loss": 0.3079,
      "num_input_tokens_seen": 31059584,
      "step": 16395
    },
    {
      "epoch": 2.675422138836773,
      "grad_norm": 0.000751836400013417,
      "learning_rate": 0.19185997897025678,
      "loss": 0.2938,
      "num_input_tokens_seen": 31067744,
      "step": 16400
    },
    {
      "epoch": 2.675422138836773,
      "eval_loss": 0.32562124729156494,
      "eval_runtime": 152.6783,
      "eval_samples_per_second": 17.848,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 31067744,
      "step": 16400
    },
    {
      "epoch": 2.6762378660575905,
      "grad_norm": 0.0011130060302093625,
      "learning_rate": 0.19180341106765672,
      "loss": 0.3001,
      "num_input_tokens_seen": 31077808,
      "step": 16405
    },
    {
      "epoch": 2.6770535932784076,
      "grad_norm": 0.0011069450993090868,
      "learning_rate": 0.19174683671844536,
      "loss": 0.3824,
      "num_input_tokens_seen": 31087936,
      "step": 16410
    },
    {
      "epoch": 2.677869320499225,
      "grad_norm": 0.0013982361415401101,
      "learning_rate": 0.19169025593134717,
      "loss": 0.3336,
      "num_input_tokens_seen": 31097312,
      "step": 16415
    },
    {
      "epoch": 2.6786850477200423,
      "grad_norm": 0.0011455011554062366,
      "learning_rate": 0.19163366871508764,
      "loss": 0.3711,
      "num_input_tokens_seen": 31106816,
      "step": 16420
    },
    {
      "epoch": 2.67950077494086,
      "grad_norm": 0.0020984960719943047,
      "learning_rate": 0.19157707507839317,
      "loss": 0.3585,
      "num_input_tokens_seen": 31116640,
      "step": 16425
    },
    {
      "epoch": 2.680316502161677,
      "grad_norm": 0.0012394206132739782,
      "learning_rate": 0.19152047502999123,
      "loss": 0.3526,
      "num_input_tokens_seen": 31126736,
      "step": 16430
    },
    {
      "epoch": 2.6811322293824946,
      "grad_norm": 0.000881916203070432,
      "learning_rate": 0.19146386857861025,
      "loss": 0.2849,
      "num_input_tokens_seen": 31136512,
      "step": 16435
    },
    {
      "epoch": 2.6819479566033118,
      "grad_norm": 0.00047507035196758807,
      "learning_rate": 0.19140725573297968,
      "loss": 0.4191,
      "num_input_tokens_seen": 31146544,
      "step": 16440
    },
    {
      "epoch": 2.6827636838241293,
      "grad_norm": 0.0012351498007774353,
      "learning_rate": 0.19135063650182987,
      "loss": 0.32,
      "num_input_tokens_seen": 31155744,
      "step": 16445
    },
    {
      "epoch": 2.6835794110449465,
      "grad_norm": 0.0019931585993617773,
      "learning_rate": 0.19129401089389234,
      "loss": 0.3259,
      "num_input_tokens_seen": 31165136,
      "step": 16450
    },
    {
      "epoch": 2.684395138265764,
      "grad_norm": 0.002073144307360053,
      "learning_rate": 0.19123737891789938,
      "loss": 0.3067,
      "num_input_tokens_seen": 31175600,
      "step": 16455
    },
    {
      "epoch": 2.685210865486581,
      "grad_norm": 0.0011359930504113436,
      "learning_rate": 0.19118074058258439,
      "loss": 0.3411,
      "num_input_tokens_seen": 31185584,
      "step": 16460
    },
    {
      "epoch": 2.6860265927073987,
      "grad_norm": 0.0008640251471661031,
      "learning_rate": 0.1911240958966816,
      "loss": 0.3207,
      "num_input_tokens_seen": 31195824,
      "step": 16465
    },
    {
      "epoch": 2.686842319928216,
      "grad_norm": 0.0012809450272470713,
      "learning_rate": 0.19106744486892652,
      "loss": 0.2916,
      "num_input_tokens_seen": 31205760,
      "step": 16470
    },
    {
      "epoch": 2.6876580471490334,
      "grad_norm": 0.0014522378332912922,
      "learning_rate": 0.1910107875080553,
      "loss": 0.2262,
      "num_input_tokens_seen": 31215824,
      "step": 16475
    },
    {
      "epoch": 2.6884737743698506,
      "grad_norm": 0.0014832157175987959,
      "learning_rate": 0.19095412382280533,
      "loss": 0.3409,
      "num_input_tokens_seen": 31224016,
      "step": 16480
    },
    {
      "epoch": 2.689289501590668,
      "grad_norm": 0.0009733329061418772,
      "learning_rate": 0.19089745382191473,
      "loss": 0.3055,
      "num_input_tokens_seen": 31233600,
      "step": 16485
    },
    {
      "epoch": 2.6901052288114853,
      "grad_norm": 0.0016847216757014394,
      "learning_rate": 0.19084077751412284,
      "loss": 0.3513,
      "num_input_tokens_seen": 31242784,
      "step": 16490
    },
    {
      "epoch": 2.690920956032303,
      "grad_norm": 0.001221417565830052,
      "learning_rate": 0.19078409490816986,
      "loss": 0.3601,
      "num_input_tokens_seen": 31252688,
      "step": 16495
    },
    {
      "epoch": 2.6917366832531204,
      "grad_norm": 0.0018079214496538043,
      "learning_rate": 0.19072740601279686,
      "loss": 0.2744,
      "num_input_tokens_seen": 31262240,
      "step": 16500
    },
    {
      "epoch": 2.6925524104739376,
      "grad_norm": 0.004075043369084597,
      "learning_rate": 0.19067071083674605,
      "loss": 0.3812,
      "num_input_tokens_seen": 31272880,
      "step": 16505
    },
    {
      "epoch": 2.6933681376947547,
      "grad_norm": 0.0015089221997186542,
      "learning_rate": 0.19061400938876052,
      "loss": 0.3561,
      "num_input_tokens_seen": 31283600,
      "step": 16510
    },
    {
      "epoch": 2.6941838649155723,
      "grad_norm": 0.002834065118804574,
      "learning_rate": 0.1905573016775844,
      "loss": 0.3351,
      "num_input_tokens_seen": 31292528,
      "step": 16515
    },
    {
      "epoch": 2.69499959213639,
      "grad_norm": 0.0009565870859660208,
      "learning_rate": 0.19050058771196263,
      "loss": 0.4556,
      "num_input_tokens_seen": 31301584,
      "step": 16520
    },
    {
      "epoch": 2.695815319357207,
      "grad_norm": 0.0009981654584407806,
      "learning_rate": 0.19044386750064132,
      "loss": 0.3342,
      "num_input_tokens_seen": 31311472,
      "step": 16525
    },
    {
      "epoch": 2.696631046578024,
      "grad_norm": 0.0012620632769539952,
      "learning_rate": 0.19038714105236737,
      "loss": 0.3939,
      "num_input_tokens_seen": 31321456,
      "step": 16530
    },
    {
      "epoch": 2.6974467737988417,
      "grad_norm": 0.0015639831544831395,
      "learning_rate": 0.19033040837588874,
      "loss": 0.3573,
      "num_input_tokens_seen": 31331472,
      "step": 16535
    },
    {
      "epoch": 2.6982625010196593,
      "grad_norm": 0.001281315111555159,
      "learning_rate": 0.1902736694799543,
      "loss": 0.3297,
      "num_input_tokens_seen": 31340416,
      "step": 16540
    },
    {
      "epoch": 2.6990782282404764,
      "grad_norm": 0.0007671655621379614,
      "learning_rate": 0.19021692437331392,
      "loss": 0.3143,
      "num_input_tokens_seen": 31348784,
      "step": 16545
    },
    {
      "epoch": 2.6998939554612935,
      "grad_norm": 0.0013645958388224244,
      "learning_rate": 0.1901601730647184,
      "loss": 0.3374,
      "num_input_tokens_seen": 31358784,
      "step": 16550
    },
    {
      "epoch": 2.700709682682111,
      "grad_norm": 0.0005948973703198135,
      "learning_rate": 0.19010341556291954,
      "loss": 0.3125,
      "num_input_tokens_seen": 31369376,
      "step": 16555
    },
    {
      "epoch": 2.7015254099029287,
      "grad_norm": 0.0019767279736697674,
      "learning_rate": 0.19004665187667,
      "loss": 0.3243,
      "num_input_tokens_seen": 31379408,
      "step": 16560
    },
    {
      "epoch": 2.702341137123746,
      "grad_norm": 0.0010666255839169025,
      "learning_rate": 0.1899898820147235,
      "loss": 0.3095,
      "num_input_tokens_seen": 31389248,
      "step": 16565
    },
    {
      "epoch": 2.703156864344563,
      "grad_norm": 0.0006641975487582386,
      "learning_rate": 0.18993310598583465,
      "loss": 0.328,
      "num_input_tokens_seen": 31398720,
      "step": 16570
    },
    {
      "epoch": 2.7039725915653805,
      "grad_norm": 0.003055349923670292,
      "learning_rate": 0.18987632379875904,
      "loss": 0.3546,
      "num_input_tokens_seen": 31409232,
      "step": 16575
    },
    {
      "epoch": 2.704788318786198,
      "grad_norm": 0.0012468713102862239,
      "learning_rate": 0.18981953546225314,
      "loss": 0.3606,
      "num_input_tokens_seen": 31419584,
      "step": 16580
    },
    {
      "epoch": 2.705604046007015,
      "grad_norm": 0.001510363188572228,
      "learning_rate": 0.18976274098507445,
      "loss": 0.3288,
      "num_input_tokens_seen": 31429088,
      "step": 16585
    },
    {
      "epoch": 2.7064197732278323,
      "grad_norm": 0.0012776539660990238,
      "learning_rate": 0.18970594037598146,
      "loss": 0.3765,
      "num_input_tokens_seen": 31437360,
      "step": 16590
    },
    {
      "epoch": 2.70723550044865,
      "grad_norm": 0.0026155125815421343,
      "learning_rate": 0.1896491336437335,
      "loss": 0.3367,
      "num_input_tokens_seen": 31445760,
      "step": 16595
    },
    {
      "epoch": 2.7080512276694675,
      "grad_norm": 0.0029376717284321785,
      "learning_rate": 0.18959232079709085,
      "loss": 0.3434,
      "num_input_tokens_seen": 31455328,
      "step": 16600
    },
    {
      "epoch": 2.7080512276694675,
      "eval_loss": 0.3325633704662323,
      "eval_runtime": 152.6518,
      "eval_samples_per_second": 17.851,
      "eval_steps_per_second": 8.929,
      "num_input_tokens_seen": 31455328,
      "step": 16600
    },
    {
      "epoch": 2.7088669548902846,
      "grad_norm": 0.0007433448918163776,
      "learning_rate": 0.18953550184481477,
      "loss": 0.3542,
      "num_input_tokens_seen": 31464944,
      "step": 16605
    },
    {
      "epoch": 2.709682682111102,
      "grad_norm": 0.0012136700097471476,
      "learning_rate": 0.18947867679566752,
      "loss": 0.3276,
      "num_input_tokens_seen": 31475088,
      "step": 16610
    },
    {
      "epoch": 2.7104984093319193,
      "grad_norm": 0.0014547595055773854,
      "learning_rate": 0.18942184565841216,
      "loss": 0.3222,
      "num_input_tokens_seen": 31484880,
      "step": 16615
    },
    {
      "epoch": 2.711314136552737,
      "grad_norm": 0.0018793356139212847,
      "learning_rate": 0.18936500844181278,
      "loss": 0.3356,
      "num_input_tokens_seen": 31495216,
      "step": 16620
    },
    {
      "epoch": 2.712129863773554,
      "grad_norm": 0.0015319723170250654,
      "learning_rate": 0.18930816515463436,
      "loss": 0.3689,
      "num_input_tokens_seen": 31505264,
      "step": 16625
    },
    {
      "epoch": 2.7129455909943716,
      "grad_norm": 0.0008373812306672335,
      "learning_rate": 0.18925131580564297,
      "loss": 0.2864,
      "num_input_tokens_seen": 31514624,
      "step": 16630
    },
    {
      "epoch": 2.7137613182151887,
      "grad_norm": 0.0023817303590476513,
      "learning_rate": 0.1891944604036054,
      "loss": 0.3439,
      "num_input_tokens_seen": 31523984,
      "step": 16635
    },
    {
      "epoch": 2.7145770454360063,
      "grad_norm": 0.001648470526561141,
      "learning_rate": 0.1891375989572895,
      "loss": 0.3269,
      "num_input_tokens_seen": 31533600,
      "step": 16640
    },
    {
      "epoch": 2.7153927726568234,
      "grad_norm": 0.0005590583896264434,
      "learning_rate": 0.18908073147546398,
      "loss": 0.3474,
      "num_input_tokens_seen": 31542208,
      "step": 16645
    },
    {
      "epoch": 2.716208499877641,
      "grad_norm": 0.0016142710810527205,
      "learning_rate": 0.18902385796689858,
      "loss": 0.3604,
      "num_input_tokens_seen": 31552112,
      "step": 16650
    },
    {
      "epoch": 2.717024227098458,
      "grad_norm": 0.0016263819998130202,
      "learning_rate": 0.18896697844036384,
      "loss": 0.3204,
      "num_input_tokens_seen": 31560864,
      "step": 16655
    },
    {
      "epoch": 2.7178399543192757,
      "grad_norm": 0.0013181234244257212,
      "learning_rate": 0.18891009290463137,
      "loss": 0.3619,
      "num_input_tokens_seen": 31570608,
      "step": 16660
    },
    {
      "epoch": 2.718655681540093,
      "grad_norm": 0.001314276596531272,
      "learning_rate": 0.18885320136847353,
      "loss": 0.3008,
      "num_input_tokens_seen": 31579536,
      "step": 16665
    },
    {
      "epoch": 2.7194714087609104,
      "grad_norm": 0.00044119407539255917,
      "learning_rate": 0.1887963038406639,
      "loss": 0.3162,
      "num_input_tokens_seen": 31588944,
      "step": 16670
    },
    {
      "epoch": 2.7202871359817276,
      "grad_norm": 0.0004935085889883339,
      "learning_rate": 0.18873940032997658,
      "loss": 0.3741,
      "num_input_tokens_seen": 31597344,
      "step": 16675
    },
    {
      "epoch": 2.721102863202545,
      "grad_norm": 0.0006684891995973885,
      "learning_rate": 0.18868249084518693,
      "loss": 0.3444,
      "num_input_tokens_seen": 31606848,
      "step": 16680
    },
    {
      "epoch": 2.7219185904233623,
      "grad_norm": 0.0013064802624285221,
      "learning_rate": 0.18862557539507102,
      "loss": 0.3117,
      "num_input_tokens_seen": 31616288,
      "step": 16685
    },
    {
      "epoch": 2.72273431764418,
      "grad_norm": 0.0009223572560586035,
      "learning_rate": 0.18856865398840605,
      "loss": 0.3224,
      "num_input_tokens_seen": 31625424,
      "step": 16690
    },
    {
      "epoch": 2.7235500448649974,
      "grad_norm": 0.0029497791547328234,
      "learning_rate": 0.18851172663396995,
      "loss": 0.3273,
      "num_input_tokens_seen": 31633376,
      "step": 16695
    },
    {
      "epoch": 2.7243657720858145,
      "grad_norm": 0.001980554312467575,
      "learning_rate": 0.1884547933405416,
      "loss": 0.3207,
      "num_input_tokens_seen": 31643520,
      "step": 16700
    },
    {
      "epoch": 2.7251814993066317,
      "grad_norm": 0.0018148425733670592,
      "learning_rate": 0.1883978541169009,
      "loss": 0.3283,
      "num_input_tokens_seen": 31653216,
      "step": 16705
    },
    {
      "epoch": 2.7259972265274492,
      "grad_norm": 0.0031807292252779007,
      "learning_rate": 0.18834090897182854,
      "loss": 0.3597,
      "num_input_tokens_seen": 31662352,
      "step": 16710
    },
    {
      "epoch": 2.726812953748267,
      "grad_norm": 0.0042901174165308475,
      "learning_rate": 0.1882839579141062,
      "loss": 0.3296,
      "num_input_tokens_seen": 31670880,
      "step": 16715
    },
    {
      "epoch": 2.727628680969084,
      "grad_norm": 0.0024332760367542505,
      "learning_rate": 0.18822700095251646,
      "loss": 0.3102,
      "num_input_tokens_seen": 31680912,
      "step": 16720
    },
    {
      "epoch": 2.728444408189901,
      "grad_norm": 0.001409720629453659,
      "learning_rate": 0.18817003809584273,
      "loss": 0.3302,
      "num_input_tokens_seen": 31690016,
      "step": 16725
    },
    {
      "epoch": 2.7292601354107187,
      "grad_norm": 0.0019859427120536566,
      "learning_rate": 0.1881130693528695,
      "loss": 0.4047,
      "num_input_tokens_seen": 31699600,
      "step": 16730
    },
    {
      "epoch": 2.7300758626315362,
      "grad_norm": 0.001634966116398573,
      "learning_rate": 0.18805609473238197,
      "loss": 0.3546,
      "num_input_tokens_seen": 31708448,
      "step": 16735
    },
    {
      "epoch": 2.7308915898523534,
      "grad_norm": 0.0012225195532664657,
      "learning_rate": 0.18799911424316643,
      "loss": 0.3114,
      "num_input_tokens_seen": 31719312,
      "step": 16740
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.0007115986081771553,
      "learning_rate": 0.18794212789400994,
      "loss": 0.308,
      "num_input_tokens_seen": 31728976,
      "step": 16745
    },
    {
      "epoch": 2.732523044293988,
      "grad_norm": 0.0006014561513438821,
      "learning_rate": 0.18788513569370052,
      "loss": 0.3123,
      "num_input_tokens_seen": 31737952,
      "step": 16750
    },
    {
      "epoch": 2.7333387715148056,
      "grad_norm": 0.001055937260389328,
      "learning_rate": 0.1878281376510271,
      "loss": 0.3492,
      "num_input_tokens_seen": 31746912,
      "step": 16755
    },
    {
      "epoch": 2.7341544987356228,
      "grad_norm": 0.002054734854027629,
      "learning_rate": 0.18777113377477941,
      "loss": 0.342,
      "num_input_tokens_seen": 31756208,
      "step": 16760
    },
    {
      "epoch": 2.73497022595644,
      "grad_norm": 0.0017505779396742582,
      "learning_rate": 0.1877141240737483,
      "loss": 0.3524,
      "num_input_tokens_seen": 31767824,
      "step": 16765
    },
    {
      "epoch": 2.7357859531772575,
      "grad_norm": 0.0016152760945260525,
      "learning_rate": 0.18765710855672527,
      "loss": 0.292,
      "num_input_tokens_seen": 31778416,
      "step": 16770
    },
    {
      "epoch": 2.736601680398075,
      "grad_norm": 0.0014569144695997238,
      "learning_rate": 0.18760008723250288,
      "loss": 0.3078,
      "num_input_tokens_seen": 31787952,
      "step": 16775
    },
    {
      "epoch": 2.737417407618892,
      "grad_norm": 0.0013332841917872429,
      "learning_rate": 0.18754306010987457,
      "loss": 0.3614,
      "num_input_tokens_seen": 31798480,
      "step": 16780
    },
    {
      "epoch": 2.7382331348397098,
      "grad_norm": 0.0012823238503187895,
      "learning_rate": 0.18748602719763457,
      "loss": 0.3293,
      "num_input_tokens_seen": 31806544,
      "step": 16785
    },
    {
      "epoch": 2.739048862060527,
      "grad_norm": 0.0012097315629944205,
      "learning_rate": 0.18742898850457804,
      "loss": 0.2959,
      "num_input_tokens_seen": 31816832,
      "step": 16790
    },
    {
      "epoch": 2.7398645892813445,
      "grad_norm": 0.0012186967069283128,
      "learning_rate": 0.1873719440395012,
      "loss": 0.3083,
      "num_input_tokens_seen": 31825312,
      "step": 16795
    },
    {
      "epoch": 2.7406803165021616,
      "grad_norm": 0.0035639822017401457,
      "learning_rate": 0.1873148938112009,
      "loss": 0.3638,
      "num_input_tokens_seen": 31833136,
      "step": 16800
    },
    {
      "epoch": 2.7406803165021616,
      "eval_loss": 0.3245711624622345,
      "eval_runtime": 152.6153,
      "eval_samples_per_second": 17.855,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 31833136,
      "step": 16800
    },
    {
      "epoch": 2.741496043722979,
      "grad_norm": 0.001343854353763163,
      "learning_rate": 0.18725783782847508,
      "loss": 0.3516,
      "num_input_tokens_seen": 31842912,
      "step": 16805
    },
    {
      "epoch": 2.7423117709437963,
      "grad_norm": 0.0022535917814821005,
      "learning_rate": 0.1872007761001224,
      "loss": 0.3189,
      "num_input_tokens_seen": 31852624,
      "step": 16810
    },
    {
      "epoch": 2.743127498164614,
      "grad_norm": 0.0016673596110194921,
      "learning_rate": 0.1871437086349426,
      "loss": 0.3643,
      "num_input_tokens_seen": 31861440,
      "step": 16815
    },
    {
      "epoch": 2.743943225385431,
      "grad_norm": 0.0015766412252560258,
      "learning_rate": 0.18708663544173615,
      "loss": 0.3168,
      "num_input_tokens_seen": 31870176,
      "step": 16820
    },
    {
      "epoch": 2.7447589526062486,
      "grad_norm": 0.0016075576422736049,
      "learning_rate": 0.18702955652930442,
      "loss": 0.3626,
      "num_input_tokens_seen": 31879856,
      "step": 16825
    },
    {
      "epoch": 2.7455746798270657,
      "grad_norm": 0.0010159928351640701,
      "learning_rate": 0.18697247190644972,
      "loss": 0.3406,
      "num_input_tokens_seen": 31889120,
      "step": 16830
    },
    {
      "epoch": 2.7463904070478833,
      "grad_norm": 0.0015976231079548597,
      "learning_rate": 0.18691538158197527,
      "loss": 0.2923,
      "num_input_tokens_seen": 31898816,
      "step": 16835
    },
    {
      "epoch": 2.7472061342687004,
      "grad_norm": 0.0012714703334495425,
      "learning_rate": 0.1868582855646851,
      "loss": 0.3313,
      "num_input_tokens_seen": 31908512,
      "step": 16840
    },
    {
      "epoch": 2.748021861489518,
      "grad_norm": 0.0015191162237897515,
      "learning_rate": 0.18680118386338404,
      "loss": 0.3045,
      "num_input_tokens_seen": 31917552,
      "step": 16845
    },
    {
      "epoch": 2.748837588710335,
      "grad_norm": 0.0019232102204114199,
      "learning_rate": 0.18674407648687794,
      "loss": 0.32,
      "num_input_tokens_seen": 31927696,
      "step": 16850
    },
    {
      "epoch": 2.7496533159311527,
      "grad_norm": 0.0021563954651355743,
      "learning_rate": 0.1866869634439736,
      "loss": 0.3378,
      "num_input_tokens_seen": 31937424,
      "step": 16855
    },
    {
      "epoch": 2.75046904315197,
      "grad_norm": 0.0024393247440457344,
      "learning_rate": 0.18662984474347838,
      "loss": 0.3301,
      "num_input_tokens_seen": 31946912,
      "step": 16860
    },
    {
      "epoch": 2.7512847703727874,
      "grad_norm": 0.001953451195731759,
      "learning_rate": 0.1865727203942008,
      "loss": 0.3975,
      "num_input_tokens_seen": 31954640,
      "step": 16865
    },
    {
      "epoch": 2.7521004975936045,
      "grad_norm": 0.001633765990845859,
      "learning_rate": 0.1865155904049501,
      "loss": 0.3483,
      "num_input_tokens_seen": 31965312,
      "step": 16870
    },
    {
      "epoch": 2.752916224814422,
      "grad_norm": 0.0009592515416443348,
      "learning_rate": 0.1864584547845365,
      "loss": 0.3334,
      "num_input_tokens_seen": 31974816,
      "step": 16875
    },
    {
      "epoch": 2.7537319520352392,
      "grad_norm": 0.00211225263774395,
      "learning_rate": 0.186401313541771,
      "loss": 0.3393,
      "num_input_tokens_seen": 31983968,
      "step": 16880
    },
    {
      "epoch": 2.754547679256057,
      "grad_norm": 0.001507022650912404,
      "learning_rate": 0.18634416668546552,
      "loss": 0.3426,
      "num_input_tokens_seen": 31992576,
      "step": 16885
    },
    {
      "epoch": 2.7553634064768744,
      "grad_norm": 0.001991522265598178,
      "learning_rate": 0.1862870142244328,
      "loss": 0.3205,
      "num_input_tokens_seen": 32000944,
      "step": 16890
    },
    {
      "epoch": 2.7561791336976915,
      "grad_norm": 0.0025539270136505365,
      "learning_rate": 0.1862298561674865,
      "loss": 0.2991,
      "num_input_tokens_seen": 32010384,
      "step": 16895
    },
    {
      "epoch": 2.7569948609185086,
      "grad_norm": 0.0014614419778808951,
      "learning_rate": 0.18617269252344104,
      "loss": 0.1899,
      "num_input_tokens_seen": 32021040,
      "step": 16900
    },
    {
      "epoch": 2.7578105881393262,
      "grad_norm": 0.002697273390367627,
      "learning_rate": 0.18611552330111186,
      "loss": 0.4771,
      "num_input_tokens_seen": 32030624,
      "step": 16905
    },
    {
      "epoch": 2.758626315360144,
      "grad_norm": 0.0012655237223953009,
      "learning_rate": 0.18605834850931507,
      "loss": 0.2634,
      "num_input_tokens_seen": 32039840,
      "step": 16910
    },
    {
      "epoch": 2.759442042580961,
      "grad_norm": 0.0029954968485981226,
      "learning_rate": 0.18600116815686787,
      "loss": 0.3259,
      "num_input_tokens_seen": 32050064,
      "step": 16915
    },
    {
      "epoch": 2.760257769801778,
      "grad_norm": 0.004809536039829254,
      "learning_rate": 0.1859439822525881,
      "loss": 0.3548,
      "num_input_tokens_seen": 32060208,
      "step": 16920
    },
    {
      "epoch": 2.7610734970225956,
      "grad_norm": 0.0012764156563207507,
      "learning_rate": 0.18588679080529455,
      "loss": 0.3714,
      "num_input_tokens_seen": 32069984,
      "step": 16925
    },
    {
      "epoch": 2.761889224243413,
      "grad_norm": 0.005230540409684181,
      "learning_rate": 0.1858295938238069,
      "loss": 0.3819,
      "num_input_tokens_seen": 32079840,
      "step": 16930
    },
    {
      "epoch": 2.7627049514642303,
      "grad_norm": 0.001498450874350965,
      "learning_rate": 0.18577239131694562,
      "loss": 0.3516,
      "num_input_tokens_seen": 32088656,
      "step": 16935
    },
    {
      "epoch": 2.7635206786850475,
      "grad_norm": 0.002527114236727357,
      "learning_rate": 0.18571518329353204,
      "loss": 0.3277,
      "num_input_tokens_seen": 32098176,
      "step": 16940
    },
    {
      "epoch": 2.764336405905865,
      "grad_norm": 0.0021655058953911066,
      "learning_rate": 0.18565796976238838,
      "loss": 0.2923,
      "num_input_tokens_seen": 32107952,
      "step": 16945
    },
    {
      "epoch": 2.7651521331266826,
      "grad_norm": 0.0024957233108580112,
      "learning_rate": 0.18560075073233764,
      "loss": 0.376,
      "num_input_tokens_seen": 32117184,
      "step": 16950
    },
    {
      "epoch": 2.7659678603474998,
      "grad_norm": 0.0006221865187399089,
      "learning_rate": 0.18554352621220377,
      "loss": 0.3425,
      "num_input_tokens_seen": 32126416,
      "step": 16955
    },
    {
      "epoch": 2.766783587568317,
      "grad_norm": 0.001155381673015654,
      "learning_rate": 0.18548629621081153,
      "loss": 0.3615,
      "num_input_tokens_seen": 32136080,
      "step": 16960
    },
    {
      "epoch": 2.7675993147891345,
      "grad_norm": 0.000566151924431324,
      "learning_rate": 0.18542906073698645,
      "loss": 0.3275,
      "num_input_tokens_seen": 32145712,
      "step": 16965
    },
    {
      "epoch": 2.768415042009952,
      "grad_norm": 0.0012740313541144133,
      "learning_rate": 0.18537181979955494,
      "loss": 0.3128,
      "num_input_tokens_seen": 32154528,
      "step": 16970
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.0017366062384098768,
      "learning_rate": 0.18531457340734434,
      "loss": 0.3661,
      "num_input_tokens_seen": 32164672,
      "step": 16975
    },
    {
      "epoch": 2.7700464964515867,
      "grad_norm": 0.0018291656160727143,
      "learning_rate": 0.1852573215691827,
      "loss": 0.2893,
      "num_input_tokens_seen": 32174544,
      "step": 16980
    },
    {
      "epoch": 2.770862223672404,
      "grad_norm": 0.002625885186716914,
      "learning_rate": 0.18520006429389904,
      "loss": 0.2853,
      "num_input_tokens_seen": 32183248,
      "step": 16985
    },
    {
      "epoch": 2.7716779508932214,
      "grad_norm": 0.0021123182959854603,
      "learning_rate": 0.1851428015903231,
      "loss": 0.3188,
      "num_input_tokens_seen": 32192768,
      "step": 16990
    },
    {
      "epoch": 2.7724936781140386,
      "grad_norm": 0.0020250072702765465,
      "learning_rate": 0.1850855334672855,
      "loss": 0.3271,
      "num_input_tokens_seen": 32202880,
      "step": 16995
    },
    {
      "epoch": 2.773309405334856,
      "grad_norm": 0.0028730176854878664,
      "learning_rate": 0.1850282599336178,
      "loss": 0.3623,
      "num_input_tokens_seen": 32213296,
      "step": 17000
    },
    {
      "epoch": 2.773309405334856,
      "eval_loss": 0.3232356607913971,
      "eval_runtime": 152.8741,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 32213296,
      "step": 17000
    },
    {
      "epoch": 2.7741251325556733,
      "grad_norm": 0.001768943970091641,
      "learning_rate": 0.18497098099815215,
      "loss": 0.3753,
      "num_input_tokens_seen": 32223056,
      "step": 17005
    },
    {
      "epoch": 2.774940859776491,
      "grad_norm": 0.0012158561730757356,
      "learning_rate": 0.18491369666972174,
      "loss": 0.3277,
      "num_input_tokens_seen": 32232336,
      "step": 17010
    },
    {
      "epoch": 2.775756586997308,
      "grad_norm": 0.001811102731153369,
      "learning_rate": 0.1848564069571606,
      "loss": 0.3384,
      "num_input_tokens_seen": 32240512,
      "step": 17015
    },
    {
      "epoch": 2.7765723142181256,
      "grad_norm": 0.0018585305660963058,
      "learning_rate": 0.18479911186930348,
      "loss": 0.341,
      "num_input_tokens_seen": 32249728,
      "step": 17020
    },
    {
      "epoch": 2.7773880414389427,
      "grad_norm": 0.00309176673181355,
      "learning_rate": 0.18474181141498597,
      "loss": 0.299,
      "num_input_tokens_seen": 32258944,
      "step": 17025
    },
    {
      "epoch": 2.7782037686597603,
      "grad_norm": 0.002313079312443733,
      "learning_rate": 0.18468450560304453,
      "loss": 0.3996,
      "num_input_tokens_seen": 32268896,
      "step": 17030
    },
    {
      "epoch": 2.7790194958805774,
      "grad_norm": 0.000877573445905,
      "learning_rate": 0.1846271944423165,
      "loss": 0.2941,
      "num_input_tokens_seen": 32279568,
      "step": 17035
    },
    {
      "epoch": 2.779835223101395,
      "grad_norm": 0.0024837094824761152,
      "learning_rate": 0.18456987794163993,
      "loss": 0.3337,
      "num_input_tokens_seen": 32288784,
      "step": 17040
    },
    {
      "epoch": 2.780650950322212,
      "grad_norm": 0.0026271366514265537,
      "learning_rate": 0.18451255610985373,
      "loss": 0.3611,
      "num_input_tokens_seen": 32298784,
      "step": 17045
    },
    {
      "epoch": 2.7814666775430297,
      "grad_norm": 0.0008116963435895741,
      "learning_rate": 0.18445522895579766,
      "loss": 0.3118,
      "num_input_tokens_seen": 32308480,
      "step": 17050
    },
    {
      "epoch": 2.782282404763847,
      "grad_norm": 0.0003506133798509836,
      "learning_rate": 0.1843978964883123,
      "loss": 0.2611,
      "num_input_tokens_seen": 32318960,
      "step": 17055
    },
    {
      "epoch": 2.7830981319846644,
      "grad_norm": 0.0012046919437125325,
      "learning_rate": 0.18434055871623906,
      "loss": 0.4231,
      "num_input_tokens_seen": 32329232,
      "step": 17060
    },
    {
      "epoch": 2.783913859205482,
      "grad_norm": 0.0017724300269037485,
      "learning_rate": 0.18428321564842007,
      "loss": 0.3503,
      "num_input_tokens_seen": 32337184,
      "step": 17065
    },
    {
      "epoch": 2.784729586426299,
      "grad_norm": 0.0026054088957607746,
      "learning_rate": 0.18422586729369841,
      "loss": 0.3353,
      "num_input_tokens_seen": 32346544,
      "step": 17070
    },
    {
      "epoch": 2.785545313647116,
      "grad_norm": 0.0018034335225820541,
      "learning_rate": 0.1841685136609179,
      "loss": 0.3517,
      "num_input_tokens_seen": 32354896,
      "step": 17075
    },
    {
      "epoch": 2.786361040867934,
      "grad_norm": 0.001837230403907597,
      "learning_rate": 0.18411115475892326,
      "loss": 0.3268,
      "num_input_tokens_seen": 32364224,
      "step": 17080
    },
    {
      "epoch": 2.7871767680887514,
      "grad_norm": 0.001814129762351513,
      "learning_rate": 0.18405379059655982,
      "loss": 0.3397,
      "num_input_tokens_seen": 32372688,
      "step": 17085
    },
    {
      "epoch": 2.7879924953095685,
      "grad_norm": 0.001581403543241322,
      "learning_rate": 0.1839964211826739,
      "loss": 0.3254,
      "num_input_tokens_seen": 32382512,
      "step": 17090
    },
    {
      "epoch": 2.7888082225303856,
      "grad_norm": 0.0023798050824552774,
      "learning_rate": 0.18393904652611265,
      "loss": 0.3321,
      "num_input_tokens_seen": 32392032,
      "step": 17095
    },
    {
      "epoch": 2.789623949751203,
      "grad_norm": 0.0020191946532577276,
      "learning_rate": 0.18388166663572392,
      "loss": 0.3365,
      "num_input_tokens_seen": 32400432,
      "step": 17100
    },
    {
      "epoch": 2.7904396769720208,
      "grad_norm": 0.0015703382669016719,
      "learning_rate": 0.18382428152035643,
      "loss": 0.3051,
      "num_input_tokens_seen": 32409520,
      "step": 17105
    },
    {
      "epoch": 2.791255404192838,
      "grad_norm": 0.00128079520072788,
      "learning_rate": 0.1837668911888596,
      "loss": 0.3236,
      "num_input_tokens_seen": 32418080,
      "step": 17110
    },
    {
      "epoch": 2.792071131413655,
      "grad_norm": 0.0007988564320839942,
      "learning_rate": 0.18370949565008388,
      "loss": 0.4718,
      "num_input_tokens_seen": 32427712,
      "step": 17115
    },
    {
      "epoch": 2.7928868586344726,
      "grad_norm": 0.0015878764679655433,
      "learning_rate": 0.1836520949128803,
      "loss": 0.3883,
      "num_input_tokens_seen": 32437504,
      "step": 17120
    },
    {
      "epoch": 2.79370258585529,
      "grad_norm": 0.0009142798371613026,
      "learning_rate": 0.18359468898610076,
      "loss": 0.3286,
      "num_input_tokens_seen": 32447456,
      "step": 17125
    },
    {
      "epoch": 2.7945183130761073,
      "grad_norm": 0.0014626809861510992,
      "learning_rate": 0.18353727787859797,
      "loss": 0.3404,
      "num_input_tokens_seen": 32455168,
      "step": 17130
    },
    {
      "epoch": 2.7953340402969244,
      "grad_norm": 0.0015386383747681975,
      "learning_rate": 0.18347986159922552,
      "loss": 0.2881,
      "num_input_tokens_seen": 32464240,
      "step": 17135
    },
    {
      "epoch": 2.796149767517742,
      "grad_norm": 0.0021403925493359566,
      "learning_rate": 0.1834224401568377,
      "loss": 0.3714,
      "num_input_tokens_seen": 32474576,
      "step": 17140
    },
    {
      "epoch": 2.7969654947385596,
      "grad_norm": 0.0013472484424710274,
      "learning_rate": 0.1833650135602896,
      "loss": 0.362,
      "num_input_tokens_seen": 32483984,
      "step": 17145
    },
    {
      "epoch": 2.7977812219593767,
      "grad_norm": 0.0005187715287320316,
      "learning_rate": 0.18330758181843707,
      "loss": 0.3337,
      "num_input_tokens_seen": 32493056,
      "step": 17150
    },
    {
      "epoch": 2.7985969491801943,
      "grad_norm": 0.0012571460101753473,
      "learning_rate": 0.18325014494013686,
      "loss": 0.3526,
      "num_input_tokens_seen": 32502960,
      "step": 17155
    },
    {
      "epoch": 2.7994126764010114,
      "grad_norm": 0.0007569692097604275,
      "learning_rate": 0.18319270293424647,
      "loss": 0.2832,
      "num_input_tokens_seen": 32512832,
      "step": 17160
    },
    {
      "epoch": 2.800228403621829,
      "grad_norm": 0.0035292450338602066,
      "learning_rate": 0.18313525580962417,
      "loss": 0.3423,
      "num_input_tokens_seen": 32522224,
      "step": 17165
    },
    {
      "epoch": 2.801044130842646,
      "grad_norm": 0.0012733038747683167,
      "learning_rate": 0.18307780357512896,
      "loss": 0.3305,
      "num_input_tokens_seen": 32530176,
      "step": 17170
    },
    {
      "epoch": 2.8018598580634637,
      "grad_norm": 0.0034950494300574064,
      "learning_rate": 0.1830203462396208,
      "loss": 0.3397,
      "num_input_tokens_seen": 32538976,
      "step": 17175
    },
    {
      "epoch": 2.802675585284281,
      "grad_norm": 0.005468964576721191,
      "learning_rate": 0.18296288381196033,
      "loss": 0.4014,
      "num_input_tokens_seen": 32548944,
      "step": 17180
    },
    {
      "epoch": 2.8034913125050984,
      "grad_norm": 0.001528833294287324,
      "learning_rate": 0.1829054163010089,
      "loss": 0.3166,
      "num_input_tokens_seen": 32558976,
      "step": 17185
    },
    {
      "epoch": 2.8043070397259156,
      "grad_norm": 0.0022784844040870667,
      "learning_rate": 0.18284794371562874,
      "loss": 0.2959,
      "num_input_tokens_seen": 32569200,
      "step": 17190
    },
    {
      "epoch": 2.805122766946733,
      "grad_norm": 0.0013660333352163434,
      "learning_rate": 0.18279046606468288,
      "loss": 0.3162,
      "num_input_tokens_seen": 32578688,
      "step": 17195
    },
    {
      "epoch": 2.8059384941675503,
      "grad_norm": 0.0013310443609952927,
      "learning_rate": 0.1827329833570351,
      "loss": 0.3703,
      "num_input_tokens_seen": 32588128,
      "step": 17200
    },
    {
      "epoch": 2.8059384941675503,
      "eval_loss": 0.33464348316192627,
      "eval_runtime": 152.6674,
      "eval_samples_per_second": 17.849,
      "eval_steps_per_second": 8.928,
      "num_input_tokens_seen": 32588128,
      "step": 17200
    },
    {
      "epoch": 2.806754221388368,
      "grad_norm": 0.001524440012872219,
      "learning_rate": 0.18267549560154991,
      "loss": 0.3661,
      "num_input_tokens_seen": 32598000,
      "step": 17205
    },
    {
      "epoch": 2.807569948609185,
      "grad_norm": 0.0026988936588168144,
      "learning_rate": 0.18261800280709267,
      "loss": 0.3243,
      "num_input_tokens_seen": 32607984,
      "step": 17210
    },
    {
      "epoch": 2.8083856758300025,
      "grad_norm": 0.002083074301481247,
      "learning_rate": 0.18256050498252957,
      "loss": 0.3349,
      "num_input_tokens_seen": 32617104,
      "step": 17215
    },
    {
      "epoch": 2.8092014030508197,
      "grad_norm": 0.002079180208966136,
      "learning_rate": 0.18250300213672735,
      "loss": 0.3265,
      "num_input_tokens_seen": 32625456,
      "step": 17220
    },
    {
      "epoch": 2.8100171302716372,
      "grad_norm": 0.0012435662792995572,
      "learning_rate": 0.18244549427855378,
      "loss": 0.3095,
      "num_input_tokens_seen": 32635280,
      "step": 17225
    },
    {
      "epoch": 2.8108328574924544,
      "grad_norm": 0.0033397681545466185,
      "learning_rate": 0.1823879814168772,
      "loss": 0.37,
      "num_input_tokens_seen": 32643136,
      "step": 17230
    },
    {
      "epoch": 2.811648584713272,
      "grad_norm": 0.0014538513496518135,
      "learning_rate": 0.18233046356056692,
      "loss": 0.3009,
      "num_input_tokens_seen": 32651600,
      "step": 17235
    },
    {
      "epoch": 2.812464311934089,
      "grad_norm": 0.0009852144867181778,
      "learning_rate": 0.18227294071849284,
      "loss": 0.3654,
      "num_input_tokens_seen": 32661264,
      "step": 17240
    },
    {
      "epoch": 2.8132800391549067,
      "grad_norm": 0.002295027021318674,
      "learning_rate": 0.18221541289952578,
      "loss": 0.3401,
      "num_input_tokens_seen": 32672032,
      "step": 17245
    },
    {
      "epoch": 2.814095766375724,
      "grad_norm": 0.00048736046301200986,
      "learning_rate": 0.18215788011253717,
      "loss": 0.3152,
      "num_input_tokens_seen": 32683024,
      "step": 17250
    },
    {
      "epoch": 2.8149114935965414,
      "grad_norm": 0.0016157825011759996,
      "learning_rate": 0.18210034236639935,
      "loss": 0.3423,
      "num_input_tokens_seen": 32692512,
      "step": 17255
    },
    {
      "epoch": 2.815727220817359,
      "grad_norm": 0.0017223075265064836,
      "learning_rate": 0.1820427996699853,
      "loss": 0.3268,
      "num_input_tokens_seen": 32702384,
      "step": 17260
    },
    {
      "epoch": 2.816542948038176,
      "grad_norm": 0.0008041823748499155,
      "learning_rate": 0.1819852520321689,
      "loss": 0.3426,
      "num_input_tokens_seen": 32712352,
      "step": 17265
    },
    {
      "epoch": 2.817358675258993,
      "grad_norm": 0.0014366005780175328,
      "learning_rate": 0.18192769946182466,
      "loss": 0.3341,
      "num_input_tokens_seen": 32722448,
      "step": 17270
    },
    {
      "epoch": 2.8181744024798108,
      "grad_norm": 0.0026530958712100983,
      "learning_rate": 0.18187014196782794,
      "loss": 0.3373,
      "num_input_tokens_seen": 32732016,
      "step": 17275
    },
    {
      "epoch": 2.8189901297006283,
      "grad_norm": 0.002821515081450343,
      "learning_rate": 0.18181257955905486,
      "loss": 0.3143,
      "num_input_tokens_seen": 32741904,
      "step": 17280
    },
    {
      "epoch": 2.8198058569214455,
      "grad_norm": 0.0030491033103317022,
      "learning_rate": 0.18175501224438217,
      "loss": 0.3084,
      "num_input_tokens_seen": 32751904,
      "step": 17285
    },
    {
      "epoch": 2.8206215841422626,
      "grad_norm": 0.0024277977645397186,
      "learning_rate": 0.18169744003268756,
      "loss": 0.3178,
      "num_input_tokens_seen": 32762096,
      "step": 17290
    },
    {
      "epoch": 2.82143731136308,
      "grad_norm": 0.0018088009674102068,
      "learning_rate": 0.18163986293284937,
      "loss": 0.3133,
      "num_input_tokens_seen": 32772272,
      "step": 17295
    },
    {
      "epoch": 2.8222530385838978,
      "grad_norm": 0.0016526132822036743,
      "learning_rate": 0.18158228095374673,
      "loss": 0.3373,
      "num_input_tokens_seen": 32781664,
      "step": 17300
    },
    {
      "epoch": 2.823068765804715,
      "grad_norm": 0.0015225186944007874,
      "learning_rate": 0.18152469410425945,
      "loss": 0.3314,
      "num_input_tokens_seen": 32792128,
      "step": 17305
    },
    {
      "epoch": 2.823884493025532,
      "grad_norm": 0.0017150212079286575,
      "learning_rate": 0.18146710239326813,
      "loss": 0.3154,
      "num_input_tokens_seen": 32801248,
      "step": 17310
    },
    {
      "epoch": 2.8247002202463496,
      "grad_norm": 0.002949811751022935,
      "learning_rate": 0.18140950582965423,
      "loss": 0.3109,
      "num_input_tokens_seen": 32810304,
      "step": 17315
    },
    {
      "epoch": 2.825515947467167,
      "grad_norm": 0.00212284573353827,
      "learning_rate": 0.1813519044222998,
      "loss": 0.3762,
      "num_input_tokens_seen": 32820496,
      "step": 17320
    },
    {
      "epoch": 2.8263316746879843,
      "grad_norm": 0.0010753689566627145,
      "learning_rate": 0.18129429818008772,
      "loss": 0.323,
      "num_input_tokens_seen": 32829968,
      "step": 17325
    },
    {
      "epoch": 2.8271474019088014,
      "grad_norm": 0.002349474234506488,
      "learning_rate": 0.18123668711190163,
      "loss": 0.3517,
      "num_input_tokens_seen": 32840432,
      "step": 17330
    },
    {
      "epoch": 2.827963129129619,
      "grad_norm": 0.003151237964630127,
      "learning_rate": 0.18117907122662583,
      "loss": 0.3187,
      "num_input_tokens_seen": 32850864,
      "step": 17335
    },
    {
      "epoch": 2.8287788563504366,
      "grad_norm": 0.000519485620316118,
      "learning_rate": 0.1811214505331454,
      "loss": 0.3036,
      "num_input_tokens_seen": 32859664,
      "step": 17340
    },
    {
      "epoch": 2.8295945835712537,
      "grad_norm": 0.0014448229921981692,
      "learning_rate": 0.1810638250403462,
      "loss": 0.3309,
      "num_input_tokens_seen": 32869008,
      "step": 17345
    },
    {
      "epoch": 2.8304103107920713,
      "grad_norm": 0.0021963375620543957,
      "learning_rate": 0.1810061947571148,
      "loss": 0.3058,
      "num_input_tokens_seen": 32878832,
      "step": 17350
    },
    {
      "epoch": 2.8312260380128884,
      "grad_norm": 0.0007642250275239348,
      "learning_rate": 0.1809485596923385,
      "loss": 0.3603,
      "num_input_tokens_seen": 32887280,
      "step": 17355
    },
    {
      "epoch": 2.832041765233706,
      "grad_norm": 0.0006760261603631079,
      "learning_rate": 0.18089091985490546,
      "loss": 0.262,
      "num_input_tokens_seen": 32896208,
      "step": 17360
    },
    {
      "epoch": 2.832857492454523,
      "grad_norm": 0.0025856904685497284,
      "learning_rate": 0.18083327525370432,
      "loss": 0.3649,
      "num_input_tokens_seen": 32905072,
      "step": 17365
    },
    {
      "epoch": 2.8336732196753407,
      "grad_norm": 0.0013519696658477187,
      "learning_rate": 0.18077562589762464,
      "loss": 0.2875,
      "num_input_tokens_seen": 32914912,
      "step": 17370
    },
    {
      "epoch": 2.834488946896158,
      "grad_norm": 0.0007719674613326788,
      "learning_rate": 0.1807179717955567,
      "loss": 0.2776,
      "num_input_tokens_seen": 32923136,
      "step": 17375
    },
    {
      "epoch": 2.8353046741169754,
      "grad_norm": 0.0020291416440159082,
      "learning_rate": 0.1806603129563915,
      "loss": 0.3995,
      "num_input_tokens_seen": 32932304,
      "step": 17380
    },
    {
      "epoch": 2.8361204013377925,
      "grad_norm": 0.0015471030492335558,
      "learning_rate": 0.1806026493890208,
      "loss": 0.3407,
      "num_input_tokens_seen": 32942016,
      "step": 17385
    },
    {
      "epoch": 2.83693612855861,
      "grad_norm": 0.002096870681270957,
      "learning_rate": 0.18054498110233688,
      "loss": 0.3285,
      "num_input_tokens_seen": 32951872,
      "step": 17390
    },
    {
      "epoch": 2.8377518557794272,
      "grad_norm": 0.0009404119336977601,
      "learning_rate": 0.1804873081052331,
      "loss": 0.359,
      "num_input_tokens_seen": 32962032,
      "step": 17395
    },
    {
      "epoch": 2.838567583000245,
      "grad_norm": 0.002330042887479067,
      "learning_rate": 0.18042963040660326,
      "loss": 0.3248,
      "num_input_tokens_seen": 32971552,
      "step": 17400
    },
    {
      "epoch": 2.838567583000245,
      "eval_loss": 0.32229700684547424,
      "eval_runtime": 152.7043,
      "eval_samples_per_second": 17.845,
      "eval_steps_per_second": 8.926,
      "num_input_tokens_seen": 32971552,
      "step": 17400
    },
    {
      "epoch": 2.839383310221062,
      "grad_norm": 0.0007003803621046245,
      "learning_rate": 0.180371948015342,
      "loss": 0.3697,
      "num_input_tokens_seen": 32981872,
      "step": 17405
    },
    {
      "epoch": 2.8401990374418795,
      "grad_norm": 0.0013251242926344275,
      "learning_rate": 0.18031426094034472,
      "loss": 0.3215,
      "num_input_tokens_seen": 32991136,
      "step": 17410
    },
    {
      "epoch": 2.8410147646626966,
      "grad_norm": 0.001845486811362207,
      "learning_rate": 0.18025656919050737,
      "loss": 0.3617,
      "num_input_tokens_seen": 33000384,
      "step": 17415
    },
    {
      "epoch": 2.841830491883514,
      "grad_norm": 0.0012184998486191034,
      "learning_rate": 0.18019887277472688,
      "loss": 0.3322,
      "num_input_tokens_seen": 33008912,
      "step": 17420
    },
    {
      "epoch": 2.8426462191043314,
      "grad_norm": 0.0030444622971117496,
      "learning_rate": 0.18014117170190067,
      "loss": 0.3367,
      "num_input_tokens_seen": 33018608,
      "step": 17425
    },
    {
      "epoch": 2.843461946325149,
      "grad_norm": 0.0012830070918425918,
      "learning_rate": 0.18008346598092703,
      "loss": 0.351,
      "num_input_tokens_seen": 33028464,
      "step": 17430
    },
    {
      "epoch": 2.8442776735459665,
      "grad_norm": 0.000886566995177418,
      "learning_rate": 0.18002575562070489,
      "loss": 0.3039,
      "num_input_tokens_seen": 33038752,
      "step": 17435
    },
    {
      "epoch": 2.8450934007667836,
      "grad_norm": 0.0011294559808447957,
      "learning_rate": 0.1799680406301339,
      "loss": 0.3204,
      "num_input_tokens_seen": 33048400,
      "step": 17440
    },
    {
      "epoch": 2.8459091279876008,
      "grad_norm": 0.002062657382339239,
      "learning_rate": 0.17991032101811447,
      "loss": 0.3606,
      "num_input_tokens_seen": 33057840,
      "step": 17445
    },
    {
      "epoch": 2.8467248552084183,
      "grad_norm": 0.001635522348806262,
      "learning_rate": 0.1798525967935476,
      "loss": 0.3065,
      "num_input_tokens_seen": 33067568,
      "step": 17450
    },
    {
      "epoch": 2.847540582429236,
      "grad_norm": 0.0005750130512751639,
      "learning_rate": 0.17979486796533517,
      "loss": 0.4045,
      "num_input_tokens_seen": 33077248,
      "step": 17455
    },
    {
      "epoch": 2.848356309650053,
      "grad_norm": 0.0046354616060853004,
      "learning_rate": 0.1797371345423797,
      "loss": 0.3578,
      "num_input_tokens_seen": 33087168,
      "step": 17460
    },
    {
      "epoch": 2.84917203687087,
      "grad_norm": 0.002580877160653472,
      "learning_rate": 0.17967939653358436,
      "loss": 0.3209,
      "num_input_tokens_seen": 33096704,
      "step": 17465
    },
    {
      "epoch": 2.8499877640916877,
      "grad_norm": 0.00296309613622725,
      "learning_rate": 0.17962165394785315,
      "loss": 0.3342,
      "num_input_tokens_seen": 33105456,
      "step": 17470
    },
    {
      "epoch": 2.8508034913125053,
      "grad_norm": 0.001300076604820788,
      "learning_rate": 0.17956390679409057,
      "loss": 0.3266,
      "num_input_tokens_seen": 33114752,
      "step": 17475
    },
    {
      "epoch": 2.8516192185333225,
      "grad_norm": 0.002533538267016411,
      "learning_rate": 0.1795061550812021,
      "loss": 0.3433,
      "num_input_tokens_seen": 33124144,
      "step": 17480
    },
    {
      "epoch": 2.8524349457541396,
      "grad_norm": 0.00136093539185822,
      "learning_rate": 0.1794483988180937,
      "loss": 0.3644,
      "num_input_tokens_seen": 33134336,
      "step": 17485
    },
    {
      "epoch": 2.853250672974957,
      "grad_norm": 0.001815672847442329,
      "learning_rate": 0.17939063801367214,
      "loss": 0.3187,
      "num_input_tokens_seen": 33143872,
      "step": 17490
    },
    {
      "epoch": 2.8540664001957747,
      "grad_norm": 0.0015507149510085583,
      "learning_rate": 0.17933287267684483,
      "loss": 0.3313,
      "num_input_tokens_seen": 33152784,
      "step": 17495
    },
    {
      "epoch": 2.854882127416592,
      "grad_norm": 0.0016589448787271976,
      "learning_rate": 0.17927510281651995,
      "loss": 0.2865,
      "num_input_tokens_seen": 33162432,
      "step": 17500
    },
    {
      "epoch": 2.855697854637409,
      "grad_norm": 0.0019703153520822525,
      "learning_rate": 0.17921732844160634,
      "loss": 0.2832,
      "num_input_tokens_seen": 33171984,
      "step": 17505
    },
    {
      "epoch": 2.8565135818582266,
      "grad_norm": 0.0014204414328560233,
      "learning_rate": 0.17915954956101351,
      "loss": 0.3464,
      "num_input_tokens_seen": 33180736,
      "step": 17510
    },
    {
      "epoch": 2.857329309079044,
      "grad_norm": 0.0029635748360306025,
      "learning_rate": 0.17910176618365165,
      "loss": 0.3699,
      "num_input_tokens_seen": 33189552,
      "step": 17515
    },
    {
      "epoch": 2.8581450362998613,
      "grad_norm": 0.002772424602881074,
      "learning_rate": 0.17904397831843177,
      "loss": 0.3357,
      "num_input_tokens_seen": 33200192,
      "step": 17520
    },
    {
      "epoch": 2.858960763520679,
      "grad_norm": 0.0018091966630890965,
      "learning_rate": 0.17898618597426547,
      "loss": 0.35,
      "num_input_tokens_seen": 33210304,
      "step": 17525
    },
    {
      "epoch": 2.859776490741496,
      "grad_norm": 0.004656167700886726,
      "learning_rate": 0.17892838916006495,
      "loss": 0.358,
      "num_input_tokens_seen": 33220400,
      "step": 17530
    },
    {
      "epoch": 2.8605922179623136,
      "grad_norm": 0.0038046101108193398,
      "learning_rate": 0.17887058788474333,
      "loss": 0.3315,
      "num_input_tokens_seen": 33229568,
      "step": 17535
    },
    {
      "epoch": 2.8614079451831307,
      "grad_norm": 0.002613077638670802,
      "learning_rate": 0.17881278215721427,
      "loss": 0.3099,
      "num_input_tokens_seen": 33240080,
      "step": 17540
    },
    {
      "epoch": 2.8622236724039483,
      "grad_norm": 0.0010955164907500148,
      "learning_rate": 0.1787549719863921,
      "loss": 0.2964,
      "num_input_tokens_seen": 33248864,
      "step": 17545
    },
    {
      "epoch": 2.8630393996247654,
      "grad_norm": 0.0013181528775021434,
      "learning_rate": 0.17869715738119188,
      "loss": 0.3245,
      "num_input_tokens_seen": 33258512,
      "step": 17550
    },
    {
      "epoch": 2.863855126845583,
      "grad_norm": 0.003251627553254366,
      "learning_rate": 0.17863933835052936,
      "loss": 0.3151,
      "num_input_tokens_seen": 33268240,
      "step": 17555
    },
    {
      "epoch": 2.8646708540664,
      "grad_norm": 0.0023538891691714525,
      "learning_rate": 0.17858151490332097,
      "loss": 0.3547,
      "num_input_tokens_seen": 33277280,
      "step": 17560
    },
    {
      "epoch": 2.8654865812872177,
      "grad_norm": 0.003114632796496153,
      "learning_rate": 0.17852368704848381,
      "loss": 0.3135,
      "num_input_tokens_seen": 33286880,
      "step": 17565
    },
    {
      "epoch": 2.866302308508035,
      "grad_norm": 0.001315200817771256,
      "learning_rate": 0.17846585479493565,
      "loss": 0.3236,
      "num_input_tokens_seen": 33297136,
      "step": 17570
    },
    {
      "epoch": 2.8671180357288524,
      "grad_norm": 0.0034644806291908026,
      "learning_rate": 0.178408018151595,
      "loss": 0.3553,
      "num_input_tokens_seen": 33306880,
      "step": 17575
    },
    {
      "epoch": 2.8679337629496695,
      "grad_norm": 0.0020303349010646343,
      "learning_rate": 0.17835017712738085,
      "loss": 0.2883,
      "num_input_tokens_seen": 33316752,
      "step": 17580
    },
    {
      "epoch": 2.868749490170487,
      "grad_norm": 0.002790703671053052,
      "learning_rate": 0.17829233173121323,
      "loss": 0.3138,
      "num_input_tokens_seen": 33326240,
      "step": 17585
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 0.0013568373396992683,
      "learning_rate": 0.17823448197201244,
      "loss": 0.3478,
      "num_input_tokens_seen": 33336320,
      "step": 17590
    },
    {
      "epoch": 2.870380944612122,
      "grad_norm": 0.0018745356937870383,
      "learning_rate": 0.1781766278586997,
      "loss": 0.3159,
      "num_input_tokens_seen": 33345584,
      "step": 17595
    },
    {
      "epoch": 2.871196671832939,
      "grad_norm": 0.0028549900744110346,
      "learning_rate": 0.1781187694001969,
      "loss": 0.3379,
      "num_input_tokens_seen": 33356064,
      "step": 17600
    },
    {
      "epoch": 2.871196671832939,
      "eval_loss": 0.32795971632003784,
      "eval_runtime": 152.5794,
      "eval_samples_per_second": 17.86,
      "eval_steps_per_second": 8.933,
      "num_input_tokens_seen": 33356064,
      "step": 17600
    },
    {
      "epoch": 2.8720123990537565,
      "grad_norm": 0.002563909161835909,
      "learning_rate": 0.1780609066054265,
      "loss": 0.3022,
      "num_input_tokens_seen": 33366688,
      "step": 17605
    },
    {
      "epoch": 2.8728281262745736,
      "grad_norm": 0.0021373811177909374,
      "learning_rate": 0.17800303948331164,
      "loss": 0.3381,
      "num_input_tokens_seen": 33376464,
      "step": 17610
    },
    {
      "epoch": 2.873643853495391,
      "grad_norm": 0.004898909013718367,
      "learning_rate": 0.1779451680427762,
      "loss": 0.3366,
      "num_input_tokens_seen": 33385872,
      "step": 17615
    },
    {
      "epoch": 2.8744595807162083,
      "grad_norm": 0.001451196731068194,
      "learning_rate": 0.17788729229274464,
      "loss": 0.3501,
      "num_input_tokens_seen": 33396432,
      "step": 17620
    },
    {
      "epoch": 2.875275307937026,
      "grad_norm": 0.0019931914284825325,
      "learning_rate": 0.17782941224214222,
      "loss": 0.3552,
      "num_input_tokens_seen": 33406208,
      "step": 17625
    },
    {
      "epoch": 2.8760910351578435,
      "grad_norm": 0.00121905782725662,
      "learning_rate": 0.17777152789989464,
      "loss": 0.3118,
      "num_input_tokens_seen": 33414944,
      "step": 17630
    },
    {
      "epoch": 2.8769067623786606,
      "grad_norm": 0.0015727704158052802,
      "learning_rate": 0.17771363927492845,
      "loss": 0.3261,
      "num_input_tokens_seen": 33424816,
      "step": 17635
    },
    {
      "epoch": 2.8777224895994777,
      "grad_norm": 0.0023076096549630165,
      "learning_rate": 0.17765574637617085,
      "loss": 0.3153,
      "num_input_tokens_seen": 33433520,
      "step": 17640
    },
    {
      "epoch": 2.8785382168202953,
      "grad_norm": 0.0020483890548348427,
      "learning_rate": 0.17759784921254962,
      "loss": 0.3491,
      "num_input_tokens_seen": 33442576,
      "step": 17645
    },
    {
      "epoch": 2.879353944041113,
      "grad_norm": 0.0013789021177217364,
      "learning_rate": 0.1775399477929932,
      "loss": 0.3365,
      "num_input_tokens_seen": 33450432,
      "step": 17650
    },
    {
      "epoch": 2.88016967126193,
      "grad_norm": 0.0018562617478892207,
      "learning_rate": 0.17748204212643076,
      "loss": 0.3638,
      "num_input_tokens_seen": 33459920,
      "step": 17655
    },
    {
      "epoch": 2.880985398482747,
      "grad_norm": 0.0023453575558960438,
      "learning_rate": 0.17742413222179204,
      "loss": 0.3375,
      "num_input_tokens_seen": 33468784,
      "step": 17660
    },
    {
      "epoch": 2.8818011257035647,
      "grad_norm": 0.0030664994847029448,
      "learning_rate": 0.17736621808800754,
      "loss": 0.3648,
      "num_input_tokens_seen": 33478080,
      "step": 17665
    },
    {
      "epoch": 2.8826168529243823,
      "grad_norm": 0.001413804362528026,
      "learning_rate": 0.17730829973400827,
      "loss": 0.334,
      "num_input_tokens_seen": 33488240,
      "step": 17670
    },
    {
      "epoch": 2.8834325801451994,
      "grad_norm": 0.001601252588443458,
      "learning_rate": 0.17725037716872602,
      "loss": 0.3328,
      "num_input_tokens_seen": 33497952,
      "step": 17675
    },
    {
      "epoch": 2.8842483073660166,
      "grad_norm": 0.0015335767529904842,
      "learning_rate": 0.17719245040109313,
      "loss": 0.3097,
      "num_input_tokens_seen": 33508048,
      "step": 17680
    },
    {
      "epoch": 2.885064034586834,
      "grad_norm": 0.0012416667304933071,
      "learning_rate": 0.17713451944004271,
      "loss": 0.3036,
      "num_input_tokens_seen": 33516464,
      "step": 17685
    },
    {
      "epoch": 2.8858797618076517,
      "grad_norm": 0.0020641398150473833,
      "learning_rate": 0.17707658429450843,
      "loss": 0.3021,
      "num_input_tokens_seen": 33526224,
      "step": 17690
    },
    {
      "epoch": 2.886695489028469,
      "grad_norm": 0.001037205453030765,
      "learning_rate": 0.1770186449734245,
      "loss": 0.2983,
      "num_input_tokens_seen": 33536672,
      "step": 17695
    },
    {
      "epoch": 2.887511216249286,
      "grad_norm": 0.0014395215548574924,
      "learning_rate": 0.17696070148572599,
      "loss": 0.37,
      "num_input_tokens_seen": 33545184,
      "step": 17700
    },
    {
      "epoch": 2.8883269434701035,
      "grad_norm": 0.0024748893920332193,
      "learning_rate": 0.17690275384034856,
      "loss": 0.3578,
      "num_input_tokens_seen": 33555392,
      "step": 17705
    },
    {
      "epoch": 2.889142670690921,
      "grad_norm": 0.001577913062646985,
      "learning_rate": 0.17684480204622835,
      "loss": 0.3429,
      "num_input_tokens_seen": 33565024,
      "step": 17710
    },
    {
      "epoch": 2.8899583979117383,
      "grad_norm": 0.0017113067442551255,
      "learning_rate": 0.1767868461123023,
      "loss": 0.3401,
      "num_input_tokens_seen": 33574192,
      "step": 17715
    },
    {
      "epoch": 2.890774125132556,
      "grad_norm": 0.0029441937804222107,
      "learning_rate": 0.176728886047508,
      "loss": 0.3486,
      "num_input_tokens_seen": 33583776,
      "step": 17720
    },
    {
      "epoch": 2.891589852353373,
      "grad_norm": 0.0012699777726083994,
      "learning_rate": 0.17667092186078362,
      "loss": 0.319,
      "num_input_tokens_seen": 33594032,
      "step": 17725
    },
    {
      "epoch": 2.8924055795741905,
      "grad_norm": 0.0013196252984926105,
      "learning_rate": 0.17661295356106785,
      "loss": 0.358,
      "num_input_tokens_seen": 33604960,
      "step": 17730
    },
    {
      "epoch": 2.8932213067950077,
      "grad_norm": 0.000766055949497968,
      "learning_rate": 0.1765549811573002,
      "loss": 0.2807,
      "num_input_tokens_seen": 33614160,
      "step": 17735
    },
    {
      "epoch": 2.8940370340158252,
      "grad_norm": 0.0021241875365376472,
      "learning_rate": 0.17649700465842078,
      "loss": 0.2405,
      "num_input_tokens_seen": 33623696,
      "step": 17740
    },
    {
      "epoch": 2.8948527612366424,
      "grad_norm": 0.0011597763514146209,
      "learning_rate": 0.17643902407337023,
      "loss": 0.3159,
      "num_input_tokens_seen": 33632672,
      "step": 17745
    },
    {
      "epoch": 2.89566848845746,
      "grad_norm": 0.0009780245600268245,
      "learning_rate": 0.17638103941108993,
      "loss": 0.395,
      "num_input_tokens_seen": 33642720,
      "step": 17750
    },
    {
      "epoch": 2.896484215678277,
      "grad_norm": 0.001064404146745801,
      "learning_rate": 0.1763230506805218,
      "loss": 0.3608,
      "num_input_tokens_seen": 33651936,
      "step": 17755
    },
    {
      "epoch": 2.8972999428990946,
      "grad_norm": 0.00078724161721766,
      "learning_rate": 0.1762650578906085,
      "loss": 0.3051,
      "num_input_tokens_seen": 33662288,
      "step": 17760
    },
    {
      "epoch": 2.898115670119912,
      "grad_norm": 0.002325265435501933,
      "learning_rate": 0.1762070610502932,
      "loss": 0.3315,
      "num_input_tokens_seen": 33670736,
      "step": 17765
    },
    {
      "epoch": 2.8989313973407294,
      "grad_norm": 0.0009193273144774139,
      "learning_rate": 0.17614906016851975,
      "loss": 0.299,
      "num_input_tokens_seen": 33680096,
      "step": 17770
    },
    {
      "epoch": 2.8997471245615465,
      "grad_norm": 0.003752124961465597,
      "learning_rate": 0.17609105525423258,
      "loss": 0.3543,
      "num_input_tokens_seen": 33690240,
      "step": 17775
    },
    {
      "epoch": 2.900562851782364,
      "grad_norm": 0.0017829489661380649,
      "learning_rate": 0.1760330463163768,
      "loss": 0.311,
      "num_input_tokens_seen": 33699680,
      "step": 17780
    },
    {
      "epoch": 2.901378579003181,
      "grad_norm": 0.0016310436185449362,
      "learning_rate": 0.17597503336389816,
      "loss": 0.3867,
      "num_input_tokens_seen": 33710656,
      "step": 17785
    },
    {
      "epoch": 2.9021943062239988,
      "grad_norm": 0.0014844348188489676,
      "learning_rate": 0.17591701640574298,
      "loss": 0.3789,
      "num_input_tokens_seen": 33720240,
      "step": 17790
    },
    {
      "epoch": 2.903010033444816,
      "grad_norm": 0.0013121009105816483,
      "learning_rate": 0.17585899545085815,
      "loss": 0.3659,
      "num_input_tokens_seen": 33730640,
      "step": 17795
    },
    {
      "epoch": 2.9038257606656335,
      "grad_norm": 0.0025931065902113914,
      "learning_rate": 0.17580097050819124,
      "loss": 0.3434,
      "num_input_tokens_seen": 33739984,
      "step": 17800
    },
    {
      "epoch": 2.9038257606656335,
      "eval_loss": 0.32441413402557373,
      "eval_runtime": 152.7164,
      "eval_samples_per_second": 17.844,
      "eval_steps_per_second": 8.925,
      "num_input_tokens_seen": 33739984,
      "step": 17800
    },
    {
      "epoch": 2.904641487886451,
      "grad_norm": 0.0007027728715911508,
      "learning_rate": 0.17574294158669046,
      "loss": 0.32,
      "num_input_tokens_seen": 33748704,
      "step": 17805
    },
    {
      "epoch": 2.905457215107268,
      "grad_norm": 0.0009873568778857589,
      "learning_rate": 0.17568490869530456,
      "loss": 0.3472,
      "num_input_tokens_seen": 33757984,
      "step": 17810
    },
    {
      "epoch": 2.9062729423280853,
      "grad_norm": 0.004158689174801111,
      "learning_rate": 0.17562687184298295,
      "loss": 0.3749,
      "num_input_tokens_seen": 33767808,
      "step": 17815
    },
    {
      "epoch": 2.907088669548903,
      "grad_norm": 0.0018632321152836084,
      "learning_rate": 0.1755688310386757,
      "loss": 0.3375,
      "num_input_tokens_seen": 33777296,
      "step": 17820
    },
    {
      "epoch": 2.9079043967697205,
      "grad_norm": 0.0006122076301835477,
      "learning_rate": 0.17551078629133335,
      "loss": 0.3496,
      "num_input_tokens_seen": 33786592,
      "step": 17825
    },
    {
      "epoch": 2.9087201239905376,
      "grad_norm": 0.001540584024041891,
      "learning_rate": 0.17545273760990718,
      "loss": 0.3477,
      "num_input_tokens_seen": 33796512,
      "step": 17830
    },
    {
      "epoch": 2.9095358512113547,
      "grad_norm": 0.0016807048814371228,
      "learning_rate": 0.17539468500334904,
      "loss": 0.3341,
      "num_input_tokens_seen": 33807280,
      "step": 17835
    },
    {
      "epoch": 2.9103515784321723,
      "grad_norm": 0.0021822878625243902,
      "learning_rate": 0.17533662848061132,
      "loss": 0.3452,
      "num_input_tokens_seen": 33816144,
      "step": 17840
    },
    {
      "epoch": 2.91116730565299,
      "grad_norm": 0.0016056066378951073,
      "learning_rate": 0.1752785680506471,
      "loss": 0.3576,
      "num_input_tokens_seen": 33824880,
      "step": 17845
    },
    {
      "epoch": 2.911983032873807,
      "grad_norm": 0.0006510275416076183,
      "learning_rate": 0.17522050372241,
      "loss": 0.2747,
      "num_input_tokens_seen": 33833920,
      "step": 17850
    },
    {
      "epoch": 2.912798760094624,
      "grad_norm": 0.001160494633950293,
      "learning_rate": 0.17516243550485425,
      "loss": 0.3471,
      "num_input_tokens_seen": 33843008,
      "step": 17855
    },
    {
      "epoch": 2.9136144873154417,
      "grad_norm": 0.0014821122167631984,
      "learning_rate": 0.17510436340693478,
      "loss": 0.3376,
      "num_input_tokens_seen": 33854464,
      "step": 17860
    },
    {
      "epoch": 2.9144302145362593,
      "grad_norm": 0.0011195610277354717,
      "learning_rate": 0.175046287437607,
      "loss": 0.3243,
      "num_input_tokens_seen": 33862800,
      "step": 17865
    },
    {
      "epoch": 2.9152459417570764,
      "grad_norm": 0.0018935349071398377,
      "learning_rate": 0.17498820760582695,
      "loss": 0.37,
      "num_input_tokens_seen": 33871536,
      "step": 17870
    },
    {
      "epoch": 2.9160616689778935,
      "grad_norm": 0.0022997786290943623,
      "learning_rate": 0.1749301239205512,
      "loss": 0.3222,
      "num_input_tokens_seen": 33879792,
      "step": 17875
    },
    {
      "epoch": 2.916877396198711,
      "grad_norm": 0.000572502554859966,
      "learning_rate": 0.1748720363907371,
      "loss": 0.3467,
      "num_input_tokens_seen": 33888960,
      "step": 17880
    },
    {
      "epoch": 2.9176931234195287,
      "grad_norm": 0.0009516754071228206,
      "learning_rate": 0.17481394502534242,
      "loss": 0.2994,
      "num_input_tokens_seen": 33898112,
      "step": 17885
    },
    {
      "epoch": 2.918508850640346,
      "grad_norm": 0.0021763977129012346,
      "learning_rate": 0.17475584983332562,
      "loss": 0.4179,
      "num_input_tokens_seen": 33907536,
      "step": 17890
    },
    {
      "epoch": 2.919324577861163,
      "grad_norm": 0.0003786629531532526,
      "learning_rate": 0.17469775082364558,
      "loss": 0.3034,
      "num_input_tokens_seen": 33916432,
      "step": 17895
    },
    {
      "epoch": 2.9201403050819805,
      "grad_norm": 0.0007732139783911407,
      "learning_rate": 0.17463964800526205,
      "loss": 0.3458,
      "num_input_tokens_seen": 33925696,
      "step": 17900
    },
    {
      "epoch": 2.920956032302798,
      "grad_norm": 0.0018326289718970656,
      "learning_rate": 0.17458154138713522,
      "loss": 0.3992,
      "num_input_tokens_seen": 33935632,
      "step": 17905
    },
    {
      "epoch": 2.9217717595236152,
      "grad_norm": 0.0012273945612832904,
      "learning_rate": 0.17452343097822576,
      "loss": 0.3253,
      "num_input_tokens_seen": 33945696,
      "step": 17910
    },
    {
      "epoch": 2.922587486744433,
      "grad_norm": 0.0008011913159862161,
      "learning_rate": 0.17446531678749497,
      "loss": 0.3491,
      "num_input_tokens_seen": 33954928,
      "step": 17915
    },
    {
      "epoch": 2.92340321396525,
      "grad_norm": 0.0020262873731553555,
      "learning_rate": 0.17440719882390496,
      "loss": 0.3239,
      "num_input_tokens_seen": 33965536,
      "step": 17920
    },
    {
      "epoch": 2.9242189411860675,
      "grad_norm": 0.0006451614899560809,
      "learning_rate": 0.17434907709641814,
      "loss": 0.3311,
      "num_input_tokens_seen": 33974336,
      "step": 17925
    },
    {
      "epoch": 2.9250346684068846,
      "grad_norm": 0.0007955402252264321,
      "learning_rate": 0.17429095161399769,
      "loss": 0.3538,
      "num_input_tokens_seen": 33983696,
      "step": 17930
    },
    {
      "epoch": 2.925850395627702,
      "grad_norm": 0.0016354183899238706,
      "learning_rate": 0.1742328223856072,
      "loss": 0.3249,
      "num_input_tokens_seen": 33993600,
      "step": 17935
    },
    {
      "epoch": 2.9266661228485193,
      "grad_norm": 0.0012282885145395994,
      "learning_rate": 0.174174689420211,
      "loss": 0.3628,
      "num_input_tokens_seen": 34003664,
      "step": 17940
    },
    {
      "epoch": 2.927481850069337,
      "grad_norm": 0.0010255496017634869,
      "learning_rate": 0.1741165527267739,
      "loss": 0.3341,
      "num_input_tokens_seen": 34013840,
      "step": 17945
    },
    {
      "epoch": 2.928297577290154,
      "grad_norm": 0.002674994757398963,
      "learning_rate": 0.17405841231426125,
      "loss": 0.3366,
      "num_input_tokens_seen": 34022576,
      "step": 17950
    },
    {
      "epoch": 2.9291133045109716,
      "grad_norm": 0.0012092981487512589,
      "learning_rate": 0.1740002681916391,
      "loss": 0.3624,
      "num_input_tokens_seen": 34032432,
      "step": 17955
    },
    {
      "epoch": 2.9299290317317888,
      "grad_norm": 0.0006246378179639578,
      "learning_rate": 0.17394212036787401,
      "loss": 0.2993,
      "num_input_tokens_seen": 34042512,
      "step": 17960
    },
    {
      "epoch": 2.9307447589526063,
      "grad_norm": 0.0025673960335552692,
      "learning_rate": 0.1738839688519331,
      "loss": 0.392,
      "num_input_tokens_seen": 34052848,
      "step": 17965
    },
    {
      "epoch": 2.9315604861734235,
      "grad_norm": 0.0019515410531312227,
      "learning_rate": 0.17382581365278402,
      "loss": 0.3314,
      "num_input_tokens_seen": 34062432,
      "step": 17970
    },
    {
      "epoch": 2.932376213394241,
      "grad_norm": 0.0012608828255906701,
      "learning_rate": 0.17376765477939507,
      "loss": 0.3195,
      "num_input_tokens_seen": 34071616,
      "step": 17975
    },
    {
      "epoch": 2.933191940615058,
      "grad_norm": 0.0010317175183445215,
      "learning_rate": 0.1737094922407351,
      "loss": 0.3671,
      "num_input_tokens_seen": 34081808,
      "step": 17980
    },
    {
      "epoch": 2.9340076678358757,
      "grad_norm": 0.0004034606390632689,
      "learning_rate": 0.1736513260457734,
      "loss": 0.3544,
      "num_input_tokens_seen": 34090704,
      "step": 17985
    },
    {
      "epoch": 2.934823395056693,
      "grad_norm": 0.0006562933558598161,
      "learning_rate": 0.17359315620348006,
      "loss": 0.3169,
      "num_input_tokens_seen": 34100624,
      "step": 17990
    },
    {
      "epoch": 2.9356391222775104,
      "grad_norm": 0.0013607111759483814,
      "learning_rate": 0.17353498272282547,
      "loss": 0.3107,
      "num_input_tokens_seen": 34110688,
      "step": 17995
    },
    {
      "epoch": 2.936454849498328,
      "grad_norm": 0.0012716397177428007,
      "learning_rate": 0.17347680561278087,
      "loss": 0.288,
      "num_input_tokens_seen": 34121824,
      "step": 18000
    },
    {
      "epoch": 2.936454849498328,
      "eval_loss": 0.32467493414878845,
      "eval_runtime": 152.7172,
      "eval_samples_per_second": 17.843,
      "eval_steps_per_second": 8.925,
      "num_input_tokens_seen": 34121824,
      "step": 18000
    },
    {
      "epoch": 2.937270576719145,
      "grad_norm": 0.0005352225853130221,
      "learning_rate": 0.1734186248823178,
      "loss": 0.3113,
      "num_input_tokens_seen": 34131984,
      "step": 18005
    },
    {
      "epoch": 2.9380863039399623,
      "grad_norm": 0.0011037569493055344,
      "learning_rate": 0.17336044054040844,
      "loss": 0.3381,
      "num_input_tokens_seen": 34141376,
      "step": 18010
    },
    {
      "epoch": 2.93890203116078,
      "grad_norm": 0.000886212510522455,
      "learning_rate": 0.1733022525960256,
      "loss": 0.2561,
      "num_input_tokens_seen": 34150864,
      "step": 18015
    },
    {
      "epoch": 2.9397177583815974,
      "grad_norm": 0.001448267255909741,
      "learning_rate": 0.1732440610581426,
      "loss": 0.3303,
      "num_input_tokens_seen": 34160640,
      "step": 18020
    },
    {
      "epoch": 2.9405334856024146,
      "grad_norm": 0.0036799018271267414,
      "learning_rate": 0.17318586593573326,
      "loss": 0.2865,
      "num_input_tokens_seen": 34170672,
      "step": 18025
    },
    {
      "epoch": 2.9413492128232317,
      "grad_norm": 0.0016223076963797212,
      "learning_rate": 0.17312766723777204,
      "loss": 0.2539,
      "num_input_tokens_seen": 34179936,
      "step": 18030
    },
    {
      "epoch": 2.9421649400440493,
      "grad_norm": 0.003400031942874193,
      "learning_rate": 0.1730694649732339,
      "loss": 0.3304,
      "num_input_tokens_seen": 34187808,
      "step": 18035
    },
    {
      "epoch": 2.942980667264867,
      "grad_norm": 0.003250806126743555,
      "learning_rate": 0.17301125915109428,
      "loss": 0.3252,
      "num_input_tokens_seen": 34197504,
      "step": 18040
    },
    {
      "epoch": 2.943796394485684,
      "grad_norm": 0.002418996999040246,
      "learning_rate": 0.17295304978032938,
      "loss": 0.3008,
      "num_input_tokens_seen": 34208048,
      "step": 18045
    },
    {
      "epoch": 2.944612121706501,
      "grad_norm": 0.0022268411703407764,
      "learning_rate": 0.17289483686991577,
      "loss": 0.3235,
      "num_input_tokens_seen": 34217616,
      "step": 18050
    },
    {
      "epoch": 2.9454278489273187,
      "grad_norm": 0.0012096383143216372,
      "learning_rate": 0.1728366204288306,
      "loss": 0.2954,
      "num_input_tokens_seen": 34227824,
      "step": 18055
    },
    {
      "epoch": 2.9462435761481363,
      "grad_norm": 0.0028247623704373837,
      "learning_rate": 0.17277840046605153,
      "loss": 0.3356,
      "num_input_tokens_seen": 34238016,
      "step": 18060
    },
    {
      "epoch": 2.9470593033689534,
      "grad_norm": 0.0038228288758546114,
      "learning_rate": 0.17272017699055686,
      "loss": 0.3232,
      "num_input_tokens_seen": 34248208,
      "step": 18065
    },
    {
      "epoch": 2.9478750305897705,
      "grad_norm": 0.0016572900349274278,
      "learning_rate": 0.17266195001132542,
      "loss": 0.3386,
      "num_input_tokens_seen": 34257472,
      "step": 18070
    },
    {
      "epoch": 2.948690757810588,
      "grad_norm": 0.0019563522655516863,
      "learning_rate": 0.17260371953733647,
      "loss": 0.3879,
      "num_input_tokens_seen": 34267344,
      "step": 18075
    },
    {
      "epoch": 2.9495064850314057,
      "grad_norm": 0.0016491764690726995,
      "learning_rate": 0.1725454855775699,
      "loss": 0.3023,
      "num_input_tokens_seen": 34276896,
      "step": 18080
    },
    {
      "epoch": 2.950322212252223,
      "grad_norm": 0.0027329709846526384,
      "learning_rate": 0.17248724814100616,
      "loss": 0.3115,
      "num_input_tokens_seen": 34284960,
      "step": 18085
    },
    {
      "epoch": 2.9511379394730404,
      "grad_norm": 0.0016763614257797599,
      "learning_rate": 0.17242900723662619,
      "loss": 0.2923,
      "num_input_tokens_seen": 34294352,
      "step": 18090
    },
    {
      "epoch": 2.9519536666938575,
      "grad_norm": 0.0008316459716297686,
      "learning_rate": 0.1723707628734114,
      "loss": 0.3524,
      "num_input_tokens_seen": 34301680,
      "step": 18095
    },
    {
      "epoch": 2.952769393914675,
      "grad_norm": 0.0006831497885286808,
      "learning_rate": 0.1723125150603438,
      "loss": 0.3043,
      "num_input_tokens_seen": 34311024,
      "step": 18100
    },
    {
      "epoch": 2.953585121135492,
      "grad_norm": 0.002378034172579646,
      "learning_rate": 0.1722542638064061,
      "loss": 0.2993,
      "num_input_tokens_seen": 34320224,
      "step": 18105
    },
    {
      "epoch": 2.95440084835631,
      "grad_norm": 0.0021778675727546215,
      "learning_rate": 0.17219600912058117,
      "loss": 0.3105,
      "num_input_tokens_seen": 34330112,
      "step": 18110
    },
    {
      "epoch": 2.955216575577127,
      "grad_norm": 0.002438077237457037,
      "learning_rate": 0.17213775101185272,
      "loss": 0.3409,
      "num_input_tokens_seen": 34339024,
      "step": 18115
    },
    {
      "epoch": 2.9560323027979445,
      "grad_norm": 0.006065982859581709,
      "learning_rate": 0.17207948948920485,
      "loss": 0.3548,
      "num_input_tokens_seen": 34348272,
      "step": 18120
    },
    {
      "epoch": 2.9568480300187616,
      "grad_norm": 0.002831718185916543,
      "learning_rate": 0.17202122456162228,
      "loss": 0.3528,
      "num_input_tokens_seen": 34358512,
      "step": 18125
    },
    {
      "epoch": 2.957663757239579,
      "grad_norm": 0.000628414039965719,
      "learning_rate": 0.17196295623809013,
      "loss": 0.3228,
      "num_input_tokens_seen": 34367184,
      "step": 18130
    },
    {
      "epoch": 2.9584794844603963,
      "grad_norm": 0.0018947422504425049,
      "learning_rate": 0.1719046845275941,
      "loss": 0.2987,
      "num_input_tokens_seen": 34377136,
      "step": 18135
    },
    {
      "epoch": 2.959295211681214,
      "grad_norm": 0.00254731229506433,
      "learning_rate": 0.17184640943912044,
      "loss": 0.2815,
      "num_input_tokens_seen": 34385568,
      "step": 18140
    },
    {
      "epoch": 2.960110938902031,
      "grad_norm": 0.005004290957003832,
      "learning_rate": 0.1717881309816559,
      "loss": 0.2637,
      "num_input_tokens_seen": 34395904,
      "step": 18145
    },
    {
      "epoch": 2.9609266661228486,
      "grad_norm": 0.001156780868768692,
      "learning_rate": 0.1717298491641878,
      "loss": 0.3301,
      "num_input_tokens_seen": 34406592,
      "step": 18150
    },
    {
      "epoch": 2.9617423933436657,
      "grad_norm": 0.001800581463612616,
      "learning_rate": 0.17167156399570385,
      "loss": 0.3378,
      "num_input_tokens_seen": 34416224,
      "step": 18155
    },
    {
      "epoch": 2.9625581205644833,
      "grad_norm": 0.003432905999943614,
      "learning_rate": 0.17161327548519242,
      "loss": 0.3521,
      "num_input_tokens_seen": 34425104,
      "step": 18160
    },
    {
      "epoch": 2.9633738477853004,
      "grad_norm": 0.0019311478827148676,
      "learning_rate": 0.1715549836416423,
      "loss": 0.3419,
      "num_input_tokens_seen": 34435040,
      "step": 18165
    },
    {
      "epoch": 2.964189575006118,
      "grad_norm": 0.0036842375993728638,
      "learning_rate": 0.17149668847404279,
      "loss": 0.3515,
      "num_input_tokens_seen": 34445680,
      "step": 18170
    },
    {
      "epoch": 2.965005302226935,
      "grad_norm": 0.001319125760346651,
      "learning_rate": 0.1714383899913838,
      "loss": 0.3153,
      "num_input_tokens_seen": 34455168,
      "step": 18175
    },
    {
      "epoch": 2.9658210294477527,
      "grad_norm": 0.0014106925809755921,
      "learning_rate": 0.17138008820265563,
      "loss": 0.3605,
      "num_input_tokens_seen": 34463904,
      "step": 18180
    },
    {
      "epoch": 2.96663675666857,
      "grad_norm": 0.002396860858425498,
      "learning_rate": 0.17132178311684917,
      "loss": 0.2912,
      "num_input_tokens_seen": 34473312,
      "step": 18185
    },
    {
      "epoch": 2.9674524838893874,
      "grad_norm": 0.0037482143379747868,
      "learning_rate": 0.1712634747429559,
      "loss": 0.3715,
      "num_input_tokens_seen": 34481680,
      "step": 18190
    },
    {
      "epoch": 2.968268211110205,
      "grad_norm": 0.0021422335412353277,
      "learning_rate": 0.17120516308996753,
      "loss": 0.3009,
      "num_input_tokens_seen": 34489472,
      "step": 18195
    },
    {
      "epoch": 2.969083938331022,
      "grad_norm": 0.0015206453390419483,
      "learning_rate": 0.17114684816687653,
      "loss": 0.3275,
      "num_input_tokens_seen": 34498368,
      "step": 18200
    },
    {
      "epoch": 2.969083938331022,
      "eval_loss": 0.34516653418540955,
      "eval_runtime": 152.6733,
      "eval_samples_per_second": 17.849,
      "eval_steps_per_second": 8.928,
      "num_input_tokens_seen": 34498368,
      "step": 18200
    },
    {
      "epoch": 2.9698996655518393,
      "grad_norm": 0.001866366364993155,
      "learning_rate": 0.17108852998267585,
      "loss": 0.3092,
      "num_input_tokens_seen": 34507008,
      "step": 18205
    },
    {
      "epoch": 2.970715392772657,
      "grad_norm": 0.0018314768094569445,
      "learning_rate": 0.17103020854635878,
      "loss": 0.3222,
      "num_input_tokens_seen": 34515840,
      "step": 18210
    },
    {
      "epoch": 2.9715311199934744,
      "grad_norm": 0.0023002526722848415,
      "learning_rate": 0.1709718838669193,
      "loss": 0.2992,
      "num_input_tokens_seen": 34525648,
      "step": 18215
    },
    {
      "epoch": 2.9723468472142915,
      "grad_norm": 0.0020182973239570856,
      "learning_rate": 0.17091355595335173,
      "loss": 0.3731,
      "num_input_tokens_seen": 34534416,
      "step": 18220
    },
    {
      "epoch": 2.9731625744351087,
      "grad_norm": 0.0012207115069031715,
      "learning_rate": 0.17085522481465107,
      "loss": 0.2223,
      "num_input_tokens_seen": 34543984,
      "step": 18225
    },
    {
      "epoch": 2.9739783016559262,
      "grad_norm": 0.0044617922976613045,
      "learning_rate": 0.17079689045981264,
      "loss": 0.4045,
      "num_input_tokens_seen": 34551984,
      "step": 18230
    },
    {
      "epoch": 2.974794028876744,
      "grad_norm": 0.0014065488940104842,
      "learning_rate": 0.17073855289783238,
      "loss": 0.458,
      "num_input_tokens_seen": 34560752,
      "step": 18235
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 0.0018817671807482839,
      "learning_rate": 0.1706802121377066,
      "loss": 0.3489,
      "num_input_tokens_seen": 34569472,
      "step": 18240
    },
    {
      "epoch": 2.976425483318378,
      "grad_norm": 0.0027209727559238672,
      "learning_rate": 0.17062186818843225,
      "loss": 0.3426,
      "num_input_tokens_seen": 34578528,
      "step": 18245
    },
    {
      "epoch": 2.9772412105391957,
      "grad_norm": 0.0026335793081671,
      "learning_rate": 0.17056352105900668,
      "loss": 0.393,
      "num_input_tokens_seen": 34588000,
      "step": 18250
    },
    {
      "epoch": 2.9780569377600132,
      "grad_norm": 0.0011886038118973374,
      "learning_rate": 0.17050517075842772,
      "loss": 0.3348,
      "num_input_tokens_seen": 34596096,
      "step": 18255
    },
    {
      "epoch": 2.9788726649808304,
      "grad_norm": 0.0009942364413291216,
      "learning_rate": 0.17044681729569375,
      "loss": 0.3302,
      "num_input_tokens_seen": 34605920,
      "step": 18260
    },
    {
      "epoch": 2.9796883922016475,
      "grad_norm": 0.00042872526682913303,
      "learning_rate": 0.17038846067980365,
      "loss": 0.2962,
      "num_input_tokens_seen": 34614752,
      "step": 18265
    },
    {
      "epoch": 2.980504119422465,
      "grad_norm": 0.0014028102159500122,
      "learning_rate": 0.17033010091975664,
      "loss": 0.3883,
      "num_input_tokens_seen": 34624112,
      "step": 18270
    },
    {
      "epoch": 2.9813198466432826,
      "grad_norm": 0.002521739574149251,
      "learning_rate": 0.17027173802455262,
      "loss": 0.3375,
      "num_input_tokens_seen": 34633456,
      "step": 18275
    },
    {
      "epoch": 2.9821355738640998,
      "grad_norm": 0.0011025656713172793,
      "learning_rate": 0.1702133720031918,
      "loss": 0.3639,
      "num_input_tokens_seen": 34642176,
      "step": 18280
    },
    {
      "epoch": 2.9829513010849174,
      "grad_norm": 0.0010226935846731067,
      "learning_rate": 0.17015500286467503,
      "loss": 0.4113,
      "num_input_tokens_seen": 34650976,
      "step": 18285
    },
    {
      "epoch": 2.9837670283057345,
      "grad_norm": 0.0015394104411825538,
      "learning_rate": 0.17009663061800354,
      "loss": 0.3776,
      "num_input_tokens_seen": 34660576,
      "step": 18290
    },
    {
      "epoch": 2.984582755526552,
      "grad_norm": 0.001565759303048253,
      "learning_rate": 0.17003825527217903,
      "loss": 0.3148,
      "num_input_tokens_seen": 34671744,
      "step": 18295
    },
    {
      "epoch": 2.985398482747369,
      "grad_norm": 0.0004731274675577879,
      "learning_rate": 0.16997987683620377,
      "loss": 0.3173,
      "num_input_tokens_seen": 34680544,
      "step": 18300
    },
    {
      "epoch": 2.9862142099681868,
      "grad_norm": 0.0016675004735589027,
      "learning_rate": 0.16992149531908043,
      "loss": 0.392,
      "num_input_tokens_seen": 34689040,
      "step": 18305
    },
    {
      "epoch": 2.987029937189004,
      "grad_norm": 0.0020203718449920416,
      "learning_rate": 0.16986311072981214,
      "loss": 0.3228,
      "num_input_tokens_seen": 34699568,
      "step": 18310
    },
    {
      "epoch": 2.9878456644098215,
      "grad_norm": 0.0022224995773285627,
      "learning_rate": 0.16980472307740255,
      "loss": 0.3575,
      "num_input_tokens_seen": 34709120,
      "step": 18315
    },
    {
      "epoch": 2.9886613916306386,
      "grad_norm": 0.0014950961340218782,
      "learning_rate": 0.1697463323708558,
      "loss": 0.3304,
      "num_input_tokens_seen": 34716496,
      "step": 18320
    },
    {
      "epoch": 2.989477118851456,
      "grad_norm": 0.0008960873237811029,
      "learning_rate": 0.16968793861917641,
      "loss": 0.3884,
      "num_input_tokens_seen": 34727248,
      "step": 18325
    },
    {
      "epoch": 2.9902928460722733,
      "grad_norm": 0.002073117531836033,
      "learning_rate": 0.16962954183136952,
      "loss": 0.3623,
      "num_input_tokens_seen": 34737760,
      "step": 18330
    },
    {
      "epoch": 2.991108573293091,
      "grad_norm": 0.0017819007625803351,
      "learning_rate": 0.16957114201644058,
      "loss": 0.3484,
      "num_input_tokens_seen": 34746800,
      "step": 18335
    },
    {
      "epoch": 2.991924300513908,
      "grad_norm": 0.0018241365905851126,
      "learning_rate": 0.16951273918339563,
      "loss": 0.3287,
      "num_input_tokens_seen": 34756624,
      "step": 18340
    },
    {
      "epoch": 2.9927400277347256,
      "grad_norm": 0.0028132018633186817,
      "learning_rate": 0.16945433334124105,
      "loss": 0.338,
      "num_input_tokens_seen": 34766224,
      "step": 18345
    },
    {
      "epoch": 2.9935557549555427,
      "grad_norm": 0.0013948263367637992,
      "learning_rate": 0.1693959244989838,
      "loss": 0.3361,
      "num_input_tokens_seen": 34775664,
      "step": 18350
    },
    {
      "epoch": 2.9943714821763603,
      "grad_norm": 0.0012147808447480202,
      "learning_rate": 0.16933751266563127,
      "loss": 0.3342,
      "num_input_tokens_seen": 34785440,
      "step": 18355
    },
    {
      "epoch": 2.9951872093971774,
      "grad_norm": 0.002549465512856841,
      "learning_rate": 0.16927909785019118,
      "loss": 0.3391,
      "num_input_tokens_seen": 34795040,
      "step": 18360
    },
    {
      "epoch": 2.996002936617995,
      "grad_norm": 0.0017326746601611376,
      "learning_rate": 0.169220680061672,
      "loss": 0.366,
      "num_input_tokens_seen": 34804080,
      "step": 18365
    },
    {
      "epoch": 2.9968186638388126,
      "grad_norm": 0.0016525943065062165,
      "learning_rate": 0.16916225930908244,
      "loss": 0.3304,
      "num_input_tokens_seen": 34813856,
      "step": 18370
    },
    {
      "epoch": 2.9976343910596297,
      "grad_norm": 0.002876714337617159,
      "learning_rate": 0.16910383560143163,
      "loss": 0.3665,
      "num_input_tokens_seen": 34823056,
      "step": 18375
    },
    {
      "epoch": 2.998450118280447,
      "grad_norm": 0.002334448741748929,
      "learning_rate": 0.16904540894772935,
      "loss": 0.3609,
      "num_input_tokens_seen": 34831376,
      "step": 18380
    },
    {
      "epoch": 2.9992658455012644,
      "grad_norm": 0.0026002204976975918,
      "learning_rate": 0.16898697935698562,
      "loss": 0.3336,
      "num_input_tokens_seen": 34839488,
      "step": 18385
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0008658687002025545,
      "learning_rate": 0.1689285468382111,
      "loss": 0.3453,
      "num_input_tokens_seen": 34846688,
      "step": 18390
    },
    {
      "epoch": 3.0008157272208176,
      "grad_norm": 0.0018540327437222004,
      "learning_rate": 0.16887011140041677,
      "loss": 0.3762,
      "num_input_tokens_seen": 34855408,
      "step": 18395
    },
    {
      "epoch": 3.0016314544416347,
      "grad_norm": 0.0008506396552547812,
      "learning_rate": 0.1688116730526141,
      "loss": 0.2901,
      "num_input_tokens_seen": 34866272,
      "step": 18400
    },
    {
      "epoch": 3.0016314544416347,
      "eval_loss": 0.324106901884079,
      "eval_runtime": 152.7111,
      "eval_samples_per_second": 17.844,
      "eval_steps_per_second": 8.925,
      "num_input_tokens_seen": 34866272,
      "step": 18400
    },
    {
      "epoch": 3.0024471816624523,
      "grad_norm": 0.0007993359467945993,
      "learning_rate": 0.1687532318038151,
      "loss": 0.3228,
      "num_input_tokens_seen": 34876304,
      "step": 18405
    },
    {
      "epoch": 3.0032629088832694,
      "grad_norm": 0.0020580808632075787,
      "learning_rate": 0.16869478766303206,
      "loss": 0.2987,
      "num_input_tokens_seen": 34884768,
      "step": 18410
    },
    {
      "epoch": 3.004078636104087,
      "grad_norm": 0.0019842004403471947,
      "learning_rate": 0.16863634063927788,
      "loss": 0.3191,
      "num_input_tokens_seen": 34895248,
      "step": 18415
    },
    {
      "epoch": 3.004894363324904,
      "grad_norm": 0.001393214799463749,
      "learning_rate": 0.16857789074156568,
      "loss": 0.3102,
      "num_input_tokens_seen": 34904032,
      "step": 18420
    },
    {
      "epoch": 3.0057100905457217,
      "grad_norm": 0.0038143021520227194,
      "learning_rate": 0.16851943797890928,
      "loss": 0.3283,
      "num_input_tokens_seen": 34914288,
      "step": 18425
    },
    {
      "epoch": 3.006525817766539,
      "grad_norm": 0.0023951551411300898,
      "learning_rate": 0.16846098236032284,
      "loss": 0.3697,
      "num_input_tokens_seen": 34923520,
      "step": 18430
    },
    {
      "epoch": 3.0073415449873564,
      "grad_norm": 0.0017278420273214579,
      "learning_rate": 0.16840252389482097,
      "loss": 0.3123,
      "num_input_tokens_seen": 34933392,
      "step": 18435
    },
    {
      "epoch": 3.0081572722081735,
      "grad_norm": 0.0011999171692878008,
      "learning_rate": 0.16834406259141857,
      "loss": 0.3312,
      "num_input_tokens_seen": 34943360,
      "step": 18440
    },
    {
      "epoch": 3.008972999428991,
      "grad_norm": 0.0021083473693579435,
      "learning_rate": 0.16828559845913124,
      "loss": 0.3069,
      "num_input_tokens_seen": 34953472,
      "step": 18445
    },
    {
      "epoch": 3.0097887266498082,
      "grad_norm": 0.0013190986355766654,
      "learning_rate": 0.16822713150697488,
      "loss": 0.3113,
      "num_input_tokens_seen": 34963536,
      "step": 18450
    },
    {
      "epoch": 3.010604453870626,
      "grad_norm": 0.0025933904107660055,
      "learning_rate": 0.16816866174396575,
      "loss": 0.377,
      "num_input_tokens_seen": 34974448,
      "step": 18455
    },
    {
      "epoch": 3.011420181091443,
      "grad_norm": 0.0015296614728868008,
      "learning_rate": 0.16811018917912057,
      "loss": 0.319,
      "num_input_tokens_seen": 34985008,
      "step": 18460
    },
    {
      "epoch": 3.0122359083122605,
      "grad_norm": 0.0011342562502250075,
      "learning_rate": 0.16805171382145673,
      "loss": 0.3463,
      "num_input_tokens_seen": 34994832,
      "step": 18465
    },
    {
      "epoch": 3.0130516355330776,
      "grad_norm": 0.0011174329556524754,
      "learning_rate": 0.16799323567999175,
      "loss": 0.2869,
      "num_input_tokens_seen": 35004544,
      "step": 18470
    },
    {
      "epoch": 3.013867362753895,
      "grad_norm": 0.0008318599429912865,
      "learning_rate": 0.16793475476374367,
      "loss": 0.348,
      "num_input_tokens_seen": 35013552,
      "step": 18475
    },
    {
      "epoch": 3.0146830899747123,
      "grad_norm": 0.0012439098209142685,
      "learning_rate": 0.1678762710817311,
      "loss": 0.3222,
      "num_input_tokens_seen": 35023312,
      "step": 18480
    },
    {
      "epoch": 3.01549881719553,
      "grad_norm": 0.0012906439369544387,
      "learning_rate": 0.1678177846429728,
      "loss": 0.3445,
      "num_input_tokens_seen": 35031568,
      "step": 18485
    },
    {
      "epoch": 3.016314544416347,
      "grad_norm": 0.0015586159424856305,
      "learning_rate": 0.16775929545648827,
      "loss": 0.3164,
      "num_input_tokens_seen": 35040784,
      "step": 18490
    },
    {
      "epoch": 3.0171302716371646,
      "grad_norm": 0.0010693592485040426,
      "learning_rate": 0.16770080353129715,
      "loss": 0.314,
      "num_input_tokens_seen": 35050592,
      "step": 18495
    },
    {
      "epoch": 3.0179459988579818,
      "grad_norm": 0.0009870724752545357,
      "learning_rate": 0.16764230887641968,
      "loss": 0.331,
      "num_input_tokens_seen": 35060784,
      "step": 18500
    },
    {
      "epoch": 3.0187617260787993,
      "grad_norm": 0.0018055211985483766,
      "learning_rate": 0.1675838115008765,
      "loss": 0.3303,
      "num_input_tokens_seen": 35070464,
      "step": 18505
    },
    {
      "epoch": 3.0195774532996165,
      "grad_norm": 0.001398032414726913,
      "learning_rate": 0.1675253114136886,
      "loss": 0.2989,
      "num_input_tokens_seen": 35081456,
      "step": 18510
    },
    {
      "epoch": 3.020393180520434,
      "grad_norm": 0.001449544564820826,
      "learning_rate": 0.16746680862387747,
      "loss": 0.4055,
      "num_input_tokens_seen": 35091680,
      "step": 18515
    },
    {
      "epoch": 3.021208907741251,
      "grad_norm": 0.0009575953590683639,
      "learning_rate": 0.16740830314046493,
      "loss": 0.2896,
      "num_input_tokens_seen": 35101216,
      "step": 18520
    },
    {
      "epoch": 3.0220246349620687,
      "grad_norm": 0.0013497694162651896,
      "learning_rate": 0.1673497949724733,
      "loss": 0.3118,
      "num_input_tokens_seen": 35111360,
      "step": 18525
    },
    {
      "epoch": 3.022840362182886,
      "grad_norm": 0.0020717706065624952,
      "learning_rate": 0.16729128412892522,
      "loss": 0.2993,
      "num_input_tokens_seen": 35121760,
      "step": 18530
    },
    {
      "epoch": 3.0236560894037035,
      "grad_norm": 0.002167154336348176,
      "learning_rate": 0.16723277061884384,
      "loss": 0.3204,
      "num_input_tokens_seen": 35131504,
      "step": 18535
    },
    {
      "epoch": 3.0244718166245206,
      "grad_norm": 0.0013438445748761296,
      "learning_rate": 0.16717425445125267,
      "loss": 0.3664,
      "num_input_tokens_seen": 35141472,
      "step": 18540
    },
    {
      "epoch": 3.025287543845338,
      "grad_norm": 0.0036672167479991913,
      "learning_rate": 0.16711573563517565,
      "loss": 0.3434,
      "num_input_tokens_seen": 35151824,
      "step": 18545
    },
    {
      "epoch": 3.0261032710661553,
      "grad_norm": 0.0013810987584292889,
      "learning_rate": 0.1670572141796371,
      "loss": 0.3505,
      "num_input_tokens_seen": 35160704,
      "step": 18550
    },
    {
      "epoch": 3.026918998286973,
      "grad_norm": 0.0022040398325771093,
      "learning_rate": 0.16699869009366175,
      "loss": 0.3802,
      "num_input_tokens_seen": 35170304,
      "step": 18555
    },
    {
      "epoch": 3.02773472550779,
      "grad_norm": 0.0016822107136249542,
      "learning_rate": 0.1669401633862748,
      "loss": 0.3319,
      "num_input_tokens_seen": 35178880,
      "step": 18560
    },
    {
      "epoch": 3.0285504527286076,
      "grad_norm": 0.0006469693616963923,
      "learning_rate": 0.16688163406650178,
      "loss": 0.3718,
      "num_input_tokens_seen": 35187104,
      "step": 18565
    },
    {
      "epoch": 3.0293661799494247,
      "grad_norm": 0.0029589771293103695,
      "learning_rate": 0.1668231021433686,
      "loss": 0.2529,
      "num_input_tokens_seen": 35198032,
      "step": 18570
    },
    {
      "epoch": 3.0301819071702423,
      "grad_norm": 0.0022336842957884073,
      "learning_rate": 0.1667645676259017,
      "loss": 0.302,
      "num_input_tokens_seen": 35208224,
      "step": 18575
    },
    {
      "epoch": 3.03099763439106,
      "grad_norm": 0.0015821823617443442,
      "learning_rate": 0.1667060305231277,
      "loss": 0.2723,
      "num_input_tokens_seen": 35218144,
      "step": 18580
    },
    {
      "epoch": 3.031813361611877,
      "grad_norm": 0.002015528967604041,
      "learning_rate": 0.16664749084407396,
      "loss": 0.4142,
      "num_input_tokens_seen": 35228720,
      "step": 18585
    },
    {
      "epoch": 3.0326290888326946,
      "grad_norm": 0.0012971735559403896,
      "learning_rate": 0.16658894859776788,
      "loss": 0.2284,
      "num_input_tokens_seen": 35238576,
      "step": 18590
    },
    {
      "epoch": 3.0334448160535117,
      "grad_norm": 0.0024163699708878994,
      "learning_rate": 0.16653040379323752,
      "loss": 0.3538,
      "num_input_tokens_seen": 35247552,
      "step": 18595
    },
    {
      "epoch": 3.0342605432743293,
      "grad_norm": 0.0015847784234210849,
      "learning_rate": 0.16647185643951107,
      "loss": 0.3238,
      "num_input_tokens_seen": 35258768,
      "step": 18600
    },
    {
      "epoch": 3.0342605432743293,
      "eval_loss": 0.3236458897590637,
      "eval_runtime": 152.7344,
      "eval_samples_per_second": 17.841,
      "eval_steps_per_second": 8.924,
      "num_input_tokens_seen": 35258768,
      "step": 18600
    },
    {
      "epoch": 3.0350762704951464,
      "grad_norm": 0.003399049863219261,
      "learning_rate": 0.1664133065456174,
      "loss": 0.3453,
      "num_input_tokens_seen": 35268160,
      "step": 18605
    },
    {
      "epoch": 3.035891997715964,
      "grad_norm": 0.0008368837297894061,
      "learning_rate": 0.1663547541205856,
      "loss": 0.3291,
      "num_input_tokens_seen": 35278176,
      "step": 18610
    },
    {
      "epoch": 3.036707724936781,
      "grad_norm": 0.0021032344084233046,
      "learning_rate": 0.16629619917344518,
      "loss": 0.3466,
      "num_input_tokens_seen": 35287856,
      "step": 18615
    },
    {
      "epoch": 3.0375234521575987,
      "grad_norm": 0.0016562244854867458,
      "learning_rate": 0.16623764171322605,
      "loss": 0.3411,
      "num_input_tokens_seen": 35298128,
      "step": 18620
    },
    {
      "epoch": 3.038339179378416,
      "grad_norm": 0.003695529652759433,
      "learning_rate": 0.1661790817489585,
      "loss": 0.3845,
      "num_input_tokens_seen": 35307312,
      "step": 18625
    },
    {
      "epoch": 3.0391549065992334,
      "grad_norm": 0.001319965347647667,
      "learning_rate": 0.16612051928967328,
      "loss": 0.3767,
      "num_input_tokens_seen": 35316608,
      "step": 18630
    },
    {
      "epoch": 3.0399706338200505,
      "grad_norm": 0.000957376032602042,
      "learning_rate": 0.16606195434440138,
      "loss": 0.2867,
      "num_input_tokens_seen": 35326512,
      "step": 18635
    },
    {
      "epoch": 3.040786361040868,
      "grad_norm": 0.0018656569300219417,
      "learning_rate": 0.16600338692217426,
      "loss": 0.3512,
      "num_input_tokens_seen": 35335504,
      "step": 18640
    },
    {
      "epoch": 3.041602088261685,
      "grad_norm": 0.0016726006288081408,
      "learning_rate": 0.16594481703202374,
      "loss": 0.3175,
      "num_input_tokens_seen": 35344800,
      "step": 18645
    },
    {
      "epoch": 3.042417815482503,
      "grad_norm": 0.001830763416364789,
      "learning_rate": 0.1658862446829821,
      "loss": 0.2828,
      "num_input_tokens_seen": 35355616,
      "step": 18650
    },
    {
      "epoch": 3.04323354270332,
      "grad_norm": 0.0018171131378039718,
      "learning_rate": 0.16582766988408187,
      "loss": 0.3112,
      "num_input_tokens_seen": 35364960,
      "step": 18655
    },
    {
      "epoch": 3.0440492699241375,
      "grad_norm": 0.0018408320611342788,
      "learning_rate": 0.16576909264435608,
      "loss": 0.3737,
      "num_input_tokens_seen": 35374544,
      "step": 18660
    },
    {
      "epoch": 3.0448649971449546,
      "grad_norm": 0.0003753071650862694,
      "learning_rate": 0.16571051297283798,
      "loss": 0.329,
      "num_input_tokens_seen": 35383392,
      "step": 18665
    },
    {
      "epoch": 3.045680724365772,
      "grad_norm": 0.0014108943287283182,
      "learning_rate": 0.16565193087856137,
      "loss": 0.2966,
      "num_input_tokens_seen": 35392320,
      "step": 18670
    },
    {
      "epoch": 3.0464964515865893,
      "grad_norm": 0.001323545235209167,
      "learning_rate": 0.16559334637056033,
      "loss": 0.2956,
      "num_input_tokens_seen": 35403248,
      "step": 18675
    },
    {
      "epoch": 3.047312178807407,
      "grad_norm": 0.0005194368422962725,
      "learning_rate": 0.16553475945786933,
      "loss": 0.3031,
      "num_input_tokens_seen": 35413248,
      "step": 18680
    },
    {
      "epoch": 3.048127906028224,
      "grad_norm": 0.000979637261480093,
      "learning_rate": 0.16547617014952318,
      "loss": 0.315,
      "num_input_tokens_seen": 35423936,
      "step": 18685
    },
    {
      "epoch": 3.0489436332490416,
      "grad_norm": 0.001467098481953144,
      "learning_rate": 0.1654175784545571,
      "loss": 0.3341,
      "num_input_tokens_seen": 35434448,
      "step": 18690
    },
    {
      "epoch": 3.0497593604698587,
      "grad_norm": 0.0017492425395175815,
      "learning_rate": 0.1653589843820067,
      "loss": 0.2694,
      "num_input_tokens_seen": 35443840,
      "step": 18695
    },
    {
      "epoch": 3.0505750876906763,
      "grad_norm": 0.0022854898124933243,
      "learning_rate": 0.1653003879409079,
      "loss": 0.3771,
      "num_input_tokens_seen": 35452672,
      "step": 18700
    },
    {
      "epoch": 3.0513908149114934,
      "grad_norm": 0.0022374792024493217,
      "learning_rate": 0.165241789140297,
      "loss": 0.3345,
      "num_input_tokens_seen": 35463072,
      "step": 18705
    },
    {
      "epoch": 3.052206542132311,
      "grad_norm": 0.0016370265511795878,
      "learning_rate": 0.16518318798921064,
      "loss": 0.4387,
      "num_input_tokens_seen": 35472640,
      "step": 18710
    },
    {
      "epoch": 3.053022269353128,
      "grad_norm": 0.0025077946484088898,
      "learning_rate": 0.16512458449668593,
      "loss": 0.3934,
      "num_input_tokens_seen": 35481744,
      "step": 18715
    },
    {
      "epoch": 3.0538379965739457,
      "grad_norm": 0.0014349103439599276,
      "learning_rate": 0.1650659786717602,
      "loss": 0.3221,
      "num_input_tokens_seen": 35492560,
      "step": 18720
    },
    {
      "epoch": 3.054653723794763,
      "grad_norm": 0.0014438353246077895,
      "learning_rate": 0.1650073705234712,
      "loss": 0.3142,
      "num_input_tokens_seen": 35500736,
      "step": 18725
    },
    {
      "epoch": 3.0554694510155804,
      "grad_norm": 0.0039938450790941715,
      "learning_rate": 0.16494876006085712,
      "loss": 0.3017,
      "num_input_tokens_seen": 35510448,
      "step": 18730
    },
    {
      "epoch": 3.0562851782363976,
      "grad_norm": 0.0025294283404946327,
      "learning_rate": 0.16489014729295634,
      "loss": 0.2907,
      "num_input_tokens_seen": 35519712,
      "step": 18735
    },
    {
      "epoch": 3.057100905457215,
      "grad_norm": 0.0016207058215513825,
      "learning_rate": 0.16483153222880775,
      "loss": 0.3392,
      "num_input_tokens_seen": 35529776,
      "step": 18740
    },
    {
      "epoch": 3.0579166326780323,
      "grad_norm": 0.001072427723556757,
      "learning_rate": 0.16477291487745052,
      "loss": 0.3226,
      "num_input_tokens_seen": 35539056,
      "step": 18745
    },
    {
      "epoch": 3.05873235989885,
      "grad_norm": 0.0017586206085979939,
      "learning_rate": 0.16471429524792416,
      "loss": 0.3119,
      "num_input_tokens_seen": 35548624,
      "step": 18750
    },
    {
      "epoch": 3.059548087119667,
      "grad_norm": 0.002421860583126545,
      "learning_rate": 0.16465567334926856,
      "loss": 0.3379,
      "num_input_tokens_seen": 35558416,
      "step": 18755
    },
    {
      "epoch": 3.0603638143404845,
      "grad_norm": 0.002172705251723528,
      "learning_rate": 0.16459704919052395,
      "loss": 0.3131,
      "num_input_tokens_seen": 35568512,
      "step": 18760
    },
    {
      "epoch": 3.0611795415613017,
      "grad_norm": 0.002387774409726262,
      "learning_rate": 0.16453842278073086,
      "loss": 0.3226,
      "num_input_tokens_seen": 35578016,
      "step": 18765
    },
    {
      "epoch": 3.0619952687821193,
      "grad_norm": 0.0008490285254083574,
      "learning_rate": 0.16447979412893038,
      "loss": 0.3815,
      "num_input_tokens_seen": 35586432,
      "step": 18770
    },
    {
      "epoch": 3.062810996002937,
      "grad_norm": 0.0019999812357127666,
      "learning_rate": 0.16442116324416367,
      "loss": 0.3133,
      "num_input_tokens_seen": 35596448,
      "step": 18775
    },
    {
      "epoch": 3.063626723223754,
      "grad_norm": 0.0036917647812515497,
      "learning_rate": 0.1643625301354723,
      "loss": 0.3256,
      "num_input_tokens_seen": 35605200,
      "step": 18780
    },
    {
      "epoch": 3.0644424504445715,
      "grad_norm": 0.0009630140848457813,
      "learning_rate": 0.16430389481189828,
      "loss": 0.3164,
      "num_input_tokens_seen": 35615184,
      "step": 18785
    },
    {
      "epoch": 3.0652581776653887,
      "grad_norm": 0.0014718524180352688,
      "learning_rate": 0.164245257282484,
      "loss": 0.3147,
      "num_input_tokens_seen": 35624224,
      "step": 18790
    },
    {
      "epoch": 3.0660739048862062,
      "grad_norm": 0.0008644790505059063,
      "learning_rate": 0.16418661755627195,
      "loss": 0.3575,
      "num_input_tokens_seen": 35634608,
      "step": 18795
    },
    {
      "epoch": 3.0668896321070234,
      "grad_norm": 0.0032676730770617723,
      "learning_rate": 0.16412797564230527,
      "loss": 0.3617,
      "num_input_tokens_seen": 35644416,
      "step": 18800
    },
    {
      "epoch": 3.0668896321070234,
      "eval_loss": 0.36542850732803345,
      "eval_runtime": 152.6805,
      "eval_samples_per_second": 17.848,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 35644416,
      "step": 18800
    },
    {
      "epoch": 3.067705359327841,
      "grad_norm": 0.0015356502262875438,
      "learning_rate": 0.16406933154962713,
      "loss": 0.2981,
      "num_input_tokens_seen": 35654160,
      "step": 18805
    },
    {
      "epoch": 3.068521086548658,
      "grad_norm": 0.0013868547976016998,
      "learning_rate": 0.16401068528728133,
      "loss": 0.3492,
      "num_input_tokens_seen": 35662480,
      "step": 18810
    },
    {
      "epoch": 3.0693368137694756,
      "grad_norm": 0.0008069119649007916,
      "learning_rate": 0.16395203686431173,
      "loss": 0.3982,
      "num_input_tokens_seen": 35671648,
      "step": 18815
    },
    {
      "epoch": 3.0701525409902928,
      "grad_norm": 0.000646220869384706,
      "learning_rate": 0.16389338628976277,
      "loss": 0.3351,
      "num_input_tokens_seen": 35681552,
      "step": 18820
    },
    {
      "epoch": 3.0709682682111104,
      "grad_norm": 0.0013283456210047007,
      "learning_rate": 0.163834733572679,
      "loss": 0.3721,
      "num_input_tokens_seen": 35691056,
      "step": 18825
    },
    {
      "epoch": 3.0717839954319275,
      "grad_norm": 0.0017521213740110397,
      "learning_rate": 0.16377607872210545,
      "loss": 0.3357,
      "num_input_tokens_seen": 35698768,
      "step": 18830
    },
    {
      "epoch": 3.072599722652745,
      "grad_norm": 0.0020325807854533195,
      "learning_rate": 0.16371742174708748,
      "loss": 0.35,
      "num_input_tokens_seen": 35707312,
      "step": 18835
    },
    {
      "epoch": 3.073415449873562,
      "grad_norm": 0.0015101138269528747,
      "learning_rate": 0.16365876265667065,
      "loss": 0.3007,
      "num_input_tokens_seen": 35716656,
      "step": 18840
    },
    {
      "epoch": 3.0742311770943798,
      "grad_norm": 0.0011976327514275908,
      "learning_rate": 0.163600101459901,
      "loss": 0.3511,
      "num_input_tokens_seen": 35727184,
      "step": 18845
    },
    {
      "epoch": 3.075046904315197,
      "grad_norm": 0.0006189662963151932,
      "learning_rate": 0.16354143816582484,
      "loss": 0.3526,
      "num_input_tokens_seen": 35736992,
      "step": 18850
    },
    {
      "epoch": 3.0758626315360145,
      "grad_norm": 0.0014373438898473978,
      "learning_rate": 0.1634827727834887,
      "loss": 0.3182,
      "num_input_tokens_seen": 35745776,
      "step": 18855
    },
    {
      "epoch": 3.0766783587568316,
      "grad_norm": 0.0015335147036239505,
      "learning_rate": 0.16342410532193954,
      "loss": 0.3413,
      "num_input_tokens_seen": 35754288,
      "step": 18860
    },
    {
      "epoch": 3.077494085977649,
      "grad_norm": 0.0019531401339918375,
      "learning_rate": 0.16336543579022464,
      "loss": 0.2927,
      "num_input_tokens_seen": 35763392,
      "step": 18865
    },
    {
      "epoch": 3.0783098131984663,
      "grad_norm": 0.002979109063744545,
      "learning_rate": 0.16330676419739157,
      "loss": 0.3118,
      "num_input_tokens_seen": 35773104,
      "step": 18870
    },
    {
      "epoch": 3.079125540419284,
      "grad_norm": 0.0013644574210047722,
      "learning_rate": 0.1632480905524883,
      "loss": 0.2698,
      "num_input_tokens_seen": 35783104,
      "step": 18875
    },
    {
      "epoch": 3.079941267640101,
      "grad_norm": 0.0012842557625845075,
      "learning_rate": 0.16318941486456293,
      "loss": 0.2737,
      "num_input_tokens_seen": 35791232,
      "step": 18880
    },
    {
      "epoch": 3.0807569948609186,
      "grad_norm": 0.00579601526260376,
      "learning_rate": 0.16313073714266405,
      "loss": 0.2478,
      "num_input_tokens_seen": 35800400,
      "step": 18885
    },
    {
      "epoch": 3.0815727220817357,
      "grad_norm": 0.003934904932975769,
      "learning_rate": 0.16307205739584052,
      "loss": 0.3355,
      "num_input_tokens_seen": 35809920,
      "step": 18890
    },
    {
      "epoch": 3.0823884493025533,
      "grad_norm": 0.0034516751766204834,
      "learning_rate": 0.16301337563314144,
      "loss": 0.3758,
      "num_input_tokens_seen": 35817984,
      "step": 18895
    },
    {
      "epoch": 3.0832041765233704,
      "grad_norm": 0.0028009223751723766,
      "learning_rate": 0.1629546918636163,
      "loss": 0.3468,
      "num_input_tokens_seen": 35828672,
      "step": 18900
    },
    {
      "epoch": 3.084019903744188,
      "grad_norm": 0.0027713123708963394,
      "learning_rate": 0.16289600609631485,
      "loss": 0.3198,
      "num_input_tokens_seen": 35838208,
      "step": 18905
    },
    {
      "epoch": 3.084835630965005,
      "grad_norm": 0.003865537466481328,
      "learning_rate": 0.16283731834028722,
      "loss": 0.3581,
      "num_input_tokens_seen": 35847504,
      "step": 18910
    },
    {
      "epoch": 3.0856513581858227,
      "grad_norm": 0.001929333433508873,
      "learning_rate": 0.16277862860458378,
      "loss": 0.3109,
      "num_input_tokens_seen": 35857136,
      "step": 18915
    },
    {
      "epoch": 3.08646708540664,
      "grad_norm": 0.003418642096221447,
      "learning_rate": 0.16271993689825526,
      "loss": 0.3617,
      "num_input_tokens_seen": 35866976,
      "step": 18920
    },
    {
      "epoch": 3.0872828126274574,
      "grad_norm": 0.0014102632412686944,
      "learning_rate": 0.1626612432303526,
      "loss": 0.3421,
      "num_input_tokens_seen": 35875744,
      "step": 18925
    },
    {
      "epoch": 3.0880985398482745,
      "grad_norm": 0.004196712281554937,
      "learning_rate": 0.1626025476099271,
      "loss": 0.3069,
      "num_input_tokens_seen": 35884480,
      "step": 18930
    },
    {
      "epoch": 3.088914267069092,
      "grad_norm": 0.0024696490727365017,
      "learning_rate": 0.1625438500460304,
      "loss": 0.3445,
      "num_input_tokens_seen": 35893408,
      "step": 18935
    },
    {
      "epoch": 3.0897299942899092,
      "grad_norm": 0.0016843982739374042,
      "learning_rate": 0.16248515054771442,
      "loss": 0.346,
      "num_input_tokens_seen": 35903488,
      "step": 18940
    },
    {
      "epoch": 3.090545721510727,
      "grad_norm": 0.0017642630264163017,
      "learning_rate": 0.16242644912403123,
      "loss": 0.3398,
      "num_input_tokens_seen": 35912160,
      "step": 18945
    },
    {
      "epoch": 3.0913614487315444,
      "grad_norm": 0.0011478071101009846,
      "learning_rate": 0.1623677457840335,
      "loss": 0.3064,
      "num_input_tokens_seen": 35921760,
      "step": 18950
    },
    {
      "epoch": 3.0921771759523615,
      "grad_norm": 0.003334126668050885,
      "learning_rate": 0.16230904053677397,
      "loss": 0.3257,
      "num_input_tokens_seen": 35932208,
      "step": 18955
    },
    {
      "epoch": 3.092992903173179,
      "grad_norm": 0.0015928401844576001,
      "learning_rate": 0.16225033339130568,
      "loss": 0.4052,
      "num_input_tokens_seen": 35940800,
      "step": 18960
    },
    {
      "epoch": 3.0938086303939962,
      "grad_norm": 0.0019077177857980132,
      "learning_rate": 0.16219162435668197,
      "loss": 0.2685,
      "num_input_tokens_seen": 35950224,
      "step": 18965
    },
    {
      "epoch": 3.094624357614814,
      "grad_norm": 0.0012568443780764937,
      "learning_rate": 0.16213291344195666,
      "loss": 0.2731,
      "num_input_tokens_seen": 35959248,
      "step": 18970
    },
    {
      "epoch": 3.095440084835631,
      "grad_norm": 0.0017527384916320443,
      "learning_rate": 0.16207420065618358,
      "loss": 0.4589,
      "num_input_tokens_seen": 35968944,
      "step": 18975
    },
    {
      "epoch": 3.0962558120564485,
      "grad_norm": 0.0032362693455070257,
      "learning_rate": 0.16201548600841706,
      "loss": 0.4059,
      "num_input_tokens_seen": 35979520,
      "step": 18980
    },
    {
      "epoch": 3.0970715392772656,
      "grad_norm": 0.0020433722529560328,
      "learning_rate": 0.16195676950771154,
      "loss": 0.3568,
      "num_input_tokens_seen": 35989056,
      "step": 18985
    },
    {
      "epoch": 3.097887266498083,
      "grad_norm": 0.0009133124258369207,
      "learning_rate": 0.16189805116312198,
      "loss": 0.3158,
      "num_input_tokens_seen": 35998480,
      "step": 18990
    },
    {
      "epoch": 3.0987029937189003,
      "grad_norm": 0.0014375337632372975,
      "learning_rate": 0.16183933098370337,
      "loss": 0.3445,
      "num_input_tokens_seen": 36008512,
      "step": 18995
    },
    {
      "epoch": 3.099518720939718,
      "grad_norm": 0.0011789847631007433,
      "learning_rate": 0.16178060897851115,
      "loss": 0.3463,
      "num_input_tokens_seen": 36017808,
      "step": 19000
    },
    {
      "epoch": 3.099518720939718,
      "eval_loss": 0.3730049133300781,
      "eval_runtime": 152.6627,
      "eval_samples_per_second": 17.85,
      "eval_steps_per_second": 8.928,
      "num_input_tokens_seen": 36017808,
      "step": 19000
    },
    {
      "epoch": 3.100334448160535,
      "grad_norm": 0.0022492073476314545,
      "learning_rate": 0.16172188515660096,
      "loss": 0.3275,
      "num_input_tokens_seen": 36027632,
      "step": 19005
    },
    {
      "epoch": 3.1011501753813526,
      "grad_norm": 0.0009211882716044784,
      "learning_rate": 0.16166315952702878,
      "loss": 0.3923,
      "num_input_tokens_seen": 36037088,
      "step": 19010
    },
    {
      "epoch": 3.1019659026021698,
      "grad_norm": 0.0006589222466573119,
      "learning_rate": 0.16160443209885084,
      "loss": 0.35,
      "num_input_tokens_seen": 36046320,
      "step": 19015
    },
    {
      "epoch": 3.1027816298229873,
      "grad_norm": 0.00044214745867066085,
      "learning_rate": 0.16154570288112363,
      "loss": 0.3,
      "num_input_tokens_seen": 36055616,
      "step": 19020
    },
    {
      "epoch": 3.1035973570438045,
      "grad_norm": 0.001610419130884111,
      "learning_rate": 0.16148697188290395,
      "loss": 0.2787,
      "num_input_tokens_seen": 36064864,
      "step": 19025
    },
    {
      "epoch": 3.104413084264622,
      "grad_norm": 0.0024368574377149343,
      "learning_rate": 0.16142823911324888,
      "loss": 0.2935,
      "num_input_tokens_seen": 36075296,
      "step": 19030
    },
    {
      "epoch": 3.105228811485439,
      "grad_norm": 0.0009993365965783596,
      "learning_rate": 0.16136950458121568,
      "loss": 0.3103,
      "num_input_tokens_seen": 36084176,
      "step": 19035
    },
    {
      "epoch": 3.1060445387062567,
      "grad_norm": 0.0027353859040886164,
      "learning_rate": 0.16131076829586205,
      "loss": 0.3408,
      "num_input_tokens_seen": 36095712,
      "step": 19040
    },
    {
      "epoch": 3.106860265927074,
      "grad_norm": 0.0026365413796156645,
      "learning_rate": 0.1612520302662457,
      "loss": 0.3174,
      "num_input_tokens_seen": 36105056,
      "step": 19045
    },
    {
      "epoch": 3.1076759931478914,
      "grad_norm": 0.000618254707660526,
      "learning_rate": 0.16119329050142497,
      "loss": 0.2818,
      "num_input_tokens_seen": 36114688,
      "step": 19050
    },
    {
      "epoch": 3.1084917203687086,
      "grad_norm": 0.0025011415127664804,
      "learning_rate": 0.16113454901045818,
      "loss": 0.3431,
      "num_input_tokens_seen": 36122784,
      "step": 19055
    },
    {
      "epoch": 3.109307447589526,
      "grad_norm": 0.002653837436810136,
      "learning_rate": 0.16107580580240397,
      "loss": 0.3686,
      "num_input_tokens_seen": 36131728,
      "step": 19060
    },
    {
      "epoch": 3.1101231748103433,
      "grad_norm": 0.0020286317449063063,
      "learning_rate": 0.16101706088632134,
      "loss": 0.3934,
      "num_input_tokens_seen": 36141728,
      "step": 19065
    },
    {
      "epoch": 3.110938902031161,
      "grad_norm": 0.0005701527115888894,
      "learning_rate": 0.16095831427126947,
      "loss": 0.3207,
      "num_input_tokens_seen": 36149680,
      "step": 19070
    },
    {
      "epoch": 3.111754629251978,
      "grad_norm": 0.0010168205481022596,
      "learning_rate": 0.16089956596630783,
      "loss": 0.2665,
      "num_input_tokens_seen": 36160112,
      "step": 19075
    },
    {
      "epoch": 3.1125703564727956,
      "grad_norm": 0.0008380547515116632,
      "learning_rate": 0.16084081598049618,
      "loss": 0.41,
      "num_input_tokens_seen": 36169056,
      "step": 19080
    },
    {
      "epoch": 3.1133860836936127,
      "grad_norm": 0.0007067843107506633,
      "learning_rate": 0.1607820643228944,
      "loss": 0.3586,
      "num_input_tokens_seen": 36177408,
      "step": 19085
    },
    {
      "epoch": 3.1142018109144303,
      "grad_norm": 0.0011998225236311555,
      "learning_rate": 0.16072331100256285,
      "loss": 0.2555,
      "num_input_tokens_seen": 36187120,
      "step": 19090
    },
    {
      "epoch": 3.1150175381352474,
      "grad_norm": 0.0033393134362995625,
      "learning_rate": 0.16066455602856197,
      "loss": 0.3466,
      "num_input_tokens_seen": 36198000,
      "step": 19095
    },
    {
      "epoch": 3.115833265356065,
      "grad_norm": 0.0019837250001728535,
      "learning_rate": 0.16060579940995257,
      "loss": 0.2965,
      "num_input_tokens_seen": 36206128,
      "step": 19100
    },
    {
      "epoch": 3.116648992576882,
      "grad_norm": 0.0021326299756765366,
      "learning_rate": 0.16054704115579557,
      "loss": 0.307,
      "num_input_tokens_seen": 36215776,
      "step": 19105
    },
    {
      "epoch": 3.1174647197976997,
      "grad_norm": 0.001543218269944191,
      "learning_rate": 0.1604882812751523,
      "loss": 0.3042,
      "num_input_tokens_seen": 36224272,
      "step": 19110
    },
    {
      "epoch": 3.118280447018517,
      "grad_norm": 0.0027090744115412235,
      "learning_rate": 0.16042951977708425,
      "loss": 0.3237,
      "num_input_tokens_seen": 36233760,
      "step": 19115
    },
    {
      "epoch": 3.1190961742393344,
      "grad_norm": 0.002379477024078369,
      "learning_rate": 0.16037075667065318,
      "loss": 0.2852,
      "num_input_tokens_seen": 36243584,
      "step": 19120
    },
    {
      "epoch": 3.1199119014601515,
      "grad_norm": 0.002495770575478673,
      "learning_rate": 0.1603119919649211,
      "loss": 0.3223,
      "num_input_tokens_seen": 36253360,
      "step": 19125
    },
    {
      "epoch": 3.120727628680969,
      "grad_norm": 0.0010735667310655117,
      "learning_rate": 0.16025322566895028,
      "loss": 0.2923,
      "num_input_tokens_seen": 36263040,
      "step": 19130
    },
    {
      "epoch": 3.121543355901786,
      "grad_norm": 0.0014754014555364847,
      "learning_rate": 0.16019445779180322,
      "loss": 0.2881,
      "num_input_tokens_seen": 36272624,
      "step": 19135
    },
    {
      "epoch": 3.122359083122604,
      "grad_norm": 0.002917239209637046,
      "learning_rate": 0.16013568834254271,
      "loss": 0.3181,
      "num_input_tokens_seen": 36281952,
      "step": 19140
    },
    {
      "epoch": 3.1231748103434214,
      "grad_norm": 0.001258554169908166,
      "learning_rate": 0.1600769173302316,
      "loss": 0.3768,
      "num_input_tokens_seen": 36291808,
      "step": 19145
    },
    {
      "epoch": 3.1239905375642385,
      "grad_norm": 0.002525918884202838,
      "learning_rate": 0.16001814476393322,
      "loss": 0.4053,
      "num_input_tokens_seen": 36300720,
      "step": 19150
    },
    {
      "epoch": 3.124806264785056,
      "grad_norm": 0.0017782417125999928,
      "learning_rate": 0.15995937065271104,
      "loss": 0.3526,
      "num_input_tokens_seen": 36309552,
      "step": 19155
    },
    {
      "epoch": 3.125621992005873,
      "grad_norm": 0.0009684105752967298,
      "learning_rate": 0.15990059500562873,
      "loss": 0.3214,
      "num_input_tokens_seen": 36319392,
      "step": 19160
    },
    {
      "epoch": 3.126437719226691,
      "grad_norm": 0.0007465930539183319,
      "learning_rate": 0.15984181783175025,
      "loss": 0.3051,
      "num_input_tokens_seen": 36327856,
      "step": 19165
    },
    {
      "epoch": 3.127253446447508,
      "grad_norm": 0.0008711701957508922,
      "learning_rate": 0.1597830391401398,
      "loss": 0.3202,
      "num_input_tokens_seen": 36337184,
      "step": 19170
    },
    {
      "epoch": 3.1280691736683255,
      "grad_norm": 0.0029066908173263073,
      "learning_rate": 0.15972425893986178,
      "loss": 0.3541,
      "num_input_tokens_seen": 36347600,
      "step": 19175
    },
    {
      "epoch": 3.1288849008891426,
      "grad_norm": 0.0018956611165776849,
      "learning_rate": 0.15966547723998084,
      "loss": 0.3246,
      "num_input_tokens_seen": 36356144,
      "step": 19180
    },
    {
      "epoch": 3.12970062810996,
      "grad_norm": 0.0017847740091383457,
      "learning_rate": 0.15960669404956176,
      "loss": 0.392,
      "num_input_tokens_seen": 36366240,
      "step": 19185
    },
    {
      "epoch": 3.1305163553307773,
      "grad_norm": 0.0014551138738170266,
      "learning_rate": 0.1595479093776698,
      "loss": 0.312,
      "num_input_tokens_seen": 36375360,
      "step": 19190
    },
    {
      "epoch": 3.131332082551595,
      "grad_norm": 0.002131106797605753,
      "learning_rate": 0.15948912323337022,
      "loss": 0.3172,
      "num_input_tokens_seen": 36385728,
      "step": 19195
    },
    {
      "epoch": 3.132147809772412,
      "grad_norm": 0.0005412581376731396,
      "learning_rate": 0.1594303356257286,
      "loss": 0.3218,
      "num_input_tokens_seen": 36393536,
      "step": 19200
    },
    {
      "epoch": 3.132147809772412,
      "eval_loss": 0.3242805600166321,
      "eval_runtime": 152.6193,
      "eval_samples_per_second": 17.855,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 36393536,
      "step": 19200
    },
    {
      "epoch": 3.1329635369932296,
      "grad_norm": 0.0007923961384221911,
      "learning_rate": 0.15937154656381072,
      "loss": 0.3455,
      "num_input_tokens_seen": 36402784,
      "step": 19205
    },
    {
      "epoch": 3.1337792642140467,
      "grad_norm": 0.00313843647018075,
      "learning_rate": 0.15931275605668258,
      "loss": 0.3563,
      "num_input_tokens_seen": 36411568,
      "step": 19210
    },
    {
      "epoch": 3.1345949914348643,
      "grad_norm": 0.0012007768964394927,
      "learning_rate": 0.1592539641134104,
      "loss": 0.3294,
      "num_input_tokens_seen": 36420160,
      "step": 19215
    },
    {
      "epoch": 3.1354107186556814,
      "grad_norm": 0.0016763736493885517,
      "learning_rate": 0.1591951707430607,
      "loss": 0.2981,
      "num_input_tokens_seen": 36429040,
      "step": 19220
    },
    {
      "epoch": 3.136226445876499,
      "grad_norm": 0.001482425956055522,
      "learning_rate": 0.15913637595470007,
      "loss": 0.3202,
      "num_input_tokens_seen": 36438112,
      "step": 19225
    },
    {
      "epoch": 3.137042173097316,
      "grad_norm": 0.0015556949656456709,
      "learning_rate": 0.15907757975739548,
      "loss": 0.3916,
      "num_input_tokens_seen": 36448080,
      "step": 19230
    },
    {
      "epoch": 3.1378579003181337,
      "grad_norm": 0.0011671254178509116,
      "learning_rate": 0.159018782160214,
      "loss": 0.3223,
      "num_input_tokens_seen": 36457568,
      "step": 19235
    },
    {
      "epoch": 3.138673627538951,
      "grad_norm": 0.0004186949518043548,
      "learning_rate": 0.158959983172223,
      "loss": 0.3279,
      "num_input_tokens_seen": 36466608,
      "step": 19240
    },
    {
      "epoch": 3.1394893547597684,
      "grad_norm": 0.0015938101569190621,
      "learning_rate": 0.15890118280249,
      "loss": 0.325,
      "num_input_tokens_seen": 36478256,
      "step": 19245
    },
    {
      "epoch": 3.1403050819805856,
      "grad_norm": 0.0013470177073031664,
      "learning_rate": 0.15884238106008275,
      "loss": 0.3709,
      "num_input_tokens_seen": 36488384,
      "step": 19250
    },
    {
      "epoch": 3.141120809201403,
      "grad_norm": 0.002590728225186467,
      "learning_rate": 0.15878357795406922,
      "loss": 0.3292,
      "num_input_tokens_seen": 36497808,
      "step": 19255
    },
    {
      "epoch": 3.1419365364222203,
      "grad_norm": 0.0007581555400975049,
      "learning_rate": 0.15872477349351757,
      "loss": 0.3363,
      "num_input_tokens_seen": 36506240,
      "step": 19260
    },
    {
      "epoch": 3.142752263643038,
      "grad_norm": 0.0008188298670575023,
      "learning_rate": 0.15866596768749622,
      "loss": 0.2881,
      "num_input_tokens_seen": 36515632,
      "step": 19265
    },
    {
      "epoch": 3.143567990863855,
      "grad_norm": 0.0016383819747716188,
      "learning_rate": 0.15860716054507373,
      "loss": 0.3963,
      "num_input_tokens_seen": 36525472,
      "step": 19270
    },
    {
      "epoch": 3.1443837180846725,
      "grad_norm": 0.003573442343622446,
      "learning_rate": 0.1585483520753189,
      "loss": 0.3448,
      "num_input_tokens_seen": 36534880,
      "step": 19275
    },
    {
      "epoch": 3.1451994453054897,
      "grad_norm": 0.0008927475428208709,
      "learning_rate": 0.1584895422873008,
      "loss": 0.2893,
      "num_input_tokens_seen": 36543744,
      "step": 19280
    },
    {
      "epoch": 3.1460151725263072,
      "grad_norm": 0.002234772080555558,
      "learning_rate": 0.1584307311900886,
      "loss": 0.3543,
      "num_input_tokens_seen": 36553328,
      "step": 19285
    },
    {
      "epoch": 3.1468308997471244,
      "grad_norm": 0.0011870195157825947,
      "learning_rate": 0.1583719187927517,
      "loss": 0.3126,
      "num_input_tokens_seen": 36562496,
      "step": 19290
    },
    {
      "epoch": 3.147646626967942,
      "grad_norm": 0.002372483490034938,
      "learning_rate": 0.15831310510435967,
      "loss": 0.3339,
      "num_input_tokens_seen": 36572704,
      "step": 19295
    },
    {
      "epoch": 3.148462354188759,
      "grad_norm": 0.001037241774611175,
      "learning_rate": 0.15825429013398243,
      "loss": 0.2619,
      "num_input_tokens_seen": 36582832,
      "step": 19300
    },
    {
      "epoch": 3.1492780814095767,
      "grad_norm": 0.0011680771131068468,
      "learning_rate": 0.15819547389068986,
      "loss": 0.3304,
      "num_input_tokens_seen": 36593024,
      "step": 19305
    },
    {
      "epoch": 3.150093808630394,
      "grad_norm": 0.0007807977963238955,
      "learning_rate": 0.1581366563835522,
      "loss": 0.2912,
      "num_input_tokens_seen": 36603744,
      "step": 19310
    },
    {
      "epoch": 3.1509095358512114,
      "grad_norm": 0.0019363636383786798,
      "learning_rate": 0.15807783762163993,
      "loss": 0.3066,
      "num_input_tokens_seen": 36613216,
      "step": 19315
    },
    {
      "epoch": 3.151725263072029,
      "grad_norm": 0.0013923613587394357,
      "learning_rate": 0.15801901761402365,
      "loss": 0.3064,
      "num_input_tokens_seen": 36623024,
      "step": 19320
    },
    {
      "epoch": 3.152540990292846,
      "grad_norm": 0.0033429889008402824,
      "learning_rate": 0.157960196369774,
      "loss": 0.3683,
      "num_input_tokens_seen": 36631616,
      "step": 19325
    },
    {
      "epoch": 3.153356717513663,
      "grad_norm": 0.0012752541806548834,
      "learning_rate": 0.157901373897962,
      "loss": 0.3352,
      "num_input_tokens_seen": 36640288,
      "step": 19330
    },
    {
      "epoch": 3.1541724447344808,
      "grad_norm": 0.0013352649984881282,
      "learning_rate": 0.15784255020765892,
      "loss": 0.3299,
      "num_input_tokens_seen": 36649424,
      "step": 19335
    },
    {
      "epoch": 3.1549881719552983,
      "grad_norm": 0.0027459729462862015,
      "learning_rate": 0.157783725307936,
      "loss": 0.3683,
      "num_input_tokens_seen": 36659728,
      "step": 19340
    },
    {
      "epoch": 3.1558038991761155,
      "grad_norm": 0.003978414461016655,
      "learning_rate": 0.15772489920786484,
      "loss": 0.3612,
      "num_input_tokens_seen": 36668544,
      "step": 19345
    },
    {
      "epoch": 3.156619626396933,
      "grad_norm": 0.0016029037069529295,
      "learning_rate": 0.15766607191651713,
      "loss": 0.2621,
      "num_input_tokens_seen": 36677168,
      "step": 19350
    },
    {
      "epoch": 3.15743535361775,
      "grad_norm": 0.0009177878382615745,
      "learning_rate": 0.1576072434429648,
      "loss": 0.3637,
      "num_input_tokens_seen": 36686688,
      "step": 19355
    },
    {
      "epoch": 3.1582510808385678,
      "grad_norm": 0.0008730568224564195,
      "learning_rate": 0.15754841379627998,
      "loss": 0.2926,
      "num_input_tokens_seen": 36696896,
      "step": 19360
    },
    {
      "epoch": 3.159066808059385,
      "grad_norm": 0.001016291556879878,
      "learning_rate": 0.15748958298553484,
      "loss": 0.2772,
      "num_input_tokens_seen": 36706688,
      "step": 19365
    },
    {
      "epoch": 3.1598825352802025,
      "grad_norm": 0.0025989802088588476,
      "learning_rate": 0.1574307510198019,
      "loss": 0.4143,
      "num_input_tokens_seen": 36716368,
      "step": 19370
    },
    {
      "epoch": 3.1606982625010196,
      "grad_norm": 0.0006595198647119105,
      "learning_rate": 0.15737191790815375,
      "loss": 0.3106,
      "num_input_tokens_seen": 36724832,
      "step": 19375
    },
    {
      "epoch": 3.161513989721837,
      "grad_norm": 0.0022104948293417692,
      "learning_rate": 0.15731308365966323,
      "loss": 0.405,
      "num_input_tokens_seen": 36733904,
      "step": 19380
    },
    {
      "epoch": 3.1623297169426543,
      "grad_norm": 0.0013469125842675567,
      "learning_rate": 0.15725424828340331,
      "loss": 0.3523,
      "num_input_tokens_seen": 36743232,
      "step": 19385
    },
    {
      "epoch": 3.163145444163472,
      "grad_norm": 0.0004751031519845128,
      "learning_rate": 0.15719541178844715,
      "loss": 0.3433,
      "num_input_tokens_seen": 36752656,
      "step": 19390
    },
    {
      "epoch": 3.163961171384289,
      "grad_norm": 0.002176135079935193,
      "learning_rate": 0.15713657418386806,
      "loss": 0.3279,
      "num_input_tokens_seen": 36761632,
      "step": 19395
    },
    {
      "epoch": 3.1647768986051066,
      "grad_norm": 0.0003370683407410979,
      "learning_rate": 0.15707773547873957,
      "loss": 0.3652,
      "num_input_tokens_seen": 36770432,
      "step": 19400
    },
    {
      "epoch": 3.1647768986051066,
      "eval_loss": 0.3416306972503662,
      "eval_runtime": 152.7725,
      "eval_samples_per_second": 17.837,
      "eval_steps_per_second": 8.922,
      "num_input_tokens_seen": 36770432,
      "step": 19400
    },
    {
      "epoch": 3.1655926258259237,
      "grad_norm": 0.0017315723234787583,
      "learning_rate": 0.1570188956821353,
      "loss": 0.3206,
      "num_input_tokens_seen": 36780016,
      "step": 19405
    },
    {
      "epoch": 3.1664083530467413,
      "grad_norm": 0.0018631417769938707,
      "learning_rate": 0.1569600548031291,
      "loss": 0.3112,
      "num_input_tokens_seen": 36790352,
      "step": 19410
    },
    {
      "epoch": 3.1672240802675584,
      "grad_norm": 0.0018958847504109144,
      "learning_rate": 0.156901212850795,
      "loss": 0.3354,
      "num_input_tokens_seen": 36800704,
      "step": 19415
    },
    {
      "epoch": 3.168039807488376,
      "grad_norm": 0.0008491704356856644,
      "learning_rate": 0.15684236983420716,
      "loss": 0.374,
      "num_input_tokens_seen": 36810368,
      "step": 19420
    },
    {
      "epoch": 3.168855534709193,
      "grad_norm": 0.0005638941656798124,
      "learning_rate": 0.1567835257624399,
      "loss": 0.2876,
      "num_input_tokens_seen": 36820224,
      "step": 19425
    },
    {
      "epoch": 3.1696712619300107,
      "grad_norm": 0.0022198467049747705,
      "learning_rate": 0.1567246806445677,
      "loss": 0.3084,
      "num_input_tokens_seen": 36828720,
      "step": 19430
    },
    {
      "epoch": 3.170486989150828,
      "grad_norm": 0.0006463272729888558,
      "learning_rate": 0.15666583448966526,
      "loss": 0.2841,
      "num_input_tokens_seen": 36838224,
      "step": 19435
    },
    {
      "epoch": 3.1713027163716454,
      "grad_norm": 0.0008541308343410492,
      "learning_rate": 0.1566069873068074,
      "loss": 0.3648,
      "num_input_tokens_seen": 36848208,
      "step": 19440
    },
    {
      "epoch": 3.1721184435924625,
      "grad_norm": 0.001502196304500103,
      "learning_rate": 0.156548139105069,
      "loss": 0.3168,
      "num_input_tokens_seen": 36857424,
      "step": 19445
    },
    {
      "epoch": 3.17293417081328,
      "grad_norm": 0.000996606657281518,
      "learning_rate": 0.15648928989352529,
      "loss": 0.3045,
      "num_input_tokens_seen": 36867088,
      "step": 19450
    },
    {
      "epoch": 3.1737498980340972,
      "grad_norm": 0.0009958968730643392,
      "learning_rate": 0.15643043968125156,
      "loss": 0.4072,
      "num_input_tokens_seen": 36877440,
      "step": 19455
    },
    {
      "epoch": 3.174565625254915,
      "grad_norm": 0.001308221137151122,
      "learning_rate": 0.15637158847732316,
      "loss": 0.3707,
      "num_input_tokens_seen": 36885712,
      "step": 19460
    },
    {
      "epoch": 3.175381352475732,
      "grad_norm": 0.003689510514959693,
      "learning_rate": 0.15631273629081582,
      "loss": 0.3011,
      "num_input_tokens_seen": 36896048,
      "step": 19465
    },
    {
      "epoch": 3.1761970796965495,
      "grad_norm": 0.001526065869256854,
      "learning_rate": 0.15625388313080518,
      "loss": 0.282,
      "num_input_tokens_seen": 36905840,
      "step": 19470
    },
    {
      "epoch": 3.1770128069173666,
      "grad_norm": 0.0016241061966866255,
      "learning_rate": 0.15619502900636714,
      "loss": 0.3376,
      "num_input_tokens_seen": 36914848,
      "step": 19475
    },
    {
      "epoch": 3.1778285341381842,
      "grad_norm": 0.002564976690337062,
      "learning_rate": 0.15613617392657783,
      "loss": 0.3956,
      "num_input_tokens_seen": 36924928,
      "step": 19480
    },
    {
      "epoch": 3.1786442613590014,
      "grad_norm": 0.0013335789553821087,
      "learning_rate": 0.15607731790051335,
      "loss": 0.3292,
      "num_input_tokens_seen": 36934672,
      "step": 19485
    },
    {
      "epoch": 3.179459988579819,
      "grad_norm": 0.0007164376438595355,
      "learning_rate": 0.15601846093725008,
      "loss": 0.3457,
      "num_input_tokens_seen": 36945056,
      "step": 19490
    },
    {
      "epoch": 3.180275715800636,
      "grad_norm": 0.001996916253119707,
      "learning_rate": 0.1559596030458645,
      "loss": 0.3078,
      "num_input_tokens_seen": 36954224,
      "step": 19495
    },
    {
      "epoch": 3.1810914430214536,
      "grad_norm": 0.001629441394470632,
      "learning_rate": 0.1559007442354333,
      "loss": 0.3002,
      "num_input_tokens_seen": 36962976,
      "step": 19500
    },
    {
      "epoch": 3.1819071702422708,
      "grad_norm": 0.0014332338469102979,
      "learning_rate": 0.15584188451503314,
      "loss": 0.3843,
      "num_input_tokens_seen": 36971360,
      "step": 19505
    },
    {
      "epoch": 3.1827228974630883,
      "grad_norm": 0.0007833793060854077,
      "learning_rate": 0.15578302389374094,
      "loss": 0.3108,
      "num_input_tokens_seen": 36980736,
      "step": 19510
    },
    {
      "epoch": 3.183538624683906,
      "grad_norm": 0.0004150441091042012,
      "learning_rate": 0.1557241623806338,
      "loss": 0.3243,
      "num_input_tokens_seen": 36990352,
      "step": 19515
    },
    {
      "epoch": 3.184354351904723,
      "grad_norm": 0.001630480750463903,
      "learning_rate": 0.15566529998478887,
      "loss": 0.302,
      "num_input_tokens_seen": 36999600,
      "step": 19520
    },
    {
      "epoch": 3.1851700791255406,
      "grad_norm": 0.0007094483007676899,
      "learning_rate": 0.15560643671528354,
      "loss": 0.2938,
      "num_input_tokens_seen": 37010272,
      "step": 19525
    },
    {
      "epoch": 3.1859858063463578,
      "grad_norm": 0.0019400573801249266,
      "learning_rate": 0.15554757258119514,
      "loss": 0.3732,
      "num_input_tokens_seen": 37020000,
      "step": 19530
    },
    {
      "epoch": 3.1868015335671753,
      "grad_norm": 0.000563108129426837,
      "learning_rate": 0.1554887075916014,
      "loss": 0.3639,
      "num_input_tokens_seen": 37030288,
      "step": 19535
    },
    {
      "epoch": 3.1876172607879925,
      "grad_norm": 0.003621466225013137,
      "learning_rate": 0.15542984175558,
      "loss": 0.2923,
      "num_input_tokens_seen": 37040752,
      "step": 19540
    },
    {
      "epoch": 3.18843298800881,
      "grad_norm": 0.0006360433180816472,
      "learning_rate": 0.1553709750822087,
      "loss": 0.3013,
      "num_input_tokens_seen": 37051344,
      "step": 19545
    },
    {
      "epoch": 3.189248715229627,
      "grad_norm": 0.0025705976877361536,
      "learning_rate": 0.15531210758056554,
      "loss": 0.2964,
      "num_input_tokens_seen": 37060224,
      "step": 19550
    },
    {
      "epoch": 3.1900644424504447,
      "grad_norm": 0.002028475282713771,
      "learning_rate": 0.15525323925972867,
      "loss": 0.2984,
      "num_input_tokens_seen": 37070576,
      "step": 19555
    },
    {
      "epoch": 3.190880169671262,
      "grad_norm": 0.0029193542432039976,
      "learning_rate": 0.15519437012877627,
      "loss": 0.3346,
      "num_input_tokens_seen": 37079616,
      "step": 19560
    },
    {
      "epoch": 3.1916958968920794,
      "grad_norm": 0.0005320710479281843,
      "learning_rate": 0.15513550019678676,
      "loss": 0.3687,
      "num_input_tokens_seen": 37088368,
      "step": 19565
    },
    {
      "epoch": 3.1925116241128966,
      "grad_norm": 0.003468334674835205,
      "learning_rate": 0.15507662947283854,
      "loss": 0.3505,
      "num_input_tokens_seen": 37098224,
      "step": 19570
    },
    {
      "epoch": 3.193327351333714,
      "grad_norm": 0.0016560754738748074,
      "learning_rate": 0.15501775796601028,
      "loss": 0.3241,
      "num_input_tokens_seen": 37107936,
      "step": 19575
    },
    {
      "epoch": 3.1941430785545313,
      "grad_norm": 0.001739446190185845,
      "learning_rate": 0.15495888568538066,
      "loss": 0.2699,
      "num_input_tokens_seen": 37117104,
      "step": 19580
    },
    {
      "epoch": 3.194958805775349,
      "grad_norm": 0.0009286319254897535,
      "learning_rate": 0.1549000126400286,
      "loss": 0.3612,
      "num_input_tokens_seen": 37125424,
      "step": 19585
    },
    {
      "epoch": 3.195774532996166,
      "grad_norm": 0.0007334935362450778,
      "learning_rate": 0.15484113883903294,
      "loss": 0.3219,
      "num_input_tokens_seen": 37134624,
      "step": 19590
    },
    {
      "epoch": 3.1965902602169836,
      "grad_norm": 0.003914317581802607,
      "learning_rate": 0.15478226429147288,
      "loss": 0.2762,
      "num_input_tokens_seen": 37144016,
      "step": 19595
    },
    {
      "epoch": 3.1974059874378007,
      "grad_norm": 0.001934947562403977,
      "learning_rate": 0.15472338900642757,
      "loss": 0.3161,
      "num_input_tokens_seen": 37152448,
      "step": 19600
    },
    {
      "epoch": 3.1974059874378007,
      "eval_loss": 0.3264324367046356,
      "eval_runtime": 152.609,
      "eval_samples_per_second": 17.856,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 37152448,
      "step": 19600
    },
    {
      "epoch": 3.1982217146586183,
      "grad_norm": 0.002515805419534445,
      "learning_rate": 0.15466451299297632,
      "loss": 0.2785,
      "num_input_tokens_seen": 37162288,
      "step": 19605
    },
    {
      "epoch": 3.1990374418794354,
      "grad_norm": 0.0011528126196935773,
      "learning_rate": 0.15460563626019852,
      "loss": 0.2704,
      "num_input_tokens_seen": 37169856,
      "step": 19610
    },
    {
      "epoch": 3.199853169100253,
      "grad_norm": 0.001742189982905984,
      "learning_rate": 0.15454675881717375,
      "loss": 0.3771,
      "num_input_tokens_seen": 37180544,
      "step": 19615
    },
    {
      "epoch": 3.20066889632107,
      "grad_norm": 0.0010596888605505228,
      "learning_rate": 0.1544878806729816,
      "loss": 0.3405,
      "num_input_tokens_seen": 37190208,
      "step": 19620
    },
    {
      "epoch": 3.2014846235418877,
      "grad_norm": 0.000880956940818578,
      "learning_rate": 0.1544290018367019,
      "loss": 0.3023,
      "num_input_tokens_seen": 37199712,
      "step": 19625
    },
    {
      "epoch": 3.202300350762705,
      "grad_norm": 0.0018516455311328173,
      "learning_rate": 0.15437012231741445,
      "loss": 0.3727,
      "num_input_tokens_seen": 37209760,
      "step": 19630
    },
    {
      "epoch": 3.2031160779835224,
      "grad_norm": 0.0023205687757581472,
      "learning_rate": 0.1543112421241992,
      "loss": 0.3691,
      "num_input_tokens_seen": 37218352,
      "step": 19635
    },
    {
      "epoch": 3.2039318052043395,
      "grad_norm": 0.001396635896526277,
      "learning_rate": 0.15425236126613626,
      "loss": 0.3514,
      "num_input_tokens_seen": 37228000,
      "step": 19640
    },
    {
      "epoch": 3.204747532425157,
      "grad_norm": 0.0015256158076226711,
      "learning_rate": 0.15419347975230577,
      "loss": 0.2946,
      "num_input_tokens_seen": 37236192,
      "step": 19645
    },
    {
      "epoch": 3.205563259645974,
      "grad_norm": 0.001455808524042368,
      "learning_rate": 0.154134597591788,
      "loss": 0.3524,
      "num_input_tokens_seen": 37245728,
      "step": 19650
    },
    {
      "epoch": 3.206378986866792,
      "grad_norm": 0.001391086378134787,
      "learning_rate": 0.1540757147936633,
      "loss": 0.3442,
      "num_input_tokens_seen": 37253904,
      "step": 19655
    },
    {
      "epoch": 3.207194714087609,
      "grad_norm": 0.001914181630127132,
      "learning_rate": 0.1540168313670122,
      "loss": 0.3192,
      "num_input_tokens_seen": 37264064,
      "step": 19660
    },
    {
      "epoch": 3.2080104413084265,
      "grad_norm": 0.0011454280465841293,
      "learning_rate": 0.1539579473209152,
      "loss": 0.3235,
      "num_input_tokens_seen": 37273072,
      "step": 19665
    },
    {
      "epoch": 3.2088261685292436,
      "grad_norm": 0.0016152027528733015,
      "learning_rate": 0.15389906266445294,
      "loss": 0.2874,
      "num_input_tokens_seen": 37282768,
      "step": 19670
    },
    {
      "epoch": 3.209641895750061,
      "grad_norm": 0.0020332555286586285,
      "learning_rate": 0.15384017740670627,
      "loss": 0.3286,
      "num_input_tokens_seen": 37291248,
      "step": 19675
    },
    {
      "epoch": 3.2104576229708783,
      "grad_norm": 0.005128134042024612,
      "learning_rate": 0.15378129155675602,
      "loss": 0.4154,
      "num_input_tokens_seen": 37302016,
      "step": 19680
    },
    {
      "epoch": 3.211273350191696,
      "grad_norm": 0.003635519417002797,
      "learning_rate": 0.15372240512368307,
      "loss": 0.3727,
      "num_input_tokens_seen": 37310880,
      "step": 19685
    },
    {
      "epoch": 3.2120890774125135,
      "grad_norm": 0.0025662051048129797,
      "learning_rate": 0.1536635181165684,
      "loss": 0.3352,
      "num_input_tokens_seen": 37321952,
      "step": 19690
    },
    {
      "epoch": 3.2129048046333306,
      "grad_norm": 0.0014010303420946002,
      "learning_rate": 0.15360463054449328,
      "loss": 0.338,
      "num_input_tokens_seen": 37331072,
      "step": 19695
    },
    {
      "epoch": 3.2137205318541477,
      "grad_norm": 0.0011608635541051626,
      "learning_rate": 0.1535457424165388,
      "loss": 0.3735,
      "num_input_tokens_seen": 37339696,
      "step": 19700
    },
    {
      "epoch": 3.2145362590749653,
      "grad_norm": 0.002158177550882101,
      "learning_rate": 0.15348685374178628,
      "loss": 0.3432,
      "num_input_tokens_seen": 37349088,
      "step": 19705
    },
    {
      "epoch": 3.215351986295783,
      "grad_norm": 0.0010054511949419975,
      "learning_rate": 0.1534279645293171,
      "loss": 0.38,
      "num_input_tokens_seen": 37359360,
      "step": 19710
    },
    {
      "epoch": 3.2161677135166,
      "grad_norm": 0.0008055234793573618,
      "learning_rate": 0.1533690747882127,
      "loss": 0.3453,
      "num_input_tokens_seen": 37368688,
      "step": 19715
    },
    {
      "epoch": 3.2169834407374176,
      "grad_norm": 0.003050709841772914,
      "learning_rate": 0.15331018452755465,
      "loss": 0.2994,
      "num_input_tokens_seen": 37378912,
      "step": 19720
    },
    {
      "epoch": 3.2177991679582347,
      "grad_norm": 0.0015319440281018615,
      "learning_rate": 0.15325129375642457,
      "loss": 0.3339,
      "num_input_tokens_seen": 37388512,
      "step": 19725
    },
    {
      "epoch": 3.2186148951790523,
      "grad_norm": 0.002166667254641652,
      "learning_rate": 0.15319240248390406,
      "loss": 0.2929,
      "num_input_tokens_seen": 37398368,
      "step": 19730
    },
    {
      "epoch": 3.2194306223998694,
      "grad_norm": 0.0016044486546888947,
      "learning_rate": 0.153133510719075,
      "loss": 0.3386,
      "num_input_tokens_seen": 37407968,
      "step": 19735
    },
    {
      "epoch": 3.220246349620687,
      "grad_norm": 0.0024197471793740988,
      "learning_rate": 0.15307461847101922,
      "loss": 0.2953,
      "num_input_tokens_seen": 37418912,
      "step": 19740
    },
    {
      "epoch": 3.221062076841504,
      "grad_norm": 0.003853290341794491,
      "learning_rate": 0.15301572574881864,
      "loss": 0.3694,
      "num_input_tokens_seen": 37429264,
      "step": 19745
    },
    {
      "epoch": 3.2218778040623217,
      "grad_norm": 0.0031689784955233335,
      "learning_rate": 0.15295683256155523,
      "loss": 0.34,
      "num_input_tokens_seen": 37438960,
      "step": 19750
    },
    {
      "epoch": 3.222693531283139,
      "grad_norm": 0.0010959766805171967,
      "learning_rate": 0.15289793891831113,
      "loss": 0.3179,
      "num_input_tokens_seen": 37448320,
      "step": 19755
    },
    {
      "epoch": 3.2235092585039564,
      "grad_norm": 0.0012204536469653249,
      "learning_rate": 0.15283904482816837,
      "loss": 0.3292,
      "num_input_tokens_seen": 37457456,
      "step": 19760
    },
    {
      "epoch": 3.2243249857247736,
      "grad_norm": 0.0016742550069466233,
      "learning_rate": 0.15278015030020928,
      "loss": 0.3138,
      "num_input_tokens_seen": 37465968,
      "step": 19765
    },
    {
      "epoch": 3.225140712945591,
      "grad_norm": 0.0010439780307933688,
      "learning_rate": 0.152721255343516,
      "loss": 0.3444,
      "num_input_tokens_seen": 37473840,
      "step": 19770
    },
    {
      "epoch": 3.2259564401664083,
      "grad_norm": 0.0008133708615787327,
      "learning_rate": 0.15266235996717098,
      "loss": 0.2606,
      "num_input_tokens_seen": 37483344,
      "step": 19775
    },
    {
      "epoch": 3.226772167387226,
      "grad_norm": 0.0029440135695040226,
      "learning_rate": 0.15260346418025664,
      "loss": 0.3201,
      "num_input_tokens_seen": 37493888,
      "step": 19780
    },
    {
      "epoch": 3.227587894608043,
      "grad_norm": 0.0016960572684183717,
      "learning_rate": 0.15254456799185537,
      "loss": 0.3704,
      "num_input_tokens_seen": 37503360,
      "step": 19785
    },
    {
      "epoch": 3.2284036218288605,
      "grad_norm": 0.000975633563939482,
      "learning_rate": 0.15248567141104974,
      "loss": 0.379,
      "num_input_tokens_seen": 37513024,
      "step": 19790
    },
    {
      "epoch": 3.2292193490496777,
      "grad_norm": 0.0008328021503984928,
      "learning_rate": 0.15242677444692232,
      "loss": 0.3526,
      "num_input_tokens_seen": 37521984,
      "step": 19795
    },
    {
      "epoch": 3.2300350762704952,
      "grad_norm": 0.001116294297389686,
      "learning_rate": 0.15236787710855584,
      "loss": 0.3487,
      "num_input_tokens_seen": 37532496,
      "step": 19800
    },
    {
      "epoch": 3.2300350762704952,
      "eval_loss": 0.34283190965652466,
      "eval_runtime": 152.5577,
      "eval_samples_per_second": 17.862,
      "eval_steps_per_second": 8.934,
      "num_input_tokens_seen": 37532496,
      "step": 19800
    },
    {
      "epoch": 3.2308508034913124,
      "grad_norm": 0.00048134688404388726,
      "learning_rate": 0.1523089794050329,
      "loss": 0.2979,
      "num_input_tokens_seen": 37540976,
      "step": 19805
    },
    {
      "epoch": 3.23166653071213,
      "grad_norm": 0.0035200007259845734,
      "learning_rate": 0.15225008134543633,
      "loss": 0.3819,
      "num_input_tokens_seen": 37550304,
      "step": 19810
    },
    {
      "epoch": 3.232482257932947,
      "grad_norm": 0.0018794640200212598,
      "learning_rate": 0.15219118293884895,
      "loss": 0.33,
      "num_input_tokens_seen": 37559600,
      "step": 19815
    },
    {
      "epoch": 3.2332979851537647,
      "grad_norm": 0.0013642478734254837,
      "learning_rate": 0.15213228419435362,
      "loss": 0.306,
      "num_input_tokens_seen": 37568576,
      "step": 19820
    },
    {
      "epoch": 3.234113712374582,
      "grad_norm": 0.002681419486179948,
      "learning_rate": 0.15207338512103327,
      "loss": 0.3457,
      "num_input_tokens_seen": 37577792,
      "step": 19825
    },
    {
      "epoch": 3.2349294395953994,
      "grad_norm": 0.001669298973865807,
      "learning_rate": 0.1520144857279709,
      "loss": 0.3187,
      "num_input_tokens_seen": 37587360,
      "step": 19830
    },
    {
      "epoch": 3.2357451668162165,
      "grad_norm": 0.0010318546555936337,
      "learning_rate": 0.1519555860242495,
      "loss": 0.3001,
      "num_input_tokens_seen": 37596816,
      "step": 19835
    },
    {
      "epoch": 3.236560894037034,
      "grad_norm": 0.002468397840857506,
      "learning_rate": 0.15189668601895218,
      "loss": 0.3294,
      "num_input_tokens_seen": 37606928,
      "step": 19840
    },
    {
      "epoch": 3.237376621257851,
      "grad_norm": 0.0010872846469283104,
      "learning_rate": 0.151837785721162,
      "loss": 0.3323,
      "num_input_tokens_seen": 37616976,
      "step": 19845
    },
    {
      "epoch": 3.2381923484786688,
      "grad_norm": 0.0009148559765890241,
      "learning_rate": 0.15177888513996218,
      "loss": 0.3643,
      "num_input_tokens_seen": 37626912,
      "step": 19850
    },
    {
      "epoch": 3.239008075699486,
      "grad_norm": 0.0007544379332102835,
      "learning_rate": 0.15171998428443592,
      "loss": 0.3179,
      "num_input_tokens_seen": 37636256,
      "step": 19855
    },
    {
      "epoch": 3.2398238029203035,
      "grad_norm": 0.0019051020499318838,
      "learning_rate": 0.1516610831636665,
      "loss": 0.3953,
      "num_input_tokens_seen": 37644048,
      "step": 19860
    },
    {
      "epoch": 3.2406395301411206,
      "grad_norm": 0.004103842191398144,
      "learning_rate": 0.15160218178673715,
      "loss": 0.3738,
      "num_input_tokens_seen": 37654064,
      "step": 19865
    },
    {
      "epoch": 3.241455257361938,
      "grad_norm": 0.003004888305440545,
      "learning_rate": 0.15154328016273122,
      "loss": 0.3112,
      "num_input_tokens_seen": 37663808,
      "step": 19870
    },
    {
      "epoch": 3.2422709845827553,
      "grad_norm": 0.0014199463184922934,
      "learning_rate": 0.1514843783007321,
      "loss": 0.3637,
      "num_input_tokens_seen": 37673888,
      "step": 19875
    },
    {
      "epoch": 3.243086711803573,
      "grad_norm": 0.0020426358096301556,
      "learning_rate": 0.15142547620982322,
      "loss": 0.3793,
      "num_input_tokens_seen": 37683312,
      "step": 19880
    },
    {
      "epoch": 3.2439024390243905,
      "grad_norm": 0.00443014083430171,
      "learning_rate": 0.15136657389908797,
      "loss": 0.3586,
      "num_input_tokens_seen": 37693696,
      "step": 19885
    },
    {
      "epoch": 3.2447181662452076,
      "grad_norm": 0.0008172177476808429,
      "learning_rate": 0.15130767137760986,
      "loss": 0.3078,
      "num_input_tokens_seen": 37702992,
      "step": 19890
    },
    {
      "epoch": 3.245533893466025,
      "grad_norm": 0.0007872836431488395,
      "learning_rate": 0.15124876865447243,
      "loss": 0.3264,
      "num_input_tokens_seen": 37712736,
      "step": 19895
    },
    {
      "epoch": 3.2463496206868423,
      "grad_norm": 0.0014380281791090965,
      "learning_rate": 0.15118986573875912,
      "loss": 0.2776,
      "num_input_tokens_seen": 37722832,
      "step": 19900
    },
    {
      "epoch": 3.24716534790766,
      "grad_norm": 0.0004313244135119021,
      "learning_rate": 0.15113096263955358,
      "loss": 0.2517,
      "num_input_tokens_seen": 37731584,
      "step": 19905
    },
    {
      "epoch": 3.247981075128477,
      "grad_norm": 0.001058106543496251,
      "learning_rate": 0.1510720593659394,
      "loss": 0.3586,
      "num_input_tokens_seen": 37741808,
      "step": 19910
    },
    {
      "epoch": 3.2487968023492946,
      "grad_norm": 0.002088902285322547,
      "learning_rate": 0.15101315592700015,
      "loss": 0.3023,
      "num_input_tokens_seen": 37751056,
      "step": 19915
    },
    {
      "epoch": 3.2496125295701117,
      "grad_norm": 0.0017389694694429636,
      "learning_rate": 0.15095425233181956,
      "loss": 0.2589,
      "num_input_tokens_seen": 37760768,
      "step": 19920
    },
    {
      "epoch": 3.2504282567909293,
      "grad_norm": 0.002798644592985511,
      "learning_rate": 0.15089534858948128,
      "loss": 0.3506,
      "num_input_tokens_seen": 37769632,
      "step": 19925
    },
    {
      "epoch": 3.2512439840117464,
      "grad_norm": 0.000491380866151303,
      "learning_rate": 0.15083644470906898,
      "loss": 0.3363,
      "num_input_tokens_seen": 37779424,
      "step": 19930
    },
    {
      "epoch": 3.252059711232564,
      "grad_norm": 0.0017101768171414733,
      "learning_rate": 0.1507775406996664,
      "loss": 0.3159,
      "num_input_tokens_seen": 37789520,
      "step": 19935
    },
    {
      "epoch": 3.252875438453381,
      "grad_norm": 0.0028178554493933916,
      "learning_rate": 0.15071863657035725,
      "loss": 0.4172,
      "num_input_tokens_seen": 37799104,
      "step": 19940
    },
    {
      "epoch": 3.2536911656741987,
      "grad_norm": 0.0007463658112101257,
      "learning_rate": 0.15065973233022534,
      "loss": 0.2947,
      "num_input_tokens_seen": 37809152,
      "step": 19945
    },
    {
      "epoch": 3.254506892895016,
      "grad_norm": 0.0017171450890600681,
      "learning_rate": 0.15060082798835442,
      "loss": 0.3263,
      "num_input_tokens_seen": 37817376,
      "step": 19950
    },
    {
      "epoch": 3.2553226201158334,
      "grad_norm": 0.001771499402821064,
      "learning_rate": 0.15054192355382823,
      "loss": 0.3359,
      "num_input_tokens_seen": 37825904,
      "step": 19955
    },
    {
      "epoch": 3.2561383473366505,
      "grad_norm": 0.0006813554209657013,
      "learning_rate": 0.15048301903573066,
      "loss": 0.3052,
      "num_input_tokens_seen": 37834928,
      "step": 19960
    },
    {
      "epoch": 3.256954074557468,
      "grad_norm": 0.0016150482697412372,
      "learning_rate": 0.15042411444314546,
      "loss": 0.301,
      "num_input_tokens_seen": 37845552,
      "step": 19965
    },
    {
      "epoch": 3.2577698017782852,
      "grad_norm": 0.0013778459979221225,
      "learning_rate": 0.1503652097851565,
      "loss": 0.3146,
      "num_input_tokens_seen": 37854688,
      "step": 19970
    },
    {
      "epoch": 3.258585528999103,
      "grad_norm": 0.002297114348039031,
      "learning_rate": 0.15030630507084758,
      "loss": 0.2929,
      "num_input_tokens_seen": 37864880,
      "step": 19975
    },
    {
      "epoch": 3.25940125621992,
      "grad_norm": 0.0018199827754870057,
      "learning_rate": 0.1502474003093026,
      "loss": 0.2946,
      "num_input_tokens_seen": 37873488,
      "step": 19980
    },
    {
      "epoch": 3.2602169834407375,
      "grad_norm": 0.0010592974722385406,
      "learning_rate": 0.15018849550960536,
      "loss": 0.3343,
      "num_input_tokens_seen": 37882240,
      "step": 19985
    },
    {
      "epoch": 3.2610327106615546,
      "grad_norm": 0.0022017478477209806,
      "learning_rate": 0.15012959068083975,
      "loss": 0.368,
      "num_input_tokens_seen": 37892592,
      "step": 19990
    },
    {
      "epoch": 3.261848437882372,
      "grad_norm": 0.002217042725533247,
      "learning_rate": 0.1500706858320896,
      "loss": 0.3346,
      "num_input_tokens_seen": 37900656,
      "step": 19995
    },
    {
      "epoch": 3.2626641651031894,
      "grad_norm": 0.0026323108468204737,
      "learning_rate": 0.15001178097243886,
      "loss": 0.3284,
      "num_input_tokens_seen": 37910480,
      "step": 20000
    },
    {
      "epoch": 3.2626641651031894,
      "eval_loss": 0.3320654332637787,
      "eval_runtime": 152.6927,
      "eval_samples_per_second": 17.846,
      "eval_steps_per_second": 8.926,
      "num_input_tokens_seen": 37910480,
      "step": 20000
    },
    {
      "epoch": 3.263479892324007,
      "grad_norm": 0.0014308097306638956,
      "learning_rate": 0.1499528761109713,
      "loss": 0.3518,
      "num_input_tokens_seen": 37919664,
      "step": 20005
    },
    {
      "epoch": 3.264295619544824,
      "grad_norm": 0.0039832573384046555,
      "learning_rate": 0.14989397125677087,
      "loss": 0.4059,
      "num_input_tokens_seen": 37928784,
      "step": 20010
    },
    {
      "epoch": 3.2651113467656416,
      "grad_norm": 0.005128995515406132,
      "learning_rate": 0.14983506641892141,
      "loss": 0.4019,
      "num_input_tokens_seen": 37938960,
      "step": 20015
    },
    {
      "epoch": 3.2659270739864588,
      "grad_norm": 0.0022391299717128277,
      "learning_rate": 0.14977616160650672,
      "loss": 0.3402,
      "num_input_tokens_seen": 37947216,
      "step": 20020
    },
    {
      "epoch": 3.2667428012072763,
      "grad_norm": 0.002348039299249649,
      "learning_rate": 0.14971725682861076,
      "loss": 0.2888,
      "num_input_tokens_seen": 37956576,
      "step": 20025
    },
    {
      "epoch": 3.2675585284280935,
      "grad_norm": 0.0033749714493751526,
      "learning_rate": 0.14965835209431738,
      "loss": 0.3502,
      "num_input_tokens_seen": 37966128,
      "step": 20030
    },
    {
      "epoch": 3.268374255648911,
      "grad_norm": 0.0014489686582237482,
      "learning_rate": 0.14959944741271036,
      "loss": 0.4273,
      "num_input_tokens_seen": 37975536,
      "step": 20035
    },
    {
      "epoch": 3.269189982869728,
      "grad_norm": 0.0031392788514494896,
      "learning_rate": 0.14954054279287363,
      "loss": 0.398,
      "num_input_tokens_seen": 37986256,
      "step": 20040
    },
    {
      "epoch": 3.2700057100905457,
      "grad_norm": 0.003361757379025221,
      "learning_rate": 0.14948163824389094,
      "loss": 0.3406,
      "num_input_tokens_seen": 37996368,
      "step": 20045
    },
    {
      "epoch": 3.270821437311363,
      "grad_norm": 0.0020818717312067747,
      "learning_rate": 0.14942273377484613,
      "loss": 0.3585,
      "num_input_tokens_seen": 38004688,
      "step": 20050
    },
    {
      "epoch": 3.2716371645321805,
      "grad_norm": 0.000962373276706785,
      "learning_rate": 0.1493638293948231,
      "loss": 0.3447,
      "num_input_tokens_seen": 38012672,
      "step": 20055
    },
    {
      "epoch": 3.272452891752998,
      "grad_norm": 0.00229767058044672,
      "learning_rate": 0.14930492511290547,
      "loss": 0.3272,
      "num_input_tokens_seen": 38022016,
      "step": 20060
    },
    {
      "epoch": 3.273268618973815,
      "grad_norm": 0.0027805936988443136,
      "learning_rate": 0.14924602093817715,
      "loss": 0.2757,
      "num_input_tokens_seen": 38031328,
      "step": 20065
    },
    {
      "epoch": 3.2740843461946323,
      "grad_norm": 0.0015725016128271818,
      "learning_rate": 0.14918711687972194,
      "loss": 0.3492,
      "num_input_tokens_seen": 38040224,
      "step": 20070
    },
    {
      "epoch": 3.27490007341545,
      "grad_norm": 0.0011255257995799184,
      "learning_rate": 0.14912821294662346,
      "loss": 0.3638,
      "num_input_tokens_seen": 38048400,
      "step": 20075
    },
    {
      "epoch": 3.2757158006362674,
      "grad_norm": 0.0017564764712005854,
      "learning_rate": 0.14906930914796554,
      "loss": 0.3809,
      "num_input_tokens_seen": 38058672,
      "step": 20080
    },
    {
      "epoch": 3.2765315278570846,
      "grad_norm": 0.0014754381263628602,
      "learning_rate": 0.14901040549283182,
      "loss": 0.3424,
      "num_input_tokens_seen": 38068688,
      "step": 20085
    },
    {
      "epoch": 3.2773472550779017,
      "grad_norm": 0.000830646138638258,
      "learning_rate": 0.148951501990306,
      "loss": 0.3136,
      "num_input_tokens_seen": 38078304,
      "step": 20090
    },
    {
      "epoch": 3.2781629822987193,
      "grad_norm": 0.0009948041988536716,
      "learning_rate": 0.14889259864947177,
      "loss": 0.3223,
      "num_input_tokens_seen": 38088688,
      "step": 20095
    },
    {
      "epoch": 3.278978709519537,
      "grad_norm": 0.0016303783049806952,
      "learning_rate": 0.14883369547941272,
      "loss": 0.3181,
      "num_input_tokens_seen": 38097680,
      "step": 20100
    },
    {
      "epoch": 3.279794436740354,
      "grad_norm": 0.002494517248123884,
      "learning_rate": 0.14877479248921247,
      "loss": 0.3399,
      "num_input_tokens_seen": 38108208,
      "step": 20105
    },
    {
      "epoch": 3.2806101639611716,
      "grad_norm": 0.002705952851101756,
      "learning_rate": 0.14871588968795468,
      "loss": 0.3471,
      "num_input_tokens_seen": 38117376,
      "step": 20110
    },
    {
      "epoch": 3.2814258911819887,
      "grad_norm": 0.001210991758853197,
      "learning_rate": 0.1486569870847228,
      "loss": 0.3075,
      "num_input_tokens_seen": 38127728,
      "step": 20115
    },
    {
      "epoch": 3.2822416184028063,
      "grad_norm": 0.001508357236161828,
      "learning_rate": 0.1485980846886004,
      "loss": 0.307,
      "num_input_tokens_seen": 38136416,
      "step": 20120
    },
    {
      "epoch": 3.2830573456236234,
      "grad_norm": 0.0016082742949947715,
      "learning_rate": 0.14853918250867096,
      "loss": 0.3153,
      "num_input_tokens_seen": 38146000,
      "step": 20125
    },
    {
      "epoch": 3.283873072844441,
      "grad_norm": 0.003171206684783101,
      "learning_rate": 0.1484802805540179,
      "loss": 0.2742,
      "num_input_tokens_seen": 38156176,
      "step": 20130
    },
    {
      "epoch": 3.284688800065258,
      "grad_norm": 0.0021353152114897966,
      "learning_rate": 0.14842137883372472,
      "loss": 0.3773,
      "num_input_tokens_seen": 38164848,
      "step": 20135
    },
    {
      "epoch": 3.2855045272860757,
      "grad_norm": 0.000986113678663969,
      "learning_rate": 0.14836247735687474,
      "loss": 0.2443,
      "num_input_tokens_seen": 38173792,
      "step": 20140
    },
    {
      "epoch": 3.286320254506893,
      "grad_norm": 0.0029055580962449312,
      "learning_rate": 0.14830357613255132,
      "loss": 0.3219,
      "num_input_tokens_seen": 38183280,
      "step": 20145
    },
    {
      "epoch": 3.2871359817277104,
      "grad_norm": 0.0013446924276649952,
      "learning_rate": 0.1482446751698378,
      "loss": 0.2982,
      "num_input_tokens_seen": 38192128,
      "step": 20150
    },
    {
      "epoch": 3.2879517089485275,
      "grad_norm": 0.003932671621441841,
      "learning_rate": 0.14818577447781744,
      "loss": 0.4042,
      "num_input_tokens_seen": 38201072,
      "step": 20155
    },
    {
      "epoch": 3.288767436169345,
      "grad_norm": 0.002732557710260153,
      "learning_rate": 0.14812687406557346,
      "loss": 0.3884,
      "num_input_tokens_seen": 38210208,
      "step": 20160
    },
    {
      "epoch": 3.289583163390162,
      "grad_norm": 0.002193657448515296,
      "learning_rate": 0.14806797394218899,
      "loss": 0.4101,
      "num_input_tokens_seen": 38219008,
      "step": 20165
    },
    {
      "epoch": 3.29039889061098,
      "grad_norm": 0.002108650514855981,
      "learning_rate": 0.1480090741167472,
      "loss": 0.3457,
      "num_input_tokens_seen": 38228512,
      "step": 20170
    },
    {
      "epoch": 3.291214617831797,
      "grad_norm": 0.0016523402882739902,
      "learning_rate": 0.1479501745983313,
      "loss": 0.3177,
      "num_input_tokens_seen": 38237872,
      "step": 20175
    },
    {
      "epoch": 3.2920303450526145,
      "grad_norm": 0.0014531429624184966,
      "learning_rate": 0.14789127539602415,
      "loss": 0.335,
      "num_input_tokens_seen": 38246896,
      "step": 20180
    },
    {
      "epoch": 3.2928460722734316,
      "grad_norm": 0.000392660207580775,
      "learning_rate": 0.14783237651890885,
      "loss": 0.3484,
      "num_input_tokens_seen": 38256208,
      "step": 20185
    },
    {
      "epoch": 3.293661799494249,
      "grad_norm": 0.0016617715591564775,
      "learning_rate": 0.14777347797606838,
      "loss": 0.3348,
      "num_input_tokens_seen": 38266448,
      "step": 20190
    },
    {
      "epoch": 3.2944775267150663,
      "grad_norm": 0.001713298843242228,
      "learning_rate": 0.14771457977658553,
      "loss": 0.3549,
      "num_input_tokens_seen": 38277136,
      "step": 20195
    },
    {
      "epoch": 3.295293253935884,
      "grad_norm": 0.0009066069032996893,
      "learning_rate": 0.14765568192954326,
      "loss": 0.3752,
      "num_input_tokens_seen": 38286080,
      "step": 20200
    },
    {
      "epoch": 3.295293253935884,
      "eval_loss": 0.32652515172958374,
      "eval_runtime": 152.7129,
      "eval_samples_per_second": 17.844,
      "eval_steps_per_second": 8.925,
      "num_input_tokens_seen": 38286080,
      "step": 20200
    },
    {
      "epoch": 3.296108981156701,
      "grad_norm": 0.001503334497101605,
      "learning_rate": 0.14759678444402421,
      "loss": 0.3041,
      "num_input_tokens_seen": 38296064,
      "step": 20205
    },
    {
      "epoch": 3.2969247083775186,
      "grad_norm": 0.0017053387127816677,
      "learning_rate": 0.14753788732911122,
      "loss": 0.3212,
      "num_input_tokens_seen": 38305104,
      "step": 20210
    },
    {
      "epoch": 3.2977404355983357,
      "grad_norm": 0.002034729113802314,
      "learning_rate": 0.147478990593887,
      "loss": 0.3101,
      "num_input_tokens_seen": 38315136,
      "step": 20215
    },
    {
      "epoch": 3.2985561628191533,
      "grad_norm": 0.0013861791230738163,
      "learning_rate": 0.14742009424743405,
      "loss": 0.2791,
      "num_input_tokens_seen": 38323280,
      "step": 20220
    },
    {
      "epoch": 3.2993718900399704,
      "grad_norm": 0.0033033534418791533,
      "learning_rate": 0.14736119829883504,
      "loss": 0.3651,
      "num_input_tokens_seen": 38332528,
      "step": 20225
    },
    {
      "epoch": 3.300187617260788,
      "grad_norm": 0.000961034675128758,
      "learning_rate": 0.14730230275717243,
      "loss": 0.2607,
      "num_input_tokens_seen": 38343216,
      "step": 20230
    },
    {
      "epoch": 3.3010033444816056,
      "grad_norm": 0.0023088103625923395,
      "learning_rate": 0.14724340763152854,
      "loss": 0.4087,
      "num_input_tokens_seen": 38353856,
      "step": 20235
    },
    {
      "epoch": 3.3018190717024227,
      "grad_norm": 0.001419688924215734,
      "learning_rate": 0.14718451293098594,
      "loss": 0.3131,
      "num_input_tokens_seen": 38363696,
      "step": 20240
    },
    {
      "epoch": 3.30263479892324,
      "grad_norm": 0.0030606482177972794,
      "learning_rate": 0.14712561866462676,
      "loss": 0.3224,
      "num_input_tokens_seen": 38372832,
      "step": 20245
    },
    {
      "epoch": 3.3034505261440574,
      "grad_norm": 0.0007845272193662822,
      "learning_rate": 0.1470667248415333,
      "loss": 0.311,
      "num_input_tokens_seen": 38383424,
      "step": 20250
    },
    {
      "epoch": 3.304266253364875,
      "grad_norm": 0.0007868824759498239,
      "learning_rate": 0.1470078314707878,
      "loss": 0.2898,
      "num_input_tokens_seen": 38393936,
      "step": 20255
    },
    {
      "epoch": 3.305081980585692,
      "grad_norm": 0.0019089375855401158,
      "learning_rate": 0.14694893856147223,
      "loss": 0.3186,
      "num_input_tokens_seen": 38402224,
      "step": 20260
    },
    {
      "epoch": 3.3058977078065093,
      "grad_norm": 0.003013055073097348,
      "learning_rate": 0.14689004612266868,
      "loss": 0.3606,
      "num_input_tokens_seen": 38411280,
      "step": 20265
    },
    {
      "epoch": 3.306713435027327,
      "grad_norm": 0.004334839526563883,
      "learning_rate": 0.14683115416345913,
      "loss": 0.3757,
      "num_input_tokens_seen": 38420384,
      "step": 20270
    },
    {
      "epoch": 3.3075291622481444,
      "grad_norm": 0.0027818004600703716,
      "learning_rate": 0.1467722626929254,
      "loss": 0.3995,
      "num_input_tokens_seen": 38429920,
      "step": 20275
    },
    {
      "epoch": 3.3083448894689615,
      "grad_norm": 0.002100664656609297,
      "learning_rate": 0.14671337172014937,
      "loss": 0.3186,
      "num_input_tokens_seen": 38439440,
      "step": 20280
    },
    {
      "epoch": 3.309160616689779,
      "grad_norm": 0.0008405867265537381,
      "learning_rate": 0.14665448125421265,
      "loss": 0.3162,
      "num_input_tokens_seen": 38448416,
      "step": 20285
    },
    {
      "epoch": 3.3099763439105963,
      "grad_norm": 0.0019882891792804003,
      "learning_rate": 0.146595591304197,
      "loss": 0.3595,
      "num_input_tokens_seen": 38458592,
      "step": 20290
    },
    {
      "epoch": 3.310792071131414,
      "grad_norm": 0.0010199459502473474,
      "learning_rate": 0.14653670187918397,
      "loss": 0.3188,
      "num_input_tokens_seen": 38468144,
      "step": 20295
    },
    {
      "epoch": 3.311607798352231,
      "grad_norm": 0.0025122042279690504,
      "learning_rate": 0.14647781298825502,
      "loss": 0.2977,
      "num_input_tokens_seen": 38478352,
      "step": 20300
    },
    {
      "epoch": 3.3124235255730485,
      "grad_norm": 0.0020389133132994175,
      "learning_rate": 0.14641892464049153,
      "loss": 0.2969,
      "num_input_tokens_seen": 38487072,
      "step": 20305
    },
    {
      "epoch": 3.3132392527938657,
      "grad_norm": 0.0005908382590860128,
      "learning_rate": 0.14636003684497495,
      "loss": 0.2968,
      "num_input_tokens_seen": 38496224,
      "step": 20310
    },
    {
      "epoch": 3.3140549800146832,
      "grad_norm": 0.0018169803079217672,
      "learning_rate": 0.14630114961078636,
      "loss": 0.3199,
      "num_input_tokens_seen": 38505616,
      "step": 20315
    },
    {
      "epoch": 3.3148707072355004,
      "grad_norm": 0.0013608128065243363,
      "learning_rate": 0.14624226294700704,
      "loss": 0.3506,
      "num_input_tokens_seen": 38515456,
      "step": 20320
    },
    {
      "epoch": 3.315686434456318,
      "grad_norm": 0.0022734981030225754,
      "learning_rate": 0.14618337686271793,
      "loss": 0.3097,
      "num_input_tokens_seen": 38525408,
      "step": 20325
    },
    {
      "epoch": 3.316502161677135,
      "grad_norm": 0.0027695195749402046,
      "learning_rate": 0.1461244913670001,
      "loss": 0.3414,
      "num_input_tokens_seen": 38534720,
      "step": 20330
    },
    {
      "epoch": 3.3173178888979526,
      "grad_norm": 0.0018068611389026046,
      "learning_rate": 0.1460656064689344,
      "loss": 0.3373,
      "num_input_tokens_seen": 38544464,
      "step": 20335
    },
    {
      "epoch": 3.31813361611877,
      "grad_norm": 0.0017917387885972857,
      "learning_rate": 0.14600672217760163,
      "loss": 0.3361,
      "num_input_tokens_seen": 38553824,
      "step": 20340
    },
    {
      "epoch": 3.3189493433395874,
      "grad_norm": 0.002764556324109435,
      "learning_rate": 0.14594783850208248,
      "loss": 0.3229,
      "num_input_tokens_seen": 38563488,
      "step": 20345
    },
    {
      "epoch": 3.3197650705604045,
      "grad_norm": 0.0018762253457680345,
      "learning_rate": 0.14588895545145758,
      "loss": 0.3169,
      "num_input_tokens_seen": 38572400,
      "step": 20350
    },
    {
      "epoch": 3.320580797781222,
      "grad_norm": 0.0007509238203056157,
      "learning_rate": 0.14583007303480738,
      "loss": 0.3399,
      "num_input_tokens_seen": 38581888,
      "step": 20355
    },
    {
      "epoch": 3.321396525002039,
      "grad_norm": 0.0014759697951376438,
      "learning_rate": 0.14577119126121235,
      "loss": 0.3309,
      "num_input_tokens_seen": 38591936,
      "step": 20360
    },
    {
      "epoch": 3.3222122522228568,
      "grad_norm": 0.0035343721974641085,
      "learning_rate": 0.14571231013975272,
      "loss": 0.3147,
      "num_input_tokens_seen": 38600624,
      "step": 20365
    },
    {
      "epoch": 3.323027979443674,
      "grad_norm": 0.0006034352700226009,
      "learning_rate": 0.1456534296795088,
      "loss": 0.2814,
      "num_input_tokens_seen": 38610080,
      "step": 20370
    },
    {
      "epoch": 3.3238437066644915,
      "grad_norm": 0.003110931720584631,
      "learning_rate": 0.14559454988956066,
      "loss": 0.3351,
      "num_input_tokens_seen": 38620480,
      "step": 20375
    },
    {
      "epoch": 3.3246594338853086,
      "grad_norm": 0.0010163624538108706,
      "learning_rate": 0.1455356707789882,
      "loss": 0.2519,
      "num_input_tokens_seen": 38629744,
      "step": 20380
    },
    {
      "epoch": 3.325475161106126,
      "grad_norm": 0.003113999729976058,
      "learning_rate": 0.14547679235687147,
      "loss": 0.3703,
      "num_input_tokens_seen": 38639264,
      "step": 20385
    },
    {
      "epoch": 3.3262908883269433,
      "grad_norm": 0.002637873636558652,
      "learning_rate": 0.14541791463229023,
      "loss": 0.3336,
      "num_input_tokens_seen": 38646976,
      "step": 20390
    },
    {
      "epoch": 3.327106615547761,
      "grad_norm": 0.002641173079609871,
      "learning_rate": 0.14535903761432406,
      "loss": 0.3089,
      "num_input_tokens_seen": 38655984,
      "step": 20395
    },
    {
      "epoch": 3.327922342768578,
      "grad_norm": 0.0010653103236109018,
      "learning_rate": 0.1453001613120527,
      "loss": 0.3216,
      "num_input_tokens_seen": 38664512,
      "step": 20400
    },
    {
      "epoch": 3.327922342768578,
      "eval_loss": 0.3320409655570984,
      "eval_runtime": 152.6829,
      "eval_samples_per_second": 17.847,
      "eval_steps_per_second": 8.927,
      "num_input_tokens_seen": 38664512,
      "step": 20400
    },
    {
      "epoch": 3.3287380699893956,
      "grad_norm": 0.0012798771494999528,
      "learning_rate": 0.14524128573455547,
      "loss": 0.4204,
      "num_input_tokens_seen": 38674416,
      "step": 20405
    },
    {
      "epoch": 3.3295537972102127,
      "grad_norm": 0.0015705039259046316,
      "learning_rate": 0.14518241089091177,
      "loss": 0.3713,
      "num_input_tokens_seen": 38684240,
      "step": 20410
    },
    {
      "epoch": 3.3303695244310303,
      "grad_norm": 0.0032501360401511192,
      "learning_rate": 0.1451235367902009,
      "loss": 0.3013,
      "num_input_tokens_seen": 38693552,
      "step": 20415
    },
    {
      "epoch": 3.3311852516518474,
      "grad_norm": 0.003588455030694604,
      "learning_rate": 0.1450646634415019,
      "loss": 0.3835,
      "num_input_tokens_seen": 38703648,
      "step": 20420
    },
    {
      "epoch": 3.332000978872665,
      "grad_norm": 0.0010941034415736794,
      "learning_rate": 0.1450057908538938,
      "loss": 0.2885,
      "num_input_tokens_seen": 38713408,
      "step": 20425
    },
    {
      "epoch": 3.3328167060934826,
      "grad_norm": 0.0008545118616893888,
      "learning_rate": 0.14494691903645557,
      "loss": 0.3148,
      "num_input_tokens_seen": 38723520,
      "step": 20430
    },
    {
      "epoch": 3.3336324333142997,
      "grad_norm": 0.0009319792152382433,
      "learning_rate": 0.14488804799826588,
      "loss": 0.3122,
      "num_input_tokens_seen": 38732736,
      "step": 20435
    },
    {
      "epoch": 3.334448160535117,
      "grad_norm": 0.0005604307516478002,
      "learning_rate": 0.14482917774840348,
      "loss": 0.3576,
      "num_input_tokens_seen": 38742816,
      "step": 20440
    },
    {
      "epoch": 3.3352638877559344,
      "grad_norm": 0.0016214079223573208,
      "learning_rate": 0.14477030829594684,
      "loss": 0.3008,
      "num_input_tokens_seen": 38752336,
      "step": 20445
    },
    {
      "epoch": 3.336079614976752,
      "grad_norm": 0.0008362341905012727,
      "learning_rate": 0.14471143964997432,
      "loss": 0.3304,
      "num_input_tokens_seen": 38760960,
      "step": 20450
    },
    {
      "epoch": 3.336895342197569,
      "grad_norm": 0.001916474662721157,
      "learning_rate": 0.14465257181956434,
      "loss": 0.3305,
      "num_input_tokens_seen": 38770592,
      "step": 20455
    },
    {
      "epoch": 3.3377110694183862,
      "grad_norm": 0.0009472124511376023,
      "learning_rate": 0.1445937048137949,
      "loss": 0.3131,
      "num_input_tokens_seen": 38780624,
      "step": 20460
    },
    {
      "epoch": 3.338526796639204,
      "grad_norm": 0.0019990303553640842,
      "learning_rate": 0.14453483864174416,
      "loss": 0.3664,
      "num_input_tokens_seen": 38791152,
      "step": 20465
    },
    {
      "epoch": 3.3393425238600214,
      "grad_norm": 0.0020263076294213533,
      "learning_rate": 0.14447597331249,
      "loss": 0.3369,
      "num_input_tokens_seen": 38800016,
      "step": 20470
    },
    {
      "epoch": 3.3401582510808385,
      "grad_norm": 0.0011247978545725346,
      "learning_rate": 0.1444171088351102,
      "loss": 0.3395,
      "num_input_tokens_seen": 38809712,
      "step": 20475
    },
    {
      "epoch": 3.340973978301656,
      "grad_norm": 0.0010774199618026614,
      "learning_rate": 0.14435824521868235,
      "loss": 0.3256,
      "num_input_tokens_seen": 38819456,
      "step": 20480
    },
    {
      "epoch": 3.3417897055224732,
      "grad_norm": 0.0009035525727085769,
      "learning_rate": 0.14429938247228397,
      "loss": 0.3175,
      "num_input_tokens_seen": 38828992,
      "step": 20485
    },
    {
      "epoch": 3.342605432743291,
      "grad_norm": 0.002999478718265891,
      "learning_rate": 0.14424052060499243,
      "loss": 0.3311,
      "num_input_tokens_seen": 38839280,
      "step": 20490
    },
    {
      "epoch": 3.343421159964108,
      "grad_norm": 0.000903166423086077,
      "learning_rate": 0.14418165962588506,
      "loss": 0.3862,
      "num_input_tokens_seen": 38848768,
      "step": 20495
    },
    {
      "epoch": 3.3442368871849255,
      "grad_norm": 0.0012977777514606714,
      "learning_rate": 0.1441227995440388,
      "loss": 0.3623,
      "num_input_tokens_seen": 38858736,
      "step": 20500
    },
    {
      "epoch": 3.3450526144057426,
      "grad_norm": 0.0008700796170160174,
      "learning_rate": 0.14406394036853082,
      "loss": 0.3505,
      "num_input_tokens_seen": 38868000,
      "step": 20505
    },
    {
      "epoch": 3.34586834162656,
      "grad_norm": 0.0012007455807179213,
      "learning_rate": 0.14400508210843774,
      "loss": 0.3122,
      "num_input_tokens_seen": 38879280,
      "step": 20510
    },
    {
      "epoch": 3.3466840688473773,
      "grad_norm": 0.0015608324902132154,
      "learning_rate": 0.1439462247728364,
      "loss": 0.3262,
      "num_input_tokens_seen": 38888928,
      "step": 20515
    },
    {
      "epoch": 3.347499796068195,
      "grad_norm": 0.0014029359444975853,
      "learning_rate": 0.14388736837080326,
      "loss": 0.3875,
      "num_input_tokens_seen": 38898032,
      "step": 20520
    },
    {
      "epoch": 3.348315523289012,
      "grad_norm": 0.0008950070478022099,
      "learning_rate": 0.14382851291141469,
      "loss": 0.2953,
      "num_input_tokens_seen": 38907408,
      "step": 20525
    },
    {
      "epoch": 3.3491312505098296,
      "grad_norm": 0.0006121888291090727,
      "learning_rate": 0.14376965840374697,
      "loss": 0.3272,
      "num_input_tokens_seen": 38916704,
      "step": 20530
    },
    {
      "epoch": 3.3499469777306468,
      "grad_norm": 0.0008863770635798573,
      "learning_rate": 0.14371080485687632,
      "loss": 0.3209,
      "num_input_tokens_seen": 38926144,
      "step": 20535
    },
    {
      "epoch": 3.3507627049514643,
      "grad_norm": 0.0011133089428767562,
      "learning_rate": 0.1436519522798785,
      "loss": 0.3212,
      "num_input_tokens_seen": 38936448,
      "step": 20540
    },
    {
      "epoch": 3.3515784321722815,
      "grad_norm": 0.002462008735165,
      "learning_rate": 0.14359310068182948,
      "loss": 0.3365,
      "num_input_tokens_seen": 38945824,
      "step": 20545
    },
    {
      "epoch": 3.352394159393099,
      "grad_norm": 0.00144184788223356,
      "learning_rate": 0.14353425007180484,
      "loss": 0.3145,
      "num_input_tokens_seen": 38956000,
      "step": 20550
    },
    {
      "epoch": 3.353209886613916,
      "grad_norm": 0.0011370102874934673,
      "learning_rate": 0.14347540045888005,
      "loss": 0.3492,
      "num_input_tokens_seen": 38966576,
      "step": 20555
    },
    {
      "epoch": 3.3540256138347337,
      "grad_norm": 0.0015678508207201958,
      "learning_rate": 0.14341655185213056,
      "loss": 0.3637,
      "num_input_tokens_seen": 38975840,
      "step": 20560
    },
    {
      "epoch": 3.354841341055551,
      "grad_norm": 0.0008127047913148999,
      "learning_rate": 0.14335770426063144,
      "loss": 0.385,
      "num_input_tokens_seen": 38985632,
      "step": 20565
    },
    {
      "epoch": 3.3556570682763684,
      "grad_norm": 0.001080697518773377,
      "learning_rate": 0.1432988576934578,
      "loss": 0.2635,
      "num_input_tokens_seen": 38995632,
      "step": 20570
    },
    {
      "epoch": 3.3564727954971856,
      "grad_norm": 0.0015209512785077095,
      "learning_rate": 0.14324001215968457,
      "loss": 0.2968,
      "num_input_tokens_seen": 39004544,
      "step": 20575
    },
    {
      "epoch": 3.357288522718003,
      "grad_norm": 0.0007269594934768975,
      "learning_rate": 0.14318116766838637,
      "loss": 0.3086,
      "num_input_tokens_seen": 39014480,
      "step": 20580
    },
    {
      "epoch": 3.3581042499388203,
      "grad_norm": 0.0007441504276357591,
      "learning_rate": 0.14312232422863788,
      "loss": 0.3012,
      "num_input_tokens_seen": 39025232,
      "step": 20585
    },
    {
      "epoch": 3.358919977159638,
      "grad_norm": 0.0027126814238727093,
      "learning_rate": 0.14306348184951334,
      "loss": 0.3047,
      "num_input_tokens_seen": 39034768,
      "step": 20590
    },
    {
      "epoch": 3.359735704380455,
      "grad_norm": 0.0012191791320219636,
      "learning_rate": 0.1430046405400871,
      "loss": 0.3418,
      "num_input_tokens_seen": 39044288,
      "step": 20595
    },
    {
      "epoch": 3.3605514316012726,
      "grad_norm": 0.0016102505614981055,
      "learning_rate": 0.14294580030943324,
      "loss": 0.3357,
      "num_input_tokens_seen": 39053472,
      "step": 20600
    },
    {
      "epoch": 3.3605514316012726,
      "eval_loss": 0.3293406069278717,
      "eval_runtime": 152.7037,
      "eval_samples_per_second": 17.845,
      "eval_steps_per_second": 8.926,
      "num_input_tokens_seen": 39053472,
      "step": 20600
    },
    {
      "epoch": 3.36136715882209,
      "grad_norm": 0.001765454187989235,
      "learning_rate": 0.14288696116662553,
      "loss": 0.4348,
      "num_input_tokens_seen": 39062560,
      "step": 20605
    },
    {
      "epoch": 3.3621828860429073,
      "grad_norm": 0.0032300001475960016,
      "learning_rate": 0.1428281231207378,
      "loss": 0.3388,
      "num_input_tokens_seen": 39072608,
      "step": 20610
    },
    {
      "epoch": 3.3629986132637244,
      "grad_norm": 0.001225082203745842,
      "learning_rate": 0.1427692861808437,
      "loss": 0.2611,
      "num_input_tokens_seen": 39081072,
      "step": 20615
    },
    {
      "epoch": 3.363814340484542,
      "grad_norm": 0.0012884795432910323,
      "learning_rate": 0.1427104503560165,
      "loss": 0.3105,
      "num_input_tokens_seen": 39090992,
      "step": 20620
    },
    {
      "epoch": 3.3646300677053596,
      "grad_norm": 0.0014906884171068668,
      "learning_rate": 0.14265161565532947,
      "loss": 0.3158,
      "num_input_tokens_seen": 39101936,
      "step": 20625
    },
    {
      "epoch": 3.3654457949261767,
      "grad_norm": 0.0011706228833645582,
      "learning_rate": 0.14259278208785564,
      "loss": 0.345,
      "num_input_tokens_seen": 39110960,
      "step": 20630
    },
    {
      "epoch": 3.366261522146994,
      "grad_norm": 0.0023107165470719337,
      "learning_rate": 0.14253394966266789,
      "loss": 0.3053,
      "num_input_tokens_seen": 39120912,
      "step": 20635
    },
    {
      "epoch": 3.3670772493678114,
      "grad_norm": 0.0016616693465039134,
      "learning_rate": 0.14247511838883894,
      "loss": 0.3069,
      "num_input_tokens_seen": 39131136,
      "step": 20640
    },
    {
      "epoch": 3.367892976588629,
      "grad_norm": 0.0012411597417667508,
      "learning_rate": 0.14241628827544126,
      "loss": 0.2595,
      "num_input_tokens_seen": 39139072,
      "step": 20645
    },
    {
      "epoch": 3.368708703809446,
      "grad_norm": 0.0010693651856854558,
      "learning_rate": 0.14235745933154723,
      "loss": 0.2979,
      "num_input_tokens_seen": 39148896,
      "step": 20650
    },
    {
      "epoch": 3.3695244310302637,
      "grad_norm": 0.002980442252010107,
      "learning_rate": 0.14229863156622907,
      "loss": 0.3219,
      "num_input_tokens_seen": 39158816,
      "step": 20655
    },
    {
      "epoch": 3.370340158251081,
      "grad_norm": 0.002171647036448121,
      "learning_rate": 0.14223980498855868,
      "loss": 0.3316,
      "num_input_tokens_seen": 39168400,
      "step": 20660
    },
    {
      "epoch": 3.3711558854718984,
      "grad_norm": 0.0014736538287252188,
      "learning_rate": 0.14218097960760792,
      "loss": 0.288,
      "num_input_tokens_seen": 39177680,
      "step": 20665
    },
    {
      "epoch": 3.3719716126927155,
      "grad_norm": 0.00170033797621727,
      "learning_rate": 0.1421221554324483,
      "loss": 0.3071,
      "num_input_tokens_seen": 39187072,
      "step": 20670
    },
    {
      "epoch": 3.372787339913533,
      "grad_norm": 0.004180192481726408,
      "learning_rate": 0.1420633324721513,
      "loss": 0.3779,
      "num_input_tokens_seen": 39197088,
      "step": 20675
    },
    {
      "epoch": 3.37360306713435,
      "grad_norm": 0.0013495474122464657,
      "learning_rate": 0.14200451073578824,
      "loss": 0.3307,
      "num_input_tokens_seen": 39206384,
      "step": 20680
    },
    {
      "epoch": 3.374418794355168,
      "grad_norm": 0.0014134731609374285,
      "learning_rate": 0.14194569023243003,
      "loss": 0.2735,
      "num_input_tokens_seen": 39216672,
      "step": 20685
    },
    {
      "epoch": 3.375234521575985,
      "grad_norm": 0.0021091855596750975,
      "learning_rate": 0.14188687097114766,
      "loss": 0.3073,
      "num_input_tokens_seen": 39227008,
      "step": 20690
    },
    {
      "epoch": 3.3760502487968025,
      "grad_norm": 0.0010977208148688078,
      "learning_rate": 0.14182805296101172,
      "loss": 0.3194,
      "num_input_tokens_seen": 39237568,
      "step": 20695
    },
    {
      "epoch": 3.3768659760176196,
      "grad_norm": 0.001181047991849482,
      "learning_rate": 0.14176923621109272,
      "loss": 0.3283,
      "num_input_tokens_seen": 39246960,
      "step": 20700
    },
    {
      "epoch": 3.377681703238437,
      "grad_norm": 0.002787907375022769,
      "learning_rate": 0.14171042073046097,
      "loss": 0.3604,
      "num_input_tokens_seen": 39256672,
      "step": 20705
    },
    {
      "epoch": 3.3784974304592543,
      "grad_norm": 0.002060172613710165,
      "learning_rate": 0.14165160652818642,
      "loss": 0.3676,
      "num_input_tokens_seen": 39266272,
      "step": 20710
    },
    {
      "epoch": 3.379313157680072,
      "grad_norm": 0.001410796889103949,
      "learning_rate": 0.14159279361333907,
      "loss": 0.3024,
      "num_input_tokens_seen": 39275632,
      "step": 20715
    },
    {
      "epoch": 3.380128884900889,
      "grad_norm": 0.0029943117406219244,
      "learning_rate": 0.14153398199498868,
      "loss": 0.3304,
      "num_input_tokens_seen": 39284144,
      "step": 20720
    },
    {
      "epoch": 3.3809446121217066,
      "grad_norm": 0.002086687134578824,
      "learning_rate": 0.14147517168220458,
      "loss": 0.3354,
      "num_input_tokens_seen": 39294000,
      "step": 20725
    },
    {
      "epoch": 3.3817603393425237,
      "grad_norm": 0.002895097713917494,
      "learning_rate": 0.14141636268405616,
      "loss": 0.3099,
      "num_input_tokens_seen": 39304224,
      "step": 20730
    },
    {
      "epoch": 3.3825760665633413,
      "grad_norm": 0.0034108245745301247,
      "learning_rate": 0.14135755500961253,
      "loss": 0.2719,
      "num_input_tokens_seen": 39313888,
      "step": 20735
    },
    {
      "epoch": 3.3833917937841584,
      "grad_norm": 0.003059527138248086,
      "learning_rate": 0.14129874866794245,
      "loss": 0.3726,
      "num_input_tokens_seen": 39322928,
      "step": 20740
    },
    {
      "epoch": 3.384207521004976,
      "grad_norm": 0.0021514303516596556,
      "learning_rate": 0.14123994366811476,
      "loss": 0.2865,
      "num_input_tokens_seen": 39331728,
      "step": 20745
    },
    {
      "epoch": 3.385023248225793,
      "grad_norm": 0.0017814739840105176,
      "learning_rate": 0.14118114001919774,
      "loss": 0.3244,
      "num_input_tokens_seen": 39340192,
      "step": 20750
    },
    {
      "epoch": 3.3858389754466107,
      "grad_norm": 0.0018109629163518548,
      "learning_rate": 0.14112233773025978,
      "loss": 0.2283,
      "num_input_tokens_seen": 39350560,
      "step": 20755
    },
    {
      "epoch": 3.386654702667428,
      "grad_norm": 0.0011060838587582111,
      "learning_rate": 0.14106353681036896,
      "loss": 0.3451,
      "num_input_tokens_seen": 39360208,
      "step": 20760
    },
    {
      "epoch": 3.3874704298882454,
      "grad_norm": 0.001038233982399106,
      "learning_rate": 0.14100473726859303,
      "loss": 0.3479,
      "num_input_tokens_seen": 39369984,
      "step": 20765
    },
    {
      "epoch": 3.3882861571090626,
      "grad_norm": 0.0008514870423823595,
      "learning_rate": 0.14094593911399964,
      "loss": 0.3171,
      "num_input_tokens_seen": 39377872,
      "step": 20770
    },
    {
      "epoch": 3.38910188432988,
      "grad_norm": 0.002430680673569441,
      "learning_rate": 0.14088714235565625,
      "loss": 0.3401,
      "num_input_tokens_seen": 39386880,
      "step": 20775
    },
    {
      "epoch": 3.3899176115506973,
      "grad_norm": 0.0019261051202192903,
      "learning_rate": 0.14082834700263,
      "loss": 0.2551,
      "num_input_tokens_seen": 39395040,
      "step": 20780
    },
    {
      "epoch": 3.390733338771515,
      "grad_norm": 0.0015857848338782787,
      "learning_rate": 0.14076955306398795,
      "loss": 0.3785,
      "num_input_tokens_seen": 39403504,
      "step": 20785
    },
    {
      "epoch": 3.391549065992332,
      "grad_norm": 0.00311967171728611,
      "learning_rate": 0.14071076054879675,
      "loss": 0.3449,
      "num_input_tokens_seen": 39413680,
      "step": 20790
    },
    {
      "epoch": 3.3923647932131495,
      "grad_norm": 0.003471443196758628,
      "learning_rate": 0.14065196946612302,
      "loss": 0.3394,
      "num_input_tokens_seen": 39423216,
      "step": 20795
    },
    {
      "epoch": 3.393180520433967,
      "grad_norm": 0.00136684556491673,
      "learning_rate": 0.1405931798250331,
      "loss": 0.3281,
      "num_input_tokens_seen": 39432032,
      "step": 20800
    },
    {
      "epoch": 3.393180520433967,
      "eval_loss": 0.3502292335033417,
      "eval_runtime": 152.7425,
      "eval_samples_per_second": 17.84,
      "eval_steps_per_second": 8.924,
      "num_input_tokens_seen": 39432032,
      "step": 20800
    },
    {
      "epoch": 3.3939962476547842,
      "grad_norm": 0.000824782473500818,
      "learning_rate": 0.14053439163459308,
      "loss": 0.2955,
      "num_input_tokens_seen": 39441056,
      "step": 20805
    },
    {
      "epoch": 3.3948119748756014,
      "grad_norm": 0.0032407627440989017,
      "learning_rate": 0.14047560490386876,
      "loss": 0.3362,
      "num_input_tokens_seen": 39450912,
      "step": 20810
    },
    {
      "epoch": 3.395627702096419,
      "grad_norm": 0.0012964539928361773,
      "learning_rate": 0.14041681964192593,
      "loss": 0.3449,
      "num_input_tokens_seen": 39459904,
      "step": 20815
    },
    {
      "epoch": 3.3964434293172365,
      "grad_norm": 0.0031893213745206594,
      "learning_rate": 0.14035803585782988,
      "loss": 0.4227,
      "num_input_tokens_seen": 39469488,
      "step": 20820
    },
    {
      "epoch": 3.3972591565380537,
      "grad_norm": 0.0015267457347363234,
      "learning_rate": 0.14029925356064593,
      "loss": 0.331,
      "num_input_tokens_seen": 39479552,
      "step": 20825
    },
    {
      "epoch": 3.398074883758871,
      "grad_norm": 0.0019856190774589777,
      "learning_rate": 0.1402404727594389,
      "loss": 0.3227,
      "num_input_tokens_seen": 39490016,
      "step": 20830
    },
    {
      "epoch": 3.3988906109796884,
      "grad_norm": 0.0006491913227364421,
      "learning_rate": 0.1401816934632737,
      "loss": 0.3591,
      "num_input_tokens_seen": 39499008,
      "step": 20835
    },
    {
      "epoch": 3.399706338200506,
      "grad_norm": 0.0010616452200338244,
      "learning_rate": 0.1401229156812147,
      "loss": 0.325,
      "num_input_tokens_seen": 39507296,
      "step": 20840
    },
    {
      "epoch": 3.400522065421323,
      "grad_norm": 0.0016943532973527908,
      "learning_rate": 0.14006413942232626,
      "loss": 0.2938,
      "num_input_tokens_seen": 39516480,
      "step": 20845
    },
    {
      "epoch": 3.4013377926421406,
      "grad_norm": 0.0007108802674338222,
      "learning_rate": 0.14000536469567235,
      "loss": 0.3437,
      "num_input_tokens_seen": 39526224,
      "step": 20850
    },
    {
      "epoch": 3.4021535198629578,
      "grad_norm": 0.001596039510332048,
      "learning_rate": 0.13994659151031685,
      "loss": 0.3309,
      "num_input_tokens_seen": 39535840,
      "step": 20855
    },
    {
      "epoch": 3.4029692470837754,
      "grad_norm": 0.002215160522609949,
      "learning_rate": 0.13988781987532323,
      "loss": 0.2867,
      "num_input_tokens_seen": 39544304,
      "step": 20860
    },
    {
      "epoch": 3.4037849743045925,
      "grad_norm": 0.0018214231822639704,
      "learning_rate": 0.1398290497997549,
      "loss": 0.3158,
      "num_input_tokens_seen": 39554368,
      "step": 20865
    },
    {
      "epoch": 3.40460070152541,
      "grad_norm": 0.001806390006095171,
      "learning_rate": 0.13977028129267488,
      "loss": 0.3509,
      "num_input_tokens_seen": 39562480,
      "step": 20870
    },
    {
      "epoch": 3.405416428746227,
      "grad_norm": 0.0016479355981573462,
      "learning_rate": 0.13971151436314605,
      "loss": 0.3093,
      "num_input_tokens_seen": 39572768,
      "step": 20875
    },
    {
      "epoch": 3.4062321559670448,
      "grad_norm": 0.0028203644324094057,
      "learning_rate": 0.13965274902023103,
      "loss": 0.3551,
      "num_input_tokens_seen": 39581968,
      "step": 20880
    },
    {
      "epoch": 3.407047883187862,
      "grad_norm": 0.003758389735594392,
      "learning_rate": 0.13959398527299208,
      "loss": 0.322,
      "num_input_tokens_seen": 39591424,
      "step": 20885
    },
    {
      "epoch": 3.4078636104086795,
      "grad_norm": 0.0011301507474854589,
      "learning_rate": 0.13953522313049138,
      "loss": 0.3158,
      "num_input_tokens_seen": 39601360,
      "step": 20890
    },
    {
      "epoch": 3.4086793376294966,
      "grad_norm": 0.0011124847223982215,
      "learning_rate": 0.13947646260179083,
      "loss": 0.2766,
      "num_input_tokens_seen": 39610528,
      "step": 20895
    },
    {
      "epoch": 3.409495064850314,
      "grad_norm": 0.003872311906889081,
      "learning_rate": 0.13941770369595194,
      "loss": 0.3848,
      "num_input_tokens_seen": 39621712,
      "step": 20900
    },
    {
      "epoch": 3.4103107920711313,
      "grad_norm": 0.0009618629119358957,
      "learning_rate": 0.1393589464220362,
      "loss": 0.3195,
      "num_input_tokens_seen": 39631296,
      "step": 20905
    },
    {
      "epoch": 3.411126519291949,
      "grad_norm": 0.0038153198547661304,
      "learning_rate": 0.13930019078910455,
      "loss": 0.3519,
      "num_input_tokens_seen": 39640144,
      "step": 20910
    },
    {
      "epoch": 3.411942246512766,
      "grad_norm": 0.0020420535001903772,
      "learning_rate": 0.139241436806218,
      "loss": 0.2869,
      "num_input_tokens_seen": 39650128,
      "step": 20915
    },
    {
      "epoch": 3.4127579737335836,
      "grad_norm": 0.0013525893446058035,
      "learning_rate": 0.13918268448243712,
      "loss": 0.3212,
      "num_input_tokens_seen": 39660768,
      "step": 20920
    },
    {
      "epoch": 3.4135737009544007,
      "grad_norm": 0.002342147519811988,
      "learning_rate": 0.13912393382682217,
      "loss": 0.3577,
      "num_input_tokens_seen": 39669472,
      "step": 20925
    },
    {
      "epoch": 3.4143894281752183,
      "grad_norm": 0.0008062539272941649,
      "learning_rate": 0.1390651848484333,
      "loss": 0.3269,
      "num_input_tokens_seen": 39679648,
      "step": 20930
    },
    {
      "epoch": 3.4152051553960354,
      "grad_norm": 0.001985859591513872,
      "learning_rate": 0.1390064375563304,
      "loss": 0.3875,
      "num_input_tokens_seen": 39689360,
      "step": 20935
    },
    {
      "epoch": 3.416020882616853,
      "grad_norm": 0.001172314747236669,
      "learning_rate": 0.13894769195957293,
      "loss": 0.3413,
      "num_input_tokens_seen": 39698112,
      "step": 20940
    },
    {
      "epoch": 3.41683660983767,
      "grad_norm": 0.0012287813005968928,
      "learning_rate": 0.13888894806722032,
      "loss": 0.3147,
      "num_input_tokens_seen": 39707616,
      "step": 20945
    },
    {
      "epoch": 3.4176523370584877,
      "grad_norm": 0.0030150171369314194,
      "learning_rate": 0.1388302058883315,
      "loss": 0.3202,
      "num_input_tokens_seen": 39716448,
      "step": 20950
    },
    {
      "epoch": 3.418468064279305,
      "grad_norm": 0.002338919322937727,
      "learning_rate": 0.13877146543196528,
      "loss": 0.3616,
      "num_input_tokens_seen": 39726512,
      "step": 20955
    },
    {
      "epoch": 3.4192837915001224,
      "grad_norm": 0.0007522573578171432,
      "learning_rate": 0.13871272670718027,
      "loss": 0.3218,
      "num_input_tokens_seen": 39735440,
      "step": 20960
    },
    {
      "epoch": 3.4200995187209395,
      "grad_norm": 0.0023052634205669165,
      "learning_rate": 0.13865398972303455,
      "loss": 0.3404,
      "num_input_tokens_seen": 39744960,
      "step": 20965
    },
    {
      "epoch": 3.420915245941757,
      "grad_norm": 0.002716091461479664,
      "learning_rate": 0.13859525448858623,
      "loss": 0.3421,
      "num_input_tokens_seen": 39755120,
      "step": 20970
    },
    {
      "epoch": 3.4217309731625742,
      "grad_norm": 0.0021020127460360527,
      "learning_rate": 0.13853652101289304,
      "loss": 0.3482,
      "num_input_tokens_seen": 39764496,
      "step": 20975
    },
    {
      "epoch": 3.422546700383392,
      "grad_norm": 0.000962516525760293,
      "learning_rate": 0.13847778930501234,
      "loss": 0.3002,
      "num_input_tokens_seen": 39774352,
      "step": 20980
    },
    {
      "epoch": 3.423362427604209,
      "grad_norm": 0.0020874522160738707,
      "learning_rate": 0.1384190593740013,
      "loss": 0.3076,
      "num_input_tokens_seen": 39785072,
      "step": 20985
    },
    {
      "epoch": 3.4241781548250265,
      "grad_norm": 0.0007369330269284546,
      "learning_rate": 0.13836033122891686,
      "loss": 0.3119,
      "num_input_tokens_seen": 39794304,
      "step": 20990
    },
    {
      "epoch": 3.424993882045844,
      "grad_norm": 0.0012444456806406379,
      "learning_rate": 0.1383016048788156,
      "loss": 0.3362,
      "num_input_tokens_seen": 39803888,
      "step": 20995
    },
    {
      "epoch": 3.4258096092666612,
      "grad_norm": 0.0024004282895475626,
      "learning_rate": 0.13824288033275392,
      "loss": 0.3459,
      "num_input_tokens_seen": 39812704,
      "step": 21000
    },
    {
      "epoch": 3.4258096092666612,
      "eval_loss": 0.322600394487381,
      "eval_runtime": 152.6198,
      "eval_samples_per_second": 17.855,
      "eval_steps_per_second": 8.931,
      "num_input_tokens_seen": 39812704,
      "step": 21000
    },
    {
      "epoch": 3.4266253364874784,
      "grad_norm": 0.0005553975352086127,
      "learning_rate": 0.1381841575997878,
      "loss": 0.296,
      "num_input_tokens_seen": 39822384,
      "step": 21005
    },
    {
      "epoch": 3.427441063708296,
      "grad_norm": 0.0016738579142838717,
      "learning_rate": 0.13812543668897306,
      "loss": 0.3342,
      "num_input_tokens_seen": 39830976,
      "step": 21010
    },
    {
      "epoch": 3.4282567909291135,
      "grad_norm": 0.0010301091242581606,
      "learning_rate": 0.13806671760936526,
      "loss": 0.3715,
      "num_input_tokens_seen": 39841264,
      "step": 21015
    },
    {
      "epoch": 3.4290725181499306,
      "grad_norm": 0.0020342788193374872,
      "learning_rate": 0.13800800037001956,
      "loss": 0.3454,
      "num_input_tokens_seen": 39850496,
      "step": 21020
    },
    {
      "epoch": 3.429888245370748,
      "grad_norm": 0.0015593577409163117,
      "learning_rate": 0.13794928497999087,
      "loss": 0.3167,
      "num_input_tokens_seen": 39858928,
      "step": 21025
    },
    {
      "epoch": 3.4307039725915653,
      "grad_norm": 0.0017030633753165603,
      "learning_rate": 0.1378905714483339,
      "loss": 0.3093,
      "num_input_tokens_seen": 39867760,
      "step": 21030
    },
    {
      "epoch": 3.431519699812383,
      "grad_norm": 0.0015049090143293142,
      "learning_rate": 0.13783185978410295,
      "loss": 0.3905,
      "num_input_tokens_seen": 39877952,
      "step": 21035
    },
    {
      "epoch": 3.4323354270332,
      "grad_norm": 0.0019626785069704056,
      "learning_rate": 0.13777314999635218,
      "loss": 0.3312,
      "num_input_tokens_seen": 39886208,
      "step": 21040
    },
    {
      "epoch": 3.4331511542540176,
      "grad_norm": 0.002565139438956976,
      "learning_rate": 0.1377144420941353,
      "loss": 0.2954,
      "num_input_tokens_seen": 39896464,
      "step": 21045
    },
    {
      "epoch": 3.4339668814748348,
      "grad_norm": 0.0005137577536515892,
      "learning_rate": 0.13765573608650586,
      "loss": 0.3154,
      "num_input_tokens_seen": 39905872,
      "step": 21050
    },
    {
      "epoch": 3.4347826086956523,
      "grad_norm": 0.0028643186669796705,
      "learning_rate": 0.13759703198251702,
      "loss": 0.3281,
      "num_input_tokens_seen": 39915552,
      "step": 21055
    },
    {
      "epoch": 3.4355983359164695,
      "grad_norm": 0.003681630827486515,
      "learning_rate": 0.13753832979122174,
      "loss": 0.2744,
      "num_input_tokens_seen": 39925344,
      "step": 21060
    },
    {
      "epoch": 3.436414063137287,
      "grad_norm": 0.002174057997763157,
      "learning_rate": 0.13747962952167264,
      "loss": 0.3489,
      "num_input_tokens_seen": 39934784,
      "step": 21065
    },
    {
      "epoch": 3.437229790358104,
      "grad_norm": 0.003021369921043515,
      "learning_rate": 0.13742093118292192,
      "loss": 0.3054,
      "num_input_tokens_seen": 39944320,
      "step": 21070
    },
    {
      "epoch": 3.4380455175789217,
      "grad_norm": 0.0013655864167958498,
      "learning_rate": 0.13736223478402174,
      "loss": 0.3391,
      "num_input_tokens_seen": 39952256,
      "step": 21075
    },
    {
      "epoch": 3.438861244799739,
      "grad_norm": 0.0034548190888017416,
      "learning_rate": 0.1373035403340238,
      "loss": 0.3104,
      "num_input_tokens_seen": 39961872,
      "step": 21080
    },
    {
      "epoch": 3.4396769720205564,
      "grad_norm": 0.0009758619708009064,
      "learning_rate": 0.13724484784197943,
      "loss": 0.3617,
      "num_input_tokens_seen": 39970848,
      "step": 21085
    },
    {
      "epoch": 3.4404926992413736,
      "grad_norm": 0.0033887107856571674,
      "learning_rate": 0.13718615731693987,
      "loss": 0.3386,
      "num_input_tokens_seen": 39979856,
      "step": 21090
    },
    {
      "epoch": 3.441308426462191,
      "grad_norm": 0.002228075871244073,
      "learning_rate": 0.13712746876795587,
      "loss": 0.3714,
      "num_input_tokens_seen": 39989072,
      "step": 21095
    },
    {
      "epoch": 3.4421241536830083,
      "grad_norm": 0.001021027099341154,
      "learning_rate": 0.13706878220407792,
      "loss": 0.3423,
      "num_input_tokens_seen": 39999040,
      "step": 21100
    },
    {
      "epoch": 3.442939880903826,
      "grad_norm": 0.0016824731137603521,
      "learning_rate": 0.13701009763435631,
      "loss": 0.3169,
      "num_input_tokens_seen": 40009264,
      "step": 21105
    },
    {
      "epoch": 3.443755608124643,
      "grad_norm": 0.0015251000877469778,
      "learning_rate": 0.13695141506784084,
      "loss": 0.3075,
      "num_input_tokens_seen": 40018992,
      "step": 21110
    },
    {
      "epoch": 3.4445713353454606,
      "grad_norm": 0.0015067339409142733,
      "learning_rate": 0.13689273451358114,
      "loss": 0.3748,
      "num_input_tokens_seen": 40028928,
      "step": 21115
    },
    {
      "epoch": 3.4453870625662777,
      "grad_norm": 0.0019718967378139496,
      "learning_rate": 0.13683405598062653,
      "loss": 0.3087,
      "num_input_tokens_seen": 40040368,
      "step": 21120
    },
    {
      "epoch": 3.4462027897870953,
      "grad_norm": 0.0032549742609262466,
      "learning_rate": 0.1367753794780259,
      "loss": 0.3275,
      "num_input_tokens_seen": 40050640,
      "step": 21125
    },
    {
      "epoch": 3.4470185170079124,
      "grad_norm": 0.0019524238305166364,
      "learning_rate": 0.13671670501482802,
      "loss": 0.3342,
      "num_input_tokens_seen": 40058928,
      "step": 21130
    },
    {
      "epoch": 3.44783424422873,
      "grad_norm": 0.0007436886662617326,
      "learning_rate": 0.1366580326000811,
      "loss": 0.3097,
      "num_input_tokens_seen": 40069056,
      "step": 21135
    },
    {
      "epoch": 3.448649971449547,
      "grad_norm": 0.0005688608507625759,
      "learning_rate": 0.1365993622428332,
      "loss": 0.3144,
      "num_input_tokens_seen": 40078896,
      "step": 21140
    },
    {
      "epoch": 3.4494656986703647,
      "grad_norm": 0.0029502888210117817,
      "learning_rate": 0.13654069395213211,
      "loss": 0.3179,
      "num_input_tokens_seen": 40088080,
      "step": 21145
    },
    {
      "epoch": 3.450281425891182,
      "grad_norm": 0.0020199341233819723,
      "learning_rate": 0.13648202773702509,
      "loss": 0.3835,
      "num_input_tokens_seen": 40097264,
      "step": 21150
    },
    {
      "epoch": 3.4510971531119994,
      "grad_norm": 0.00351854064501822,
      "learning_rate": 0.13642336360655927,
      "loss": 0.3545,
      "num_input_tokens_seen": 40106448,
      "step": 21155
    },
    {
      "epoch": 3.4519128803328165,
      "grad_norm": 0.0020587611943483353,
      "learning_rate": 0.13636470156978145,
      "loss": 0.326,
      "num_input_tokens_seen": 40115712,
      "step": 21160
    },
    {
      "epoch": 3.452728607553634,
      "grad_norm": 0.0009676878107711673,
      "learning_rate": 0.13630604163573798,
      "loss": 0.3484,
      "num_input_tokens_seen": 40125312,
      "step": 21165
    },
    {
      "epoch": 3.4535443347744517,
      "grad_norm": 0.0006443327874876559,
      "learning_rate": 0.13624738381347495,
      "loss": 0.3874,
      "num_input_tokens_seen": 40135392,
      "step": 21170
    },
    {
      "epoch": 3.454360061995269,
      "grad_norm": 0.00110877794213593,
      "learning_rate": 0.1361887281120382,
      "loss": 0.3288,
      "num_input_tokens_seen": 40145232,
      "step": 21175
    },
    {
      "epoch": 3.455175789216086,
      "grad_norm": 0.0014728239038959146,
      "learning_rate": 0.13613007454047307,
      "loss": 0.345,
      "num_input_tokens_seen": 40154416,
      "step": 21180
    },
    {
      "epoch": 3.4559915164369035,
      "grad_norm": 0.002803630894050002,
      "learning_rate": 0.13607142310782486,
      "loss": 0.3584,
      "num_input_tokens_seen": 40164064,
      "step": 21185
    },
    {
      "epoch": 3.456807243657721,
      "grad_norm": 0.001741896034218371,
      "learning_rate": 0.13601277382313814,
      "loss": 0.3217,
      "num_input_tokens_seen": 40172528,
      "step": 21190
    },
    {
      "epoch": 3.457622970878538,
      "grad_norm": 0.0014588999329134822,
      "learning_rate": 0.1359541266954575,
      "loss": 0.3716,
      "num_input_tokens_seen": 40181392,
      "step": 21195
    },
    {
      "epoch": 3.4584386980993553,
      "grad_norm": 0.0017119516851380467,
      "learning_rate": 0.13589548173382707,
      "loss": 0.3314,
      "num_input_tokens_seen": 40191088,
      "step": 21200
    },
    {
      "epoch": 3.4584386980993553,
      "eval_loss": 0.32770687341690063,
      "eval_runtime": 152.7178,
      "eval_samples_per_second": 17.843,
      "eval_steps_per_second": 8.925,
      "num_input_tokens_seen": 40191088,
      "step": 21200
    },
    {
      "epoch": 3.459254425320173,
      "grad_norm": 0.0015161341289058328,
      "learning_rate": 0.1358368389472906,
      "loss": 0.3834,
      "num_input_tokens_seen": 40200864,
      "step": 21205
    },
    {
      "epoch": 3.4600701525409905,
      "grad_norm": 0.0022740946151316166,
      "learning_rate": 0.13577819834489155,
      "loss": 0.3936,
      "num_input_tokens_seen": 40210288,
      "step": 21210
    },
    {
      "epoch": 3.4608858797618076,
      "grad_norm": 0.0022624521516263485,
      "learning_rate": 0.135719559935673,
      "loss": 0.3312,
      "num_input_tokens_seen": 40219248,
      "step": 21215
    },
    {
      "epoch": 3.461701606982625,
      "grad_norm": 0.0021273482125252485,
      "learning_rate": 0.13566092372867775,
      "loss": 0.3066,
      "num_input_tokens_seen": 40229440,
      "step": 21220
    },
    {
      "epoch": 3.4625173342034423,
      "grad_norm": 0.0004383650957606733,
      "learning_rate": 0.13560228973294833,
      "loss": 0.3315,
      "num_input_tokens_seen": 40238384,
      "step": 21225
    },
    {
      "epoch": 3.46333306142426,
      "grad_norm": 0.0016341402661055326,
      "learning_rate": 0.13554365795752668,
      "loss": 0.3314,
      "num_input_tokens_seen": 40247120,
      "step": 21230
    },
    {
      "epoch": 3.464148788645077,
      "grad_norm": 0.0021341396495699883,
      "learning_rate": 0.1354850284114547,
      "loss": 0.3709,
      "num_input_tokens_seen": 40255712,
      "step": 21235
    },
    {
      "epoch": 3.4649645158658946,
      "grad_norm": 0.0027394501958042383,
      "learning_rate": 0.13542640110377374,
      "loss": 0.4263,
      "num_input_tokens_seen": 40263872,
      "step": 21240
    },
    {
      "epoch": 3.4657802430867117,
      "grad_norm": 0.0009520603343844414,
      "learning_rate": 0.13536777604352487,
      "loss": 0.3351,
      "num_input_tokens_seen": 40273984,
      "step": 21245
    },
    {
      "epoch": 3.4665959703075293,
      "grad_norm": 0.0019385816995054483,
      "learning_rate": 0.13530915323974887,
      "loss": 0.2617,
      "num_input_tokens_seen": 40283440,
      "step": 21250
    },
    {
      "epoch": 3.4674116975283464,
      "grad_norm": 0.002061351900920272,
      "learning_rate": 0.13525053270148596,
      "loss": 0.3652,
      "num_input_tokens_seen": 40292464,
      "step": 21255
    },
    {
      "epoch": 3.468227424749164,
      "grad_norm": 0.002047430258244276,
      "learning_rate": 0.13519191443777628,
      "loss": 0.2733,
      "num_input_tokens_seen": 40302384,
      "step": 21260
    },
    {
      "epoch": 3.469043151969981,
      "grad_norm": 0.0014777849428355694,
      "learning_rate": 0.13513329845765953,
      "loss": 0.4391,
      "num_input_tokens_seen": 40311024,
      "step": 21265
    },
    {
      "epoch": 3.4698588791907987,
      "grad_norm": 0.0014671551762148738,
      "learning_rate": 0.13507468477017495,
      "loss": 0.3075,
      "num_input_tokens_seen": 40320672,
      "step": 21270
    },
    {
      "epoch": 3.470674606411616,
      "grad_norm": 0.002332135336473584,
      "learning_rate": 0.13501607338436153,
      "loss": 0.3404,
      "num_input_tokens_seen": 40330720,
      "step": 21275
    },
    {
      "epoch": 3.4714903336324334,
      "grad_norm": 0.001846630941145122,
      "learning_rate": 0.13495746430925798,
      "loss": 0.3641,
      "num_input_tokens_seen": 40339824,
      "step": 21280
    },
    {
      "epoch": 3.4723060608532506,
      "grad_norm": 0.001147815608419478,
      "learning_rate": 0.13489885755390238,
      "loss": 0.3283,
      "num_input_tokens_seen": 40350720,
      "step": 21285
    },
    {
      "epoch": 3.473121788074068,
      "grad_norm": 0.0008716462762095034,
      "learning_rate": 0.13484025312733275,
      "loss": 0.235,
      "num_input_tokens_seen": 40358608,
      "step": 21290
    },
    {
      "epoch": 3.4739375152948853,
      "grad_norm": 0.004013642203062773,
      "learning_rate": 0.13478165103858658,
      "loss": 0.3328,
      "num_input_tokens_seen": 40368336,
      "step": 21295
    },
    {
      "epoch": 3.474753242515703,
      "grad_norm": 0.002905050991103053,
      "learning_rate": 0.13472305129670106,
      "loss": 0.3596,
      "num_input_tokens_seen": 40377904,
      "step": 21300
    },
    {
      "epoch": 3.47556896973652,
      "grad_norm": 0.002074390184134245,
      "learning_rate": 0.13466445391071305,
      "loss": 0.3536,
      "num_input_tokens_seen": 40387920,
      "step": 21305
    },
    {
      "epoch": 3.4763846969573375,
      "grad_norm": 0.0008377204067073762,
      "learning_rate": 0.13460585888965895,
      "loss": 0.2852,
      "num_input_tokens_seen": 40396672,
      "step": 21310
    },
    {
      "epoch": 3.4772004241781547,
      "grad_norm": 0.0016532978042960167,
      "learning_rate": 0.13454726624257482,
      "loss": 0.3065,
      "num_input_tokens_seen": 40407200,
      "step": 21315
    },
    {
      "epoch": 3.4780161513989722,
      "grad_norm": 0.001809928216971457,
      "learning_rate": 0.1344886759784965,
      "loss": 0.2869,
      "num_input_tokens_seen": 40416112,
      "step": 21320
    },
    {
      "epoch": 3.4788318786197894,
      "grad_norm": 0.0012369026662781835,
      "learning_rate": 0.13443008810645923,
      "loss": 0.3433,
      "num_input_tokens_seen": 40425072,
      "step": 21325
    },
    {
      "epoch": 3.479647605840607,
      "grad_norm": 0.001977680716663599,
      "learning_rate": 0.13437150263549807,
      "loss": 0.3328,
      "num_input_tokens_seen": 40434944,
      "step": 21330
    },
    {
      "epoch": 3.480463333061424,
      "grad_norm": 0.0005474185454659164,
      "learning_rate": 0.13431291957464755,
      "loss": 0.3389,
      "num_input_tokens_seen": 40445120,
      "step": 21335
    },
    {
      "epoch": 3.4812790602822417,
      "grad_norm": 0.001147269969806075,
      "learning_rate": 0.13425433893294197,
      "loss": 0.2626,
      "num_input_tokens_seen": 40453696,
      "step": 21340
    },
    {
      "epoch": 3.482094787503059,
      "grad_norm": 0.0007949058781377971,
      "learning_rate": 0.13419576071941525,
      "loss": 0.4085,
      "num_input_tokens_seen": 40463536,
      "step": 21345
    },
    {
      "epoch": 3.4829105147238764,
      "grad_norm": 0.0011747806565836072,
      "learning_rate": 0.1341371849431008,
      "loss": 0.3729,
      "num_input_tokens_seen": 40474464,
      "step": 21350
    },
    {
      "epoch": 3.4837262419446935,
      "grad_norm": 0.0017259630840271711,
      "learning_rate": 0.13407861161303178,
      "loss": 0.3335,
      "num_input_tokens_seen": 40484384,
      "step": 21355
    },
    {
      "epoch": 3.484541969165511,
      "grad_norm": 0.0022580658551305532,
      "learning_rate": 0.13402004073824098,
      "loss": 0.3821,
      "num_input_tokens_seen": 40493088,
      "step": 21360
    },
    {
      "epoch": 3.4853576963863286,
      "grad_norm": 0.003036948386579752,
      "learning_rate": 0.13396147232776062,
      "loss": 0.382,
      "num_input_tokens_seen": 40501968,
      "step": 21365
    },
    {
      "epoch": 3.4861734236071458,
      "grad_norm": 0.0017444804543629289,
      "learning_rate": 0.13390290639062288,
      "loss": 0.314,
      "num_input_tokens_seen": 40511776,
      "step": 21370
    },
    {
      "epoch": 3.486989150827963,
      "grad_norm": 0.0012476976262405515,
      "learning_rate": 0.13384434293585917,
      "loss": 0.3735,
      "num_input_tokens_seen": 40522000,
      "step": 21375
    },
    {
      "epoch": 3.4878048780487805,
      "grad_norm": 0.002340944716706872,
      "learning_rate": 0.13378578197250088,
      "loss": 0.343,
      "num_input_tokens_seen": 40531440,
      "step": 21380
    },
    {
      "epoch": 3.488620605269598,
      "grad_norm": 0.0017004274995997548,
      "learning_rate": 0.13372722350957872,
      "loss": 0.3447,
      "num_input_tokens_seen": 40540752,
      "step": 21385
    },
    {
      "epoch": 3.489436332490415,
      "grad_norm": 0.00199405406601727,
      "learning_rate": 0.13366866755612322,
      "loss": 0.3246,
      "num_input_tokens_seen": 40550080,
      "step": 21390
    },
    {
      "epoch": 3.4902520597112328,
      "grad_norm": 0.00048172997776418924,
      "learning_rate": 0.13361011412116436,
      "loss": 0.3083,
      "num_input_tokens_seen": 40559120,
      "step": 21395
    },
    {
      "epoch": 3.49106778693205,
      "grad_norm": 0.002043427200987935,
      "learning_rate": 0.13355156321373196,
      "loss": 0.422,
      "num_input_tokens_seen": 40567216,
      "step": 21400
    },
    {
      "epoch": 3.49106778693205,
      "eval_loss": 0.32822924852371216,
      "eval_runtime": 152.9647,
      "eval_samples_per_second": 17.815,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 40567216,
      "step": 21400
    },
    {
      "epoch": 3.4918835141528675,
      "grad_norm": 0.0011454281629994512,
      "learning_rate": 0.13349301484285514,
      "loss": 0.3195,
      "num_input_tokens_seen": 40577056,
      "step": 21405
    },
    {
      "epoch": 3.4926992413736846,
      "grad_norm": 0.0018644172232598066,
      "learning_rate": 0.13343446901756295,
      "loss": 0.318,
      "num_input_tokens_seen": 40585776,
      "step": 21410
    },
    {
      "epoch": 3.493514968594502,
      "grad_norm": 0.0015371468616649508,
      "learning_rate": 0.13337592574688376,
      "loss": 0.3037,
      "num_input_tokens_seen": 40594320,
      "step": 21415
    },
    {
      "epoch": 3.4943306958153193,
      "grad_norm": 0.001601106021553278,
      "learning_rate": 0.13331738503984572,
      "loss": 0.3572,
      "num_input_tokens_seen": 40602976,
      "step": 21420
    },
    {
      "epoch": 3.495146423036137,
      "grad_norm": 0.0015010049100965261,
      "learning_rate": 0.1332588469054766,
      "loss": 0.3772,
      "num_input_tokens_seen": 40613200,
      "step": 21425
    },
    {
      "epoch": 3.495962150256954,
      "grad_norm": 0.0019368692301213741,
      "learning_rate": 0.1332003113528036,
      "loss": 0.3254,
      "num_input_tokens_seen": 40622128,
      "step": 21430
    },
    {
      "epoch": 3.4967778774777716,
      "grad_norm": 0.00048187284846790135,
      "learning_rate": 0.13314177839085373,
      "loss": 0.3165,
      "num_input_tokens_seen": 40631936,
      "step": 21435
    },
    {
      "epoch": 3.4975936046985887,
      "grad_norm": 0.0007507581613026559,
      "learning_rate": 0.13308324802865354,
      "loss": 0.35,
      "num_input_tokens_seen": 40641776,
      "step": 21440
    },
    {
      "epoch": 3.4984093319194063,
      "grad_norm": 0.002346765948459506,
      "learning_rate": 0.13302472027522905,
      "loss": 0.3243,
      "num_input_tokens_seen": 40651632,
      "step": 21445
    },
    {
      "epoch": 3.4992250591402234,
      "grad_norm": 0.0014380980283021927,
      "learning_rate": 0.13296619513960606,
      "loss": 0.3267,
      "num_input_tokens_seen": 40661584,
      "step": 21450
    },
    {
      "epoch": 3.500040786361041,
      "grad_norm": 0.0004583485424518585,
      "learning_rate": 0.1329076726308098,
      "loss": 0.3376,
      "num_input_tokens_seen": 40671120,
      "step": 21455
    },
    {
      "epoch": 3.500856513581858,
      "grad_norm": 0.0011860700324177742,
      "learning_rate": 0.13284915275786519,
      "loss": 0.3671,
      "num_input_tokens_seen": 40679456,
      "step": 21460
    },
    {
      "epoch": 3.5016722408026757,
      "grad_norm": 0.0023049446754157543,
      "learning_rate": 0.1327906355297968,
      "loss": 0.3518,
      "num_input_tokens_seen": 40688752,
      "step": 21465
    },
    {
      "epoch": 3.502487968023493,
      "grad_norm": 0.0006128916866146028,
      "learning_rate": 0.13273212095562867,
      "loss": 0.3578,
      "num_input_tokens_seen": 40698112,
      "step": 21470
    },
    {
      "epoch": 3.5033036952443104,
      "grad_norm": 0.0010280560236424208,
      "learning_rate": 0.13267360904438444,
      "loss": 0.2918,
      "num_input_tokens_seen": 40707664,
      "step": 21475
    },
    {
      "epoch": 3.5041194224651275,
      "grad_norm": 0.0005823133978992701,
      "learning_rate": 0.1326150998050875,
      "loss": 0.333,
      "num_input_tokens_seen": 40716528,
      "step": 21480
    },
    {
      "epoch": 3.504935149685945,
      "grad_norm": 0.0021871880162507296,
      "learning_rate": 0.1325565932467606,
      "loss": 0.3362,
      "num_input_tokens_seen": 40725200,
      "step": 21485
    },
    {
      "epoch": 3.5057508769067622,
      "grad_norm": 0.0015569949755445123,
      "learning_rate": 0.13249808937842628,
      "loss": 0.3431,
      "num_input_tokens_seen": 40735200,
      "step": 21490
    },
    {
      "epoch": 3.50656660412758,
      "grad_norm": 0.002679965225979686,
      "learning_rate": 0.1324395882091065,
      "loss": 0.3455,
      "num_input_tokens_seen": 40745008,
      "step": 21495
    },
    {
      "epoch": 3.507382331348397,
      "grad_norm": 0.001385559793561697,
      "learning_rate": 0.13238108974782284,
      "loss": 0.3496,
      "num_input_tokens_seen": 40755216,
      "step": 21500
    },
    {
      "epoch": 3.5081980585692145,
      "grad_norm": 0.0013344889739528298,
      "learning_rate": 0.13232259400359664,
      "loss": 0.2746,
      "num_input_tokens_seen": 40765392,
      "step": 21505
    },
    {
      "epoch": 3.5090137857900316,
      "grad_norm": 0.0014374643797054887,
      "learning_rate": 0.13226410098544852,
      "loss": 0.3123,
      "num_input_tokens_seen": 40775392,
      "step": 21510
    },
    {
      "epoch": 3.5098295130108492,
      "grad_norm": 0.0007713611703366041,
      "learning_rate": 0.13220561070239892,
      "loss": 0.4121,
      "num_input_tokens_seen": 40784400,
      "step": 21515
    },
    {
      "epoch": 3.510645240231667,
      "grad_norm": 0.00239378004334867,
      "learning_rate": 0.13214712316346783,
      "loss": 0.2924,
      "num_input_tokens_seen": 40794048,
      "step": 21520
    },
    {
      "epoch": 3.511460967452484,
      "grad_norm": 0.0015462951269000769,
      "learning_rate": 0.13208863837767465,
      "loss": 0.319,
      "num_input_tokens_seen": 40803344,
      "step": 21525
    },
    {
      "epoch": 3.512276694673301,
      "grad_norm": 0.0003169701958540827,
      "learning_rate": 0.13203015635403856,
      "loss": 0.3117,
      "num_input_tokens_seen": 40812688,
      "step": 21530
    },
    {
      "epoch": 3.5130924218941186,
      "grad_norm": 0.0015351480105891824,
      "learning_rate": 0.13197167710157817,
      "loss": 0.3004,
      "num_input_tokens_seen": 40822288,
      "step": 21535
    },
    {
      "epoch": 3.513908149114936,
      "grad_norm": 0.0023046559654176235,
      "learning_rate": 0.13191320062931167,
      "loss": 0.3654,
      "num_input_tokens_seen": 40832272,
      "step": 21540
    },
    {
      "epoch": 3.5147238763357533,
      "grad_norm": 0.0008767014951445162,
      "learning_rate": 0.13185472694625702,
      "loss": 0.3333,
      "num_input_tokens_seen": 40842704,
      "step": 21545
    },
    {
      "epoch": 3.5155396035565705,
      "grad_norm": 0.0016622188268229365,
      "learning_rate": 0.13179625606143142,
      "loss": 0.3455,
      "num_input_tokens_seen": 40852064,
      "step": 21550
    },
    {
      "epoch": 3.516355330777388,
      "grad_norm": 0.0005137319676578045,
      "learning_rate": 0.13173778798385188,
      "loss": 0.342,
      "num_input_tokens_seen": 40861936,
      "step": 21555
    },
    {
      "epoch": 3.5171710579982056,
      "grad_norm": 0.001592441345565021,
      "learning_rate": 0.13167932272253505,
      "loss": 0.3665,
      "num_input_tokens_seen": 40871056,
      "step": 21560
    },
    {
      "epoch": 3.5179867852190227,
      "grad_norm": 0.0018430331256240606,
      "learning_rate": 0.1316208602864968,
      "loss": 0.3292,
      "num_input_tokens_seen": 40881088,
      "step": 21565
    },
    {
      "epoch": 3.51880251243984,
      "grad_norm": 0.001572383800521493,
      "learning_rate": 0.13156240068475292,
      "loss": 0.3356,
      "num_input_tokens_seen": 40889360,
      "step": 21570
    },
    {
      "epoch": 3.5196182396606575,
      "grad_norm": 0.0023521860130131245,
      "learning_rate": 0.1315039439263185,
      "loss": 0.3652,
      "num_input_tokens_seen": 40900048,
      "step": 21575
    },
    {
      "epoch": 3.520433966881475,
      "grad_norm": 0.002089299727231264,
      "learning_rate": 0.13144549002020833,
      "loss": 0.3644,
      "num_input_tokens_seen": 40909552,
      "step": 21580
    },
    {
      "epoch": 3.521249694102292,
      "grad_norm": 0.0008530145278200507,
      "learning_rate": 0.13138703897543688,
      "loss": 0.2882,
      "num_input_tokens_seen": 40917760,
      "step": 21585
    },
    {
      "epoch": 3.5220654213231093,
      "grad_norm": 0.0019527393160387874,
      "learning_rate": 0.1313285908010178,
      "loss": 0.3108,
      "num_input_tokens_seen": 40927936,
      "step": 21590
    },
    {
      "epoch": 3.522881148543927,
      "grad_norm": 0.0028816356789320707,
      "learning_rate": 0.13127014550596475,
      "loss": 0.3521,
      "num_input_tokens_seen": 40938016,
      "step": 21595
    },
    {
      "epoch": 3.5236968757647444,
      "grad_norm": 0.0014776792377233505,
      "learning_rate": 0.1312117030992906,
      "loss": 0.3273,
      "num_input_tokens_seen": 40947696,
      "step": 21600
    },
    {
      "epoch": 3.5236968757647444,
      "eval_loss": 0.3568617105484009,
      "eval_runtime": 152.9243,
      "eval_samples_per_second": 17.819,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 40947696,
      "step": 21600
    },
    {
      "epoch": 3.5245126029855616,
      "grad_norm": 0.0016393144614994526,
      "learning_rate": 0.13115326359000795,
      "loss": 0.3494,
      "num_input_tokens_seen": 40956576,
      "step": 21605
    },
    {
      "epoch": 3.525328330206379,
      "grad_norm": 0.002987222746014595,
      "learning_rate": 0.13109482698712896,
      "loss": 0.3547,
      "num_input_tokens_seen": 40965040,
      "step": 21610
    },
    {
      "epoch": 3.5261440574271963,
      "grad_norm": 0.004911592695862055,
      "learning_rate": 0.1310363932996651,
      "loss": 0.3668,
      "num_input_tokens_seen": 40974672,
      "step": 21615
    },
    {
      "epoch": 3.526959784648014,
      "grad_norm": 0.0013463489012792706,
      "learning_rate": 0.13097796253662775,
      "loss": 0.3194,
      "num_input_tokens_seen": 40983024,
      "step": 21620
    },
    {
      "epoch": 3.527775511868831,
      "grad_norm": 0.0004133988404646516,
      "learning_rate": 0.1309195347070277,
      "loss": 0.2992,
      "num_input_tokens_seen": 40992432,
      "step": 21625
    },
    {
      "epoch": 3.5285912390896486,
      "grad_norm": 0.001380839734338224,
      "learning_rate": 0.13086110981987506,
      "loss": 0.3244,
      "num_input_tokens_seen": 41002400,
      "step": 21630
    },
    {
      "epoch": 3.5294069663104657,
      "grad_norm": 0.002523483708500862,
      "learning_rate": 0.13080268788417987,
      "loss": 0.347,
      "num_input_tokens_seen": 41010976,
      "step": 21635
    },
    {
      "epoch": 3.5302226935312833,
      "grad_norm": 0.0018452556105330586,
      "learning_rate": 0.1307442689089515,
      "loss": 0.3362,
      "num_input_tokens_seen": 41020224,
      "step": 21640
    },
    {
      "epoch": 3.5310384207521004,
      "grad_norm": 0.0015152550768107176,
      "learning_rate": 0.13068585290319873,
      "loss": 0.2679,
      "num_input_tokens_seen": 41029936,
      "step": 21645
    },
    {
      "epoch": 3.531854147972918,
      "grad_norm": 0.002553661586716771,
      "learning_rate": 0.13062743987593026,
      "loss": 0.3514,
      "num_input_tokens_seen": 41039152,
      "step": 21650
    },
    {
      "epoch": 3.532669875193735,
      "grad_norm": 0.001700489199720323,
      "learning_rate": 0.13056902983615395,
      "loss": 0.3433,
      "num_input_tokens_seen": 41048704,
      "step": 21655
    },
    {
      "epoch": 3.5334856024145527,
      "grad_norm": 0.0020460430532693863,
      "learning_rate": 0.13051062279287742,
      "loss": 0.3094,
      "num_input_tokens_seen": 41059040,
      "step": 21660
    },
    {
      "epoch": 3.53430132963537,
      "grad_norm": 0.001314252964220941,
      "learning_rate": 0.13045221875510782,
      "loss": 0.3617,
      "num_input_tokens_seen": 41068160,
      "step": 21665
    },
    {
      "epoch": 3.5351170568561874,
      "grad_norm": 0.0031172060407698154,
      "learning_rate": 0.13039381773185174,
      "loss": 0.3383,
      "num_input_tokens_seen": 41077872,
      "step": 21670
    },
    {
      "epoch": 3.5359327840770045,
      "grad_norm": 0.0007103524403646588,
      "learning_rate": 0.1303354197321153,
      "loss": 0.3233,
      "num_input_tokens_seen": 41088320,
      "step": 21675
    },
    {
      "epoch": 3.536748511297822,
      "grad_norm": 0.002117183757945895,
      "learning_rate": 0.13027702476490433,
      "loss": 0.2942,
      "num_input_tokens_seen": 41098608,
      "step": 21680
    },
    {
      "epoch": 3.537564238518639,
      "grad_norm": 0.0009481548913754523,
      "learning_rate": 0.1302186328392239,
      "loss": 0.3105,
      "num_input_tokens_seen": 41108528,
      "step": 21685
    },
    {
      "epoch": 3.538379965739457,
      "grad_norm": 0.0009720821399241686,
      "learning_rate": 0.130160243964079,
      "loss": 0.2875,
      "num_input_tokens_seen": 41118208,
      "step": 21690
    },
    {
      "epoch": 3.539195692960274,
      "grad_norm": 0.001275290036574006,
      "learning_rate": 0.13010185814847372,
      "loss": 0.3417,
      "num_input_tokens_seen": 41127776,
      "step": 21695
    },
    {
      "epoch": 3.5400114201810915,
      "grad_norm": 0.0027775801718235016,
      "learning_rate": 0.13004347540141192,
      "loss": 0.474,
      "num_input_tokens_seen": 41138176,
      "step": 21700
    },
    {
      "epoch": 3.5408271474019086,
      "grad_norm": 0.0006419835845008492,
      "learning_rate": 0.12998509573189712,
      "loss": 0.3708,
      "num_input_tokens_seen": 41147520,
      "step": 21705
    },
    {
      "epoch": 3.541642874622726,
      "grad_norm": 0.001323764561675489,
      "learning_rate": 0.12992671914893203,
      "loss": 0.3462,
      "num_input_tokens_seen": 41156560,
      "step": 21710
    },
    {
      "epoch": 3.5424586018435438,
      "grad_norm": 0.0013225845759734511,
      "learning_rate": 0.12986834566151909,
      "loss": 0.3722,
      "num_input_tokens_seen": 41166816,
      "step": 21715
    },
    {
      "epoch": 3.543274329064361,
      "grad_norm": 0.0011284787906333804,
      "learning_rate": 0.12980997527866028,
      "loss": 0.3056,
      "num_input_tokens_seen": 41176576,
      "step": 21720
    },
    {
      "epoch": 3.544090056285178,
      "grad_norm": 0.001627498073503375,
      "learning_rate": 0.12975160800935692,
      "loss": 0.323,
      "num_input_tokens_seen": 41184880,
      "step": 21725
    },
    {
      "epoch": 3.5449057835059956,
      "grad_norm": 0.0008115966920740902,
      "learning_rate": 0.12969324386261016,
      "loss": 0.2973,
      "num_input_tokens_seen": 41195472,
      "step": 21730
    },
    {
      "epoch": 3.545721510726813,
      "grad_norm": 0.0014096249360591173,
      "learning_rate": 0.12963488284742034,
      "loss": 0.3588,
      "num_input_tokens_seen": 41204752,
      "step": 21735
    },
    {
      "epoch": 3.5465372379476303,
      "grad_norm": 0.002141540637239814,
      "learning_rate": 0.12957652497278752,
      "loss": 0.3431,
      "num_input_tokens_seen": 41213568,
      "step": 21740
    },
    {
      "epoch": 3.5473529651684474,
      "grad_norm": 0.003271761815994978,
      "learning_rate": 0.12951817024771117,
      "loss": 0.3061,
      "num_input_tokens_seen": 41224032,
      "step": 21745
    },
    {
      "epoch": 3.548168692389265,
      "grad_norm": 0.0015043345047160983,
      "learning_rate": 0.12945981868119041,
      "loss": 0.3231,
      "num_input_tokens_seen": 41233216,
      "step": 21750
    },
    {
      "epoch": 3.5489844196100826,
      "grad_norm": 0.0018960246816277504,
      "learning_rate": 0.12940147028222376,
      "loss": 0.3035,
      "num_input_tokens_seen": 41242960,
      "step": 21755
    },
    {
      "epoch": 3.5498001468308997,
      "grad_norm": 0.001543715363368392,
      "learning_rate": 0.12934312505980916,
      "loss": 0.3097,
      "num_input_tokens_seen": 41252256,
      "step": 21760
    },
    {
      "epoch": 3.550615874051717,
      "grad_norm": 0.0021722777746617794,
      "learning_rate": 0.1292847830229443,
      "loss": 0.3577,
      "num_input_tokens_seen": 41260336,
      "step": 21765
    },
    {
      "epoch": 3.5514316012725344,
      "grad_norm": 0.0012311430182307959,
      "learning_rate": 0.12922644418062626,
      "loss": 0.3079,
      "num_input_tokens_seen": 41270464,
      "step": 21770
    },
    {
      "epoch": 3.552247328493352,
      "grad_norm": 0.001337196328677237,
      "learning_rate": 0.1291681085418515,
      "loss": 0.2913,
      "num_input_tokens_seen": 41280432,
      "step": 21775
    },
    {
      "epoch": 3.553063055714169,
      "grad_norm": 0.0006020268192514777,
      "learning_rate": 0.12910977611561628,
      "loss": 0.3522,
      "num_input_tokens_seen": 41289616,
      "step": 21780
    },
    {
      "epoch": 3.5538787829349863,
      "grad_norm": 0.0012338581727817655,
      "learning_rate": 0.1290514469109161,
      "loss": 0.2996,
      "num_input_tokens_seen": 41300272,
      "step": 21785
    },
    {
      "epoch": 3.554694510155804,
      "grad_norm": 0.0015776855871081352,
      "learning_rate": 0.128993120936746,
      "loss": 0.3291,
      "num_input_tokens_seen": 41310352,
      "step": 21790
    },
    {
      "epoch": 3.5555102373766214,
      "grad_norm": 0.0011975447414442897,
      "learning_rate": 0.12893479820210071,
      "loss": 0.3263,
      "num_input_tokens_seen": 41320816,
      "step": 21795
    },
    {
      "epoch": 3.5563259645974385,
      "grad_norm": 0.0005237573641352355,
      "learning_rate": 0.1288764787159742,
      "loss": 0.2933,
      "num_input_tokens_seen": 41330624,
      "step": 21800
    },
    {
      "epoch": 3.5563259645974385,
      "eval_loss": 0.3223639726638794,
      "eval_runtime": 152.9137,
      "eval_samples_per_second": 17.821,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 41330624,
      "step": 21800
    },
    {
      "epoch": 3.557141691818256,
      "grad_norm": 0.0028855318669229746,
      "learning_rate": 0.1288181624873601,
      "loss": 0.3039,
      "num_input_tokens_seen": 41340608,
      "step": 21805
    },
    {
      "epoch": 3.5579574190390733,
      "grad_norm": 0.0027580701280385256,
      "learning_rate": 0.12875984952525163,
      "loss": 0.3882,
      "num_input_tokens_seen": 41350240,
      "step": 21810
    },
    {
      "epoch": 3.558773146259891,
      "grad_norm": 0.003185815643519163,
      "learning_rate": 0.12870153983864122,
      "loss": 0.3569,
      "num_input_tokens_seen": 41360064,
      "step": 21815
    },
    {
      "epoch": 3.559588873480708,
      "grad_norm": 0.0015748623991385102,
      "learning_rate": 0.12864323343652104,
      "loss": 0.2942,
      "num_input_tokens_seen": 41368912,
      "step": 21820
    },
    {
      "epoch": 3.5604046007015255,
      "grad_norm": 0.0023390124551951885,
      "learning_rate": 0.12858493032788268,
      "loss": 0.3034,
      "num_input_tokens_seen": 41379072,
      "step": 21825
    },
    {
      "epoch": 3.5612203279223427,
      "grad_norm": 0.0014430294977501035,
      "learning_rate": 0.12852663052171714,
      "loss": 0.3318,
      "num_input_tokens_seen": 41389056,
      "step": 21830
    },
    {
      "epoch": 3.5620360551431602,
      "grad_norm": 0.002615954028442502,
      "learning_rate": 0.12846833402701507,
      "loss": 0.3178,
      "num_input_tokens_seen": 41398656,
      "step": 21835
    },
    {
      "epoch": 3.5628517823639774,
      "grad_norm": 0.0018404320580884814,
      "learning_rate": 0.12841004085276642,
      "loss": 0.3369,
      "num_input_tokens_seen": 41408576,
      "step": 21840
    },
    {
      "epoch": 3.563667509584795,
      "grad_norm": 0.0022392706014215946,
      "learning_rate": 0.12835175100796076,
      "loss": 0.3333,
      "num_input_tokens_seen": 41417872,
      "step": 21845
    },
    {
      "epoch": 3.564483236805612,
      "grad_norm": 0.0021804345306009054,
      "learning_rate": 0.12829346450158724,
      "loss": 0.3116,
      "num_input_tokens_seen": 41426560,
      "step": 21850
    },
    {
      "epoch": 3.5652989640264297,
      "grad_norm": 0.0019656214863061905,
      "learning_rate": 0.12823518134263423,
      "loss": 0.3044,
      "num_input_tokens_seen": 41436544,
      "step": 21855
    },
    {
      "epoch": 3.566114691247247,
      "grad_norm": 0.0016019881004467607,
      "learning_rate": 0.12817690154008973,
      "loss": 0.345,
      "num_input_tokens_seen": 41446704,
      "step": 21860
    },
    {
      "epoch": 3.5669304184680644,
      "grad_norm": 0.0019723137374967337,
      "learning_rate": 0.12811862510294134,
      "loss": 0.3173,
      "num_input_tokens_seen": 41455872,
      "step": 21865
    },
    {
      "epoch": 3.5677461456888815,
      "grad_norm": 0.0012743358965963125,
      "learning_rate": 0.12806035204017585,
      "loss": 0.3365,
      "num_input_tokens_seen": 41465520,
      "step": 21870
    },
    {
      "epoch": 3.568561872909699,
      "grad_norm": 0.0003505925997160375,
      "learning_rate": 0.12800208236077987,
      "loss": 0.3267,
      "num_input_tokens_seen": 41474416,
      "step": 21875
    },
    {
      "epoch": 3.569377600130516,
      "grad_norm": 0.00115779263433069,
      "learning_rate": 0.12794381607373917,
      "loss": 0.3537,
      "num_input_tokens_seen": 41485136,
      "step": 21880
    },
    {
      "epoch": 3.5701933273513338,
      "grad_norm": 0.002460829447954893,
      "learning_rate": 0.12788555318803924,
      "loss": 0.3754,
      "num_input_tokens_seen": 41495024,
      "step": 21885
    },
    {
      "epoch": 3.5710090545721513,
      "grad_norm": 0.002128991764038801,
      "learning_rate": 0.1278272937126649,
      "loss": 0.3283,
      "num_input_tokens_seen": 41503584,
      "step": 21890
    },
    {
      "epoch": 3.5718247817929685,
      "grad_norm": 0.0020813399460166693,
      "learning_rate": 0.1277690376566005,
      "loss": 0.3241,
      "num_input_tokens_seen": 41513984,
      "step": 21895
    },
    {
      "epoch": 3.5726405090137856,
      "grad_norm": 0.000530576566234231,
      "learning_rate": 0.12771078502882985,
      "loss": 0.3421,
      "num_input_tokens_seen": 41523392,
      "step": 21900
    },
    {
      "epoch": 3.573456236234603,
      "grad_norm": 0.001911128987558186,
      "learning_rate": 0.12765253583833633,
      "loss": 0.3324,
      "num_input_tokens_seen": 41532880,
      "step": 21905
    },
    {
      "epoch": 3.5742719634554208,
      "grad_norm": 0.001256487681530416,
      "learning_rate": 0.12759429009410256,
      "loss": 0.3106,
      "num_input_tokens_seen": 41541952,
      "step": 21910
    },
    {
      "epoch": 3.575087690676238,
      "grad_norm": 0.0012387585593387485,
      "learning_rate": 0.12753604780511085,
      "loss": 0.3245,
      "num_input_tokens_seen": 41550352,
      "step": 21915
    },
    {
      "epoch": 3.575903417897055,
      "grad_norm": 0.0013206058647483587,
      "learning_rate": 0.12747780898034283,
      "loss": 0.3009,
      "num_input_tokens_seen": 41560080,
      "step": 21920
    },
    {
      "epoch": 3.5767191451178726,
      "grad_norm": 0.003034047083929181,
      "learning_rate": 0.12741957362877973,
      "loss": 0.3771,
      "num_input_tokens_seen": 41569408,
      "step": 21925
    },
    {
      "epoch": 3.57753487233869,
      "grad_norm": 0.001465849345549941,
      "learning_rate": 0.12736134175940214,
      "loss": 0.3458,
      "num_input_tokens_seen": 41578192,
      "step": 21930
    },
    {
      "epoch": 3.5783505995595073,
      "grad_norm": 0.0011485186405479908,
      "learning_rate": 0.12730311338119016,
      "loss": 0.332,
      "num_input_tokens_seen": 41588096,
      "step": 21935
    },
    {
      "epoch": 3.5791663267803244,
      "grad_norm": 0.0017411267617717385,
      "learning_rate": 0.12724488850312327,
      "loss": 0.3042,
      "num_input_tokens_seen": 41597680,
      "step": 21940
    },
    {
      "epoch": 3.579982054001142,
      "grad_norm": 0.0018234781455248594,
      "learning_rate": 0.1271866671341806,
      "loss": 0.2959,
      "num_input_tokens_seen": 41605712,
      "step": 21945
    },
    {
      "epoch": 3.5807977812219596,
      "grad_norm": 0.0011751858983188868,
      "learning_rate": 0.12712844928334047,
      "loss": 0.3526,
      "num_input_tokens_seen": 41614816,
      "step": 21950
    },
    {
      "epoch": 3.5816135084427767,
      "grad_norm": 0.0015094991540536284,
      "learning_rate": 0.12707023495958095,
      "loss": 0.3286,
      "num_input_tokens_seen": 41624560,
      "step": 21955
    },
    {
      "epoch": 3.582429235663594,
      "grad_norm": 0.0013193958438932896,
      "learning_rate": 0.12701202417187932,
      "loss": 0.3106,
      "num_input_tokens_seen": 41634496,
      "step": 21960
    },
    {
      "epoch": 3.5832449628844114,
      "grad_norm": 0.0011708101956173778,
      "learning_rate": 0.12695381692921243,
      "loss": 0.3112,
      "num_input_tokens_seen": 41643312,
      "step": 21965
    },
    {
      "epoch": 3.584060690105229,
      "grad_norm": 0.0024050588253885508,
      "learning_rate": 0.12689561324055665,
      "loss": 0.3475,
      "num_input_tokens_seen": 41653472,
      "step": 21970
    },
    {
      "epoch": 3.584876417326046,
      "grad_norm": 0.0016309628263115883,
      "learning_rate": 0.12683741311488758,
      "loss": 0.3132,
      "num_input_tokens_seen": 41662608,
      "step": 21975
    },
    {
      "epoch": 3.5856921445468637,
      "grad_norm": 0.0013008012901991606,
      "learning_rate": 0.1267792165611805,
      "loss": 0.263,
      "num_input_tokens_seen": 41671776,
      "step": 21980
    },
    {
      "epoch": 3.586507871767681,
      "grad_norm": 0.0015447356272488832,
      "learning_rate": 0.1267210235884101,
      "loss": 0.3508,
      "num_input_tokens_seen": 41680272,
      "step": 21985
    },
    {
      "epoch": 3.5873235989884984,
      "grad_norm": 0.0015654272865504026,
      "learning_rate": 0.12666283420555033,
      "loss": 0.3105,
      "num_input_tokens_seen": 41690080,
      "step": 21990
    },
    {
      "epoch": 3.5881393262093155,
      "grad_norm": 0.0026453672908246517,
      "learning_rate": 0.12660464842157487,
      "loss": 0.3739,
      "num_input_tokens_seen": 41699168,
      "step": 21995
    },
    {
      "epoch": 3.588955053430133,
      "grad_norm": 0.003317045047879219,
      "learning_rate": 0.1265464662454566,
      "loss": 0.4252,
      "num_input_tokens_seen": 41708800,
      "step": 22000
    },
    {
      "epoch": 3.588955053430133,
      "eval_loss": 0.32555049657821655,
      "eval_runtime": 152.876,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 41708800,
      "step": 22000
    },
    {
      "epoch": 3.5897707806509502,
      "grad_norm": 0.0007601291872560978,
      "learning_rate": 0.12648828768616793,
      "loss": 0.3915,
      "num_input_tokens_seen": 41719280,
      "step": 22005
    },
    {
      "epoch": 3.590586507871768,
      "grad_norm": 0.003411050420254469,
      "learning_rate": 0.12643011275268085,
      "loss": 0.3454,
      "num_input_tokens_seen": 41727968,
      "step": 22010
    },
    {
      "epoch": 3.591402235092585,
      "grad_norm": 0.0009662867523729801,
      "learning_rate": 0.1263719414539665,
      "loss": 0.293,
      "num_input_tokens_seen": 41738160,
      "step": 22015
    },
    {
      "epoch": 3.5922179623134025,
      "grad_norm": 0.001024851342663169,
      "learning_rate": 0.1263137737989957,
      "loss": 0.3048,
      "num_input_tokens_seen": 41746832,
      "step": 22020
    },
    {
      "epoch": 3.5930336895342196,
      "grad_norm": 0.0026571836788207293,
      "learning_rate": 0.1262556097967387,
      "loss": 0.4207,
      "num_input_tokens_seen": 41756304,
      "step": 22025
    },
    {
      "epoch": 3.593849416755037,
      "grad_norm": 0.0029251340311020613,
      "learning_rate": 0.126197449456165,
      "loss": 0.3272,
      "num_input_tokens_seen": 41765952,
      "step": 22030
    },
    {
      "epoch": 3.5946651439758543,
      "grad_norm": 0.001745953457430005,
      "learning_rate": 0.12613929278624378,
      "loss": 0.3438,
      "num_input_tokens_seen": 41776080,
      "step": 22035
    },
    {
      "epoch": 3.595480871196672,
      "grad_norm": 0.001525577623397112,
      "learning_rate": 0.12608113979594343,
      "loss": 0.3315,
      "num_input_tokens_seen": 41786448,
      "step": 22040
    },
    {
      "epoch": 3.596296598417489,
      "grad_norm": 0.0013471459969878197,
      "learning_rate": 0.1260229904942319,
      "loss": 0.4052,
      "num_input_tokens_seen": 41796848,
      "step": 22045
    },
    {
      "epoch": 3.5971123256383066,
      "grad_norm": 0.0010217578383162618,
      "learning_rate": 0.12596484489007662,
      "loss": 0.3998,
      "num_input_tokens_seen": 41805520,
      "step": 22050
    },
    {
      "epoch": 3.5979280528591238,
      "grad_norm": 0.0023457773495465517,
      "learning_rate": 0.1259067029924442,
      "loss": 0.3113,
      "num_input_tokens_seen": 41813168,
      "step": 22055
    },
    {
      "epoch": 3.5987437800799413,
      "grad_norm": 0.0026380911003798246,
      "learning_rate": 0.12584856481030096,
      "loss": 0.3713,
      "num_input_tokens_seen": 41822768,
      "step": 22060
    },
    {
      "epoch": 3.5995595073007585,
      "grad_norm": 0.0011655581183731556,
      "learning_rate": 0.12579043035261261,
      "loss": 0.314,
      "num_input_tokens_seen": 41832624,
      "step": 22065
    },
    {
      "epoch": 3.600375234521576,
      "grad_norm": 0.0012211876455694437,
      "learning_rate": 0.1257322996283441,
      "loss": 0.2864,
      "num_input_tokens_seen": 41840624,
      "step": 22070
    },
    {
      "epoch": 3.601190961742393,
      "grad_norm": 0.0023370205890387297,
      "learning_rate": 0.12567417264645994,
      "loss": 0.3746,
      "num_input_tokens_seen": 41848944,
      "step": 22075
    },
    {
      "epoch": 3.6020066889632107,
      "grad_norm": 0.0021902211010456085,
      "learning_rate": 0.12561604941592408,
      "loss": 0.2955,
      "num_input_tokens_seen": 41858752,
      "step": 22080
    },
    {
      "epoch": 3.6028224161840283,
      "grad_norm": 0.001123281428590417,
      "learning_rate": 0.12555792994569978,
      "loss": 0.2829,
      "num_input_tokens_seen": 41869216,
      "step": 22085
    },
    {
      "epoch": 3.6036381434048455,
      "grad_norm": 0.0016824065241962671,
      "learning_rate": 0.1254998142447499,
      "loss": 0.2878,
      "num_input_tokens_seen": 41878464,
      "step": 22090
    },
    {
      "epoch": 3.6044538706256626,
      "grad_norm": 0.0012690442381426692,
      "learning_rate": 0.1254417023220365,
      "loss": 0.3137,
      "num_input_tokens_seen": 41888528,
      "step": 22095
    },
    {
      "epoch": 3.60526959784648,
      "grad_norm": 0.0012541264295578003,
      "learning_rate": 0.12538359418652126,
      "loss": 0.3761,
      "num_input_tokens_seen": 41897824,
      "step": 22100
    },
    {
      "epoch": 3.6060853250672977,
      "grad_norm": 0.0020408087875694036,
      "learning_rate": 0.12532548984716513,
      "loss": 0.3086,
      "num_input_tokens_seen": 41906640,
      "step": 22105
    },
    {
      "epoch": 3.606901052288115,
      "grad_norm": 0.001446430222131312,
      "learning_rate": 0.12526738931292855,
      "loss": 0.3252,
      "num_input_tokens_seen": 41915728,
      "step": 22110
    },
    {
      "epoch": 3.607716779508932,
      "grad_norm": 0.0025177530478686094,
      "learning_rate": 0.1252092925927714,
      "loss": 0.3845,
      "num_input_tokens_seen": 41925968,
      "step": 22115
    },
    {
      "epoch": 3.6085325067297496,
      "grad_norm": 0.0009946993086487055,
      "learning_rate": 0.12515119969565278,
      "loss": 0.3483,
      "num_input_tokens_seen": 41935504,
      "step": 22120
    },
    {
      "epoch": 3.609348233950567,
      "grad_norm": 0.0031790914945304394,
      "learning_rate": 0.12509311063053144,
      "loss": 0.3188,
      "num_input_tokens_seen": 41944544,
      "step": 22125
    },
    {
      "epoch": 3.6101639611713843,
      "grad_norm": 0.001998345600441098,
      "learning_rate": 0.1250350254063655,
      "loss": 0.3206,
      "num_input_tokens_seen": 41954880,
      "step": 22130
    },
    {
      "epoch": 3.6109796883922014,
      "grad_norm": 0.0015543048502877355,
      "learning_rate": 0.1249769440321123,
      "loss": 0.3403,
      "num_input_tokens_seen": 41963808,
      "step": 22135
    },
    {
      "epoch": 3.611795415613019,
      "grad_norm": 0.002973835216835141,
      "learning_rate": 0.12491886651672884,
      "loss": 0.3075,
      "num_input_tokens_seen": 41973120,
      "step": 22140
    },
    {
      "epoch": 3.6126111428338366,
      "grad_norm": 0.0019509789999574423,
      "learning_rate": 0.12486079286917139,
      "loss": 0.4214,
      "num_input_tokens_seen": 41982960,
      "step": 22145
    },
    {
      "epoch": 3.6134268700546537,
      "grad_norm": 0.0010771133238449693,
      "learning_rate": 0.12480272309839553,
      "loss": 0.2917,
      "num_input_tokens_seen": 41993008,
      "step": 22150
    },
    {
      "epoch": 3.614242597275471,
      "grad_norm": 0.0007664227741770446,
      "learning_rate": 0.12474465721335648,
      "loss": 0.3295,
      "num_input_tokens_seen": 42003056,
      "step": 22155
    },
    {
      "epoch": 3.6150583244962884,
      "grad_norm": 0.001325107179582119,
      "learning_rate": 0.12468659522300861,
      "loss": 0.3451,
      "num_input_tokens_seen": 42013200,
      "step": 22160
    },
    {
      "epoch": 3.615874051717106,
      "grad_norm": 0.0011171726509928703,
      "learning_rate": 0.12462853713630584,
      "loss": 0.3604,
      "num_input_tokens_seen": 42022272,
      "step": 22165
    },
    {
      "epoch": 3.616689778937923,
      "grad_norm": 0.0018116896972060204,
      "learning_rate": 0.12457048296220156,
      "loss": 0.3328,
      "num_input_tokens_seen": 42031376,
      "step": 22170
    },
    {
      "epoch": 3.6175055061587407,
      "grad_norm": 0.001814192859455943,
      "learning_rate": 0.12451243270964832,
      "loss": 0.3666,
      "num_input_tokens_seen": 42041376,
      "step": 22175
    },
    {
      "epoch": 3.618321233379558,
      "grad_norm": 0.0012160680489614606,
      "learning_rate": 0.12445438638759827,
      "loss": 0.3578,
      "num_input_tokens_seen": 42051808,
      "step": 22180
    },
    {
      "epoch": 3.6191369606003754,
      "grad_norm": 0.0010335355764254928,
      "learning_rate": 0.1243963440050029,
      "loss": 0.3091,
      "num_input_tokens_seen": 42061392,
      "step": 22185
    },
    {
      "epoch": 3.6199526878211925,
      "grad_norm": 0.00046340248081833124,
      "learning_rate": 0.12433830557081298,
      "loss": 0.3468,
      "num_input_tokens_seen": 42070320,
      "step": 22190
    },
    {
      "epoch": 3.62076841504201,
      "grad_norm": 0.001345719676464796,
      "learning_rate": 0.12428027109397889,
      "loss": 0.3343,
      "num_input_tokens_seen": 42079280,
      "step": 22195
    },
    {
      "epoch": 3.621584142262827,
      "grad_norm": 0.0015602048952132463,
      "learning_rate": 0.12422224058345015,
      "loss": 0.3007,
      "num_input_tokens_seen": 42087824,
      "step": 22200
    },
    {
      "epoch": 3.621584142262827,
      "eval_loss": 0.3228893280029297,
      "eval_runtime": 153.0622,
      "eval_samples_per_second": 17.803,
      "eval_steps_per_second": 8.905,
      "num_input_tokens_seen": 42087824,
      "step": 22200
    },
    {
      "epoch": 3.622399869483645,
      "grad_norm": 0.0018532327376306057,
      "learning_rate": 0.12416421404817583,
      "loss": 0.3371,
      "num_input_tokens_seen": 42097584,
      "step": 22205
    },
    {
      "epoch": 3.623215596704462,
      "grad_norm": 0.001609833212569356,
      "learning_rate": 0.12410619149710447,
      "loss": 0.382,
      "num_input_tokens_seen": 42107744,
      "step": 22210
    },
    {
      "epoch": 3.6240313239252795,
      "grad_norm": 0.0019767240155488253,
      "learning_rate": 0.12404817293918374,
      "loss": 0.3185,
      "num_input_tokens_seen": 42116448,
      "step": 22215
    },
    {
      "epoch": 3.6248470511460966,
      "grad_norm": 0.0019214458297938108,
      "learning_rate": 0.12399015838336086,
      "loss": 0.3028,
      "num_input_tokens_seen": 42126144,
      "step": 22220
    },
    {
      "epoch": 3.625662778366914,
      "grad_norm": 0.0006634854944422841,
      "learning_rate": 0.12393214783858246,
      "loss": 0.2988,
      "num_input_tokens_seen": 42135072,
      "step": 22225
    },
    {
      "epoch": 3.6264785055877313,
      "grad_norm": 0.001533630071207881,
      "learning_rate": 0.1238741413137944,
      "loss": 0.2954,
      "num_input_tokens_seen": 42144752,
      "step": 22230
    },
    {
      "epoch": 3.627294232808549,
      "grad_norm": 0.0013490957207977772,
      "learning_rate": 0.12381613881794212,
      "loss": 0.3562,
      "num_input_tokens_seen": 42154880,
      "step": 22235
    },
    {
      "epoch": 3.628109960029366,
      "grad_norm": 0.0030376084614545107,
      "learning_rate": 0.12375814035997022,
      "loss": 0.3127,
      "num_input_tokens_seen": 42164560,
      "step": 22240
    },
    {
      "epoch": 3.6289256872501836,
      "grad_norm": 0.0007161649409681559,
      "learning_rate": 0.12370014594882285,
      "loss": 0.3298,
      "num_input_tokens_seen": 42174016,
      "step": 22245
    },
    {
      "epoch": 3.6297414144710007,
      "grad_norm": 0.0016729068011045456,
      "learning_rate": 0.12364215559344356,
      "loss": 0.3511,
      "num_input_tokens_seen": 42181104,
      "step": 22250
    },
    {
      "epoch": 3.6305571416918183,
      "grad_norm": 0.0014386363327503204,
      "learning_rate": 0.12358416930277506,
      "loss": 0.3485,
      "num_input_tokens_seen": 42189456,
      "step": 22255
    },
    {
      "epoch": 3.631372868912636,
      "grad_norm": 0.0004820072208531201,
      "learning_rate": 0.1235261870857596,
      "loss": 0.3117,
      "num_input_tokens_seen": 42198912,
      "step": 22260
    },
    {
      "epoch": 3.632188596133453,
      "grad_norm": 0.0005406636628322303,
      "learning_rate": 0.12346820895133884,
      "loss": 0.3702,
      "num_input_tokens_seen": 42208352,
      "step": 22265
    },
    {
      "epoch": 3.63300432335427,
      "grad_norm": 0.0022884774953126907,
      "learning_rate": 0.12341023490845361,
      "loss": 0.3587,
      "num_input_tokens_seen": 42218304,
      "step": 22270
    },
    {
      "epoch": 3.6338200505750877,
      "grad_norm": 0.0011145304888486862,
      "learning_rate": 0.12335226496604437,
      "loss": 0.2998,
      "num_input_tokens_seen": 42226928,
      "step": 22275
    },
    {
      "epoch": 3.6346357777959053,
      "grad_norm": 0.0011814304161816835,
      "learning_rate": 0.12329429913305069,
      "loss": 0.3657,
      "num_input_tokens_seen": 42236464,
      "step": 22280
    },
    {
      "epoch": 3.6354515050167224,
      "grad_norm": 0.001509045367129147,
      "learning_rate": 0.12323633741841171,
      "loss": 0.3046,
      "num_input_tokens_seen": 42245280,
      "step": 22285
    },
    {
      "epoch": 3.6362672322375396,
      "grad_norm": 0.0010021340567618608,
      "learning_rate": 0.12317837983106583,
      "loss": 0.359,
      "num_input_tokens_seen": 42253840,
      "step": 22290
    },
    {
      "epoch": 3.637082959458357,
      "grad_norm": 0.0021045678295195103,
      "learning_rate": 0.12312042637995087,
      "loss": 0.348,
      "num_input_tokens_seen": 42262736,
      "step": 22295
    },
    {
      "epoch": 3.6378986866791747,
      "grad_norm": 0.0015356888761743903,
      "learning_rate": 0.12306247707400389,
      "loss": 0.3711,
      "num_input_tokens_seen": 42271664,
      "step": 22300
    },
    {
      "epoch": 3.638714413899992,
      "grad_norm": 0.0008913541096262634,
      "learning_rate": 0.12300453192216154,
      "loss": 0.3225,
      "num_input_tokens_seen": 42281312,
      "step": 22305
    },
    {
      "epoch": 3.639530141120809,
      "grad_norm": 0.0020343903452157974,
      "learning_rate": 0.12294659093335956,
      "loss": 0.363,
      "num_input_tokens_seen": 42290640,
      "step": 22310
    },
    {
      "epoch": 3.6403458683416265,
      "grad_norm": 0.0032703480683267117,
      "learning_rate": 0.12288865411653327,
      "loss": 0.3737,
      "num_input_tokens_seen": 42300976,
      "step": 22315
    },
    {
      "epoch": 3.641161595562444,
      "grad_norm": 0.002845400245860219,
      "learning_rate": 0.12283072148061717,
      "loss": 0.3498,
      "num_input_tokens_seen": 42310896,
      "step": 22320
    },
    {
      "epoch": 3.6419773227832613,
      "grad_norm": 0.000716956565156579,
      "learning_rate": 0.12277279303454529,
      "loss": 0.3151,
      "num_input_tokens_seen": 42320784,
      "step": 22325
    },
    {
      "epoch": 3.6427930500040784,
      "grad_norm": 0.002419508760794997,
      "learning_rate": 0.12271486878725091,
      "loss": 0.3771,
      "num_input_tokens_seen": 42331072,
      "step": 22330
    },
    {
      "epoch": 3.643608777224896,
      "grad_norm": 0.0025008907541632652,
      "learning_rate": 0.12265694874766658,
      "loss": 0.3733,
      "num_input_tokens_seen": 42340688,
      "step": 22335
    },
    {
      "epoch": 3.6444245044457135,
      "grad_norm": 0.0007897811592556536,
      "learning_rate": 0.12259903292472435,
      "loss": 0.3823,
      "num_input_tokens_seen": 42349312,
      "step": 22340
    },
    {
      "epoch": 3.6452402316665307,
      "grad_norm": 0.0011441196547821164,
      "learning_rate": 0.12254112132735567,
      "loss": 0.3791,
      "num_input_tokens_seen": 42357200,
      "step": 22345
    },
    {
      "epoch": 3.6460559588873482,
      "grad_norm": 0.0008648961083963513,
      "learning_rate": 0.12248321396449108,
      "loss": 0.3617,
      "num_input_tokens_seen": 42366688,
      "step": 22350
    },
    {
      "epoch": 3.6468716861081654,
      "grad_norm": 0.0012176537420600653,
      "learning_rate": 0.12242531084506075,
      "loss": 0.335,
      "num_input_tokens_seen": 42375920,
      "step": 22355
    },
    {
      "epoch": 3.647687413328983,
      "grad_norm": 0.0006443989113904536,
      "learning_rate": 0.122367411977994,
      "loss": 0.3275,
      "num_input_tokens_seen": 42385520,
      "step": 22360
    },
    {
      "epoch": 3.6485031405498,
      "grad_norm": 0.0033035078085958958,
      "learning_rate": 0.12230951737221954,
      "loss": 0.3515,
      "num_input_tokens_seen": 42395584,
      "step": 22365
    },
    {
      "epoch": 3.6493188677706176,
      "grad_norm": 0.0005906449514441192,
      "learning_rate": 0.12225162703666555,
      "loss": 0.3389,
      "num_input_tokens_seen": 42404256,
      "step": 22370
    },
    {
      "epoch": 3.6501345949914348,
      "grad_norm": 0.00076024339068681,
      "learning_rate": 0.1221937409802593,
      "loss": 0.3317,
      "num_input_tokens_seen": 42414016,
      "step": 22375
    },
    {
      "epoch": 3.6509503222122524,
      "grad_norm": 0.0005890795146115124,
      "learning_rate": 0.12213585921192768,
      "loss": 0.314,
      "num_input_tokens_seen": 42423840,
      "step": 22380
    },
    {
      "epoch": 3.6517660494330695,
      "grad_norm": 0.001686744624748826,
      "learning_rate": 0.1220779817405967,
      "loss": 0.3625,
      "num_input_tokens_seen": 42433568,
      "step": 22385
    },
    {
      "epoch": 3.652581776653887,
      "grad_norm": 0.0018619182519614697,
      "learning_rate": 0.12202010857519181,
      "loss": 0.3419,
      "num_input_tokens_seen": 42442976,
      "step": 22390
    },
    {
      "epoch": 3.653397503874704,
      "grad_norm": 0.0012704274849966168,
      "learning_rate": 0.12196223972463785,
      "loss": 0.3005,
      "num_input_tokens_seen": 42451600,
      "step": 22395
    },
    {
      "epoch": 3.6542132310955218,
      "grad_norm": 0.0007373142288997769,
      "learning_rate": 0.12190437519785885,
      "loss": 0.324,
      "num_input_tokens_seen": 42461936,
      "step": 22400
    },
    {
      "epoch": 3.6542132310955218,
      "eval_loss": 0.3247263431549072,
      "eval_runtime": 152.914,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 42461936,
      "step": 22400
    },
    {
      "epoch": 3.655028958316339,
      "grad_norm": 0.0019244520226493478,
      "learning_rate": 0.12184651500377823,
      "loss": 0.3442,
      "num_input_tokens_seen": 42471968,
      "step": 22405
    },
    {
      "epoch": 3.6558446855371565,
      "grad_norm": 0.0016226420411840081,
      "learning_rate": 0.12178865915131885,
      "loss": 0.2979,
      "num_input_tokens_seen": 42480288,
      "step": 22410
    },
    {
      "epoch": 3.6566604127579736,
      "grad_norm": 0.0016843624180182815,
      "learning_rate": 0.1217308076494027,
      "loss": 0.3492,
      "num_input_tokens_seen": 42489504,
      "step": 22415
    },
    {
      "epoch": 3.657476139978791,
      "grad_norm": 0.0018916713306680322,
      "learning_rate": 0.12167296050695134,
      "loss": 0.3386,
      "num_input_tokens_seen": 42499440,
      "step": 22420
    },
    {
      "epoch": 3.6582918671996083,
      "grad_norm": 0.0015029554488137364,
      "learning_rate": 0.12161511773288536,
      "loss": 0.3212,
      "num_input_tokens_seen": 42509088,
      "step": 22425
    },
    {
      "epoch": 3.659107594420426,
      "grad_norm": 0.001961385365575552,
      "learning_rate": 0.121557279336125,
      "loss": 0.4101,
      "num_input_tokens_seen": 42519328,
      "step": 22430
    },
    {
      "epoch": 3.659923321641243,
      "grad_norm": 0.0010584562551230192,
      "learning_rate": 0.12149944532558957,
      "loss": 0.3097,
      "num_input_tokens_seen": 42528976,
      "step": 22435
    },
    {
      "epoch": 3.6607390488620606,
      "grad_norm": 0.0017378502525389194,
      "learning_rate": 0.12144161571019785,
      "loss": 0.3182,
      "num_input_tokens_seen": 42539424,
      "step": 22440
    },
    {
      "epoch": 3.6615547760828777,
      "grad_norm": 0.0021168235689401627,
      "learning_rate": 0.12138379049886781,
      "loss": 0.323,
      "num_input_tokens_seen": 42548640,
      "step": 22445
    },
    {
      "epoch": 3.6623705033036953,
      "grad_norm": 0.0012579887406900525,
      "learning_rate": 0.12132596970051697,
      "loss": 0.2917,
      "num_input_tokens_seen": 42557808,
      "step": 22450
    },
    {
      "epoch": 3.663186230524513,
      "grad_norm": 0.0012507830979302526,
      "learning_rate": 0.12126815332406189,
      "loss": 0.2805,
      "num_input_tokens_seen": 42567632,
      "step": 22455
    },
    {
      "epoch": 3.66400195774533,
      "grad_norm": 0.0022000742610543966,
      "learning_rate": 0.12121034137841868,
      "loss": 0.3312,
      "num_input_tokens_seen": 42576720,
      "step": 22460
    },
    {
      "epoch": 3.664817684966147,
      "grad_norm": 0.0020420115906745195,
      "learning_rate": 0.12115253387250258,
      "loss": 0.3694,
      "num_input_tokens_seen": 42586832,
      "step": 22465
    },
    {
      "epoch": 3.6656334121869647,
      "grad_norm": 0.0015780952526256442,
      "learning_rate": 0.12109473081522831,
      "loss": 0.3119,
      "num_input_tokens_seen": 42595312,
      "step": 22470
    },
    {
      "epoch": 3.6664491394077823,
      "grad_norm": 0.0008969815680757165,
      "learning_rate": 0.12103693221550982,
      "loss": 0.3318,
      "num_input_tokens_seen": 42604400,
      "step": 22475
    },
    {
      "epoch": 3.6672648666285994,
      "grad_norm": 0.0024083410389721394,
      "learning_rate": 0.12097913808226027,
      "loss": 0.3469,
      "num_input_tokens_seen": 42614272,
      "step": 22480
    },
    {
      "epoch": 3.6680805938494165,
      "grad_norm": 0.0023732748813927174,
      "learning_rate": 0.12092134842439234,
      "loss": 0.3617,
      "num_input_tokens_seen": 42622240,
      "step": 22485
    },
    {
      "epoch": 3.668896321070234,
      "grad_norm": 0.0007820558384992182,
      "learning_rate": 0.12086356325081798,
      "loss": 0.2835,
      "num_input_tokens_seen": 42632752,
      "step": 22490
    },
    {
      "epoch": 3.6697120482910517,
      "grad_norm": 0.0014952969504520297,
      "learning_rate": 0.12080578257044824,
      "loss": 0.3428,
      "num_input_tokens_seen": 42643168,
      "step": 22495
    },
    {
      "epoch": 3.670527775511869,
      "grad_norm": 0.002059066668152809,
      "learning_rate": 0.12074800639219378,
      "loss": 0.319,
      "num_input_tokens_seen": 42652672,
      "step": 22500
    },
    {
      "epoch": 3.671343502732686,
      "grad_norm": 0.0009909174405038357,
      "learning_rate": 0.12069023472496428,
      "loss": 0.3299,
      "num_input_tokens_seen": 42661472,
      "step": 22505
    },
    {
      "epoch": 3.6721592299535035,
      "grad_norm": 0.0015873024240136147,
      "learning_rate": 0.12063246757766893,
      "loss": 0.3852,
      "num_input_tokens_seen": 42671264,
      "step": 22510
    },
    {
      "epoch": 3.672974957174321,
      "grad_norm": 0.0018602233612909913,
      "learning_rate": 0.12057470495921618,
      "loss": 0.3387,
      "num_input_tokens_seen": 42680144,
      "step": 22515
    },
    {
      "epoch": 3.6737906843951382,
      "grad_norm": 0.0012149004032835364,
      "learning_rate": 0.12051694687851364,
      "loss": 0.3529,
      "num_input_tokens_seen": 42688848,
      "step": 22520
    },
    {
      "epoch": 3.6746064116159554,
      "grad_norm": 0.001029850565828383,
      "learning_rate": 0.12045919334446839,
      "loss": 0.2972,
      "num_input_tokens_seen": 42698672,
      "step": 22525
    },
    {
      "epoch": 3.675422138836773,
      "grad_norm": 0.0012229790445417166,
      "learning_rate": 0.12040144436598683,
      "loss": 0.3021,
      "num_input_tokens_seen": 42708720,
      "step": 22530
    },
    {
      "epoch": 3.6762378660575905,
      "grad_norm": 0.0015181305352598429,
      "learning_rate": 0.12034369995197444,
      "loss": 0.3262,
      "num_input_tokens_seen": 42718560,
      "step": 22535
    },
    {
      "epoch": 3.6770535932784076,
      "grad_norm": 0.0011895345523953438,
      "learning_rate": 0.12028596011133627,
      "loss": 0.3223,
      "num_input_tokens_seen": 42727984,
      "step": 22540
    },
    {
      "epoch": 3.677869320499225,
      "grad_norm": 0.0007101083174347878,
      "learning_rate": 0.12022822485297643,
      "loss": 0.3062,
      "num_input_tokens_seen": 42737968,
      "step": 22545
    },
    {
      "epoch": 3.6786850477200423,
      "grad_norm": 0.0014448221772909164,
      "learning_rate": 0.12017049418579843,
      "loss": 0.3876,
      "num_input_tokens_seen": 42747328,
      "step": 22550
    },
    {
      "epoch": 3.67950077494086,
      "grad_norm": 0.0011822328669950366,
      "learning_rate": 0.12011276811870514,
      "loss": 0.3131,
      "num_input_tokens_seen": 42755168,
      "step": 22555
    },
    {
      "epoch": 3.680316502161677,
      "grad_norm": 0.0022509226109832525,
      "learning_rate": 0.12005504666059852,
      "loss": 0.2935,
      "num_input_tokens_seen": 42765680,
      "step": 22560
    },
    {
      "epoch": 3.6811322293824946,
      "grad_norm": 0.0024838000535964966,
      "learning_rate": 0.11999732982038003,
      "loss": 0.3129,
      "num_input_tokens_seen": 42775024,
      "step": 22565
    },
    {
      "epoch": 3.6819479566033118,
      "grad_norm": 0.0018451990326866508,
      "learning_rate": 0.11993961760695038,
      "loss": 0.3112,
      "num_input_tokens_seen": 42784464,
      "step": 22570
    },
    {
      "epoch": 3.6827636838241293,
      "grad_norm": 0.0024405557196587324,
      "learning_rate": 0.11988191002920942,
      "loss": 0.3403,
      "num_input_tokens_seen": 42793168,
      "step": 22575
    },
    {
      "epoch": 3.6835794110449465,
      "grad_norm": 0.0014704151544719934,
      "learning_rate": 0.11982420709605641,
      "loss": 0.3097,
      "num_input_tokens_seen": 42804160,
      "step": 22580
    },
    {
      "epoch": 3.684395138265764,
      "grad_norm": 0.0018186394590884447,
      "learning_rate": 0.11976650881638991,
      "loss": 0.2982,
      "num_input_tokens_seen": 42814464,
      "step": 22585
    },
    {
      "epoch": 3.685210865486581,
      "grad_norm": 0.0019068169640377164,
      "learning_rate": 0.11970881519910764,
      "loss": 0.3143,
      "num_input_tokens_seen": 42824848,
      "step": 22590
    },
    {
      "epoch": 3.6860265927073987,
      "grad_norm": 0.002847161842510104,
      "learning_rate": 0.1196511262531068,
      "loss": 0.2981,
      "num_input_tokens_seen": 42835248,
      "step": 22595
    },
    {
      "epoch": 3.686842319928216,
      "grad_norm": 0.001642127986997366,
      "learning_rate": 0.11959344198728361,
      "loss": 0.3796,
      "num_input_tokens_seen": 42843696,
      "step": 22600
    },
    {
      "epoch": 3.686842319928216,
      "eval_loss": 0.32404837012290955,
      "eval_runtime": 152.8314,
      "eval_samples_per_second": 17.83,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 42843696,
      "step": 22600
    },
    {
      "epoch": 3.6876580471490334,
      "grad_norm": 0.001727459253743291,
      "learning_rate": 0.11953576241053378,
      "loss": 0.3333,
      "num_input_tokens_seen": 42851664,
      "step": 22605
    },
    {
      "epoch": 3.6884737743698506,
      "grad_norm": 0.0019989414140582085,
      "learning_rate": 0.11947808753175228,
      "loss": 0.3719,
      "num_input_tokens_seen": 42862480,
      "step": 22610
    },
    {
      "epoch": 3.689289501590668,
      "grad_norm": 0.0014192144153639674,
      "learning_rate": 0.1194204173598332,
      "loss": 0.3425,
      "num_input_tokens_seen": 42871584,
      "step": 22615
    },
    {
      "epoch": 3.6901052288114853,
      "grad_norm": 0.002097612712532282,
      "learning_rate": 0.11936275190367007,
      "loss": 0.3776,
      "num_input_tokens_seen": 42882400,
      "step": 22620
    },
    {
      "epoch": 3.690920956032303,
      "grad_norm": 0.0017353793373331428,
      "learning_rate": 0.11930509117215563,
      "loss": 0.3112,
      "num_input_tokens_seen": 42892064,
      "step": 22625
    },
    {
      "epoch": 3.6917366832531204,
      "grad_norm": 0.0014297566376626492,
      "learning_rate": 0.11924743517418179,
      "loss": 0.338,
      "num_input_tokens_seen": 42902720,
      "step": 22630
    },
    {
      "epoch": 3.6925524104739376,
      "grad_norm": 0.0008811501320451498,
      "learning_rate": 0.11918978391864,
      "loss": 0.38,
      "num_input_tokens_seen": 42910928,
      "step": 22635
    },
    {
      "epoch": 3.6933681376947547,
      "grad_norm": 0.0008429898880422115,
      "learning_rate": 0.11913213741442065,
      "loss": 0.3406,
      "num_input_tokens_seen": 42921472,
      "step": 22640
    },
    {
      "epoch": 3.6941838649155723,
      "grad_norm": 0.001406872645020485,
      "learning_rate": 0.11907449567041364,
      "loss": 0.32,
      "num_input_tokens_seen": 42931328,
      "step": 22645
    },
    {
      "epoch": 3.69499959213639,
      "grad_norm": 0.0024254953023046255,
      "learning_rate": 0.11901685869550803,
      "loss": 0.3617,
      "num_input_tokens_seen": 42939808,
      "step": 22650
    },
    {
      "epoch": 3.695815319357207,
      "grad_norm": 0.0018050147918984294,
      "learning_rate": 0.1189592264985922,
      "loss": 0.3408,
      "num_input_tokens_seen": 42949136,
      "step": 22655
    },
    {
      "epoch": 3.696631046578024,
      "grad_norm": 0.0012442274019122124,
      "learning_rate": 0.11890159908855373,
      "loss": 0.3768,
      "num_input_tokens_seen": 42958720,
      "step": 22660
    },
    {
      "epoch": 3.6974467737988417,
      "grad_norm": 0.0016417308943346143,
      "learning_rate": 0.11884397647427941,
      "loss": 0.3214,
      "num_input_tokens_seen": 42968560,
      "step": 22665
    },
    {
      "epoch": 3.6982625010196593,
      "grad_norm": 0.0013291840441524982,
      "learning_rate": 0.11878635866465546,
      "loss": 0.3433,
      "num_input_tokens_seen": 42978352,
      "step": 22670
    },
    {
      "epoch": 3.6990782282404764,
      "grad_norm": 0.0010263302829116583,
      "learning_rate": 0.11872874566856734,
      "loss": 0.3175,
      "num_input_tokens_seen": 42987072,
      "step": 22675
    },
    {
      "epoch": 3.6998939554612935,
      "grad_norm": 0.0032206238247454166,
      "learning_rate": 0.11867113749489955,
      "loss": 0.3221,
      "num_input_tokens_seen": 42996384,
      "step": 22680
    },
    {
      "epoch": 3.700709682682111,
      "grad_norm": 0.0018433415098115802,
      "learning_rate": 0.11861353415253607,
      "loss": 0.3402,
      "num_input_tokens_seen": 43006016,
      "step": 22685
    },
    {
      "epoch": 3.7015254099029287,
      "grad_norm": 0.0007987777353264391,
      "learning_rate": 0.11855593565036011,
      "loss": 0.3407,
      "num_input_tokens_seen": 43014816,
      "step": 22690
    },
    {
      "epoch": 3.702341137123746,
      "grad_norm": 0.0013802428729832172,
      "learning_rate": 0.11849834199725394,
      "loss": 0.3053,
      "num_input_tokens_seen": 43024912,
      "step": 22695
    },
    {
      "epoch": 3.703156864344563,
      "grad_norm": 0.0012638611951842904,
      "learning_rate": 0.1184407532020994,
      "loss": 0.3306,
      "num_input_tokens_seen": 43034864,
      "step": 22700
    },
    {
      "epoch": 3.7039725915653805,
      "grad_norm": 0.001982063287869096,
      "learning_rate": 0.11838316927377723,
      "loss": 0.3913,
      "num_input_tokens_seen": 43045168,
      "step": 22705
    },
    {
      "epoch": 3.704788318786198,
      "grad_norm": 0.0014736454468220472,
      "learning_rate": 0.11832559022116766,
      "loss": 0.3429,
      "num_input_tokens_seen": 43054208,
      "step": 22710
    },
    {
      "epoch": 3.705604046007015,
      "grad_norm": 0.0018478252459317446,
      "learning_rate": 0.11826801605315022,
      "loss": 0.3006,
      "num_input_tokens_seen": 43063824,
      "step": 22715
    },
    {
      "epoch": 3.7064197732278323,
      "grad_norm": 0.0007018604665063322,
      "learning_rate": 0.1182104467786034,
      "loss": 0.3292,
      "num_input_tokens_seen": 43073872,
      "step": 22720
    },
    {
      "epoch": 3.70723550044865,
      "grad_norm": 0.0010986606357619166,
      "learning_rate": 0.1181528824064052,
      "loss": 0.3612,
      "num_input_tokens_seen": 43082672,
      "step": 22725
    },
    {
      "epoch": 3.7080512276694675,
      "grad_norm": 0.0011363154044374824,
      "learning_rate": 0.11809532294543279,
      "loss": 0.3755,
      "num_input_tokens_seen": 43092384,
      "step": 22730
    },
    {
      "epoch": 3.7088669548902846,
      "grad_norm": 0.0010339139262214303,
      "learning_rate": 0.11803776840456245,
      "loss": 0.3099,
      "num_input_tokens_seen": 43101408,
      "step": 22735
    },
    {
      "epoch": 3.709682682111102,
      "grad_norm": 0.0009426730102859437,
      "learning_rate": 0.11798021879266997,
      "loss": 0.3364,
      "num_input_tokens_seen": 43111344,
      "step": 22740
    },
    {
      "epoch": 3.7104984093319193,
      "grad_norm": 0.0019666391890496016,
      "learning_rate": 0.11792267411863006,
      "loss": 0.3609,
      "num_input_tokens_seen": 43119344,
      "step": 22745
    },
    {
      "epoch": 3.711314136552737,
      "grad_norm": 0.0019539606291800737,
      "learning_rate": 0.1178651343913169,
      "loss": 0.3736,
      "num_input_tokens_seen": 43129152,
      "step": 22750
    },
    {
      "epoch": 3.712129863773554,
      "grad_norm": 0.0011962292483076453,
      "learning_rate": 0.11780759961960392,
      "loss": 0.3073,
      "num_input_tokens_seen": 43138080,
      "step": 22755
    },
    {
      "epoch": 3.7129455909943716,
      "grad_norm": 0.0010507082333788276,
      "learning_rate": 0.1177500698123636,
      "loss": 0.3465,
      "num_input_tokens_seen": 43146624,
      "step": 22760
    },
    {
      "epoch": 3.7137613182151887,
      "grad_norm": 0.0012161016929894686,
      "learning_rate": 0.11769254497846778,
      "loss": 0.3014,
      "num_input_tokens_seen": 43155456,
      "step": 22765
    },
    {
      "epoch": 3.7145770454360063,
      "grad_norm": 0.002795193577185273,
      "learning_rate": 0.11763502512678758,
      "loss": 0.3318,
      "num_input_tokens_seen": 43164800,
      "step": 22770
    },
    {
      "epoch": 3.7153927726568234,
      "grad_norm": 0.002262481953948736,
      "learning_rate": 0.11757751026619315,
      "loss": 0.3601,
      "num_input_tokens_seen": 43174960,
      "step": 22775
    },
    {
      "epoch": 3.716208499877641,
      "grad_norm": 0.001985644455999136,
      "learning_rate": 0.11752000040555416,
      "loss": 0.2811,
      "num_input_tokens_seen": 43183488,
      "step": 22780
    },
    {
      "epoch": 3.717024227098458,
      "grad_norm": 0.002257696120068431,
      "learning_rate": 0.11746249555373921,
      "loss": 0.328,
      "num_input_tokens_seen": 43194192,
      "step": 22785
    },
    {
      "epoch": 3.7178399543192757,
      "grad_norm": 0.0017079999670386314,
      "learning_rate": 0.11740499571961638,
      "loss": 0.3296,
      "num_input_tokens_seen": 43202528,
      "step": 22790
    },
    {
      "epoch": 3.718655681540093,
      "grad_norm": 0.0015804098220542073,
      "learning_rate": 0.11734750091205279,
      "loss": 0.3341,
      "num_input_tokens_seen": 43211664,
      "step": 22795
    },
    {
      "epoch": 3.7194714087609104,
      "grad_norm": 0.0005415595951490104,
      "learning_rate": 0.11729001113991493,
      "loss": 0.2865,
      "num_input_tokens_seen": 43221120,
      "step": 22800
    },
    {
      "epoch": 3.7194714087609104,
      "eval_loss": 0.3220099210739136,
      "eval_runtime": 152.8207,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 43221120,
      "step": 22800
    },
    {
      "epoch": 3.7202871359817276,
      "grad_norm": 0.003503777552396059,
      "learning_rate": 0.11723252641206837,
      "loss": 0.3162,
      "num_input_tokens_seen": 43231152,
      "step": 22805
    },
    {
      "epoch": 3.721102863202545,
      "grad_norm": 0.0015577030135318637,
      "learning_rate": 0.11717504673737808,
      "loss": 0.3239,
      "num_input_tokens_seen": 43240384,
      "step": 22810
    },
    {
      "epoch": 3.7219185904233623,
      "grad_norm": 0.0015630971174687147,
      "learning_rate": 0.11711757212470802,
      "loss": 0.3107,
      "num_input_tokens_seen": 43249712,
      "step": 22815
    },
    {
      "epoch": 3.72273431764418,
      "grad_norm": 0.0029249300714582205,
      "learning_rate": 0.11706010258292165,
      "loss": 0.3315,
      "num_input_tokens_seen": 43259136,
      "step": 22820
    },
    {
      "epoch": 3.7235500448649974,
      "grad_norm": 0.001562788151204586,
      "learning_rate": 0.11700263812088131,
      "loss": 0.3131,
      "num_input_tokens_seen": 43269408,
      "step": 22825
    },
    {
      "epoch": 3.7243657720858145,
      "grad_norm": 0.001470971736125648,
      "learning_rate": 0.11694517874744892,
      "loss": 0.2602,
      "num_input_tokens_seen": 43280288,
      "step": 22830
    },
    {
      "epoch": 3.7251814993066317,
      "grad_norm": 0.0016128895804286003,
      "learning_rate": 0.11688772447148532,
      "loss": 0.3912,
      "num_input_tokens_seen": 43289856,
      "step": 22835
    },
    {
      "epoch": 3.7259972265274492,
      "grad_norm": 0.0015518027357757092,
      "learning_rate": 0.11683027530185074,
      "loss": 0.3227,
      "num_input_tokens_seen": 43300368,
      "step": 22840
    },
    {
      "epoch": 3.726812953748267,
      "grad_norm": 0.0016276187961921096,
      "learning_rate": 0.11677283124740451,
      "loss": 0.3114,
      "num_input_tokens_seen": 43309664,
      "step": 22845
    },
    {
      "epoch": 3.727628680969084,
      "grad_norm": 0.0005327638355083764,
      "learning_rate": 0.11671539231700531,
      "loss": 0.3301,
      "num_input_tokens_seen": 43319152,
      "step": 22850
    },
    {
      "epoch": 3.728444408189901,
      "grad_norm": 0.0025680840481072664,
      "learning_rate": 0.11665795851951084,
      "loss": 0.3239,
      "num_input_tokens_seen": 43328160,
      "step": 22855
    },
    {
      "epoch": 3.7292601354107187,
      "grad_norm": 0.002499379450455308,
      "learning_rate": 0.11660052986377825,
      "loss": 0.3631,
      "num_input_tokens_seen": 43337648,
      "step": 22860
    },
    {
      "epoch": 3.7300758626315362,
      "grad_norm": 0.002251829719170928,
      "learning_rate": 0.1165431063586636,
      "loss": 0.367,
      "num_input_tokens_seen": 43347872,
      "step": 22865
    },
    {
      "epoch": 3.7308915898523534,
      "grad_norm": 0.001676475047133863,
      "learning_rate": 0.11648568801302245,
      "loss": 0.3313,
      "num_input_tokens_seen": 43357728,
      "step": 22870
    },
    {
      "epoch": 3.7317073170731705,
      "grad_norm": 0.0021186391822993755,
      "learning_rate": 0.11642827483570937,
      "loss": 0.2884,
      "num_input_tokens_seen": 43366544,
      "step": 22875
    },
    {
      "epoch": 3.732523044293988,
      "grad_norm": 0.003554231021553278,
      "learning_rate": 0.11637086683557815,
      "loss": 0.328,
      "num_input_tokens_seen": 43375824,
      "step": 22880
    },
    {
      "epoch": 3.7333387715148056,
      "grad_norm": 0.0013245706213638186,
      "learning_rate": 0.11631346402148188,
      "loss": 0.3044,
      "num_input_tokens_seen": 43385376,
      "step": 22885
    },
    {
      "epoch": 3.7341544987356228,
      "grad_norm": 0.0018878959817811847,
      "learning_rate": 0.11625606640227285,
      "loss": 0.3499,
      "num_input_tokens_seen": 43394320,
      "step": 22890
    },
    {
      "epoch": 3.73497022595644,
      "grad_norm": 0.0020325854420661926,
      "learning_rate": 0.11619867398680238,
      "loss": 0.4298,
      "num_input_tokens_seen": 43402688,
      "step": 22895
    },
    {
      "epoch": 3.7357859531772575,
      "grad_norm": 0.0021092179231345654,
      "learning_rate": 0.11614128678392119,
      "loss": 0.3593,
      "num_input_tokens_seen": 43411872,
      "step": 22900
    },
    {
      "epoch": 3.736601680398075,
      "grad_norm": 0.0016686159651726484,
      "learning_rate": 0.11608390480247906,
      "loss": 0.3123,
      "num_input_tokens_seen": 43420944,
      "step": 22905
    },
    {
      "epoch": 3.737417407618892,
      "grad_norm": 0.0016182645922526717,
      "learning_rate": 0.11602652805132499,
      "loss": 0.3266,
      "num_input_tokens_seen": 43429824,
      "step": 22910
    },
    {
      "epoch": 3.7382331348397098,
      "grad_norm": 0.0023415868636220694,
      "learning_rate": 0.11596915653930731,
      "loss": 0.368,
      "num_input_tokens_seen": 43438576,
      "step": 22915
    },
    {
      "epoch": 3.739048862060527,
      "grad_norm": 0.0014094669604673982,
      "learning_rate": 0.11591179027527328,
      "loss": 0.3193,
      "num_input_tokens_seen": 43448576,
      "step": 22920
    },
    {
      "epoch": 3.7398645892813445,
      "grad_norm": 0.001365010510198772,
      "learning_rate": 0.11585442926806956,
      "loss": 0.3306,
      "num_input_tokens_seen": 43458224,
      "step": 22925
    },
    {
      "epoch": 3.7406803165021616,
      "grad_norm": 0.001295529305934906,
      "learning_rate": 0.11579707352654202,
      "loss": 0.3768,
      "num_input_tokens_seen": 43465888,
      "step": 22930
    },
    {
      "epoch": 3.741496043722979,
      "grad_norm": 0.0011144106974825263,
      "learning_rate": 0.11573972305953548,
      "loss": 0.3491,
      "num_input_tokens_seen": 43475216,
      "step": 22935
    },
    {
      "epoch": 3.7423117709437963,
      "grad_norm": 0.002253815298900008,
      "learning_rate": 0.11568237787589426,
      "loss": 0.2783,
      "num_input_tokens_seen": 43484992,
      "step": 22940
    },
    {
      "epoch": 3.743127498164614,
      "grad_norm": 0.0007651429041288793,
      "learning_rate": 0.11562503798446161,
      "loss": 0.3155,
      "num_input_tokens_seen": 43493712,
      "step": 22945
    },
    {
      "epoch": 3.743943225385431,
      "grad_norm": 0.0009211970609612763,
      "learning_rate": 0.11556770339408005,
      "loss": 0.2987,
      "num_input_tokens_seen": 43502880,
      "step": 22950
    },
    {
      "epoch": 3.7447589526062486,
      "grad_norm": 0.00164534873329103,
      "learning_rate": 0.1155103741135914,
      "loss": 0.3561,
      "num_input_tokens_seen": 43511920,
      "step": 22955
    },
    {
      "epoch": 3.7455746798270657,
      "grad_norm": 0.0007480328204110265,
      "learning_rate": 0.1154530501518364,
      "loss": 0.3154,
      "num_input_tokens_seen": 43521312,
      "step": 22960
    },
    {
      "epoch": 3.7463904070478833,
      "grad_norm": 0.0014563172589987516,
      "learning_rate": 0.11539573151765523,
      "loss": 0.3641,
      "num_input_tokens_seen": 43531680,
      "step": 22965
    },
    {
      "epoch": 3.7472061342687004,
      "grad_norm": 0.0024716064799576998,
      "learning_rate": 0.11533841821988719,
      "loss": 0.3257,
      "num_input_tokens_seen": 43541008,
      "step": 22970
    },
    {
      "epoch": 3.748021861489518,
      "grad_norm": 0.002937988145276904,
      "learning_rate": 0.11528111026737059,
      "loss": 0.3346,
      "num_input_tokens_seen": 43551168,
      "step": 22975
    },
    {
      "epoch": 3.748837588710335,
      "grad_norm": 0.002128890249878168,
      "learning_rate": 0.11522380766894312,
      "loss": 0.3598,
      "num_input_tokens_seen": 43560640,
      "step": 22980
    },
    {
      "epoch": 3.7496533159311527,
      "grad_norm": 0.0009078148868866265,
      "learning_rate": 0.11516651043344152,
      "loss": 0.329,
      "num_input_tokens_seen": 43569472,
      "step": 22985
    },
    {
      "epoch": 3.75046904315197,
      "grad_norm": 0.0005963940056972206,
      "learning_rate": 0.11510921856970172,
      "loss": 0.3069,
      "num_input_tokens_seen": 43578720,
      "step": 22990
    },
    {
      "epoch": 3.7512847703727874,
      "grad_norm": 0.0020838978234678507,
      "learning_rate": 0.11505193208655895,
      "loss": 0.3218,
      "num_input_tokens_seen": 43588112,
      "step": 22995
    },
    {
      "epoch": 3.7521004975936045,
      "grad_norm": 0.0014978526160120964,
      "learning_rate": 0.11499465099284738,
      "loss": 0.3379,
      "num_input_tokens_seen": 43597776,
      "step": 23000
    },
    {
      "epoch": 3.7521004975936045,
      "eval_loss": 0.32120680809020996,
      "eval_runtime": 152.7617,
      "eval_samples_per_second": 17.838,
      "eval_steps_per_second": 8.922,
      "num_input_tokens_seen": 43597776,
      "step": 23000
    },
    {
      "epoch": 3.752916224814422,
      "grad_norm": 0.0019121362129226327,
      "learning_rate": 0.1149373752974006,
      "loss": 0.2986,
      "num_input_tokens_seen": 43606992,
      "step": 23005
    },
    {
      "epoch": 3.7537319520352392,
      "grad_norm": 0.0024880466517060995,
      "learning_rate": 0.11488010500905109,
      "loss": 0.3475,
      "num_input_tokens_seen": 43617072,
      "step": 23010
    },
    {
      "epoch": 3.754547679256057,
      "grad_norm": 0.0015823739813640714,
      "learning_rate": 0.11482284013663077,
      "loss": 0.3574,
      "num_input_tokens_seen": 43627424,
      "step": 23015
    },
    {
      "epoch": 3.7553634064768744,
      "grad_norm": 0.0009513837867416441,
      "learning_rate": 0.11476558068897061,
      "loss": 0.2943,
      "num_input_tokens_seen": 43637728,
      "step": 23020
    },
    {
      "epoch": 3.7561791336976915,
      "grad_norm": 0.0012954541016370058,
      "learning_rate": 0.11470832667490061,
      "loss": 0.4507,
      "num_input_tokens_seen": 43646784,
      "step": 23025
    },
    {
      "epoch": 3.7569948609185086,
      "grad_norm": 0.0011287034722045064,
      "learning_rate": 0.11465107810325013,
      "loss": 0.3385,
      "num_input_tokens_seen": 43656416,
      "step": 23030
    },
    {
      "epoch": 3.7578105881393262,
      "grad_norm": 0.0008253912092186511,
      "learning_rate": 0.11459383498284771,
      "loss": 0.3447,
      "num_input_tokens_seen": 43665632,
      "step": 23035
    },
    {
      "epoch": 3.758626315360144,
      "grad_norm": 0.0013090100837871432,
      "learning_rate": 0.11453659732252082,
      "loss": 0.3391,
      "num_input_tokens_seen": 43674624,
      "step": 23040
    },
    {
      "epoch": 3.759442042580961,
      "grad_norm": 0.001885438454337418,
      "learning_rate": 0.11447936513109633,
      "loss": 0.3387,
      "num_input_tokens_seen": 43684416,
      "step": 23045
    },
    {
      "epoch": 3.760257769801778,
      "grad_norm": 0.0007665348239243031,
      "learning_rate": 0.11442213841740011,
      "loss": 0.2848,
      "num_input_tokens_seen": 43692928,
      "step": 23050
    },
    {
      "epoch": 3.7610734970225956,
      "grad_norm": 0.0014041912509128451,
      "learning_rate": 0.1143649171902572,
      "loss": 0.368,
      "num_input_tokens_seen": 43703456,
      "step": 23055
    },
    {
      "epoch": 3.761889224243413,
      "grad_norm": 0.0016130743315443397,
      "learning_rate": 0.11430770145849194,
      "loss": 0.2683,
      "num_input_tokens_seen": 43711632,
      "step": 23060
    },
    {
      "epoch": 3.7627049514642303,
      "grad_norm": 0.0018049435457214713,
      "learning_rate": 0.11425049123092756,
      "loss": 0.3094,
      "num_input_tokens_seen": 43721008,
      "step": 23065
    },
    {
      "epoch": 3.7635206786850475,
      "grad_norm": 0.0011454178020358086,
      "learning_rate": 0.11419328651638674,
      "loss": 0.3802,
      "num_input_tokens_seen": 43730736,
      "step": 23070
    },
    {
      "epoch": 3.764336405905865,
      "grad_norm": 0.0024524489417672157,
      "learning_rate": 0.11413608732369115,
      "loss": 0.2654,
      "num_input_tokens_seen": 43741072,
      "step": 23075
    },
    {
      "epoch": 3.7651521331266826,
      "grad_norm": 0.0029239682480692863,
      "learning_rate": 0.11407889366166153,
      "loss": 0.3602,
      "num_input_tokens_seen": 43751296,
      "step": 23080
    },
    {
      "epoch": 3.7659678603474998,
      "grad_norm": 0.0010980911320075393,
      "learning_rate": 0.11402170553911797,
      "loss": 0.3152,
      "num_input_tokens_seen": 43760304,
      "step": 23085
    },
    {
      "epoch": 3.766783587568317,
      "grad_norm": 0.0021628239192068577,
      "learning_rate": 0.11396452296487955,
      "loss": 0.3894,
      "num_input_tokens_seen": 43769808,
      "step": 23090
    },
    {
      "epoch": 3.7675993147891345,
      "grad_norm": 0.0022727802861481905,
      "learning_rate": 0.11390734594776449,
      "loss": 0.3543,
      "num_input_tokens_seen": 43778528,
      "step": 23095
    },
    {
      "epoch": 3.768415042009952,
      "grad_norm": 0.0011819108622148633,
      "learning_rate": 0.11385017449659031,
      "loss": 0.315,
      "num_input_tokens_seen": 43789632,
      "step": 23100
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 0.0015528061194345355,
      "learning_rate": 0.11379300862017344,
      "loss": 0.3647,
      "num_input_tokens_seen": 43799344,
      "step": 23105
    },
    {
      "epoch": 3.7700464964515867,
      "grad_norm": 0.0008351426804438233,
      "learning_rate": 0.11373584832732966,
      "loss": 0.3225,
      "num_input_tokens_seen": 43808656,
      "step": 23110
    },
    {
      "epoch": 3.770862223672404,
      "grad_norm": 0.0018051665974780917,
      "learning_rate": 0.11367869362687386,
      "loss": 0.2773,
      "num_input_tokens_seen": 43817504,
      "step": 23115
    },
    {
      "epoch": 3.7716779508932214,
      "grad_norm": 0.001237542019225657,
      "learning_rate": 0.11362154452761988,
      "loss": 0.3137,
      "num_input_tokens_seen": 43827200,
      "step": 23120
    },
    {
      "epoch": 3.7724936781140386,
      "grad_norm": 0.0029575140215456486,
      "learning_rate": 0.11356440103838095,
      "loss": 0.3593,
      "num_input_tokens_seen": 43836976,
      "step": 23125
    },
    {
      "epoch": 3.773309405334856,
      "grad_norm": 0.0011740530608221889,
      "learning_rate": 0.11350726316796922,
      "loss": 0.2982,
      "num_input_tokens_seen": 43846768,
      "step": 23130
    },
    {
      "epoch": 3.7741251325556733,
      "grad_norm": 0.0008700908510945737,
      "learning_rate": 0.11345013092519607,
      "loss": 0.2855,
      "num_input_tokens_seen": 43855056,
      "step": 23135
    },
    {
      "epoch": 3.774940859776491,
      "grad_norm": 0.0017002723179757595,
      "learning_rate": 0.11339300431887213,
      "loss": 0.3637,
      "num_input_tokens_seen": 43864032,
      "step": 23140
    },
    {
      "epoch": 3.775756586997308,
      "grad_norm": 0.0013573506148532033,
      "learning_rate": 0.11333588335780687,
      "loss": 0.429,
      "num_input_tokens_seen": 43874160,
      "step": 23145
    },
    {
      "epoch": 3.7765723142181256,
      "grad_norm": 0.002169795800000429,
      "learning_rate": 0.11327876805080916,
      "loss": 0.3233,
      "num_input_tokens_seen": 43883200,
      "step": 23150
    },
    {
      "epoch": 3.7773880414389427,
      "grad_norm": 0.0014827981358394027,
      "learning_rate": 0.11322165840668696,
      "loss": 0.3349,
      "num_input_tokens_seen": 43893152,
      "step": 23155
    },
    {
      "epoch": 3.7782037686597603,
      "grad_norm": 0.0018250124994665384,
      "learning_rate": 0.11316455443424717,
      "loss": 0.3479,
      "num_input_tokens_seen": 43903408,
      "step": 23160
    },
    {
      "epoch": 3.7790194958805774,
      "grad_norm": 0.0018769848393276334,
      "learning_rate": 0.11310745614229603,
      "loss": 0.3151,
      "num_input_tokens_seen": 43913568,
      "step": 23165
    },
    {
      "epoch": 3.779835223101395,
      "grad_norm": 0.0011568805202841759,
      "learning_rate": 0.1130503635396387,
      "loss": 0.2961,
      "num_input_tokens_seen": 43923296,
      "step": 23170
    },
    {
      "epoch": 3.780650950322212,
      "grad_norm": 0.002911169547587633,
      "learning_rate": 0.11299327663507966,
      "loss": 0.3447,
      "num_input_tokens_seen": 43932880,
      "step": 23175
    },
    {
      "epoch": 3.7814666775430297,
      "grad_norm": 0.001877893810160458,
      "learning_rate": 0.11293619543742246,
      "loss": 0.3583,
      "num_input_tokens_seen": 43942448,
      "step": 23180
    },
    {
      "epoch": 3.782282404763847,
      "grad_norm": 0.0018297478090971708,
      "learning_rate": 0.11287911995546965,
      "loss": 0.3401,
      "num_input_tokens_seen": 43951376,
      "step": 23185
    },
    {
      "epoch": 3.7830981319846644,
      "grad_norm": 0.0009428522898815572,
      "learning_rate": 0.11282205019802308,
      "loss": 0.3647,
      "num_input_tokens_seen": 43960688,
      "step": 23190
    },
    {
      "epoch": 3.783913859205482,
      "grad_norm": 0.0009388491162098944,
      "learning_rate": 0.11276498617388354,
      "loss": 0.3052,
      "num_input_tokens_seen": 43969728,
      "step": 23195
    },
    {
      "epoch": 3.784729586426299,
      "grad_norm": 0.0003787367604672909,
      "learning_rate": 0.11270792789185109,
      "loss": 0.3036,
      "num_input_tokens_seen": 43979312,
      "step": 23200
    },
    {
      "epoch": 3.784729586426299,
      "eval_loss": 0.3241947591304779,
      "eval_runtime": 152.881,
      "eval_samples_per_second": 17.824,
      "eval_steps_per_second": 8.915,
      "num_input_tokens_seen": 43979312,
      "step": 23200
    },
    {
      "epoch": 3.785545313647116,
      "grad_norm": 0.0029898073989897966,
      "learning_rate": 0.11265087536072482,
      "loss": 0.3442,
      "num_input_tokens_seen": 43989600,
      "step": 23205
    },
    {
      "epoch": 3.786361040867934,
      "grad_norm": 0.0007267753826454282,
      "learning_rate": 0.11259382858930288,
      "loss": 0.3119,
      "num_input_tokens_seen": 43999376,
      "step": 23210
    },
    {
      "epoch": 3.7871767680887514,
      "grad_norm": 0.0012964395573362708,
      "learning_rate": 0.11253678758638262,
      "loss": 0.3587,
      "num_input_tokens_seen": 44008400,
      "step": 23215
    },
    {
      "epoch": 3.7879924953095685,
      "grad_norm": 0.002892403630539775,
      "learning_rate": 0.11247975236076059,
      "loss": 0.3403,
      "num_input_tokens_seen": 44017520,
      "step": 23220
    },
    {
      "epoch": 3.7888082225303856,
      "grad_norm": 0.002671052236109972,
      "learning_rate": 0.11242272292123218,
      "loss": 0.3467,
      "num_input_tokens_seen": 44027424,
      "step": 23225
    },
    {
      "epoch": 3.789623949751203,
      "grad_norm": 0.002495348919183016,
      "learning_rate": 0.11236569927659217,
      "loss": 0.3278,
      "num_input_tokens_seen": 44037312,
      "step": 23230
    },
    {
      "epoch": 3.7904396769720208,
      "grad_norm": 0.0019077523611485958,
      "learning_rate": 0.11230868143563429,
      "loss": 0.3614,
      "num_input_tokens_seen": 44047488,
      "step": 23235
    },
    {
      "epoch": 3.791255404192838,
      "grad_norm": 0.0032134673092514277,
      "learning_rate": 0.11225166940715131,
      "loss": 0.3456,
      "num_input_tokens_seen": 44057232,
      "step": 23240
    },
    {
      "epoch": 3.792071131413655,
      "grad_norm": 0.004433479160070419,
      "learning_rate": 0.11219466319993537,
      "loss": 0.3685,
      "num_input_tokens_seen": 44066704,
      "step": 23245
    },
    {
      "epoch": 3.7928868586344726,
      "grad_norm": 0.0020936671644449234,
      "learning_rate": 0.11213766282277739,
      "loss": 0.3433,
      "num_input_tokens_seen": 44076432,
      "step": 23250
    },
    {
      "epoch": 3.79370258585529,
      "grad_norm": 0.0009458170970901847,
      "learning_rate": 0.11208066828446761,
      "loss": 0.3221,
      "num_input_tokens_seen": 44084848,
      "step": 23255
    },
    {
      "epoch": 3.7945183130761073,
      "grad_norm": 0.0027802479453384876,
      "learning_rate": 0.11202367959379537,
      "loss": 0.2907,
      "num_input_tokens_seen": 44094208,
      "step": 23260
    },
    {
      "epoch": 3.7953340402969244,
      "grad_norm": 0.0015997940208762884,
      "learning_rate": 0.11196669675954894,
      "loss": 0.3298,
      "num_input_tokens_seen": 44103904,
      "step": 23265
    },
    {
      "epoch": 3.796149767517742,
      "grad_norm": 0.0009327677544206381,
      "learning_rate": 0.1119097197905158,
      "loss": 0.2806,
      "num_input_tokens_seen": 44113760,
      "step": 23270
    },
    {
      "epoch": 3.7969654947385596,
      "grad_norm": 0.0006034907419234514,
      "learning_rate": 0.11185274869548259,
      "loss": 0.3705,
      "num_input_tokens_seen": 44123312,
      "step": 23275
    },
    {
      "epoch": 3.7977812219593767,
      "grad_norm": 0.0013638610253110528,
      "learning_rate": 0.11179578348323486,
      "loss": 0.2845,
      "num_input_tokens_seen": 44131984,
      "step": 23280
    },
    {
      "epoch": 3.7985969491801943,
      "grad_norm": 0.001319423085078597,
      "learning_rate": 0.1117388241625575,
      "loss": 0.3304,
      "num_input_tokens_seen": 44141504,
      "step": 23285
    },
    {
      "epoch": 3.7994126764010114,
      "grad_norm": 0.001391467871144414,
      "learning_rate": 0.11168187074223421,
      "loss": 0.3163,
      "num_input_tokens_seen": 44151120,
      "step": 23290
    },
    {
      "epoch": 3.800228403621829,
      "grad_norm": 0.0017861009109765291,
      "learning_rate": 0.11162492323104796,
      "loss": 0.3879,
      "num_input_tokens_seen": 44160752,
      "step": 23295
    },
    {
      "epoch": 3.801044130842646,
      "grad_norm": 0.0018359172390773892,
      "learning_rate": 0.11156798163778091,
      "loss": 0.3191,
      "num_input_tokens_seen": 44169072,
      "step": 23300
    },
    {
      "epoch": 3.8018598580634637,
      "grad_norm": 0.0017156116664409637,
      "learning_rate": 0.11151104597121399,
      "loss": 0.3242,
      "num_input_tokens_seen": 44178528,
      "step": 23305
    },
    {
      "epoch": 3.802675585284281,
      "grad_norm": 0.0019262614659965038,
      "learning_rate": 0.11145411624012742,
      "loss": 0.3115,
      "num_input_tokens_seen": 44189536,
      "step": 23310
    },
    {
      "epoch": 3.8034913125050984,
      "grad_norm": 0.002032379386946559,
      "learning_rate": 0.11139719245330063,
      "loss": 0.4191,
      "num_input_tokens_seen": 44198144,
      "step": 23315
    },
    {
      "epoch": 3.8043070397259156,
      "grad_norm": 0.0015435984823852777,
      "learning_rate": 0.11134027461951179,
      "loss": 0.3469,
      "num_input_tokens_seen": 44207168,
      "step": 23320
    },
    {
      "epoch": 3.805122766946733,
      "grad_norm": 0.0021973508410155773,
      "learning_rate": 0.11128336274753849,
      "loss": 0.3487,
      "num_input_tokens_seen": 44216816,
      "step": 23325
    },
    {
      "epoch": 3.8059384941675503,
      "grad_norm": 0.002074124291539192,
      "learning_rate": 0.11122645684615715,
      "loss": 0.3214,
      "num_input_tokens_seen": 44225632,
      "step": 23330
    },
    {
      "epoch": 3.806754221388368,
      "grad_norm": 0.0009243948734365404,
      "learning_rate": 0.11116955692414345,
      "loss": 0.3234,
      "num_input_tokens_seen": 44234928,
      "step": 23335
    },
    {
      "epoch": 3.807569948609185,
      "grad_norm": 0.00168658047914505,
      "learning_rate": 0.11111266299027203,
      "loss": 0.3076,
      "num_input_tokens_seen": 44244224,
      "step": 23340
    },
    {
      "epoch": 3.8083856758300025,
      "grad_norm": 0.0015232901787385345,
      "learning_rate": 0.11105577505331668,
      "loss": 0.3272,
      "num_input_tokens_seen": 44253248,
      "step": 23345
    },
    {
      "epoch": 3.8092014030508197,
      "grad_norm": 0.0016471486305817962,
      "learning_rate": 0.11099889312205018,
      "loss": 0.3365,
      "num_input_tokens_seen": 44261168,
      "step": 23350
    },
    {
      "epoch": 3.8100171302716372,
      "grad_norm": 0.0030925353057682514,
      "learning_rate": 0.11094201720524455,
      "loss": 0.3897,
      "num_input_tokens_seen": 44270832,
      "step": 23355
    },
    {
      "epoch": 3.8108328574924544,
      "grad_norm": 0.001328104524873197,
      "learning_rate": 0.11088514731167064,
      "loss": 0.3346,
      "num_input_tokens_seen": 44280896,
      "step": 23360
    },
    {
      "epoch": 3.811648584713272,
      "grad_norm": 0.0014640935696661472,
      "learning_rate": 0.11082828345009862,
      "loss": 0.3321,
      "num_input_tokens_seen": 44291280,
      "step": 23365
    },
    {
      "epoch": 3.812464311934089,
      "grad_norm": 0.0026275143027305603,
      "learning_rate": 0.11077142562929748,
      "loss": 0.3862,
      "num_input_tokens_seen": 44300400,
      "step": 23370
    },
    {
      "epoch": 3.8132800391549067,
      "grad_norm": 0.0013328019995242357,
      "learning_rate": 0.11071457385803554,
      "loss": 0.3203,
      "num_input_tokens_seen": 44309488,
      "step": 23375
    },
    {
      "epoch": 3.814095766375724,
      "grad_norm": 0.002264600247144699,
      "learning_rate": 0.11065772814508001,
      "loss": 0.2908,
      "num_input_tokens_seen": 44318944,
      "step": 23380
    },
    {
      "epoch": 3.8149114935965414,
      "grad_norm": 0.0015858119586482644,
      "learning_rate": 0.11060088849919715,
      "loss": 0.3396,
      "num_input_tokens_seen": 44328112,
      "step": 23385
    },
    {
      "epoch": 3.815727220817359,
      "grad_norm": 0.0013651904882863164,
      "learning_rate": 0.11054405492915244,
      "loss": 0.3388,
      "num_input_tokens_seen": 44336960,
      "step": 23390
    },
    {
      "epoch": 3.816542948038176,
      "grad_norm": 0.0027406844310462475,
      "learning_rate": 0.11048722744371031,
      "loss": 0.3415,
      "num_input_tokens_seen": 44345600,
      "step": 23395
    },
    {
      "epoch": 3.817358675258993,
      "grad_norm": 0.0037799240089952946,
      "learning_rate": 0.1104304060516342,
      "loss": 0.3774,
      "num_input_tokens_seen": 44354480,
      "step": 23400
    },
    {
      "epoch": 3.817358675258993,
      "eval_loss": 0.3327629566192627,
      "eval_runtime": 152.8142,
      "eval_samples_per_second": 17.832,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 44354480,
      "step": 23400
    },
    {
      "epoch": 3.8181744024798108,
      "grad_norm": 0.001453216653317213,
      "learning_rate": 0.11037359076168682,
      "loss": 0.2888,
      "num_input_tokens_seen": 44363680,
      "step": 23405
    },
    {
      "epoch": 3.8189901297006283,
      "grad_norm": 0.0015431493520736694,
      "learning_rate": 0.11031678158262966,
      "loss": 0.3162,
      "num_input_tokens_seen": 44372704,
      "step": 23410
    },
    {
      "epoch": 3.8198058569214455,
      "grad_norm": 0.0017223951872438192,
      "learning_rate": 0.11025997852322349,
      "loss": 0.3376,
      "num_input_tokens_seen": 44381104,
      "step": 23415
    },
    {
      "epoch": 3.8206215841422626,
      "grad_norm": 0.0023081665858626366,
      "learning_rate": 0.11020318159222807,
      "loss": 0.3033,
      "num_input_tokens_seen": 44388736,
      "step": 23420
    },
    {
      "epoch": 3.82143731136308,
      "grad_norm": 0.0014033755287528038,
      "learning_rate": 0.1101463907984021,
      "loss": 0.3648,
      "num_input_tokens_seen": 44398160,
      "step": 23425
    },
    {
      "epoch": 3.8222530385838978,
      "grad_norm": 0.001998164225369692,
      "learning_rate": 0.11008960615050352,
      "loss": 0.2589,
      "num_input_tokens_seen": 44408176,
      "step": 23430
    },
    {
      "epoch": 3.823068765804715,
      "grad_norm": 0.0019995849579572678,
      "learning_rate": 0.11003282765728925,
      "loss": 0.333,
      "num_input_tokens_seen": 44418496,
      "step": 23435
    },
    {
      "epoch": 3.823884493025532,
      "grad_norm": 0.0025092875584959984,
      "learning_rate": 0.10997605532751518,
      "loss": 0.2806,
      "num_input_tokens_seen": 44429584,
      "step": 23440
    },
    {
      "epoch": 3.8247002202463496,
      "grad_norm": 0.0012650584103539586,
      "learning_rate": 0.1099192891699364,
      "loss": 0.3623,
      "num_input_tokens_seen": 44439568,
      "step": 23445
    },
    {
      "epoch": 3.825515947467167,
      "grad_norm": 0.0019774548709392548,
      "learning_rate": 0.10986252919330687,
      "loss": 0.4001,
      "num_input_tokens_seen": 44449264,
      "step": 23450
    },
    {
      "epoch": 3.8263316746879843,
      "grad_norm": 0.0020326103549450636,
      "learning_rate": 0.10980577540637973,
      "loss": 0.37,
      "num_input_tokens_seen": 44458368,
      "step": 23455
    },
    {
      "epoch": 3.8271474019088014,
      "grad_norm": 0.0017390528228133917,
      "learning_rate": 0.10974902781790719,
      "loss": 0.2897,
      "num_input_tokens_seen": 44467136,
      "step": 23460
    },
    {
      "epoch": 3.827963129129619,
      "grad_norm": 0.0023488581646233797,
      "learning_rate": 0.10969228643664032,
      "loss": 0.3336,
      "num_input_tokens_seen": 44477584,
      "step": 23465
    },
    {
      "epoch": 3.8287788563504366,
      "grad_norm": 0.001137827872298658,
      "learning_rate": 0.10963555127132942,
      "loss": 0.3112,
      "num_input_tokens_seen": 44487392,
      "step": 23470
    },
    {
      "epoch": 3.8295945835712537,
      "grad_norm": 0.0019557750783860683,
      "learning_rate": 0.10957882233072382,
      "loss": 0.303,
      "num_input_tokens_seen": 44496528,
      "step": 23475
    },
    {
      "epoch": 3.8304103107920713,
      "grad_norm": 0.001986923860386014,
      "learning_rate": 0.10952209962357176,
      "loss": 0.3773,
      "num_input_tokens_seen": 44504880,
      "step": 23480
    },
    {
      "epoch": 3.8312260380128884,
      "grad_norm": 0.0025890052784234285,
      "learning_rate": 0.10946538315862062,
      "loss": 0.3275,
      "num_input_tokens_seen": 44514464,
      "step": 23485
    },
    {
      "epoch": 3.832041765233706,
      "grad_norm": 0.002366316271945834,
      "learning_rate": 0.10940867294461679,
      "loss": 0.3267,
      "num_input_tokens_seen": 44523792,
      "step": 23490
    },
    {
      "epoch": 3.832857492454523,
      "grad_norm": 0.0017083869315683842,
      "learning_rate": 0.10935196899030565,
      "loss": 0.3052,
      "num_input_tokens_seen": 44532944,
      "step": 23495
    },
    {
      "epoch": 3.8336732196753407,
      "grad_norm": 0.002230934100225568,
      "learning_rate": 0.10929527130443177,
      "loss": 0.3555,
      "num_input_tokens_seen": 44542608,
      "step": 23500
    },
    {
      "epoch": 3.834488946896158,
      "grad_norm": 0.0025350030045956373,
      "learning_rate": 0.1092385798957385,
      "loss": 0.3755,
      "num_input_tokens_seen": 44552000,
      "step": 23505
    },
    {
      "epoch": 3.8353046741169754,
      "grad_norm": 0.000665290979668498,
      "learning_rate": 0.10918189477296848,
      "loss": 0.3485,
      "num_input_tokens_seen": 44560976,
      "step": 23510
    },
    {
      "epoch": 3.8361204013377925,
      "grad_norm": 0.0016542780213057995,
      "learning_rate": 0.1091252159448633,
      "loss": 0.3428,
      "num_input_tokens_seen": 44570256,
      "step": 23515
    },
    {
      "epoch": 3.83693612855861,
      "grad_norm": 0.0008189543732441962,
      "learning_rate": 0.10906854342016345,
      "loss": 0.2939,
      "num_input_tokens_seen": 44579968,
      "step": 23520
    },
    {
      "epoch": 3.8377518557794272,
      "grad_norm": 0.0013051561545580626,
      "learning_rate": 0.10901187720760858,
      "loss": 0.3196,
      "num_input_tokens_seen": 44587680,
      "step": 23525
    },
    {
      "epoch": 3.838567583000245,
      "grad_norm": 0.0010055802995339036,
      "learning_rate": 0.10895521731593734,
      "loss": 0.3111,
      "num_input_tokens_seen": 44596112,
      "step": 23530
    },
    {
      "epoch": 3.839383310221062,
      "grad_norm": 0.0018135817954316735,
      "learning_rate": 0.10889856375388733,
      "loss": 0.343,
      "num_input_tokens_seen": 44606288,
      "step": 23535
    },
    {
      "epoch": 3.8401990374418795,
      "grad_norm": 0.0009594417060725391,
      "learning_rate": 0.1088419165301954,
      "loss": 0.2715,
      "num_input_tokens_seen": 44616016,
      "step": 23540
    },
    {
      "epoch": 3.8410147646626966,
      "grad_norm": 0.0014990858035162091,
      "learning_rate": 0.1087852756535971,
      "loss": 0.3227,
      "num_input_tokens_seen": 44626176,
      "step": 23545
    },
    {
      "epoch": 3.841830491883514,
      "grad_norm": 0.0023516893852502108,
      "learning_rate": 0.10872864113282725,
      "loss": 0.2819,
      "num_input_tokens_seen": 44634304,
      "step": 23550
    },
    {
      "epoch": 3.8426462191043314,
      "grad_norm": 0.0009451790247112513,
      "learning_rate": 0.10867201297661958,
      "loss": 0.3575,
      "num_input_tokens_seen": 44642672,
      "step": 23555
    },
    {
      "epoch": 3.843461946325149,
      "grad_norm": 0.001673916820436716,
      "learning_rate": 0.10861539119370689,
      "loss": 0.3675,
      "num_input_tokens_seen": 44651408,
      "step": 23560
    },
    {
      "epoch": 3.8442776735459665,
      "grad_norm": 0.002313221339136362,
      "learning_rate": 0.10855877579282096,
      "loss": 0.3065,
      "num_input_tokens_seen": 44660400,
      "step": 23565
    },
    {
      "epoch": 3.8450934007667836,
      "grad_norm": 0.0016938206972554326,
      "learning_rate": 0.10850216678269252,
      "loss": 0.3465,
      "num_input_tokens_seen": 44669536,
      "step": 23570
    },
    {
      "epoch": 3.8459091279876008,
      "grad_norm": 0.0010593902552500367,
      "learning_rate": 0.10844556417205146,
      "loss": 0.3301,
      "num_input_tokens_seen": 44677536,
      "step": 23575
    },
    {
      "epoch": 3.8467248552084183,
      "grad_norm": 0.0028323365841060877,
      "learning_rate": 0.10838896796962669,
      "loss": 0.2837,
      "num_input_tokens_seen": 44687600,
      "step": 23580
    },
    {
      "epoch": 3.847540582429236,
      "grad_norm": 0.001688170013949275,
      "learning_rate": 0.1083323781841459,
      "loss": 0.3314,
      "num_input_tokens_seen": 44697248,
      "step": 23585
    },
    {
      "epoch": 3.848356309650053,
      "grad_norm": 0.00083962018834427,
      "learning_rate": 0.10827579482433607,
      "loss": 0.3365,
      "num_input_tokens_seen": 44706640,
      "step": 23590
    },
    {
      "epoch": 3.84917203687087,
      "grad_norm": 0.0009756334475241601,
      "learning_rate": 0.10821921789892304,
      "loss": 0.2711,
      "num_input_tokens_seen": 44717520,
      "step": 23595
    },
    {
      "epoch": 3.8499877640916877,
      "grad_norm": 0.0011350270360708237,
      "learning_rate": 0.10816264741663158,
      "loss": 0.3691,
      "num_input_tokens_seen": 44727696,
      "step": 23600
    },
    {
      "epoch": 3.8499877640916877,
      "eval_loss": 0.32497772574424744,
      "eval_runtime": 152.8578,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 44727696,
      "step": 23600
    },
    {
      "epoch": 3.8508034913125053,
      "grad_norm": 0.001380355330184102,
      "learning_rate": 0.10810608338618573,
      "loss": 0.3641,
      "num_input_tokens_seen": 44736704,
      "step": 23605
    },
    {
      "epoch": 3.8516192185333225,
      "grad_norm": 0.001566216815263033,
      "learning_rate": 0.10804952581630821,
      "loss": 0.3308,
      "num_input_tokens_seen": 44747200,
      "step": 23610
    },
    {
      "epoch": 3.8524349457541396,
      "grad_norm": 0.0012752393959090114,
      "learning_rate": 0.10799297471572102,
      "loss": 0.385,
      "num_input_tokens_seen": 44756272,
      "step": 23615
    },
    {
      "epoch": 3.853250672974957,
      "grad_norm": 0.0020533574279397726,
      "learning_rate": 0.10793643009314507,
      "loss": 0.316,
      "num_input_tokens_seen": 44766800,
      "step": 23620
    },
    {
      "epoch": 3.8540664001957747,
      "grad_norm": 0.0014036123175173998,
      "learning_rate": 0.10787989195730015,
      "loss": 0.3004,
      "num_input_tokens_seen": 44774992,
      "step": 23625
    },
    {
      "epoch": 3.854882127416592,
      "grad_norm": 0.0015250422293320298,
      "learning_rate": 0.10782336031690525,
      "loss": 0.3746,
      "num_input_tokens_seen": 44785104,
      "step": 23630
    },
    {
      "epoch": 3.855697854637409,
      "grad_norm": 0.0009473295649513602,
      "learning_rate": 0.10776683518067821,
      "loss": 0.2971,
      "num_input_tokens_seen": 44794320,
      "step": 23635
    },
    {
      "epoch": 3.8565135818582266,
      "grad_norm": 0.0013687964528799057,
      "learning_rate": 0.10771031655733587,
      "loss": 0.3062,
      "num_input_tokens_seen": 44804160,
      "step": 23640
    },
    {
      "epoch": 3.857329309079044,
      "grad_norm": 0.0015560073079541326,
      "learning_rate": 0.10765380445559422,
      "loss": 0.3579,
      "num_input_tokens_seen": 44812752,
      "step": 23645
    },
    {
      "epoch": 3.8581450362998613,
      "grad_norm": 0.0019878819584846497,
      "learning_rate": 0.10759729888416801,
      "loss": 0.3367,
      "num_input_tokens_seen": 44822848,
      "step": 23650
    },
    {
      "epoch": 3.858960763520679,
      "grad_norm": 0.0012171223061159253,
      "learning_rate": 0.10754079985177119,
      "loss": 0.319,
      "num_input_tokens_seen": 44832224,
      "step": 23655
    },
    {
      "epoch": 3.859776490741496,
      "grad_norm": 0.0014267676742747426,
      "learning_rate": 0.10748430736711667,
      "loss": 0.2935,
      "num_input_tokens_seen": 44841120,
      "step": 23660
    },
    {
      "epoch": 3.8605922179623136,
      "grad_norm": 0.001762240077368915,
      "learning_rate": 0.10742782143891623,
      "loss": 0.3114,
      "num_input_tokens_seen": 44851424,
      "step": 23665
    },
    {
      "epoch": 3.8614079451831307,
      "grad_norm": 0.0013986044796183705,
      "learning_rate": 0.10737134207588069,
      "loss": 0.3286,
      "num_input_tokens_seen": 44861360,
      "step": 23670
    },
    {
      "epoch": 3.8622236724039483,
      "grad_norm": 0.0008306295494548976,
      "learning_rate": 0.10731486928671992,
      "loss": 0.3682,
      "num_input_tokens_seen": 44871616,
      "step": 23675
    },
    {
      "epoch": 3.8630393996247654,
      "grad_norm": 0.001893728389404714,
      "learning_rate": 0.10725840308014269,
      "loss": 0.2937,
      "num_input_tokens_seen": 44881488,
      "step": 23680
    },
    {
      "epoch": 3.863855126845583,
      "grad_norm": 0.0016992958262562752,
      "learning_rate": 0.10720194346485688,
      "loss": 0.3377,
      "num_input_tokens_seen": 44891104,
      "step": 23685
    },
    {
      "epoch": 3.8646708540664,
      "grad_norm": 0.0018323679687455297,
      "learning_rate": 0.10714549044956918,
      "loss": 0.3397,
      "num_input_tokens_seen": 44900704,
      "step": 23690
    },
    {
      "epoch": 3.8654865812872177,
      "grad_norm": 0.003165541449561715,
      "learning_rate": 0.10708904404298542,
      "loss": 0.3689,
      "num_input_tokens_seen": 44909968,
      "step": 23695
    },
    {
      "epoch": 3.866302308508035,
      "grad_norm": 0.0016366904601454735,
      "learning_rate": 0.1070326042538103,
      "loss": 0.3209,
      "num_input_tokens_seen": 44919584,
      "step": 23700
    },
    {
      "epoch": 3.8671180357288524,
      "grad_norm": 0.0013793697580695152,
      "learning_rate": 0.10697617109074758,
      "loss": 0.3183,
      "num_input_tokens_seen": 44928736,
      "step": 23705
    },
    {
      "epoch": 3.8679337629496695,
      "grad_norm": 0.002024526707828045,
      "learning_rate": 0.10691974456249999,
      "loss": 0.3467,
      "num_input_tokens_seen": 44938304,
      "step": 23710
    },
    {
      "epoch": 3.868749490170487,
      "grad_norm": 0.0005984354647807777,
      "learning_rate": 0.10686332467776909,
      "loss": 0.3446,
      "num_input_tokens_seen": 44947616,
      "step": 23715
    },
    {
      "epoch": 3.869565217391304,
      "grad_norm": 0.002082446124404669,
      "learning_rate": 0.10680691144525563,
      "loss": 0.3961,
      "num_input_tokens_seen": 44956560,
      "step": 23720
    },
    {
      "epoch": 3.870380944612122,
      "grad_norm": 0.0014188257046043873,
      "learning_rate": 0.10675050487365928,
      "loss": 0.3884,
      "num_input_tokens_seen": 44965376,
      "step": 23725
    },
    {
      "epoch": 3.871196671832939,
      "grad_norm": 0.0014003276592120528,
      "learning_rate": 0.10669410497167851,
      "loss": 0.3602,
      "num_input_tokens_seen": 44975344,
      "step": 23730
    },
    {
      "epoch": 3.8720123990537565,
      "grad_norm": 0.0012149678077548742,
      "learning_rate": 0.10663771174801102,
      "loss": 0.3276,
      "num_input_tokens_seen": 44984464,
      "step": 23735
    },
    {
      "epoch": 3.8728281262745736,
      "grad_norm": 0.0010843289783224463,
      "learning_rate": 0.10658132521135329,
      "loss": 0.3494,
      "num_input_tokens_seen": 44994048,
      "step": 23740
    },
    {
      "epoch": 3.873643853495391,
      "grad_norm": 0.0014994582161307335,
      "learning_rate": 0.10652494537040084,
      "loss": 0.3556,
      "num_input_tokens_seen": 45003616,
      "step": 23745
    },
    {
      "epoch": 3.8744595807162083,
      "grad_norm": 0.0010191194014623761,
      "learning_rate": 0.1064685722338482,
      "loss": 0.3403,
      "num_input_tokens_seen": 45013152,
      "step": 23750
    },
    {
      "epoch": 3.875275307937026,
      "grad_norm": 0.0014731461415067315,
      "learning_rate": 0.10641220581038871,
      "loss": 0.3394,
      "num_input_tokens_seen": 45023616,
      "step": 23755
    },
    {
      "epoch": 3.8760910351578435,
      "grad_norm": 0.0014252405380830169,
      "learning_rate": 0.10635584610871483,
      "loss": 0.3117,
      "num_input_tokens_seen": 45032576,
      "step": 23760
    },
    {
      "epoch": 3.8769067623786606,
      "grad_norm": 0.0010496334871277213,
      "learning_rate": 0.10629949313751803,
      "loss": 0.3419,
      "num_input_tokens_seen": 45041312,
      "step": 23765
    },
    {
      "epoch": 3.8777224895994777,
      "grad_norm": 0.0008985652821138501,
      "learning_rate": 0.10624314690548849,
      "loss": 0.3155,
      "num_input_tokens_seen": 45051040,
      "step": 23770
    },
    {
      "epoch": 3.8785382168202953,
      "grad_norm": 0.002116999588906765,
      "learning_rate": 0.1061868074213156,
      "loss": 0.3442,
      "num_input_tokens_seen": 45061088,
      "step": 23775
    },
    {
      "epoch": 3.879353944041113,
      "grad_norm": 0.0009879368590191007,
      "learning_rate": 0.10613047469368765,
      "loss": 0.3409,
      "num_input_tokens_seen": 45070912,
      "step": 23780
    },
    {
      "epoch": 3.88016967126193,
      "grad_norm": 0.0012836280511692166,
      "learning_rate": 0.10607414873129171,
      "loss": 0.2746,
      "num_input_tokens_seen": 45080368,
      "step": 23785
    },
    {
      "epoch": 3.880985398482747,
      "grad_norm": 0.0012642167275771499,
      "learning_rate": 0.10601782954281413,
      "loss": 0.2946,
      "num_input_tokens_seen": 45089392,
      "step": 23790
    },
    {
      "epoch": 3.8818011257035647,
      "grad_norm": 0.0009717497741803527,
      "learning_rate": 0.1059615171369399,
      "loss": 0.3357,
      "num_input_tokens_seen": 45098688,
      "step": 23795
    },
    {
      "epoch": 3.8826168529243823,
      "grad_norm": 0.0005050708423368633,
      "learning_rate": 0.10590521152235312,
      "loss": 0.361,
      "num_input_tokens_seen": 45108608,
      "step": 23800
    },
    {
      "epoch": 3.8826168529243823,
      "eval_loss": 0.32399341464042664,
      "eval_runtime": 152.7907,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 45108608,
      "step": 23800
    },
    {
      "epoch": 3.8834325801451994,
      "grad_norm": 0.002539447508752346,
      "learning_rate": 0.1058489127077369,
      "loss": 0.3667,
      "num_input_tokens_seen": 45117984,
      "step": 23805
    },
    {
      "epoch": 3.8842483073660166,
      "grad_norm": 0.0009270839509554207,
      "learning_rate": 0.1057926207017732,
      "loss": 0.2818,
      "num_input_tokens_seen": 45125056,
      "step": 23810
    },
    {
      "epoch": 3.885064034586834,
      "grad_norm": 0.0024351561442017555,
      "learning_rate": 0.10573633551314285,
      "loss": 0.4084,
      "num_input_tokens_seen": 45135680,
      "step": 23815
    },
    {
      "epoch": 3.8858797618076517,
      "grad_norm": 0.0018599331378936768,
      "learning_rate": 0.1056800571505259,
      "loss": 0.2797,
      "num_input_tokens_seen": 45145664,
      "step": 23820
    },
    {
      "epoch": 3.886695489028469,
      "grad_norm": 0.003335540881380439,
      "learning_rate": 0.10562378562260105,
      "loss": 0.3031,
      "num_input_tokens_seen": 45154832,
      "step": 23825
    },
    {
      "epoch": 3.887511216249286,
      "grad_norm": 0.002027421724051237,
      "learning_rate": 0.10556752093804615,
      "loss": 0.4021,
      "num_input_tokens_seen": 45164768,
      "step": 23830
    },
    {
      "epoch": 3.8883269434701035,
      "grad_norm": 0.0012560702161863446,
      "learning_rate": 0.10551126310553786,
      "loss": 0.3561,
      "num_input_tokens_seen": 45174096,
      "step": 23835
    },
    {
      "epoch": 3.889142670690921,
      "grad_norm": 0.0014153898227959871,
      "learning_rate": 0.10545501213375187,
      "loss": 0.3138,
      "num_input_tokens_seen": 45183680,
      "step": 23840
    },
    {
      "epoch": 3.8899583979117383,
      "grad_norm": 0.0021438393741846085,
      "learning_rate": 0.10539876803136287,
      "loss": 0.3923,
      "num_input_tokens_seen": 45193392,
      "step": 23845
    },
    {
      "epoch": 3.890774125132556,
      "grad_norm": 0.0011724402429535985,
      "learning_rate": 0.10534253080704428,
      "loss": 0.32,
      "num_input_tokens_seen": 45201056,
      "step": 23850
    },
    {
      "epoch": 3.891589852353373,
      "grad_norm": 0.0033126715570688248,
      "learning_rate": 0.10528630046946862,
      "loss": 0.3329,
      "num_input_tokens_seen": 45209392,
      "step": 23855
    },
    {
      "epoch": 3.8924055795741905,
      "grad_norm": 0.0015022065490484238,
      "learning_rate": 0.1052300770273074,
      "loss": 0.3478,
      "num_input_tokens_seen": 45219584,
      "step": 23860
    },
    {
      "epoch": 3.8932213067950077,
      "grad_norm": 0.0005838942597620189,
      "learning_rate": 0.10517386048923086,
      "loss": 0.329,
      "num_input_tokens_seen": 45228864,
      "step": 23865
    },
    {
      "epoch": 3.8940370340158252,
      "grad_norm": 0.0031067980453372,
      "learning_rate": 0.10511765086390841,
      "loss": 0.3431,
      "num_input_tokens_seen": 45238176,
      "step": 23870
    },
    {
      "epoch": 3.8948527612366424,
      "grad_norm": 0.001357305678538978,
      "learning_rate": 0.10506144816000816,
      "loss": 0.3063,
      "num_input_tokens_seen": 45248352,
      "step": 23875
    },
    {
      "epoch": 3.89566848845746,
      "grad_norm": 0.003340239869430661,
      "learning_rate": 0.10500525238619736,
      "loss": 0.3667,
      "num_input_tokens_seen": 45256352,
      "step": 23880
    },
    {
      "epoch": 3.896484215678277,
      "grad_norm": 0.001935178879648447,
      "learning_rate": 0.10494906355114209,
      "loss": 0.347,
      "num_input_tokens_seen": 45265840,
      "step": 23885
    },
    {
      "epoch": 3.8972999428990946,
      "grad_norm": 0.0009784818394109607,
      "learning_rate": 0.10489288166350737,
      "loss": 0.2922,
      "num_input_tokens_seen": 45275104,
      "step": 23890
    },
    {
      "epoch": 3.898115670119912,
      "grad_norm": 0.0016783918254077435,
      "learning_rate": 0.10483670673195711,
      "loss": 0.3213,
      "num_input_tokens_seen": 45285600,
      "step": 23895
    },
    {
      "epoch": 3.8989313973407294,
      "grad_norm": 0.0011113700456917286,
      "learning_rate": 0.10478053876515431,
      "loss": 0.346,
      "num_input_tokens_seen": 45294848,
      "step": 23900
    },
    {
      "epoch": 3.8997471245615465,
      "grad_norm": 0.0018458918202668428,
      "learning_rate": 0.10472437777176061,
      "loss": 0.3275,
      "num_input_tokens_seen": 45303648,
      "step": 23905
    },
    {
      "epoch": 3.900562851782364,
      "grad_norm": 0.0012288423022255301,
      "learning_rate": 0.1046682237604369,
      "loss": 0.3712,
      "num_input_tokens_seen": 45311568,
      "step": 23910
    },
    {
      "epoch": 3.901378579003181,
      "grad_norm": 0.002049647504463792,
      "learning_rate": 0.1046120767398427,
      "loss": 0.3644,
      "num_input_tokens_seen": 45320896,
      "step": 23915
    },
    {
      "epoch": 3.9021943062239988,
      "grad_norm": 0.003738614497706294,
      "learning_rate": 0.10455593671863667,
      "loss": 0.3529,
      "num_input_tokens_seen": 45330304,
      "step": 23920
    },
    {
      "epoch": 3.903010033444816,
      "grad_norm": 0.002802918665111065,
      "learning_rate": 0.1044998037054763,
      "loss": 0.3569,
      "num_input_tokens_seen": 45339008,
      "step": 23925
    },
    {
      "epoch": 3.9038257606656335,
      "grad_norm": 0.0009312286274507642,
      "learning_rate": 0.10444367770901794,
      "loss": 0.3171,
      "num_input_tokens_seen": 45348768,
      "step": 23930
    },
    {
      "epoch": 3.904641487886451,
      "grad_norm": 0.0015814137877896428,
      "learning_rate": 0.10438755873791698,
      "loss": 0.3101,
      "num_input_tokens_seen": 45357904,
      "step": 23935
    },
    {
      "epoch": 3.905457215107268,
      "grad_norm": 0.002969019114971161,
      "learning_rate": 0.10433144680082775,
      "loss": 0.384,
      "num_input_tokens_seen": 45368192,
      "step": 23940
    },
    {
      "epoch": 3.9062729423280853,
      "grad_norm": 0.0007805818459019065,
      "learning_rate": 0.10427534190640322,
      "loss": 0.3683,
      "num_input_tokens_seen": 45378240,
      "step": 23945
    },
    {
      "epoch": 3.907088669548903,
      "grad_norm": 0.0006311176693998277,
      "learning_rate": 0.10421924406329568,
      "loss": 0.3157,
      "num_input_tokens_seen": 45387184,
      "step": 23950
    },
    {
      "epoch": 3.9079043967697205,
      "grad_norm": 0.00046705565182492137,
      "learning_rate": 0.10416315328015598,
      "loss": 0.303,
      "num_input_tokens_seen": 45396368,
      "step": 23955
    },
    {
      "epoch": 3.9087201239905376,
      "grad_norm": 0.0007231780327856541,
      "learning_rate": 0.10410706956563402,
      "loss": 0.3503,
      "num_input_tokens_seen": 45406192,
      "step": 23960
    },
    {
      "epoch": 3.9095358512113547,
      "grad_norm": 0.001034803339280188,
      "learning_rate": 0.10405099292837874,
      "loss": 0.3483,
      "num_input_tokens_seen": 45414848,
      "step": 23965
    },
    {
      "epoch": 3.9103515784321723,
      "grad_norm": 0.0016023563221096992,
      "learning_rate": 0.10399492337703771,
      "loss": 0.3696,
      "num_input_tokens_seen": 45424944,
      "step": 23970
    },
    {
      "epoch": 3.91116730565299,
      "grad_norm": 0.0013558033388108015,
      "learning_rate": 0.10393886092025764,
      "loss": 0.3144,
      "num_input_tokens_seen": 45435840,
      "step": 23975
    },
    {
      "epoch": 3.911983032873807,
      "grad_norm": 0.0011934516951441765,
      "learning_rate": 0.10388280556668412,
      "loss": 0.3404,
      "num_input_tokens_seen": 45446352,
      "step": 23980
    },
    {
      "epoch": 3.912798760094624,
      "grad_norm": 0.0008344654925167561,
      "learning_rate": 0.10382675732496145,
      "loss": 0.3495,
      "num_input_tokens_seen": 45455632,
      "step": 23985
    },
    {
      "epoch": 3.9136144873154417,
      "grad_norm": 0.0014527601888403296,
      "learning_rate": 0.10377071620373311,
      "loss": 0.3009,
      "num_input_tokens_seen": 45464864,
      "step": 23990
    },
    {
      "epoch": 3.9144302145362593,
      "grad_norm": 0.001436682534404099,
      "learning_rate": 0.10371468221164128,
      "loss": 0.352,
      "num_input_tokens_seen": 45473360,
      "step": 23995
    },
    {
      "epoch": 3.9152459417570764,
      "grad_norm": 0.002727256389334798,
      "learning_rate": 0.10365865535732706,
      "loss": 0.3729,
      "num_input_tokens_seen": 45482928,
      "step": 24000
    },
    {
      "epoch": 3.9152459417570764,
      "eval_loss": 0.3245013654232025,
      "eval_runtime": 152.9132,
      "eval_samples_per_second": 17.821,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 45482928,
      "step": 24000
    },
    {
      "epoch": 3.9160616689778935,
      "grad_norm": 0.0003415180544834584,
      "learning_rate": 0.10360263564943062,
      "loss": 0.2857,
      "num_input_tokens_seen": 45491600,
      "step": 24005
    },
    {
      "epoch": 3.916877396198711,
      "grad_norm": 0.001220689620822668,
      "learning_rate": 0.10354662309659075,
      "loss": 0.3488,
      "num_input_tokens_seen": 45501840,
      "step": 24010
    },
    {
      "epoch": 3.9176931234195287,
      "grad_norm": 0.001665335614234209,
      "learning_rate": 0.10349061770744537,
      "loss": 0.327,
      "num_input_tokens_seen": 45509264,
      "step": 24015
    },
    {
      "epoch": 3.918508850640346,
      "grad_norm": 0.001613801927305758,
      "learning_rate": 0.10343461949063128,
      "loss": 0.3407,
      "num_input_tokens_seen": 45518288,
      "step": 24020
    },
    {
      "epoch": 3.919324577861163,
      "grad_norm": 0.0009388944017700851,
      "learning_rate": 0.103378628454784,
      "loss": 0.3176,
      "num_input_tokens_seen": 45528096,
      "step": 24025
    },
    {
      "epoch": 3.9201403050819805,
      "grad_norm": 0.0018527384381741285,
      "learning_rate": 0.10332264460853811,
      "loss": 0.3196,
      "num_input_tokens_seen": 45537776,
      "step": 24030
    },
    {
      "epoch": 3.920956032302798,
      "grad_norm": 0.000725470541510731,
      "learning_rate": 0.10326666796052701,
      "loss": 0.357,
      "num_input_tokens_seen": 45546864,
      "step": 24035
    },
    {
      "epoch": 3.9217717595236152,
      "grad_norm": 0.0015422999858856201,
      "learning_rate": 0.10321069851938296,
      "loss": 0.2991,
      "num_input_tokens_seen": 45556848,
      "step": 24040
    },
    {
      "epoch": 3.922587486744433,
      "grad_norm": 0.0016062272479757667,
      "learning_rate": 0.10315473629373724,
      "loss": 0.2973,
      "num_input_tokens_seen": 45565344,
      "step": 24045
    },
    {
      "epoch": 3.92340321396525,
      "grad_norm": 0.0008627940551377833,
      "learning_rate": 0.10309878129221982,
      "loss": 0.3552,
      "num_input_tokens_seen": 45575216,
      "step": 24050
    },
    {
      "epoch": 3.9242189411860675,
      "grad_norm": 0.0013956687180325389,
      "learning_rate": 0.10304283352345973,
      "loss": 0.3601,
      "num_input_tokens_seen": 45584848,
      "step": 24055
    },
    {
      "epoch": 3.9250346684068846,
      "grad_norm": 0.0018157255835831165,
      "learning_rate": 0.10298689299608486,
      "loss": 0.2957,
      "num_input_tokens_seen": 45594192,
      "step": 24060
    },
    {
      "epoch": 3.925850395627702,
      "grad_norm": 0.0006183324148878455,
      "learning_rate": 0.10293095971872188,
      "loss": 0.2821,
      "num_input_tokens_seen": 45603728,
      "step": 24065
    },
    {
      "epoch": 3.9266661228485193,
      "grad_norm": 0.0009990835096687078,
      "learning_rate": 0.10287503369999645,
      "loss": 0.2982,
      "num_input_tokens_seen": 45613088,
      "step": 24070
    },
    {
      "epoch": 3.927481850069337,
      "grad_norm": 0.002000782173126936,
      "learning_rate": 0.10281911494853295,
      "loss": 0.3171,
      "num_input_tokens_seen": 45622448,
      "step": 24075
    },
    {
      "epoch": 3.928297577290154,
      "grad_norm": 0.0008417614153586328,
      "learning_rate": 0.10276320347295485,
      "loss": 0.292,
      "num_input_tokens_seen": 45632752,
      "step": 24080
    },
    {
      "epoch": 3.9291133045109716,
      "grad_norm": 0.0014106897870078683,
      "learning_rate": 0.10270729928188446,
      "loss": 0.2928,
      "num_input_tokens_seen": 45642272,
      "step": 24085
    },
    {
      "epoch": 3.9299290317317888,
      "grad_norm": 0.0024779532104730606,
      "learning_rate": 0.10265140238394276,
      "loss": 0.2942,
      "num_input_tokens_seen": 45651888,
      "step": 24090
    },
    {
      "epoch": 3.9307447589526063,
      "grad_norm": 0.0015109919477254152,
      "learning_rate": 0.10259551278774988,
      "loss": 0.3768,
      "num_input_tokens_seen": 45661952,
      "step": 24095
    },
    {
      "epoch": 3.9315604861734235,
      "grad_norm": 0.001947817043401301,
      "learning_rate": 0.10253963050192462,
      "loss": 0.396,
      "num_input_tokens_seen": 45672176,
      "step": 24100
    },
    {
      "epoch": 3.932376213394241,
      "grad_norm": 0.0011841731611639261,
      "learning_rate": 0.10248375553508478,
      "loss": 0.3065,
      "num_input_tokens_seen": 45682080,
      "step": 24105
    },
    {
      "epoch": 3.933191940615058,
      "grad_norm": 0.003046129597350955,
      "learning_rate": 0.102427887895847,
      "loss": 0.2872,
      "num_input_tokens_seen": 45692256,
      "step": 24110
    },
    {
      "epoch": 3.9340076678358757,
      "grad_norm": 0.0022852434776723385,
      "learning_rate": 0.10237202759282668,
      "loss": 0.3829,
      "num_input_tokens_seen": 45702512,
      "step": 24115
    },
    {
      "epoch": 3.934823395056693,
      "grad_norm": 0.0013436974259093404,
      "learning_rate": 0.10231617463463821,
      "loss": 0.2975,
      "num_input_tokens_seen": 45712432,
      "step": 24120
    },
    {
      "epoch": 3.9356391222775104,
      "grad_norm": 0.002624347573146224,
      "learning_rate": 0.10226032902989492,
      "loss": 0.3279,
      "num_input_tokens_seen": 45721920,
      "step": 24125
    },
    {
      "epoch": 3.936454849498328,
      "grad_norm": 0.0016711961943656206,
      "learning_rate": 0.10220449078720877,
      "loss": 0.3317,
      "num_input_tokens_seen": 45732160,
      "step": 24130
    },
    {
      "epoch": 3.937270576719145,
      "grad_norm": 0.002537130145356059,
      "learning_rate": 0.1021486599151908,
      "loss": 0.3502,
      "num_input_tokens_seen": 45742096,
      "step": 24135
    },
    {
      "epoch": 3.9380863039399623,
      "grad_norm": 0.001992165809497237,
      "learning_rate": 0.10209283642245084,
      "loss": 0.3424,
      "num_input_tokens_seen": 45750176,
      "step": 24140
    },
    {
      "epoch": 3.93890203116078,
      "grad_norm": 0.0013853848213329911,
      "learning_rate": 0.10203702031759748,
      "loss": 0.3274,
      "num_input_tokens_seen": 45759744,
      "step": 24145
    },
    {
      "epoch": 3.9397177583815974,
      "grad_norm": 0.0006208218401297927,
      "learning_rate": 0.1019812116092384,
      "loss": 0.2867,
      "num_input_tokens_seen": 45768896,
      "step": 24150
    },
    {
      "epoch": 3.9405334856024146,
      "grad_norm": 0.003189650597050786,
      "learning_rate": 0.10192541030597986,
      "loss": 0.3196,
      "num_input_tokens_seen": 45777184,
      "step": 24155
    },
    {
      "epoch": 3.9413492128232317,
      "grad_norm": 0.0017227655043825507,
      "learning_rate": 0.1018696164164272,
      "loss": 0.3906,
      "num_input_tokens_seen": 45786928,
      "step": 24160
    },
    {
      "epoch": 3.9421649400440493,
      "grad_norm": 0.0018308988073840737,
      "learning_rate": 0.10181382994918459,
      "loss": 0.3402,
      "num_input_tokens_seen": 45797680,
      "step": 24165
    },
    {
      "epoch": 3.942980667264867,
      "grad_norm": 0.0018497431883588433,
      "learning_rate": 0.10175805091285492,
      "loss": 0.3139,
      "num_input_tokens_seen": 45806384,
      "step": 24170
    },
    {
      "epoch": 3.943796394485684,
      "grad_norm": 0.0014910422032698989,
      "learning_rate": 0.10170227931603999,
      "loss": 0.3147,
      "num_input_tokens_seen": 45816576,
      "step": 24175
    },
    {
      "epoch": 3.944612121706501,
      "grad_norm": 0.002181346295401454,
      "learning_rate": 0.10164651516734062,
      "loss": 0.3163,
      "num_input_tokens_seen": 45826192,
      "step": 24180
    },
    {
      "epoch": 3.9454278489273187,
      "grad_norm": 0.0016895100707188249,
      "learning_rate": 0.1015907584753562,
      "loss": 0.3707,
      "num_input_tokens_seen": 45835856,
      "step": 24185
    },
    {
      "epoch": 3.9462435761481363,
      "grad_norm": 0.0006568885874003172,
      "learning_rate": 0.10153500924868523,
      "loss": 0.3608,
      "num_input_tokens_seen": 45845488,
      "step": 24190
    },
    {
      "epoch": 3.9470593033689534,
      "grad_norm": 0.001730348914861679,
      "learning_rate": 0.10147926749592483,
      "loss": 0.3086,
      "num_input_tokens_seen": 45854112,
      "step": 24195
    },
    {
      "epoch": 3.9478750305897705,
      "grad_norm": 0.0008959663682617247,
      "learning_rate": 0.10142353322567112,
      "loss": 0.3025,
      "num_input_tokens_seen": 45861584,
      "step": 24200
    },
    {
      "epoch": 3.9478750305897705,
      "eval_loss": 0.32283222675323486,
      "eval_runtime": 152.7847,
      "eval_samples_per_second": 17.836,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 45861584,
      "step": 24200
    },
    {
      "epoch": 3.948690757810588,
      "grad_norm": 0.0017440904630348086,
      "learning_rate": 0.1013678064465191,
      "loss": 0.3318,
      "num_input_tokens_seen": 45872688,
      "step": 24205
    },
    {
      "epoch": 3.9495064850314057,
      "grad_norm": 0.0028134072199463844,
      "learning_rate": 0.10131208716706244,
      "loss": 0.3169,
      "num_input_tokens_seen": 45882272,
      "step": 24210
    },
    {
      "epoch": 3.950322212252223,
      "grad_norm": 0.0006367588066495955,
      "learning_rate": 0.10125637539589379,
      "loss": 0.3379,
      "num_input_tokens_seen": 45892672,
      "step": 24215
    },
    {
      "epoch": 3.9511379394730404,
      "grad_norm": 0.0014326616656035185,
      "learning_rate": 0.10120067114160464,
      "loss": 0.3329,
      "num_input_tokens_seen": 45902592,
      "step": 24220
    },
    {
      "epoch": 3.9519536666938575,
      "grad_norm": 0.002312663709744811,
      "learning_rate": 0.10114497441278517,
      "loss": 0.3227,
      "num_input_tokens_seen": 45911904,
      "step": 24225
    },
    {
      "epoch": 3.952769393914675,
      "grad_norm": 0.000847596034873277,
      "learning_rate": 0.10108928521802468,
      "loss": 0.3264,
      "num_input_tokens_seen": 45921664,
      "step": 24230
    },
    {
      "epoch": 3.953585121135492,
      "grad_norm": 0.0013982634991407394,
      "learning_rate": 0.101033603565911,
      "loss": 0.254,
      "num_input_tokens_seen": 45930896,
      "step": 24235
    },
    {
      "epoch": 3.95440084835631,
      "grad_norm": 0.003919609356671572,
      "learning_rate": 0.10097792946503102,
      "loss": 0.3385,
      "num_input_tokens_seen": 45940272,
      "step": 24240
    },
    {
      "epoch": 3.955216575577127,
      "grad_norm": 0.0023050615563988686,
      "learning_rate": 0.10092226292397039,
      "loss": 0.3544,
      "num_input_tokens_seen": 45950256,
      "step": 24245
    },
    {
      "epoch": 3.9560323027979445,
      "grad_norm": 0.002250116551294923,
      "learning_rate": 0.10086660395131354,
      "loss": 0.3287,
      "num_input_tokens_seen": 45959424,
      "step": 24250
    },
    {
      "epoch": 3.9568480300187616,
      "grad_norm": 0.0013807173818349838,
      "learning_rate": 0.10081095255564385,
      "loss": 0.2642,
      "num_input_tokens_seen": 45968304,
      "step": 24255
    },
    {
      "epoch": 3.957663757239579,
      "grad_norm": 0.001465797540731728,
      "learning_rate": 0.10075530874554335,
      "loss": 0.3217,
      "num_input_tokens_seen": 45977280,
      "step": 24260
    },
    {
      "epoch": 3.9584794844603963,
      "grad_norm": 0.0016364241018891335,
      "learning_rate": 0.10069967252959311,
      "loss": 0.2973,
      "num_input_tokens_seen": 45987168,
      "step": 24265
    },
    {
      "epoch": 3.959295211681214,
      "grad_norm": 0.002106500556692481,
      "learning_rate": 0.10064404391637297,
      "loss": 0.3288,
      "num_input_tokens_seen": 45996272,
      "step": 24270
    },
    {
      "epoch": 3.960110938902031,
      "grad_norm": 0.0036362374667078257,
      "learning_rate": 0.10058842291446145,
      "loss": 0.3166,
      "num_input_tokens_seen": 46005328,
      "step": 24275
    },
    {
      "epoch": 3.9609266661228486,
      "grad_norm": 0.0021112088579684496,
      "learning_rate": 0.10053280953243608,
      "loss": 0.382,
      "num_input_tokens_seen": 46014704,
      "step": 24280
    },
    {
      "epoch": 3.9617423933436657,
      "grad_norm": 0.0009070041705854237,
      "learning_rate": 0.10047720377887315,
      "loss": 0.3195,
      "num_input_tokens_seen": 46023600,
      "step": 24285
    },
    {
      "epoch": 3.9625581205644833,
      "grad_norm": 0.0033615855500102043,
      "learning_rate": 0.10042160566234767,
      "loss": 0.3873,
      "num_input_tokens_seen": 46032592,
      "step": 24290
    },
    {
      "epoch": 3.9633738477853004,
      "grad_norm": 0.0016168189467862248,
      "learning_rate": 0.10036601519143372,
      "loss": 0.3393,
      "num_input_tokens_seen": 46042432,
      "step": 24295
    },
    {
      "epoch": 3.964189575006118,
      "grad_norm": 0.001424730638973415,
      "learning_rate": 0.1003104323747039,
      "loss": 0.2721,
      "num_input_tokens_seen": 46052048,
      "step": 24300
    },
    {
      "epoch": 3.965005302226935,
      "grad_norm": 0.0012770071625709534,
      "learning_rate": 0.10025485722072984,
      "loss": 0.3021,
      "num_input_tokens_seen": 46060704,
      "step": 24305
    },
    {
      "epoch": 3.9658210294477527,
      "grad_norm": 0.0010064226808026433,
      "learning_rate": 0.10019928973808201,
      "loss": 0.3568,
      "num_input_tokens_seen": 46071120,
      "step": 24310
    },
    {
      "epoch": 3.96663675666857,
      "grad_norm": 0.0021152584813535213,
      "learning_rate": 0.10014372993532945,
      "loss": 0.3448,
      "num_input_tokens_seen": 46081376,
      "step": 24315
    },
    {
      "epoch": 3.9674524838893874,
      "grad_norm": 0.0026907548308372498,
      "learning_rate": 0.1000881778210403,
      "loss": 0.3308,
      "num_input_tokens_seen": 46090192,
      "step": 24320
    },
    {
      "epoch": 3.968268211110205,
      "grad_norm": 0.0027851415798068047,
      "learning_rate": 0.10003263340378142,
      "loss": 0.3078,
      "num_input_tokens_seen": 46097552,
      "step": 24325
    },
    {
      "epoch": 3.969083938331022,
      "grad_norm": 0.0022005694918334484,
      "learning_rate": 0.09997709669211834,
      "loss": 0.3462,
      "num_input_tokens_seen": 46108784,
      "step": 24330
    },
    {
      "epoch": 3.9698996655518393,
      "grad_norm": 0.0009793693898245692,
      "learning_rate": 0.0999215676946156,
      "loss": 0.3306,
      "num_input_tokens_seen": 46118464,
      "step": 24335
    },
    {
      "epoch": 3.970715392772657,
      "grad_norm": 0.0029265121556818485,
      "learning_rate": 0.0998660464198364,
      "loss": 0.3113,
      "num_input_tokens_seen": 46127952,
      "step": 24340
    },
    {
      "epoch": 3.9715311199934744,
      "grad_norm": 0.0019378263968974352,
      "learning_rate": 0.09981053287634288,
      "loss": 0.3371,
      "num_input_tokens_seen": 46137504,
      "step": 24345
    },
    {
      "epoch": 3.9723468472142915,
      "grad_norm": 0.0006855620304122567,
      "learning_rate": 0.09975502707269596,
      "loss": 0.3376,
      "num_input_tokens_seen": 46147120,
      "step": 24350
    },
    {
      "epoch": 3.9731625744351087,
      "grad_norm": 0.0011704231146723032,
      "learning_rate": 0.09969952901745524,
      "loss": 0.299,
      "num_input_tokens_seen": 46156512,
      "step": 24355
    },
    {
      "epoch": 3.9739783016559262,
      "grad_norm": 0.0023232470266520977,
      "learning_rate": 0.09964403871917925,
      "loss": 0.3845,
      "num_input_tokens_seen": 46166768,
      "step": 24360
    },
    {
      "epoch": 3.974794028876744,
      "grad_norm": 0.0011840396327897906,
      "learning_rate": 0.09958855618642536,
      "loss": 0.2656,
      "num_input_tokens_seen": 46175744,
      "step": 24365
    },
    {
      "epoch": 3.975609756097561,
      "grad_norm": 0.0018401517299935222,
      "learning_rate": 0.09953308142774955,
      "loss": 0.3617,
      "num_input_tokens_seen": 46185440,
      "step": 24370
    },
    {
      "epoch": 3.976425483318378,
      "grad_norm": 0.0026531049516052008,
      "learning_rate": 0.09947761445170686,
      "loss": 0.389,
      "num_input_tokens_seen": 46195936,
      "step": 24375
    },
    {
      "epoch": 3.9772412105391957,
      "grad_norm": 0.0015249678399413824,
      "learning_rate": 0.09942215526685086,
      "loss": 0.3441,
      "num_input_tokens_seen": 46205152,
      "step": 24380
    },
    {
      "epoch": 3.9780569377600132,
      "grad_norm": 0.001313672517426312,
      "learning_rate": 0.09936670388173414,
      "loss": 0.2666,
      "num_input_tokens_seen": 46213888,
      "step": 24385
    },
    {
      "epoch": 3.9788726649808304,
      "grad_norm": 0.0006567632663063705,
      "learning_rate": 0.09931126030490799,
      "loss": 0.3385,
      "num_input_tokens_seen": 46224000,
      "step": 24390
    },
    {
      "epoch": 3.9796883922016475,
      "grad_norm": 0.0008844695403240621,
      "learning_rate": 0.0992558245449225,
      "loss": 0.3098,
      "num_input_tokens_seen": 46234064,
      "step": 24395
    },
    {
      "epoch": 3.980504119422465,
      "grad_norm": 0.0011935518123209476,
      "learning_rate": 0.09920039661032651,
      "loss": 0.3069,
      "num_input_tokens_seen": 46243072,
      "step": 24400
    },
    {
      "epoch": 3.980504119422465,
      "eval_loss": 0.32205745577812195,
      "eval_runtime": 152.7579,
      "eval_samples_per_second": 17.839,
      "eval_steps_per_second": 8.923,
      "num_input_tokens_seen": 46243072,
      "step": 24400
    },
    {
      "epoch": 3.9813198466432826,
      "grad_norm": 0.0017284145578742027,
      "learning_rate": 0.09914497650966782,
      "loss": 0.3395,
      "num_input_tokens_seen": 46253840,
      "step": 24405
    },
    {
      "epoch": 3.9821355738640998,
      "grad_norm": 0.001478784834034741,
      "learning_rate": 0.09908956425149276,
      "loss": 0.333,
      "num_input_tokens_seen": 46262208,
      "step": 24410
    },
    {
      "epoch": 3.9829513010849174,
      "grad_norm": 0.0011444177944213152,
      "learning_rate": 0.09903415984434677,
      "loss": 0.2889,
      "num_input_tokens_seen": 46271504,
      "step": 24415
    },
    {
      "epoch": 3.9837670283057345,
      "grad_norm": 0.0014504734426736832,
      "learning_rate": 0.09897876329677373,
      "loss": 0.3102,
      "num_input_tokens_seen": 46281264,
      "step": 24420
    },
    {
      "epoch": 3.984582755526552,
      "grad_norm": 0.0024791129399091005,
      "learning_rate": 0.09892337461731658,
      "loss": 0.3063,
      "num_input_tokens_seen": 46290544,
      "step": 24425
    },
    {
      "epoch": 3.985398482747369,
      "grad_norm": 0.002900657244026661,
      "learning_rate": 0.09886799381451693,
      "loss": 0.3414,
      "num_input_tokens_seen": 46299600,
      "step": 24430
    },
    {
      "epoch": 3.9862142099681868,
      "grad_norm": 0.0013974873581901193,
      "learning_rate": 0.09881262089691521,
      "loss": 0.3203,
      "num_input_tokens_seen": 46308928,
      "step": 24435
    },
    {
      "epoch": 3.987029937189004,
      "grad_norm": 0.0011201153974980116,
      "learning_rate": 0.09875725587305059,
      "loss": 0.39,
      "num_input_tokens_seen": 46318896,
      "step": 24440
    },
    {
      "epoch": 3.9878456644098215,
      "grad_norm": 0.0012111305259168148,
      "learning_rate": 0.09870189875146111,
      "loss": 0.3335,
      "num_input_tokens_seen": 46328048,
      "step": 24445
    },
    {
      "epoch": 3.9886613916306386,
      "grad_norm": 0.0012678515631705523,
      "learning_rate": 0.09864654954068346,
      "loss": 0.2914,
      "num_input_tokens_seen": 46336864,
      "step": 24450
    },
    {
      "epoch": 3.989477118851456,
      "grad_norm": 0.002041151514276862,
      "learning_rate": 0.09859120824925326,
      "loss": 0.3768,
      "num_input_tokens_seen": 46345920,
      "step": 24455
    },
    {
      "epoch": 3.9902928460722733,
      "grad_norm": 0.0021377692464739084,
      "learning_rate": 0.09853587488570474,
      "loss": 0.2979,
      "num_input_tokens_seen": 46355056,
      "step": 24460
    },
    {
      "epoch": 3.991108573293091,
      "grad_norm": 0.002555760322138667,
      "learning_rate": 0.09848054945857107,
      "loss": 0.2899,
      "num_input_tokens_seen": 46364608,
      "step": 24465
    },
    {
      "epoch": 3.991924300513908,
      "grad_norm": 0.0011242483742535114,
      "learning_rate": 0.09842523197638416,
      "loss": 0.3034,
      "num_input_tokens_seen": 46374368,
      "step": 24470
    },
    {
      "epoch": 3.9927400277347256,
      "grad_norm": 0.002855862258002162,
      "learning_rate": 0.09836992244767452,
      "loss": 0.3643,
      "num_input_tokens_seen": 46383728,
      "step": 24475
    },
    {
      "epoch": 3.9935557549555427,
      "grad_norm": 0.0019479473121464252,
      "learning_rate": 0.09831462088097168,
      "loss": 0.2734,
      "num_input_tokens_seen": 46393472,
      "step": 24480
    },
    {
      "epoch": 3.9943714821763603,
      "grad_norm": 0.0037959839683026075,
      "learning_rate": 0.09825932728480385,
      "loss": 0.3559,
      "num_input_tokens_seen": 46401776,
      "step": 24485
    },
    {
      "epoch": 3.9951872093971774,
      "grad_norm": 0.0020482453983277082,
      "learning_rate": 0.09820404166769794,
      "loss": 0.2717,
      "num_input_tokens_seen": 46411216,
      "step": 24490
    },
    {
      "epoch": 3.996002936617995,
      "grad_norm": 0.0019949160050600767,
      "learning_rate": 0.09814876403817978,
      "loss": 0.3279,
      "num_input_tokens_seen": 46421072,
      "step": 24495
    },
    {
      "epoch": 3.9968186638388126,
      "grad_norm": 0.0013663314748555422,
      "learning_rate": 0.09809349440477376,
      "loss": 0.3395,
      "num_input_tokens_seen": 46431232,
      "step": 24500
    },
    {
      "epoch": 3.9976343910596297,
      "grad_norm": 0.0031576731707900763,
      "learning_rate": 0.09803823277600317,
      "loss": 0.3251,
      "num_input_tokens_seen": 46440464,
      "step": 24505
    },
    {
      "epoch": 3.998450118280447,
      "grad_norm": 0.002778294961899519,
      "learning_rate": 0.09798297916039014,
      "loss": 0.3652,
      "num_input_tokens_seen": 46449488,
      "step": 24510
    },
    {
      "epoch": 3.9992658455012644,
      "grad_norm": 0.0011934806825593114,
      "learning_rate": 0.09792773356645534,
      "loss": 0.3126,
      "num_input_tokens_seen": 46459168,
      "step": 24515
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0008677042787894607,
      "learning_rate": 0.09787249600271843,
      "loss": 0.3207,
      "num_input_tokens_seen": 46467712,
      "step": 24520
    },
    {
      "epoch": 4.000815727220817,
      "grad_norm": 0.0014109687181189656,
      "learning_rate": 0.09781726647769776,
      "loss": 0.3668,
      "num_input_tokens_seen": 46478256,
      "step": 24525
    },
    {
      "epoch": 4.001631454441635,
      "grad_norm": 0.001966217067092657,
      "learning_rate": 0.0977620449999103,
      "loss": 0.3183,
      "num_input_tokens_seen": 46488416,
      "step": 24530
    },
    {
      "epoch": 4.002447181662452,
      "grad_norm": 0.0008705994114279747,
      "learning_rate": 0.09770683157787204,
      "loss": 0.3361,
      "num_input_tokens_seen": 46495952,
      "step": 24535
    },
    {
      "epoch": 4.003262908883269,
      "grad_norm": 0.001266937586478889,
      "learning_rate": 0.09765162622009745,
      "loss": 0.2986,
      "num_input_tokens_seen": 46503760,
      "step": 24540
    },
    {
      "epoch": 4.0040786361040865,
      "grad_norm": 0.0010410617105662823,
      "learning_rate": 0.09759642893509995,
      "loss": 0.3503,
      "num_input_tokens_seen": 46513904,
      "step": 24545
    },
    {
      "epoch": 4.004894363324905,
      "grad_norm": 0.0019114474998787045,
      "learning_rate": 0.09754123973139169,
      "loss": 0.2995,
      "num_input_tokens_seen": 46523760,
      "step": 24550
    },
    {
      "epoch": 4.005710090545722,
      "grad_norm": 0.0013520788634195924,
      "learning_rate": 0.09748605861748345,
      "loss": 0.3436,
      "num_input_tokens_seen": 46533056,
      "step": 24555
    },
    {
      "epoch": 4.006525817766539,
      "grad_norm": 0.0043513099662959576,
      "learning_rate": 0.0974308856018849,
      "loss": 0.378,
      "num_input_tokens_seen": 46541552,
      "step": 24560
    },
    {
      "epoch": 4.007341544987356,
      "grad_norm": 0.0024478270206600428,
      "learning_rate": 0.09737572069310449,
      "loss": 0.3079,
      "num_input_tokens_seen": 46551024,
      "step": 24565
    },
    {
      "epoch": 4.008157272208174,
      "grad_norm": 0.0022812061943113804,
      "learning_rate": 0.09732056389964922,
      "loss": 0.3408,
      "num_input_tokens_seen": 46561648,
      "step": 24570
    },
    {
      "epoch": 4.008972999428991,
      "grad_norm": 0.0017913987394422293,
      "learning_rate": 0.097265415230025,
      "loss": 0.3495,
      "num_input_tokens_seen": 46571776,
      "step": 24575
    },
    {
      "epoch": 4.009788726649808,
      "grad_norm": 0.001894142129458487,
      "learning_rate": 0.09721027469273648,
      "loss": 0.3629,
      "num_input_tokens_seen": 46581248,
      "step": 24580
    },
    {
      "epoch": 4.010604453870625,
      "grad_norm": 0.0012085071066394448,
      "learning_rate": 0.09715514229628695,
      "loss": 0.3403,
      "num_input_tokens_seen": 46591248,
      "step": 24585
    },
    {
      "epoch": 4.011420181091443,
      "grad_norm": 0.0013665435835719109,
      "learning_rate": 0.09710001804917864,
      "loss": 0.3414,
      "num_input_tokens_seen": 46600160,
      "step": 24590
    },
    {
      "epoch": 4.0122359083122605,
      "grad_norm": 0.0010597265791147947,
      "learning_rate": 0.09704490195991226,
      "loss": 0.3119,
      "num_input_tokens_seen": 46609344,
      "step": 24595
    },
    {
      "epoch": 4.013051635533078,
      "grad_norm": 0.0011623934842646122,
      "learning_rate": 0.09698979403698753,
      "loss": 0.338,
      "num_input_tokens_seen": 46619680,
      "step": 24600
    },
    {
      "epoch": 4.013051635533078,
      "eval_loss": 0.32052186131477356,
      "eval_runtime": 152.8974,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 46619680,
      "step": 24600
    },
    {
      "epoch": 4.013867362753895,
      "grad_norm": 0.002022149972617626,
      "learning_rate": 0.0969346942889027,
      "loss": 0.3345,
      "num_input_tokens_seen": 46630064,
      "step": 24605
    },
    {
      "epoch": 4.014683089974713,
      "grad_norm": 0.0008831148152239621,
      "learning_rate": 0.09687960272415487,
      "loss": 0.3606,
      "num_input_tokens_seen": 46639104,
      "step": 24610
    },
    {
      "epoch": 4.01549881719553,
      "grad_norm": 0.0010007803793996572,
      "learning_rate": 0.0968245193512399,
      "loss": 0.2607,
      "num_input_tokens_seen": 46648752,
      "step": 24615
    },
    {
      "epoch": 4.016314544416347,
      "grad_norm": 0.000888364331331104,
      "learning_rate": 0.09676944417865221,
      "loss": 0.2721,
      "num_input_tokens_seen": 46657712,
      "step": 24620
    },
    {
      "epoch": 4.017130271637164,
      "grad_norm": 0.0014541358686983585,
      "learning_rate": 0.09671437721488517,
      "loss": 0.3534,
      "num_input_tokens_seen": 46666992,
      "step": 24625
    },
    {
      "epoch": 4.017945998857982,
      "grad_norm": 0.000786645570769906,
      "learning_rate": 0.09665931846843086,
      "loss": 0.2597,
      "num_input_tokens_seen": 46676704,
      "step": 24630
    },
    {
      "epoch": 4.018761726078799,
      "grad_norm": 0.0017008199356496334,
      "learning_rate": 0.0966042679477799,
      "loss": 0.3551,
      "num_input_tokens_seen": 46687504,
      "step": 24635
    },
    {
      "epoch": 4.0195774532996165,
      "grad_norm": 0.0009466272895224392,
      "learning_rate": 0.09654922566142186,
      "loss": 0.3666,
      "num_input_tokens_seen": 46697712,
      "step": 24640
    },
    {
      "epoch": 4.020393180520434,
      "grad_norm": 0.0008950736955739558,
      "learning_rate": 0.09649419161784498,
      "loss": 0.336,
      "num_input_tokens_seen": 46707408,
      "step": 24645
    },
    {
      "epoch": 4.021208907741252,
      "grad_norm": 0.0027740178629755974,
      "learning_rate": 0.09643916582553606,
      "loss": 0.3095,
      "num_input_tokens_seen": 46716320,
      "step": 24650
    },
    {
      "epoch": 4.022024634962069,
      "grad_norm": 0.0006912582903169096,
      "learning_rate": 0.09638414829298093,
      "loss": 0.3877,
      "num_input_tokens_seen": 46725104,
      "step": 24655
    },
    {
      "epoch": 4.022840362182886,
      "grad_norm": 0.001172421034425497,
      "learning_rate": 0.09632913902866386,
      "loss": 0.2853,
      "num_input_tokens_seen": 46734592,
      "step": 24660
    },
    {
      "epoch": 4.023656089403703,
      "grad_norm": 0.0012391067575663328,
      "learning_rate": 0.096274138041068,
      "loss": 0.3083,
      "num_input_tokens_seen": 46744480,
      "step": 24665
    },
    {
      "epoch": 4.024471816624521,
      "grad_norm": 0.0014543297002092004,
      "learning_rate": 0.09621914533867527,
      "loss": 0.3277,
      "num_input_tokens_seen": 46754496,
      "step": 24670
    },
    {
      "epoch": 4.025287543845338,
      "grad_norm": 0.0013836530270054936,
      "learning_rate": 0.09616416092996616,
      "loss": 0.3263,
      "num_input_tokens_seen": 46764208,
      "step": 24675
    },
    {
      "epoch": 4.026103271066155,
      "grad_norm": 0.000747251498978585,
      "learning_rate": 0.09610918482342,
      "loss": 0.3003,
      "num_input_tokens_seen": 46773664,
      "step": 24680
    },
    {
      "epoch": 4.026918998286972,
      "grad_norm": 0.0013915507588535547,
      "learning_rate": 0.09605421702751478,
      "loss": 0.3197,
      "num_input_tokens_seen": 46784160,
      "step": 24685
    },
    {
      "epoch": 4.02773472550779,
      "grad_norm": 0.0007602981058880687,
      "learning_rate": 0.09599925755072718,
      "loss": 0.3383,
      "num_input_tokens_seen": 46795296,
      "step": 24690
    },
    {
      "epoch": 4.028550452728608,
      "grad_norm": 0.0022837852593511343,
      "learning_rate": 0.09594430640153273,
      "loss": 0.2725,
      "num_input_tokens_seen": 46804816,
      "step": 24695
    },
    {
      "epoch": 4.029366179949425,
      "grad_norm": 0.002553181489929557,
      "learning_rate": 0.09588936358840547,
      "loss": 0.3123,
      "num_input_tokens_seen": 46814768,
      "step": 24700
    },
    {
      "epoch": 4.030181907170242,
      "grad_norm": 0.0021422835998237133,
      "learning_rate": 0.09583442911981836,
      "loss": 0.3759,
      "num_input_tokens_seen": 46824448,
      "step": 24705
    },
    {
      "epoch": 4.03099763439106,
      "grad_norm": 0.000521911249961704,
      "learning_rate": 0.09577950300424302,
      "loss": 0.3353,
      "num_input_tokens_seen": 46833776,
      "step": 24710
    },
    {
      "epoch": 4.031813361611877,
      "grad_norm": 0.0005865658167749643,
      "learning_rate": 0.09572458525014967,
      "loss": 0.2393,
      "num_input_tokens_seen": 46842592,
      "step": 24715
    },
    {
      "epoch": 4.032629088832694,
      "grad_norm": 0.001426059054210782,
      "learning_rate": 0.0956696758660073,
      "loss": 0.3597,
      "num_input_tokens_seen": 46851952,
      "step": 24720
    },
    {
      "epoch": 4.033444816053512,
      "grad_norm": 0.0018268218263983727,
      "learning_rate": 0.09561477486028373,
      "loss": 0.3664,
      "num_input_tokens_seen": 46860912,
      "step": 24725
    },
    {
      "epoch": 4.034260543274329,
      "grad_norm": 0.0015606384258717299,
      "learning_rate": 0.09555988224144528,
      "loss": 0.3419,
      "num_input_tokens_seen": 46869552,
      "step": 24730
    },
    {
      "epoch": 4.035076270495146,
      "grad_norm": 0.002179388189688325,
      "learning_rate": 0.09550499801795717,
      "loss": 0.3182,
      "num_input_tokens_seen": 46880208,
      "step": 24735
    },
    {
      "epoch": 4.0358919977159635,
      "grad_norm": 0.0011073910864070058,
      "learning_rate": 0.09545012219828314,
      "loss": 0.3347,
      "num_input_tokens_seen": 46890032,
      "step": 24740
    },
    {
      "epoch": 4.0367077249367815,
      "grad_norm": 0.0014867876889184117,
      "learning_rate": 0.09539525479088577,
      "loss": 0.3199,
      "num_input_tokens_seen": 46899056,
      "step": 24745
    },
    {
      "epoch": 4.037523452157599,
      "grad_norm": 0.0022113004233688116,
      "learning_rate": 0.0953403958042264,
      "loss": 0.3052,
      "num_input_tokens_seen": 46909760,
      "step": 24750
    },
    {
      "epoch": 4.038339179378416,
      "grad_norm": 0.0009454285027459264,
      "learning_rate": 0.09528554524676484,
      "loss": 0.3264,
      "num_input_tokens_seen": 46917664,
      "step": 24755
    },
    {
      "epoch": 4.039154906599233,
      "grad_norm": 0.000588312977924943,
      "learning_rate": 0.09523070312695978,
      "loss": 0.3237,
      "num_input_tokens_seen": 46928608,
      "step": 24760
    },
    {
      "epoch": 4.039970633820051,
      "grad_norm": 0.001859127776697278,
      "learning_rate": 0.09517586945326863,
      "loss": 0.3105,
      "num_input_tokens_seen": 46938480,
      "step": 24765
    },
    {
      "epoch": 4.040786361040868,
      "grad_norm": 0.0028057473246008158,
      "learning_rate": 0.0951210442341473,
      "loss": 0.3187,
      "num_input_tokens_seen": 46949392,
      "step": 24770
    },
    {
      "epoch": 4.041602088261685,
      "grad_norm": 0.0007094775210134685,
      "learning_rate": 0.09506622747805066,
      "loss": 0.3544,
      "num_input_tokens_seen": 46958240,
      "step": 24775
    },
    {
      "epoch": 4.042417815482502,
      "grad_norm": 0.001483274158090353,
      "learning_rate": 0.09501141919343203,
      "loss": 0.2931,
      "num_input_tokens_seen": 46967392,
      "step": 24780
    },
    {
      "epoch": 4.04323354270332,
      "grad_norm": 0.0013172993203625083,
      "learning_rate": 0.09495661938874361,
      "loss": 0.2725,
      "num_input_tokens_seen": 46976992,
      "step": 24785
    },
    {
      "epoch": 4.0440492699241375,
      "grad_norm": 0.0015708605060353875,
      "learning_rate": 0.0949018280724362,
      "loss": 0.2892,
      "num_input_tokens_seen": 46986992,
      "step": 24790
    },
    {
      "epoch": 4.044864997144955,
      "grad_norm": 0.0008327856776304543,
      "learning_rate": 0.09484704525295934,
      "loss": 0.2974,
      "num_input_tokens_seen": 46997120,
      "step": 24795
    },
    {
      "epoch": 4.045680724365772,
      "grad_norm": 0.0012319778325036168,
      "learning_rate": 0.09479227093876112,
      "loss": 0.3089,
      "num_input_tokens_seen": 47007360,
      "step": 24800
    },
    {
      "epoch": 4.045680724365772,
      "eval_loss": 0.32104378938674927,
      "eval_runtime": 152.9154,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 47007360,
      "step": 24800
    },
    {
      "epoch": 4.04649645158659,
      "grad_norm": 0.001168697141110897,
      "learning_rate": 0.0947375051382886,
      "loss": 0.3332,
      "num_input_tokens_seen": 47015968,
      "step": 24805
    },
    {
      "epoch": 4.047312178807407,
      "grad_norm": 0.0019158804789185524,
      "learning_rate": 0.09468274785998718,
      "loss": 0.2782,
      "num_input_tokens_seen": 47025392,
      "step": 24810
    },
    {
      "epoch": 4.048127906028224,
      "grad_norm": 0.0029949217569082975,
      "learning_rate": 0.09462799911230127,
      "loss": 0.2916,
      "num_input_tokens_seen": 47035200,
      "step": 24815
    },
    {
      "epoch": 4.048943633249041,
      "grad_norm": 0.001937147811986506,
      "learning_rate": 0.0945732589036737,
      "loss": 0.3465,
      "num_input_tokens_seen": 47045792,
      "step": 24820
    },
    {
      "epoch": 4.049759360469859,
      "grad_norm": 0.0034338543191552162,
      "learning_rate": 0.09451852724254614,
      "loss": 0.2616,
      "num_input_tokens_seen": 47054064,
      "step": 24825
    },
    {
      "epoch": 4.050575087690676,
      "grad_norm": 0.0028316315729171038,
      "learning_rate": 0.09446380413735894,
      "loss": 0.3865,
      "num_input_tokens_seen": 47064352,
      "step": 24830
    },
    {
      "epoch": 4.051390814911493,
      "grad_norm": 0.002808941062539816,
      "learning_rate": 0.09440908959655099,
      "loss": 0.3237,
      "num_input_tokens_seen": 47075232,
      "step": 24835
    },
    {
      "epoch": 4.052206542132311,
      "grad_norm": 0.0010950586292892694,
      "learning_rate": 0.09435438362856004,
      "loss": 0.3556,
      "num_input_tokens_seen": 47084544,
      "step": 24840
    },
    {
      "epoch": 4.053022269353129,
      "grad_norm": 0.0018424531444907188,
      "learning_rate": 0.0942996862418225,
      "loss": 0.3741,
      "num_input_tokens_seen": 47093888,
      "step": 24845
    },
    {
      "epoch": 4.053837996573946,
      "grad_norm": 0.0033473316580057144,
      "learning_rate": 0.09424499744477322,
      "loss": 0.3092,
      "num_input_tokens_seen": 47102720,
      "step": 24850
    },
    {
      "epoch": 4.054653723794763,
      "grad_norm": 0.0022043311037123203,
      "learning_rate": 0.09419031724584608,
      "loss": 0.3163,
      "num_input_tokens_seen": 47112240,
      "step": 24855
    },
    {
      "epoch": 4.05546945101558,
      "grad_norm": 0.0007555453921668231,
      "learning_rate": 0.09413564565347331,
      "loss": 0.3388,
      "num_input_tokens_seen": 47122704,
      "step": 24860
    },
    {
      "epoch": 4.056285178236398,
      "grad_norm": 0.0024694986641407013,
      "learning_rate": 0.094080982676086,
      "loss": 0.3263,
      "num_input_tokens_seen": 47132592,
      "step": 24865
    },
    {
      "epoch": 4.057100905457215,
      "grad_norm": 0.002248683711513877,
      "learning_rate": 0.09402632832211395,
      "loss": 0.3636,
      "num_input_tokens_seen": 47142000,
      "step": 24870
    },
    {
      "epoch": 4.057916632678032,
      "grad_norm": 0.0008215125999413431,
      "learning_rate": 0.09397168259998541,
      "loss": 0.2694,
      "num_input_tokens_seen": 47152464,
      "step": 24875
    },
    {
      "epoch": 4.058732359898849,
      "grad_norm": 0.002377008553594351,
      "learning_rate": 0.09391704551812759,
      "loss": 0.3006,
      "num_input_tokens_seen": 47161552,
      "step": 24880
    },
    {
      "epoch": 4.059548087119667,
      "grad_norm": 0.0022127702832221985,
      "learning_rate": 0.09386241708496605,
      "loss": 0.3414,
      "num_input_tokens_seen": 47170368,
      "step": 24885
    },
    {
      "epoch": 4.0603638143404845,
      "grad_norm": 0.0019893001299351454,
      "learning_rate": 0.09380779730892527,
      "loss": 0.434,
      "num_input_tokens_seen": 47179872,
      "step": 24890
    },
    {
      "epoch": 4.061179541561302,
      "grad_norm": 0.0017885963898152113,
      "learning_rate": 0.09375318619842836,
      "loss": 0.3805,
      "num_input_tokens_seen": 47189808,
      "step": 24895
    },
    {
      "epoch": 4.06199526878212,
      "grad_norm": 0.0008346422691829503,
      "learning_rate": 0.09369858376189696,
      "loss": 0.3462,
      "num_input_tokens_seen": 47199712,
      "step": 24900
    },
    {
      "epoch": 4.062810996002937,
      "grad_norm": 0.000830533099360764,
      "learning_rate": 0.09364399000775143,
      "loss": 0.3008,
      "num_input_tokens_seen": 47208688,
      "step": 24905
    },
    {
      "epoch": 4.063626723223754,
      "grad_norm": 0.0013558818027377129,
      "learning_rate": 0.09358940494441093,
      "loss": 0.3259,
      "num_input_tokens_seen": 47218144,
      "step": 24910
    },
    {
      "epoch": 4.064442450444571,
      "grad_norm": 0.0007328902720473707,
      "learning_rate": 0.09353482858029301,
      "loss": 0.3619,
      "num_input_tokens_seen": 47227232,
      "step": 24915
    },
    {
      "epoch": 4.065258177665389,
      "grad_norm": 0.0006101785693317652,
      "learning_rate": 0.09348026092381419,
      "loss": 0.363,
      "num_input_tokens_seen": 47236320,
      "step": 24920
    },
    {
      "epoch": 4.066073904886206,
      "grad_norm": 0.002177478978410363,
      "learning_rate": 0.09342570198338931,
      "loss": 0.2544,
      "num_input_tokens_seen": 47244784,
      "step": 24925
    },
    {
      "epoch": 4.066889632107023,
      "grad_norm": 0.0007038626354187727,
      "learning_rate": 0.0933711517674322,
      "loss": 0.3255,
      "num_input_tokens_seen": 47255168,
      "step": 24930
    },
    {
      "epoch": 4.0677053593278405,
      "grad_norm": 0.0006846914766356349,
      "learning_rate": 0.09331661028435513,
      "loss": 0.3247,
      "num_input_tokens_seen": 47265136,
      "step": 24935
    },
    {
      "epoch": 4.0685210865486585,
      "grad_norm": 0.0015366524457931519,
      "learning_rate": 0.09326207754256909,
      "loss": 0.3399,
      "num_input_tokens_seen": 47274016,
      "step": 24940
    },
    {
      "epoch": 4.069336813769476,
      "grad_norm": 0.0025515977758914232,
      "learning_rate": 0.09320755355048366,
      "loss": 0.3213,
      "num_input_tokens_seen": 47283440,
      "step": 24945
    },
    {
      "epoch": 4.070152540990293,
      "grad_norm": 0.0014488950837403536,
      "learning_rate": 0.09315303831650722,
      "loss": 0.3074,
      "num_input_tokens_seen": 47292576,
      "step": 24950
    },
    {
      "epoch": 4.07096826821111,
      "grad_norm": 0.0015699643408879638,
      "learning_rate": 0.09309853184904661,
      "loss": 0.3625,
      "num_input_tokens_seen": 47303168,
      "step": 24955
    },
    {
      "epoch": 4.071783995431928,
      "grad_norm": 0.002484799362719059,
      "learning_rate": 0.09304403415650753,
      "loss": 0.3164,
      "num_input_tokens_seen": 47312656,
      "step": 24960
    },
    {
      "epoch": 4.072599722652745,
      "grad_norm": 0.001724529080092907,
      "learning_rate": 0.09298954524729405,
      "loss": 0.3195,
      "num_input_tokens_seen": 47322016,
      "step": 24965
    },
    {
      "epoch": 4.073415449873562,
      "grad_norm": 0.002641767030581832,
      "learning_rate": 0.09293506512980916,
      "loss": 0.333,
      "num_input_tokens_seen": 47331792,
      "step": 24970
    },
    {
      "epoch": 4.074231177094379,
      "grad_norm": 0.0014555910602211952,
      "learning_rate": 0.0928805938124544,
      "loss": 0.3569,
      "num_input_tokens_seen": 47341312,
      "step": 24975
    },
    {
      "epoch": 4.075046904315197,
      "grad_norm": 0.0009322161204181612,
      "learning_rate": 0.09282613130362982,
      "loss": 0.3207,
      "num_input_tokens_seen": 47351040,
      "step": 24980
    },
    {
      "epoch": 4.0758626315360145,
      "grad_norm": 0.001966955605894327,
      "learning_rate": 0.09277167761173427,
      "loss": 0.2995,
      "num_input_tokens_seen": 47361648,
      "step": 24985
    },
    {
      "epoch": 4.076678358756832,
      "grad_norm": 0.003203087020665407,
      "learning_rate": 0.0927172327451653,
      "loss": 0.3307,
      "num_input_tokens_seen": 47371968,
      "step": 24990
    },
    {
      "epoch": 4.077494085977649,
      "grad_norm": 0.0015209022676572204,
      "learning_rate": 0.09266279671231882,
      "loss": 0.3153,
      "num_input_tokens_seen": 47381968,
      "step": 24995
    },
    {
      "epoch": 4.078309813198467,
      "grad_norm": 0.0015174332074820995,
      "learning_rate": 0.09260836952158967,
      "loss": 0.2882,
      "num_input_tokens_seen": 47391600,
      "step": 25000
    },
    {
      "epoch": 4.078309813198467,
      "eval_loss": 0.3291836678981781,
      "eval_runtime": 152.769,
      "eval_samples_per_second": 17.837,
      "eval_steps_per_second": 8.922,
      "num_input_tokens_seen": 47391600,
      "step": 25000
    },
    {
      "epoch": 4.079125540419284,
      "grad_norm": 0.002798877889290452,
      "learning_rate": 0.09255395118137114,
      "loss": 0.3672,
      "num_input_tokens_seen": 47401120,
      "step": 25005
    },
    {
      "epoch": 4.079941267640101,
      "grad_norm": 0.0029236632399260998,
      "learning_rate": 0.09249954170005527,
      "loss": 0.2874,
      "num_input_tokens_seen": 47410432,
      "step": 25010
    },
    {
      "epoch": 4.080756994860918,
      "grad_norm": 0.0012734585907310247,
      "learning_rate": 0.0924451410860327,
      "loss": 0.3345,
      "num_input_tokens_seen": 47420336,
      "step": 25015
    },
    {
      "epoch": 4.081572722081736,
      "grad_norm": 0.0026448643766343594,
      "learning_rate": 0.09239074934769258,
      "loss": 0.296,
      "num_input_tokens_seen": 47429744,
      "step": 25020
    },
    {
      "epoch": 4.082388449302553,
      "grad_norm": 0.001200798898935318,
      "learning_rate": 0.09233636649342288,
      "loss": 0.3118,
      "num_input_tokens_seen": 47438400,
      "step": 25025
    },
    {
      "epoch": 4.08320417652337,
      "grad_norm": 0.0015231765573844314,
      "learning_rate": 0.09228199253161017,
      "loss": 0.3761,
      "num_input_tokens_seen": 47447216,
      "step": 25030
    },
    {
      "epoch": 4.0840199037441876,
      "grad_norm": 0.001450921525247395,
      "learning_rate": 0.09222762747063949,
      "loss": 0.3113,
      "num_input_tokens_seen": 47457840,
      "step": 25035
    },
    {
      "epoch": 4.084835630965006,
      "grad_norm": 0.0014930529287084937,
      "learning_rate": 0.09217327131889473,
      "loss": 0.3343,
      "num_input_tokens_seen": 47467280,
      "step": 25040
    },
    {
      "epoch": 4.085651358185823,
      "grad_norm": 0.0010648495517671108,
      "learning_rate": 0.09211892408475818,
      "loss": 0.3033,
      "num_input_tokens_seen": 47476656,
      "step": 25045
    },
    {
      "epoch": 4.08646708540664,
      "grad_norm": 0.0010960439685732126,
      "learning_rate": 0.09206458577661089,
      "loss": 0.3437,
      "num_input_tokens_seen": 47485568,
      "step": 25050
    },
    {
      "epoch": 4.087282812627457,
      "grad_norm": 0.0013315753312781453,
      "learning_rate": 0.09201025640283263,
      "loss": 0.308,
      "num_input_tokens_seen": 47494832,
      "step": 25055
    },
    {
      "epoch": 4.088098539848275,
      "grad_norm": 0.0011818180792033672,
      "learning_rate": 0.09195593597180148,
      "loss": 0.3226,
      "num_input_tokens_seen": 47504416,
      "step": 25060
    },
    {
      "epoch": 4.088914267069092,
      "grad_norm": 0.0023003271780908108,
      "learning_rate": 0.09190162449189444,
      "loss": 0.3335,
      "num_input_tokens_seen": 47514544,
      "step": 25065
    },
    {
      "epoch": 4.089729994289909,
      "grad_norm": 0.0021976407151669264,
      "learning_rate": 0.09184732197148705,
      "loss": 0.3366,
      "num_input_tokens_seen": 47522992,
      "step": 25070
    },
    {
      "epoch": 4.090545721510727,
      "grad_norm": 0.0021969641093164682,
      "learning_rate": 0.09179302841895343,
      "loss": 0.34,
      "num_input_tokens_seen": 47532256,
      "step": 25075
    },
    {
      "epoch": 4.091361448731544,
      "grad_norm": 0.002432528417557478,
      "learning_rate": 0.09173874384266625,
      "loss": 0.3053,
      "num_input_tokens_seen": 47543424,
      "step": 25080
    },
    {
      "epoch": 4.0921771759523615,
      "grad_norm": 0.0008847577264532447,
      "learning_rate": 0.09168446825099695,
      "loss": 0.3221,
      "num_input_tokens_seen": 47553104,
      "step": 25085
    },
    {
      "epoch": 4.092992903173179,
      "grad_norm": 0.001905154320411384,
      "learning_rate": 0.09163020165231545,
      "loss": 0.3076,
      "num_input_tokens_seen": 47562704,
      "step": 25090
    },
    {
      "epoch": 4.093808630393997,
      "grad_norm": 0.001407297095283866,
      "learning_rate": 0.09157594405499044,
      "loss": 0.3642,
      "num_input_tokens_seen": 47572928,
      "step": 25095
    },
    {
      "epoch": 4.094624357614814,
      "grad_norm": 0.0008034990751184523,
      "learning_rate": 0.09152169546738899,
      "loss": 0.3103,
      "num_input_tokens_seen": 47581712,
      "step": 25100
    },
    {
      "epoch": 4.095440084835631,
      "grad_norm": 0.0036121055018156767,
      "learning_rate": 0.09146745589787698,
      "loss": 0.3654,
      "num_input_tokens_seen": 47590640,
      "step": 25105
    },
    {
      "epoch": 4.096255812056448,
      "grad_norm": 0.0028683138079941273,
      "learning_rate": 0.09141322535481891,
      "loss": 0.3135,
      "num_input_tokens_seen": 47599376,
      "step": 25110
    },
    {
      "epoch": 4.097071539277266,
      "grad_norm": 0.0007732211379334331,
      "learning_rate": 0.0913590038465777,
      "loss": 0.3497,
      "num_input_tokens_seen": 47607952,
      "step": 25115
    },
    {
      "epoch": 4.097887266498083,
      "grad_norm": 0.0026436508633196354,
      "learning_rate": 0.09130479138151505,
      "loss": 0.2538,
      "num_input_tokens_seen": 47618736,
      "step": 25120
    },
    {
      "epoch": 4.0987029937189,
      "grad_norm": 0.0015202508075162768,
      "learning_rate": 0.09125058796799114,
      "loss": 0.3602,
      "num_input_tokens_seen": 47627312,
      "step": 25125
    },
    {
      "epoch": 4.0995187209397175,
      "grad_norm": 0.0015180185437202454,
      "learning_rate": 0.09119639361436485,
      "loss": 0.3628,
      "num_input_tokens_seen": 47637808,
      "step": 25130
    },
    {
      "epoch": 4.1003344481605355,
      "grad_norm": 0.002506241900846362,
      "learning_rate": 0.09114220832899368,
      "loss": 0.3666,
      "num_input_tokens_seen": 47647360,
      "step": 25135
    },
    {
      "epoch": 4.101150175381353,
      "grad_norm": 0.0005449295276775956,
      "learning_rate": 0.0910880321202336,
      "loss": 0.27,
      "num_input_tokens_seen": 47657312,
      "step": 25140
    },
    {
      "epoch": 4.10196590260217,
      "grad_norm": 0.003172550816088915,
      "learning_rate": 0.09103386499643933,
      "loss": 0.3131,
      "num_input_tokens_seen": 47666592,
      "step": 25145
    },
    {
      "epoch": 4.102781629822987,
      "grad_norm": 0.002501977141946554,
      "learning_rate": 0.09097970696596407,
      "loss": 0.3176,
      "num_input_tokens_seen": 47676096,
      "step": 25150
    },
    {
      "epoch": 4.103597357043805,
      "grad_norm": 0.0014188488712534308,
      "learning_rate": 0.09092555803715971,
      "loss": 0.3108,
      "num_input_tokens_seen": 47684928,
      "step": 25155
    },
    {
      "epoch": 4.104413084264622,
      "grad_norm": 0.0014701626496389508,
      "learning_rate": 0.0908714182183767,
      "loss": 0.2821,
      "num_input_tokens_seen": 47693872,
      "step": 25160
    },
    {
      "epoch": 4.105228811485439,
      "grad_norm": 0.0016633706400170922,
      "learning_rate": 0.090817287517964,
      "loss": 0.3371,
      "num_input_tokens_seen": 47702560,
      "step": 25165
    },
    {
      "epoch": 4.106044538706256,
      "grad_norm": 0.0010742767481133342,
      "learning_rate": 0.09076316594426931,
      "loss": 0.2567,
      "num_input_tokens_seen": 47711248,
      "step": 25170
    },
    {
      "epoch": 4.106860265927074,
      "grad_norm": 0.0027955968398600817,
      "learning_rate": 0.09070905350563888,
      "loss": 0.36,
      "num_input_tokens_seen": 47719968,
      "step": 25175
    },
    {
      "epoch": 4.1076759931478914,
      "grad_norm": 0.0008851542370393872,
      "learning_rate": 0.09065495021041745,
      "loss": 0.2954,
      "num_input_tokens_seen": 47729712,
      "step": 25180
    },
    {
      "epoch": 4.108491720368709,
      "grad_norm": 0.0011509894393384457,
      "learning_rate": 0.09060085606694851,
      "loss": 0.2759,
      "num_input_tokens_seen": 47739632,
      "step": 25185
    },
    {
      "epoch": 4.109307447589526,
      "grad_norm": 0.0016929756384342909,
      "learning_rate": 0.09054677108357405,
      "loss": 0.2676,
      "num_input_tokens_seen": 47748576,
      "step": 25190
    },
    {
      "epoch": 4.110123174810344,
      "grad_norm": 0.0023212574888020754,
      "learning_rate": 0.09049269526863457,
      "loss": 0.3549,
      "num_input_tokens_seen": 47758384,
      "step": 25195
    },
    {
      "epoch": 4.110938902031161,
      "grad_norm": 0.0012146926019340754,
      "learning_rate": 0.09043862863046935,
      "loss": 0.3406,
      "num_input_tokens_seen": 47768320,
      "step": 25200
    },
    {
      "epoch": 4.110938902031161,
      "eval_loss": 0.32235240936279297,
      "eval_runtime": 152.8169,
      "eval_samples_per_second": 17.832,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 47768320,
      "step": 25200
    },
    {
      "epoch": 4.111754629251978,
      "grad_norm": 0.00514410063624382,
      "learning_rate": 0.09038457117741602,
      "loss": 0.3229,
      "num_input_tokens_seen": 47776336,
      "step": 25205
    },
    {
      "epoch": 4.112570356472795,
      "grad_norm": 0.0014531584456562996,
      "learning_rate": 0.09033052291781099,
      "loss": 0.382,
      "num_input_tokens_seen": 47786576,
      "step": 25210
    },
    {
      "epoch": 4.113386083693613,
      "grad_norm": 0.0015645052772015333,
      "learning_rate": 0.09027648385998926,
      "loss": 0.376,
      "num_input_tokens_seen": 47794544,
      "step": 25215
    },
    {
      "epoch": 4.11420181091443,
      "grad_norm": 0.0008500678814016283,
      "learning_rate": 0.09022245401228417,
      "loss": 0.3374,
      "num_input_tokens_seen": 47805456,
      "step": 25220
    },
    {
      "epoch": 4.115017538135247,
      "grad_norm": 0.0016327726189047098,
      "learning_rate": 0.09016843338302792,
      "loss": 0.3168,
      "num_input_tokens_seen": 47814288,
      "step": 25225
    },
    {
      "epoch": 4.1158332653560645,
      "grad_norm": 0.0013926998944953084,
      "learning_rate": 0.09011442198055115,
      "loss": 0.3801,
      "num_input_tokens_seen": 47823648,
      "step": 25230
    },
    {
      "epoch": 4.1166489925768825,
      "grad_norm": 0.0008437062497250736,
      "learning_rate": 0.09006041981318305,
      "loss": 0.3108,
      "num_input_tokens_seen": 47833440,
      "step": 25235
    },
    {
      "epoch": 4.1174647197977,
      "grad_norm": 0.0012684748508036137,
      "learning_rate": 0.09000642688925149,
      "loss": 0.3075,
      "num_input_tokens_seen": 47843424,
      "step": 25240
    },
    {
      "epoch": 4.118280447018517,
      "grad_norm": 0.0009191606659442186,
      "learning_rate": 0.0899524432170828,
      "loss": 0.3263,
      "num_input_tokens_seen": 47853264,
      "step": 25245
    },
    {
      "epoch": 4.119096174239334,
      "grad_norm": 0.0025846727658063173,
      "learning_rate": 0.08989846880500196,
      "loss": 0.3329,
      "num_input_tokens_seen": 47863312,
      "step": 25250
    },
    {
      "epoch": 4.119911901460152,
      "grad_norm": 0.0019656214863061905,
      "learning_rate": 0.08984450366133256,
      "loss": 0.2875,
      "num_input_tokens_seen": 47873392,
      "step": 25255
    },
    {
      "epoch": 4.120727628680969,
      "grad_norm": 0.0020012592431157827,
      "learning_rate": 0.08979054779439664,
      "loss": 0.2457,
      "num_input_tokens_seen": 47883728,
      "step": 25260
    },
    {
      "epoch": 4.121543355901786,
      "grad_norm": 0.001914084772579372,
      "learning_rate": 0.08973660121251485,
      "loss": 0.371,
      "num_input_tokens_seen": 47892032,
      "step": 25265
    },
    {
      "epoch": 4.122359083122603,
      "grad_norm": 0.0008049597381614149,
      "learning_rate": 0.08968266392400655,
      "loss": 0.3317,
      "num_input_tokens_seen": 47902080,
      "step": 25270
    },
    {
      "epoch": 4.123174810343421,
      "grad_norm": 0.0011283416533842683,
      "learning_rate": 0.0896287359371894,
      "loss": 0.3019,
      "num_input_tokens_seen": 47910784,
      "step": 25275
    },
    {
      "epoch": 4.1239905375642385,
      "grad_norm": 0.002412092871963978,
      "learning_rate": 0.08957481726037989,
      "loss": 0.3802,
      "num_input_tokens_seen": 47919824,
      "step": 25280
    },
    {
      "epoch": 4.124806264785056,
      "grad_norm": 0.0014916539657860994,
      "learning_rate": 0.08952090790189286,
      "loss": 0.3387,
      "num_input_tokens_seen": 47928720,
      "step": 25285
    },
    {
      "epoch": 4.125621992005874,
      "grad_norm": 0.0012447787448763847,
      "learning_rate": 0.08946700787004187,
      "loss": 0.3523,
      "num_input_tokens_seen": 47937904,
      "step": 25290
    },
    {
      "epoch": 4.126437719226691,
      "grad_norm": 0.001855355454608798,
      "learning_rate": 0.08941311717313899,
      "loss": 0.3192,
      "num_input_tokens_seen": 47947984,
      "step": 25295
    },
    {
      "epoch": 4.127253446447508,
      "grad_norm": 0.0015659505734220147,
      "learning_rate": 0.08935923581949483,
      "loss": 0.3144,
      "num_input_tokens_seen": 47957792,
      "step": 25300
    },
    {
      "epoch": 4.128069173668325,
      "grad_norm": 0.0008826687699183822,
      "learning_rate": 0.0893053638174185,
      "loss": 0.3044,
      "num_input_tokens_seen": 47967696,
      "step": 25305
    },
    {
      "epoch": 4.128884900889143,
      "grad_norm": 0.0016528143314644694,
      "learning_rate": 0.0892515011752179,
      "loss": 0.3441,
      "num_input_tokens_seen": 47975520,
      "step": 25310
    },
    {
      "epoch": 4.12970062810996,
      "grad_norm": 0.0017465499695390463,
      "learning_rate": 0.08919764790119918,
      "loss": 0.3326,
      "num_input_tokens_seen": 47984624,
      "step": 25315
    },
    {
      "epoch": 4.130516355330777,
      "grad_norm": 0.0011237080907449126,
      "learning_rate": 0.08914380400366727,
      "loss": 0.3492,
      "num_input_tokens_seen": 47994512,
      "step": 25320
    },
    {
      "epoch": 4.1313320825515945,
      "grad_norm": 0.0033654463477432728,
      "learning_rate": 0.08908996949092551,
      "loss": 0.3083,
      "num_input_tokens_seen": 48003200,
      "step": 25325
    },
    {
      "epoch": 4.1321478097724125,
      "grad_norm": 0.002162331249564886,
      "learning_rate": 0.08903614437127592,
      "loss": 0.3016,
      "num_input_tokens_seen": 48012128,
      "step": 25330
    },
    {
      "epoch": 4.13296353699323,
      "grad_norm": 0.0013323865132406354,
      "learning_rate": 0.088982328653019,
      "loss": 0.2963,
      "num_input_tokens_seen": 48021680,
      "step": 25335
    },
    {
      "epoch": 4.133779264214047,
      "grad_norm": 0.0014727842062711716,
      "learning_rate": 0.0889285223444538,
      "loss": 0.3807,
      "num_input_tokens_seen": 48031360,
      "step": 25340
    },
    {
      "epoch": 4.134594991434864,
      "grad_norm": 0.0021524957846850157,
      "learning_rate": 0.08887472545387787,
      "loss": 0.3475,
      "num_input_tokens_seen": 48040416,
      "step": 25345
    },
    {
      "epoch": 4.135410718655682,
      "grad_norm": 0.0016199396923184395,
      "learning_rate": 0.08882093798958751,
      "loss": 0.2906,
      "num_input_tokens_seen": 48049408,
      "step": 25350
    },
    {
      "epoch": 4.136226445876499,
      "grad_norm": 0.0015710268635302782,
      "learning_rate": 0.08876715995987726,
      "loss": 0.3548,
      "num_input_tokens_seen": 48058176,
      "step": 25355
    },
    {
      "epoch": 4.137042173097316,
      "grad_norm": 0.0026997707318514585,
      "learning_rate": 0.08871339137304052,
      "loss": 0.3433,
      "num_input_tokens_seen": 48066992,
      "step": 25360
    },
    {
      "epoch": 4.137857900318133,
      "grad_norm": 0.002843734808266163,
      "learning_rate": 0.0886596322373689,
      "loss": 0.3853,
      "num_input_tokens_seen": 48076864,
      "step": 25365
    },
    {
      "epoch": 4.138673627538951,
      "grad_norm": 0.0011827383423224092,
      "learning_rate": 0.08860588256115293,
      "loss": 0.3563,
      "num_input_tokens_seen": 48086496,
      "step": 25370
    },
    {
      "epoch": 4.139489354759768,
      "grad_norm": 0.002336567034944892,
      "learning_rate": 0.0885521423526814,
      "loss": 0.312,
      "num_input_tokens_seen": 48095840,
      "step": 25375
    },
    {
      "epoch": 4.1403050819805856,
      "grad_norm": 0.0016677199164405465,
      "learning_rate": 0.08849841162024165,
      "loss": 0.3336,
      "num_input_tokens_seen": 48105264,
      "step": 25380
    },
    {
      "epoch": 4.141120809201403,
      "grad_norm": 0.001079595647752285,
      "learning_rate": 0.08844469037211973,
      "loss": 0.3206,
      "num_input_tokens_seen": 48115088,
      "step": 25385
    },
    {
      "epoch": 4.141936536422221,
      "grad_norm": 0.002093397779390216,
      "learning_rate": 0.08839097861660014,
      "loss": 0.3324,
      "num_input_tokens_seen": 48123456,
      "step": 25390
    },
    {
      "epoch": 4.142752263643038,
      "grad_norm": 0.0011259193997830153,
      "learning_rate": 0.08833727636196585,
      "loss": 0.3275,
      "num_input_tokens_seen": 48133488,
      "step": 25395
    },
    {
      "epoch": 4.143567990863855,
      "grad_norm": 0.0006926616770215333,
      "learning_rate": 0.08828358361649848,
      "loss": 0.3066,
      "num_input_tokens_seen": 48143424,
      "step": 25400
    },
    {
      "epoch": 4.143567990863855,
      "eval_loss": 0.3206457197666168,
      "eval_runtime": 152.7995,
      "eval_samples_per_second": 17.834,
      "eval_steps_per_second": 8.92,
      "num_input_tokens_seen": 48143424,
      "step": 25400
    },
    {
      "epoch": 4.144383718084672,
      "grad_norm": 0.001519506680779159,
      "learning_rate": 0.08822990038847807,
      "loss": 0.31,
      "num_input_tokens_seen": 48152160,
      "step": 25405
    },
    {
      "epoch": 4.14519944530549,
      "grad_norm": 0.002089761896058917,
      "learning_rate": 0.08817622668618325,
      "loss": 0.3365,
      "num_input_tokens_seen": 48161312,
      "step": 25410
    },
    {
      "epoch": 4.146015172526307,
      "grad_norm": 0.0022798837162554264,
      "learning_rate": 0.08812256251789125,
      "loss": 0.3444,
      "num_input_tokens_seen": 48171008,
      "step": 25415
    },
    {
      "epoch": 4.146830899747124,
      "grad_norm": 0.0021477146074175835,
      "learning_rate": 0.08806890789187766,
      "loss": 0.3274,
      "num_input_tokens_seen": 48180368,
      "step": 25420
    },
    {
      "epoch": 4.1476466269679415,
      "grad_norm": 0.0036298437044024467,
      "learning_rate": 0.08801526281641672,
      "loss": 0.3537,
      "num_input_tokens_seen": 48190144,
      "step": 25425
    },
    {
      "epoch": 4.1484623541887595,
      "grad_norm": 0.0007637437665835023,
      "learning_rate": 0.0879616272997813,
      "loss": 0.335,
      "num_input_tokens_seen": 48200672,
      "step": 25430
    },
    {
      "epoch": 4.149278081409577,
      "grad_norm": 0.002319985069334507,
      "learning_rate": 0.08790800135024247,
      "loss": 0.3503,
      "num_input_tokens_seen": 48210464,
      "step": 25435
    },
    {
      "epoch": 4.150093808630394,
      "grad_norm": 0.0026183214504271746,
      "learning_rate": 0.08785438497607023,
      "loss": 0.3439,
      "num_input_tokens_seen": 48218944,
      "step": 25440
    },
    {
      "epoch": 4.150909535851211,
      "grad_norm": 0.0009952801046893,
      "learning_rate": 0.08780077818553277,
      "loss": 0.3161,
      "num_input_tokens_seen": 48227888,
      "step": 25445
    },
    {
      "epoch": 4.151725263072029,
      "grad_norm": 0.001074901781976223,
      "learning_rate": 0.0877471809868969,
      "loss": 0.2596,
      "num_input_tokens_seen": 48237472,
      "step": 25450
    },
    {
      "epoch": 4.152540990292846,
      "grad_norm": 0.0038844679947942495,
      "learning_rate": 0.08769359338842811,
      "loss": 0.3043,
      "num_input_tokens_seen": 48247872,
      "step": 25455
    },
    {
      "epoch": 4.153356717513663,
      "grad_norm": 0.0020328531973063946,
      "learning_rate": 0.08764001539839016,
      "loss": 0.2749,
      "num_input_tokens_seen": 48257008,
      "step": 25460
    },
    {
      "epoch": 4.154172444734481,
      "grad_norm": 0.001997065031901002,
      "learning_rate": 0.08758644702504548,
      "loss": 0.3691,
      "num_input_tokens_seen": 48266048,
      "step": 25465
    },
    {
      "epoch": 4.154988171955298,
      "grad_norm": 0.001983924536034465,
      "learning_rate": 0.0875328882766551,
      "loss": 0.3197,
      "num_input_tokens_seen": 48276048,
      "step": 25470
    },
    {
      "epoch": 4.1558038991761155,
      "grad_norm": 0.0010204854188486934,
      "learning_rate": 0.08747933916147828,
      "loss": 0.2901,
      "num_input_tokens_seen": 48286432,
      "step": 25475
    },
    {
      "epoch": 4.156619626396933,
      "grad_norm": 0.0011196652194485068,
      "learning_rate": 0.0874257996877731,
      "loss": 0.2982,
      "num_input_tokens_seen": 48296944,
      "step": 25480
    },
    {
      "epoch": 4.157435353617751,
      "grad_norm": 0.004672668408602476,
      "learning_rate": 0.08737226986379593,
      "loss": 0.3455,
      "num_input_tokens_seen": 48306064,
      "step": 25485
    },
    {
      "epoch": 4.158251080838568,
      "grad_norm": 0.001068328507244587,
      "learning_rate": 0.08731874969780173,
      "loss": 0.255,
      "num_input_tokens_seen": 48315696,
      "step": 25490
    },
    {
      "epoch": 4.159066808059385,
      "grad_norm": 0.001662823255173862,
      "learning_rate": 0.08726523919804412,
      "loss": 0.3369,
      "num_input_tokens_seen": 48323808,
      "step": 25495
    },
    {
      "epoch": 4.159882535280202,
      "grad_norm": 0.0011025797575712204,
      "learning_rate": 0.08721173837277492,
      "loss": 0.3496,
      "num_input_tokens_seen": 48333632,
      "step": 25500
    },
    {
      "epoch": 4.16069826250102,
      "grad_norm": 0.002354647498577833,
      "learning_rate": 0.08715824723024479,
      "loss": 0.3905,
      "num_input_tokens_seen": 48343888,
      "step": 25505
    },
    {
      "epoch": 4.161513989721837,
      "grad_norm": 0.002429486019536853,
      "learning_rate": 0.08710476577870258,
      "loss": 0.3112,
      "num_input_tokens_seen": 48353104,
      "step": 25510
    },
    {
      "epoch": 4.162329716942654,
      "grad_norm": 0.0018036699621006846,
      "learning_rate": 0.08705129402639587,
      "loss": 0.388,
      "num_input_tokens_seen": 48362304,
      "step": 25515
    },
    {
      "epoch": 4.163145444163471,
      "grad_norm": 0.00154482526704669,
      "learning_rate": 0.08699783198157078,
      "loss": 0.3608,
      "num_input_tokens_seen": 48371824,
      "step": 25520
    },
    {
      "epoch": 4.1639611713842895,
      "grad_norm": 0.0009129964746534824,
      "learning_rate": 0.08694437965247163,
      "loss": 0.3152,
      "num_input_tokens_seen": 48381808,
      "step": 25525
    },
    {
      "epoch": 4.164776898605107,
      "grad_norm": 0.0012803891440853477,
      "learning_rate": 0.08689093704734165,
      "loss": 0.3594,
      "num_input_tokens_seen": 48391296,
      "step": 25530
    },
    {
      "epoch": 4.165592625825924,
      "grad_norm": 0.001136497943662107,
      "learning_rate": 0.08683750417442222,
      "loss": 0.3056,
      "num_input_tokens_seen": 48400656,
      "step": 25535
    },
    {
      "epoch": 4.166408353046741,
      "grad_norm": 0.0007804796914570034,
      "learning_rate": 0.08678408104195334,
      "loss": 0.3519,
      "num_input_tokens_seen": 48409744,
      "step": 25540
    },
    {
      "epoch": 4.167224080267559,
      "grad_norm": 0.0013349377550184727,
      "learning_rate": 0.08673066765817365,
      "loss": 0.3809,
      "num_input_tokens_seen": 48418608,
      "step": 25545
    },
    {
      "epoch": 4.168039807488376,
      "grad_norm": 0.0023493270855396986,
      "learning_rate": 0.08667726403132005,
      "loss": 0.3217,
      "num_input_tokens_seen": 48427104,
      "step": 25550
    },
    {
      "epoch": 4.168855534709193,
      "grad_norm": 0.0022237494122236967,
      "learning_rate": 0.0866238701696281,
      "loss": 0.3268,
      "num_input_tokens_seen": 48437584,
      "step": 25555
    },
    {
      "epoch": 4.16967126193001,
      "grad_norm": 0.0016564377583563328,
      "learning_rate": 0.08657048608133185,
      "loss": 0.3552,
      "num_input_tokens_seen": 48447664,
      "step": 25560
    },
    {
      "epoch": 4.170486989150828,
      "grad_norm": 0.002070801565423608,
      "learning_rate": 0.08651711177466369,
      "loss": 0.3689,
      "num_input_tokens_seen": 48458288,
      "step": 25565
    },
    {
      "epoch": 4.171302716371645,
      "grad_norm": 0.001289336127229035,
      "learning_rate": 0.08646374725785466,
      "loss": 0.2759,
      "num_input_tokens_seen": 48466096,
      "step": 25570
    },
    {
      "epoch": 4.1721184435924625,
      "grad_norm": 0.001323985867202282,
      "learning_rate": 0.08641039253913434,
      "loss": 0.3024,
      "num_input_tokens_seen": 48475888,
      "step": 25575
    },
    {
      "epoch": 4.17293417081328,
      "grad_norm": 0.0021184280049055815,
      "learning_rate": 0.08635704762673052,
      "loss": 0.3465,
      "num_input_tokens_seen": 48485648,
      "step": 25580
    },
    {
      "epoch": 4.173749898034098,
      "grad_norm": 0.0003601339121814817,
      "learning_rate": 0.08630371252886981,
      "loss": 0.3523,
      "num_input_tokens_seen": 48494864,
      "step": 25585
    },
    {
      "epoch": 4.174565625254915,
      "grad_norm": 0.003730779280886054,
      "learning_rate": 0.08625038725377704,
      "loss": 0.3005,
      "num_input_tokens_seen": 48503824,
      "step": 25590
    },
    {
      "epoch": 4.175381352475732,
      "grad_norm": 0.001365776639431715,
      "learning_rate": 0.08619707180967566,
      "loss": 0.3118,
      "num_input_tokens_seen": 48514432,
      "step": 25595
    },
    {
      "epoch": 4.176197079696549,
      "grad_norm": 0.0017495050560683012,
      "learning_rate": 0.08614376620478768,
      "loss": 0.4355,
      "num_input_tokens_seen": 48524368,
      "step": 25600
    },
    {
      "epoch": 4.176197079696549,
      "eval_loss": 0.32397526502609253,
      "eval_runtime": 152.8825,
      "eval_samples_per_second": 17.824,
      "eval_steps_per_second": 8.915,
      "num_input_tokens_seen": 48524368,
      "step": 25600
    },
    {
      "epoch": 4.177012806917367,
      "grad_norm": 0.002043413929641247,
      "learning_rate": 0.08609047044733344,
      "loss": 0.3012,
      "num_input_tokens_seen": 48533120,
      "step": 25605
    },
    {
      "epoch": 4.177828534138184,
      "grad_norm": 0.0014029524754732847,
      "learning_rate": 0.08603718454553168,
      "loss": 0.3207,
      "num_input_tokens_seen": 48542848,
      "step": 25610
    },
    {
      "epoch": 4.178644261359001,
      "grad_norm": 0.0011302487691864371,
      "learning_rate": 0.08598390850759997,
      "loss": 0.3224,
      "num_input_tokens_seen": 48552192,
      "step": 25615
    },
    {
      "epoch": 4.1794599885798185,
      "grad_norm": 0.0006606329116038978,
      "learning_rate": 0.08593064234175397,
      "loss": 0.2791,
      "num_input_tokens_seen": 48561792,
      "step": 25620
    },
    {
      "epoch": 4.1802757158006365,
      "grad_norm": 0.001529049826785922,
      "learning_rate": 0.08587738605620815,
      "loss": 0.3304,
      "num_input_tokens_seen": 48571376,
      "step": 25625
    },
    {
      "epoch": 4.181091443021454,
      "grad_norm": 0.0007670525228604674,
      "learning_rate": 0.08582413965917512,
      "loss": 0.3719,
      "num_input_tokens_seen": 48581984,
      "step": 25630
    },
    {
      "epoch": 4.181907170242271,
      "grad_norm": 0.0013860033359378576,
      "learning_rate": 0.08577090315886628,
      "loss": 0.3351,
      "num_input_tokens_seen": 48591072,
      "step": 25635
    },
    {
      "epoch": 4.182722897463089,
      "grad_norm": 0.0011776476167142391,
      "learning_rate": 0.08571767656349136,
      "loss": 0.2885,
      "num_input_tokens_seen": 48599776,
      "step": 25640
    },
    {
      "epoch": 4.183538624683906,
      "grad_norm": 0.0015591079136356711,
      "learning_rate": 0.08566445988125847,
      "loss": 0.3775,
      "num_input_tokens_seen": 48608960,
      "step": 25645
    },
    {
      "epoch": 4.184354351904723,
      "grad_norm": 0.002781753661110997,
      "learning_rate": 0.08561125312037436,
      "loss": 0.3483,
      "num_input_tokens_seen": 48617696,
      "step": 25650
    },
    {
      "epoch": 4.18517007912554,
      "grad_norm": 0.0025500725023448467,
      "learning_rate": 0.08555805628904424,
      "loss": 0.3026,
      "num_input_tokens_seen": 48627008,
      "step": 25655
    },
    {
      "epoch": 4.185985806346358,
      "grad_norm": 0.0006357983220368624,
      "learning_rate": 0.08550486939547161,
      "loss": 0.3304,
      "num_input_tokens_seen": 48636096,
      "step": 25660
    },
    {
      "epoch": 4.186801533567175,
      "grad_norm": 0.003013307461515069,
      "learning_rate": 0.08545169244785869,
      "loss": 0.3262,
      "num_input_tokens_seen": 48645376,
      "step": 25665
    },
    {
      "epoch": 4.1876172607879925,
      "grad_norm": 0.0015338960802182555,
      "learning_rate": 0.08539852545440589,
      "loss": 0.3541,
      "num_input_tokens_seen": 48653632,
      "step": 25670
    },
    {
      "epoch": 4.18843298800881,
      "grad_norm": 0.001215217518620193,
      "learning_rate": 0.08534536842331235,
      "loss": 0.3363,
      "num_input_tokens_seen": 48664544,
      "step": 25675
    },
    {
      "epoch": 4.189248715229628,
      "grad_norm": 0.0011073206551373005,
      "learning_rate": 0.08529222136277545,
      "loss": 0.3863,
      "num_input_tokens_seen": 48673744,
      "step": 25680
    },
    {
      "epoch": 4.190064442450445,
      "grad_norm": 0.0019354888936504722,
      "learning_rate": 0.08523908428099125,
      "loss": 0.3296,
      "num_input_tokens_seen": 48684112,
      "step": 25685
    },
    {
      "epoch": 4.190880169671262,
      "grad_norm": 0.0018792469054460526,
      "learning_rate": 0.08518595718615402,
      "loss": 0.3367,
      "num_input_tokens_seen": 48693376,
      "step": 25690
    },
    {
      "epoch": 4.191695896892079,
      "grad_norm": 0.0023990129120647907,
      "learning_rate": 0.08513284008645675,
      "loss": 0.2913,
      "num_input_tokens_seen": 48702832,
      "step": 25695
    },
    {
      "epoch": 4.192511624112897,
      "grad_norm": 0.002243145601823926,
      "learning_rate": 0.08507973299009065,
      "loss": 0.2959,
      "num_input_tokens_seen": 48711488,
      "step": 25700
    },
    {
      "epoch": 4.193327351333714,
      "grad_norm": 0.0020176314283162355,
      "learning_rate": 0.08502663590524563,
      "loss": 0.2741,
      "num_input_tokens_seen": 48720992,
      "step": 25705
    },
    {
      "epoch": 4.194143078554531,
      "grad_norm": 0.0026842313818633556,
      "learning_rate": 0.08497354884010981,
      "loss": 0.298,
      "num_input_tokens_seen": 48731520,
      "step": 25710
    },
    {
      "epoch": 4.194958805775348,
      "grad_norm": 0.0018480295548215508,
      "learning_rate": 0.0849204718028699,
      "loss": 0.3559,
      "num_input_tokens_seen": 48741472,
      "step": 25715
    },
    {
      "epoch": 4.195774532996166,
      "grad_norm": 0.0012481344165280461,
      "learning_rate": 0.08486740480171118,
      "loss": 0.3267,
      "num_input_tokens_seen": 48750640,
      "step": 25720
    },
    {
      "epoch": 4.196590260216984,
      "grad_norm": 0.0011547787580639124,
      "learning_rate": 0.08481434784481706,
      "loss": 0.3157,
      "num_input_tokens_seen": 48759792,
      "step": 25725
    },
    {
      "epoch": 4.197405987437801,
      "grad_norm": 0.003102495800703764,
      "learning_rate": 0.08476130094036968,
      "loss": 0.2368,
      "num_input_tokens_seen": 48769952,
      "step": 25730
    },
    {
      "epoch": 4.198221714658618,
      "grad_norm": 0.0015606519300490618,
      "learning_rate": 0.08470826409654961,
      "loss": 0.3214,
      "num_input_tokens_seen": 48779808,
      "step": 25735
    },
    {
      "epoch": 4.199037441879436,
      "grad_norm": 0.0008958242833614349,
      "learning_rate": 0.08465523732153564,
      "loss": 0.2699,
      "num_input_tokens_seen": 48788176,
      "step": 25740
    },
    {
      "epoch": 4.199853169100253,
      "grad_norm": 0.0018494714749976993,
      "learning_rate": 0.08460222062350532,
      "loss": 0.3055,
      "num_input_tokens_seen": 48796976,
      "step": 25745
    },
    {
      "epoch": 4.20066889632107,
      "grad_norm": 0.0015754009364172816,
      "learning_rate": 0.08454921401063442,
      "loss": 0.294,
      "num_input_tokens_seen": 48806496,
      "step": 25750
    },
    {
      "epoch": 4.201484623541887,
      "grad_norm": 0.002187164966017008,
      "learning_rate": 0.08449621749109716,
      "loss": 0.3839,
      "num_input_tokens_seen": 48816464,
      "step": 25755
    },
    {
      "epoch": 4.202300350762705,
      "grad_norm": 0.0011925315484404564,
      "learning_rate": 0.08444323107306641,
      "loss": 0.2804,
      "num_input_tokens_seen": 48826208,
      "step": 25760
    },
    {
      "epoch": 4.203116077983522,
      "grad_norm": 0.001887403312139213,
      "learning_rate": 0.0843902547647132,
      "loss": 0.3239,
      "num_input_tokens_seen": 48834848,
      "step": 25765
    },
    {
      "epoch": 4.2039318052043395,
      "grad_norm": 0.001914519933052361,
      "learning_rate": 0.0843372885742072,
      "loss": 0.3752,
      "num_input_tokens_seen": 48843184,
      "step": 25770
    },
    {
      "epoch": 4.204747532425157,
      "grad_norm": 0.001106573035940528,
      "learning_rate": 0.08428433250971652,
      "loss": 0.3418,
      "num_input_tokens_seen": 48853488,
      "step": 25775
    },
    {
      "epoch": 4.205563259645975,
      "grad_norm": 0.0013946056133136153,
      "learning_rate": 0.08423138657940757,
      "loss": 0.3098,
      "num_input_tokens_seen": 48863248,
      "step": 25780
    },
    {
      "epoch": 4.206378986866792,
      "grad_norm": 0.0041793459095060825,
      "learning_rate": 0.08417845079144536,
      "loss": 0.3793,
      "num_input_tokens_seen": 48872224,
      "step": 25785
    },
    {
      "epoch": 4.207194714087609,
      "grad_norm": 0.0011668757069855928,
      "learning_rate": 0.08412552515399314,
      "loss": 0.2968,
      "num_input_tokens_seen": 48881312,
      "step": 25790
    },
    {
      "epoch": 4.208010441308426,
      "grad_norm": 0.001762059866450727,
      "learning_rate": 0.08407260967521278,
      "loss": 0.3392,
      "num_input_tokens_seen": 48890608,
      "step": 25795
    },
    {
      "epoch": 4.208826168529244,
      "grad_norm": 0.0009456704137846828,
      "learning_rate": 0.08401970436326454,
      "loss": 0.3275,
      "num_input_tokens_seen": 48899856,
      "step": 25800
    },
    {
      "epoch": 4.208826168529244,
      "eval_loss": 0.3204803764820099,
      "eval_runtime": 152.9336,
      "eval_samples_per_second": 17.818,
      "eval_steps_per_second": 8.912,
      "num_input_tokens_seen": 48899856,
      "step": 25800
    },
    {
      "epoch": 4.209641895750061,
      "grad_norm": 0.0021835314109921455,
      "learning_rate": 0.08396680922630702,
      "loss": 0.3346,
      "num_input_tokens_seen": 48909040,
      "step": 25805
    },
    {
      "epoch": 4.210457622970878,
      "grad_norm": 0.0021337561774998903,
      "learning_rate": 0.08391392427249732,
      "loss": 0.3303,
      "num_input_tokens_seen": 48918816,
      "step": 25810
    },
    {
      "epoch": 4.211273350191696,
      "grad_norm": 0.0015687092673033476,
      "learning_rate": 0.08386104950999107,
      "loss": 0.337,
      "num_input_tokens_seen": 48928816,
      "step": 25815
    },
    {
      "epoch": 4.2120890774125135,
      "grad_norm": 0.0019472759449854493,
      "learning_rate": 0.0838081849469421,
      "loss": 0.3098,
      "num_input_tokens_seen": 48937760,
      "step": 25820
    },
    {
      "epoch": 4.212904804633331,
      "grad_norm": 0.0013718759873881936,
      "learning_rate": 0.08375533059150281,
      "loss": 0.3112,
      "num_input_tokens_seen": 48948208,
      "step": 25825
    },
    {
      "epoch": 4.213720531854148,
      "grad_norm": 0.0012090974487364292,
      "learning_rate": 0.08370248645182406,
      "loss": 0.2855,
      "num_input_tokens_seen": 48956656,
      "step": 25830
    },
    {
      "epoch": 4.214536259074965,
      "grad_norm": 0.0028884708881378174,
      "learning_rate": 0.083649652536055,
      "loss": 0.3721,
      "num_input_tokens_seen": 48965360,
      "step": 25835
    },
    {
      "epoch": 4.215351986295783,
      "grad_norm": 0.0011948475148528814,
      "learning_rate": 0.08359682885234339,
      "loss": 0.3125,
      "num_input_tokens_seen": 48975568,
      "step": 25840
    },
    {
      "epoch": 4.2161677135166,
      "grad_norm": 0.0032551861368119717,
      "learning_rate": 0.08354401540883516,
      "loss": 0.3115,
      "num_input_tokens_seen": 48984880,
      "step": 25845
    },
    {
      "epoch": 4.216983440737417,
      "grad_norm": 0.0022084901574999094,
      "learning_rate": 0.0834912122136749,
      "loss": 0.3194,
      "num_input_tokens_seen": 48993872,
      "step": 25850
    },
    {
      "epoch": 4.217799167958235,
      "grad_norm": 0.0028030253015458584,
      "learning_rate": 0.0834384192750056,
      "loss": 0.3333,
      "num_input_tokens_seen": 49003584,
      "step": 25855
    },
    {
      "epoch": 4.218614895179052,
      "grad_norm": 0.0015131481923162937,
      "learning_rate": 0.08338563660096844,
      "loss": 0.3523,
      "num_input_tokens_seen": 49013456,
      "step": 25860
    },
    {
      "epoch": 4.219430622399869,
      "grad_norm": 0.0010800834279507399,
      "learning_rate": 0.08333286419970329,
      "loss": 0.3018,
      "num_input_tokens_seen": 49024288,
      "step": 25865
    },
    {
      "epoch": 4.220246349620687,
      "grad_norm": 0.001390600111335516,
      "learning_rate": 0.08328010207934824,
      "loss": 0.2912,
      "num_input_tokens_seen": 49034080,
      "step": 25870
    },
    {
      "epoch": 4.221062076841505,
      "grad_norm": 0.0004958797944709659,
      "learning_rate": 0.08322735024803989,
      "loss": 0.2966,
      "num_input_tokens_seen": 49043856,
      "step": 25875
    },
    {
      "epoch": 4.221877804062322,
      "grad_norm": 0.003269537352025509,
      "learning_rate": 0.08317460871391331,
      "loss": 0.3208,
      "num_input_tokens_seen": 49054272,
      "step": 25880
    },
    {
      "epoch": 4.222693531283139,
      "grad_norm": 0.002623138017952442,
      "learning_rate": 0.08312187748510179,
      "loss": 0.3177,
      "num_input_tokens_seen": 49063696,
      "step": 25885
    },
    {
      "epoch": 4.223509258503956,
      "grad_norm": 0.0017445869743824005,
      "learning_rate": 0.08306915656973726,
      "loss": 0.3544,
      "num_input_tokens_seen": 49072624,
      "step": 25890
    },
    {
      "epoch": 4.224324985724774,
      "grad_norm": 0.0020118525717407465,
      "learning_rate": 0.08301644597594988,
      "loss": 0.3382,
      "num_input_tokens_seen": 49082144,
      "step": 25895
    },
    {
      "epoch": 4.225140712945591,
      "grad_norm": 0.0026867857668548822,
      "learning_rate": 0.08296374571186826,
      "loss": 0.3547,
      "num_input_tokens_seen": 49092880,
      "step": 25900
    },
    {
      "epoch": 4.225956440166408,
      "grad_norm": 0.004224561620503664,
      "learning_rate": 0.08291105578561955,
      "loss": 0.2979,
      "num_input_tokens_seen": 49102784,
      "step": 25905
    },
    {
      "epoch": 4.226772167387225,
      "grad_norm": 0.0006111544789746404,
      "learning_rate": 0.08285837620532904,
      "loss": 0.3365,
      "num_input_tokens_seen": 49111664,
      "step": 25910
    },
    {
      "epoch": 4.227587894608043,
      "grad_norm": 0.0038300456944853067,
      "learning_rate": 0.0828057069791207,
      "loss": 0.4037,
      "num_input_tokens_seen": 49121216,
      "step": 25915
    },
    {
      "epoch": 4.2284036218288605,
      "grad_norm": 0.0025387718342244625,
      "learning_rate": 0.0827530481151168,
      "loss": 0.3103,
      "num_input_tokens_seen": 49130576,
      "step": 25920
    },
    {
      "epoch": 4.229219349049678,
      "grad_norm": 0.0016241909470409155,
      "learning_rate": 0.08270039962143792,
      "loss": 0.3496,
      "num_input_tokens_seen": 49140528,
      "step": 25925
    },
    {
      "epoch": 4.230035076270495,
      "grad_norm": 0.0019439853494986892,
      "learning_rate": 0.08264776150620314,
      "loss": 0.3402,
      "num_input_tokens_seen": 49151008,
      "step": 25930
    },
    {
      "epoch": 4.230850803491313,
      "grad_norm": 0.003990562632679939,
      "learning_rate": 0.08259513377753,
      "loss": 0.3007,
      "num_input_tokens_seen": 49160992,
      "step": 25935
    },
    {
      "epoch": 4.23166653071213,
      "grad_norm": 0.0010791243985295296,
      "learning_rate": 0.08254251644353423,
      "loss": 0.3101,
      "num_input_tokens_seen": 49170192,
      "step": 25940
    },
    {
      "epoch": 4.232482257932947,
      "grad_norm": 0.0031398222781717777,
      "learning_rate": 0.08248990951233022,
      "loss": 0.3513,
      "num_input_tokens_seen": 49178736,
      "step": 25945
    },
    {
      "epoch": 4.233297985153764,
      "grad_norm": 0.001936398446559906,
      "learning_rate": 0.08243731299203048,
      "loss": 0.2971,
      "num_input_tokens_seen": 49188704,
      "step": 25950
    },
    {
      "epoch": 4.234113712374582,
      "grad_norm": 0.0024267728440463543,
      "learning_rate": 0.08238472689074612,
      "loss": 0.2928,
      "num_input_tokens_seen": 49198064,
      "step": 25955
    },
    {
      "epoch": 4.234929439595399,
      "grad_norm": 0.0014901255490258336,
      "learning_rate": 0.08233215121658666,
      "loss": 0.339,
      "num_input_tokens_seen": 49207632,
      "step": 25960
    },
    {
      "epoch": 4.2357451668162165,
      "grad_norm": 0.0010208917083218694,
      "learning_rate": 0.08227958597765982,
      "loss": 0.304,
      "num_input_tokens_seen": 49216416,
      "step": 25965
    },
    {
      "epoch": 4.236560894037034,
      "grad_norm": 0.0010287775658071041,
      "learning_rate": 0.08222703118207181,
      "loss": 0.3497,
      "num_input_tokens_seen": 49224256,
      "step": 25970
    },
    {
      "epoch": 4.237376621257852,
      "grad_norm": 0.0020156381651759148,
      "learning_rate": 0.08217448683792734,
      "loss": 0.2948,
      "num_input_tokens_seen": 49234880,
      "step": 25975
    },
    {
      "epoch": 4.238192348478669,
      "grad_norm": 0.001615807181224227,
      "learning_rate": 0.08212195295332926,
      "loss": 0.3284,
      "num_input_tokens_seen": 49244544,
      "step": 25980
    },
    {
      "epoch": 4.239008075699486,
      "grad_norm": 0.0008325764792971313,
      "learning_rate": 0.08206942953637915,
      "loss": 0.3608,
      "num_input_tokens_seen": 49252992,
      "step": 25985
    },
    {
      "epoch": 4.239823802920303,
      "grad_norm": 0.00248791859485209,
      "learning_rate": 0.08201691659517658,
      "loss": 0.3146,
      "num_input_tokens_seen": 49261904,
      "step": 25990
    },
    {
      "epoch": 4.240639530141121,
      "grad_norm": 0.0016211681067943573,
      "learning_rate": 0.08196441413781981,
      "loss": 0.3093,
      "num_input_tokens_seen": 49270640,
      "step": 25995
    },
    {
      "epoch": 4.241455257361938,
      "grad_norm": 0.001954183913767338,
      "learning_rate": 0.08191192217240544,
      "loss": 0.2758,
      "num_input_tokens_seen": 49280208,
      "step": 26000
    },
    {
      "epoch": 4.241455257361938,
      "eval_loss": 0.3293290138244629,
      "eval_runtime": 152.788,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 49280208,
      "step": 26000
    },
    {
      "epoch": 4.242270984582755,
      "grad_norm": 0.001744957990013063,
      "learning_rate": 0.08185944070702823,
      "loss": 0.329,
      "num_input_tokens_seen": 49288880,
      "step": 26005
    },
    {
      "epoch": 4.243086711803572,
      "grad_norm": 0.0014932325575500727,
      "learning_rate": 0.08180696974978159,
      "loss": 0.2949,
      "num_input_tokens_seen": 49299408,
      "step": 26010
    },
    {
      "epoch": 4.2439024390243905,
      "grad_norm": 0.003111523576080799,
      "learning_rate": 0.08175450930875724,
      "loss": 0.3309,
      "num_input_tokens_seen": 49309360,
      "step": 26015
    },
    {
      "epoch": 4.244718166245208,
      "grad_norm": 0.001963652204722166,
      "learning_rate": 0.08170205939204513,
      "loss": 0.2892,
      "num_input_tokens_seen": 49319056,
      "step": 26020
    },
    {
      "epoch": 4.245533893466025,
      "grad_norm": 0.0017723814817145467,
      "learning_rate": 0.08164962000773379,
      "loss": 0.3674,
      "num_input_tokens_seen": 49328272,
      "step": 26025
    },
    {
      "epoch": 4.246349620686843,
      "grad_norm": 0.0010222869459539652,
      "learning_rate": 0.08159719116390995,
      "loss": 0.3328,
      "num_input_tokens_seen": 49337264,
      "step": 26030
    },
    {
      "epoch": 4.24716534790766,
      "grad_norm": 0.003177608596161008,
      "learning_rate": 0.08154477286865887,
      "loss": 0.3649,
      "num_input_tokens_seen": 49346160,
      "step": 26035
    },
    {
      "epoch": 4.247981075128477,
      "grad_norm": 0.0011338881449773908,
      "learning_rate": 0.08149236513006404,
      "loss": 0.3256,
      "num_input_tokens_seen": 49356112,
      "step": 26040
    },
    {
      "epoch": 4.248796802349294,
      "grad_norm": 0.001421342371031642,
      "learning_rate": 0.08143996795620746,
      "loss": 0.3005,
      "num_input_tokens_seen": 49365408,
      "step": 26045
    },
    {
      "epoch": 4.249612529570112,
      "grad_norm": 0.0014633856480941176,
      "learning_rate": 0.08138758135516938,
      "loss": 0.3082,
      "num_input_tokens_seen": 49376240,
      "step": 26050
    },
    {
      "epoch": 4.250428256790929,
      "grad_norm": 0.000742425094358623,
      "learning_rate": 0.08133520533502851,
      "loss": 0.295,
      "num_input_tokens_seen": 49386144,
      "step": 26055
    },
    {
      "epoch": 4.251243984011746,
      "grad_norm": 0.0017234114930033684,
      "learning_rate": 0.08128283990386184,
      "loss": 0.3251,
      "num_input_tokens_seen": 49395392,
      "step": 26060
    },
    {
      "epoch": 4.2520597112325635,
      "grad_norm": 0.0025665992870926857,
      "learning_rate": 0.08123048506974488,
      "loss": 0.3065,
      "num_input_tokens_seen": 49405856,
      "step": 26065
    },
    {
      "epoch": 4.252875438453382,
      "grad_norm": 0.0030323003884404898,
      "learning_rate": 0.08117814084075124,
      "loss": 0.3425,
      "num_input_tokens_seen": 49415504,
      "step": 26070
    },
    {
      "epoch": 4.253691165674199,
      "grad_norm": 0.0012094670673832297,
      "learning_rate": 0.08112580722495318,
      "loss": 0.3846,
      "num_input_tokens_seen": 49426032,
      "step": 26075
    },
    {
      "epoch": 4.254506892895016,
      "grad_norm": 0.0017875401536002755,
      "learning_rate": 0.08107348423042122,
      "loss": 0.2812,
      "num_input_tokens_seen": 49434432,
      "step": 26080
    },
    {
      "epoch": 4.255322620115833,
      "grad_norm": 0.0007317587733268738,
      "learning_rate": 0.08102117186522413,
      "loss": 0.335,
      "num_input_tokens_seen": 49444688,
      "step": 26085
    },
    {
      "epoch": 4.256138347336651,
      "grad_norm": 0.001554250018671155,
      "learning_rate": 0.08096887013742916,
      "loss": 0.319,
      "num_input_tokens_seen": 49453776,
      "step": 26090
    },
    {
      "epoch": 4.256954074557468,
      "grad_norm": 0.0018000510754063725,
      "learning_rate": 0.08091657905510198,
      "loss": 0.3318,
      "num_input_tokens_seen": 49461968,
      "step": 26095
    },
    {
      "epoch": 4.257769801778285,
      "grad_norm": 0.0007960131042636931,
      "learning_rate": 0.08086429862630642,
      "loss": 0.2809,
      "num_input_tokens_seen": 49472768,
      "step": 26100
    },
    {
      "epoch": 4.258585528999102,
      "grad_norm": 0.0014775024028494954,
      "learning_rate": 0.08081202885910488,
      "loss": 0.3206,
      "num_input_tokens_seen": 49482368,
      "step": 26105
    },
    {
      "epoch": 4.25940125621992,
      "grad_norm": 0.0025290821213275194,
      "learning_rate": 0.08075976976155795,
      "loss": 0.2935,
      "num_input_tokens_seen": 49491680,
      "step": 26110
    },
    {
      "epoch": 4.2602169834407375,
      "grad_norm": 0.0023676902055740356,
      "learning_rate": 0.08070752134172461,
      "loss": 0.2602,
      "num_input_tokens_seen": 49500880,
      "step": 26115
    },
    {
      "epoch": 4.261032710661555,
      "grad_norm": 0.0031701219268143177,
      "learning_rate": 0.08065528360766229,
      "loss": 0.3006,
      "num_input_tokens_seen": 49509472,
      "step": 26120
    },
    {
      "epoch": 4.261848437882372,
      "grad_norm": 0.0040496280416846275,
      "learning_rate": 0.08060305656742664,
      "loss": 0.3564,
      "num_input_tokens_seen": 49517824,
      "step": 26125
    },
    {
      "epoch": 4.26266416510319,
      "grad_norm": 0.0011752774007618427,
      "learning_rate": 0.08055084022907182,
      "loss": 0.353,
      "num_input_tokens_seen": 49526448,
      "step": 26130
    },
    {
      "epoch": 4.263479892324007,
      "grad_norm": 0.0021623442880809307,
      "learning_rate": 0.08049863460065014,
      "loss": 0.327,
      "num_input_tokens_seen": 49535664,
      "step": 26135
    },
    {
      "epoch": 4.264295619544824,
      "grad_norm": 0.002612708369269967,
      "learning_rate": 0.0804464396902124,
      "loss": 0.3162,
      "num_input_tokens_seen": 49544848,
      "step": 26140
    },
    {
      "epoch": 4.265111346765641,
      "grad_norm": 0.002523548901081085,
      "learning_rate": 0.08039425550580777,
      "loss": 0.401,
      "num_input_tokens_seen": 49552592,
      "step": 26145
    },
    {
      "epoch": 4.265927073986459,
      "grad_norm": 0.000888580281753093,
      "learning_rate": 0.08034208205548363,
      "loss": 0.3086,
      "num_input_tokens_seen": 49562560,
      "step": 26150
    },
    {
      "epoch": 4.266742801207276,
      "grad_norm": 0.0031600843649357557,
      "learning_rate": 0.08028991934728581,
      "loss": 0.3248,
      "num_input_tokens_seen": 49572512,
      "step": 26155
    },
    {
      "epoch": 4.2675585284280935,
      "grad_norm": 0.0012454130919650197,
      "learning_rate": 0.0802377673892585,
      "loss": 0.3684,
      "num_input_tokens_seen": 49580656,
      "step": 26160
    },
    {
      "epoch": 4.268374255648911,
      "grad_norm": 0.0016364016337320209,
      "learning_rate": 0.0801856261894441,
      "loss": 0.3233,
      "num_input_tokens_seen": 49591392,
      "step": 26165
    },
    {
      "epoch": 4.269189982869729,
      "grad_norm": 0.003063984215259552,
      "learning_rate": 0.08013349575588354,
      "loss": 0.318,
      "num_input_tokens_seen": 49602048,
      "step": 26170
    },
    {
      "epoch": 4.270005710090546,
      "grad_norm": 0.0012660189531743526,
      "learning_rate": 0.08008137609661586,
      "loss": 0.3432,
      "num_input_tokens_seen": 49610944,
      "step": 26175
    },
    {
      "epoch": 4.270821437311363,
      "grad_norm": 0.0016324162716045976,
      "learning_rate": 0.08002926721967872,
      "loss": 0.2928,
      "num_input_tokens_seen": 49619728,
      "step": 26180
    },
    {
      "epoch": 4.27163716453218,
      "grad_norm": 0.001093649072572589,
      "learning_rate": 0.07997716913310782,
      "loss": 0.3055,
      "num_input_tokens_seen": 49629680,
      "step": 26185
    },
    {
      "epoch": 4.272452891752998,
      "grad_norm": 0.0026113097555935383,
      "learning_rate": 0.07992508184493745,
      "loss": 0.3605,
      "num_input_tokens_seen": 49639488,
      "step": 26190
    },
    {
      "epoch": 4.273268618973815,
      "grad_norm": 0.0028106900863349438,
      "learning_rate": 0.07987300536320001,
      "loss": 0.3578,
      "num_input_tokens_seen": 49648992,
      "step": 26195
    },
    {
      "epoch": 4.274084346194632,
      "grad_norm": 0.00140187528450042,
      "learning_rate": 0.07982093969592649,
      "loss": 0.3199,
      "num_input_tokens_seen": 49658080,
      "step": 26200
    },
    {
      "epoch": 4.274084346194632,
      "eval_loss": 0.3263053596019745,
      "eval_runtime": 152.8391,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 49658080,
      "step": 26200
    },
    {
      "epoch": 4.27490007341545,
      "grad_norm": 0.001842053490690887,
      "learning_rate": 0.07976888485114592,
      "loss": 0.3589,
      "num_input_tokens_seen": 49667248,
      "step": 26205
    },
    {
      "epoch": 4.275715800636267,
      "grad_norm": 0.0014943333808332682,
      "learning_rate": 0.07971684083688595,
      "loss": 0.3286,
      "num_input_tokens_seen": 49675952,
      "step": 26210
    },
    {
      "epoch": 4.276531527857085,
      "grad_norm": 0.002006806433200836,
      "learning_rate": 0.0796648076611723,
      "loss": 0.2771,
      "num_input_tokens_seen": 49685664,
      "step": 26215
    },
    {
      "epoch": 4.277347255077902,
      "grad_norm": 0.0014077811501920223,
      "learning_rate": 0.07961278533202922,
      "loss": 0.3101,
      "num_input_tokens_seen": 49695008,
      "step": 26220
    },
    {
      "epoch": 4.27816298229872,
      "grad_norm": 0.0018555353162810206,
      "learning_rate": 0.07956077385747919,
      "loss": 0.3186,
      "num_input_tokens_seen": 49704448,
      "step": 26225
    },
    {
      "epoch": 4.278978709519537,
      "grad_norm": 0.0012749187881127,
      "learning_rate": 0.079508773245543,
      "loss": 0.3269,
      "num_input_tokens_seen": 49713248,
      "step": 26230
    },
    {
      "epoch": 4.279794436740354,
      "grad_norm": 0.002521687187254429,
      "learning_rate": 0.07945678350423982,
      "loss": 0.3447,
      "num_input_tokens_seen": 49722752,
      "step": 26235
    },
    {
      "epoch": 4.280610163961171,
      "grad_norm": 0.0019540656358003616,
      "learning_rate": 0.07940480464158717,
      "loss": 0.3094,
      "num_input_tokens_seen": 49732736,
      "step": 26240
    },
    {
      "epoch": 4.281425891181989,
      "grad_norm": 0.0011742018396034837,
      "learning_rate": 0.07935283666560076,
      "loss": 0.3261,
      "num_input_tokens_seen": 49742768,
      "step": 26245
    },
    {
      "epoch": 4.282241618402806,
      "grad_norm": 0.0031256440561264753,
      "learning_rate": 0.07930087958429478,
      "loss": 0.3168,
      "num_input_tokens_seen": 49752192,
      "step": 26250
    },
    {
      "epoch": 4.283057345623623,
      "grad_norm": 0.0011421741219237447,
      "learning_rate": 0.07924893340568159,
      "loss": 0.3111,
      "num_input_tokens_seen": 49761728,
      "step": 26255
    },
    {
      "epoch": 4.2838730728444405,
      "grad_norm": 0.0011258921585977077,
      "learning_rate": 0.07919699813777205,
      "loss": 0.2906,
      "num_input_tokens_seen": 49771568,
      "step": 26260
    },
    {
      "epoch": 4.2846888000652585,
      "grad_norm": 0.0023420799989253283,
      "learning_rate": 0.07914507378857515,
      "loss": 0.3768,
      "num_input_tokens_seen": 49780976,
      "step": 26265
    },
    {
      "epoch": 4.285504527286076,
      "grad_norm": 0.0022041359916329384,
      "learning_rate": 0.07909316036609822,
      "loss": 0.3346,
      "num_input_tokens_seen": 49791120,
      "step": 26270
    },
    {
      "epoch": 4.286320254506893,
      "grad_norm": 0.003064095973968506,
      "learning_rate": 0.07904125787834704,
      "loss": 0.3767,
      "num_input_tokens_seen": 49799952,
      "step": 26275
    },
    {
      "epoch": 4.28713598172771,
      "grad_norm": 0.0018453917000442743,
      "learning_rate": 0.07898936633332569,
      "loss": 0.3063,
      "num_input_tokens_seen": 49809008,
      "step": 26280
    },
    {
      "epoch": 4.287951708948528,
      "grad_norm": 0.0021941184531897306,
      "learning_rate": 0.07893748573903635,
      "loss": 0.421,
      "num_input_tokens_seen": 49818992,
      "step": 26285
    },
    {
      "epoch": 4.288767436169345,
      "grad_norm": 0.0033685157541185617,
      "learning_rate": 0.0788856161034798,
      "loss": 0.3669,
      "num_input_tokens_seen": 49828976,
      "step": 26290
    },
    {
      "epoch": 4.289583163390162,
      "grad_norm": 0.0035801257472485304,
      "learning_rate": 0.07883375743465487,
      "loss": 0.3231,
      "num_input_tokens_seen": 49837712,
      "step": 26295
    },
    {
      "epoch": 4.290398890610979,
      "grad_norm": 0.0015137314330786467,
      "learning_rate": 0.07878190974055888,
      "loss": 0.3531,
      "num_input_tokens_seen": 49846128,
      "step": 26300
    },
    {
      "epoch": 4.291214617831797,
      "grad_norm": 0.0017067287117242813,
      "learning_rate": 0.07873007302918746,
      "loss": 0.3318,
      "num_input_tokens_seen": 49856192,
      "step": 26305
    },
    {
      "epoch": 4.2920303450526145,
      "grad_norm": 0.003151815617457032,
      "learning_rate": 0.07867824730853433,
      "loss": 0.3367,
      "num_input_tokens_seen": 49865936,
      "step": 26310
    },
    {
      "epoch": 4.292846072273432,
      "grad_norm": 0.0017699110321700573,
      "learning_rate": 0.07862643258659176,
      "loss": 0.3664,
      "num_input_tokens_seen": 49875488,
      "step": 26315
    },
    {
      "epoch": 4.293661799494249,
      "grad_norm": 0.0016016808804124594,
      "learning_rate": 0.07857462887135026,
      "loss": 0.3315,
      "num_input_tokens_seen": 49884800,
      "step": 26320
    },
    {
      "epoch": 4.294477526715067,
      "grad_norm": 0.003206356894224882,
      "learning_rate": 0.0785228361707986,
      "loss": 0.4239,
      "num_input_tokens_seen": 49894176,
      "step": 26325
    },
    {
      "epoch": 4.295293253935884,
      "grad_norm": 0.0020775310695171356,
      "learning_rate": 0.07847105449292378,
      "loss": 0.3501,
      "num_input_tokens_seen": 49903488,
      "step": 26330
    },
    {
      "epoch": 4.296108981156701,
      "grad_norm": 0.0023868440184742212,
      "learning_rate": 0.0784192838457113,
      "loss": 0.3225,
      "num_input_tokens_seen": 49912096,
      "step": 26335
    },
    {
      "epoch": 4.296924708377518,
      "grad_norm": 0.0010834578424692154,
      "learning_rate": 0.07836752423714473,
      "loss": 0.2936,
      "num_input_tokens_seen": 49921408,
      "step": 26340
    },
    {
      "epoch": 4.297740435598336,
      "grad_norm": 0.0009710023878142238,
      "learning_rate": 0.07831577567520616,
      "loss": 0.3478,
      "num_input_tokens_seen": 49931168,
      "step": 26345
    },
    {
      "epoch": 4.298556162819153,
      "grad_norm": 0.0010188412852585316,
      "learning_rate": 0.07826403816787579,
      "loss": 0.3646,
      "num_input_tokens_seen": 49941024,
      "step": 26350
    },
    {
      "epoch": 4.2993718900399704,
      "grad_norm": 0.0005183416651561856,
      "learning_rate": 0.0782123117231322,
      "loss": 0.3389,
      "num_input_tokens_seen": 49949232,
      "step": 26355
    },
    {
      "epoch": 4.300187617260788,
      "grad_norm": 0.000900274608284235,
      "learning_rate": 0.07816059634895237,
      "loss": 0.3466,
      "num_input_tokens_seen": 49959072,
      "step": 26360
    },
    {
      "epoch": 4.301003344481606,
      "grad_norm": 0.0013111617881804705,
      "learning_rate": 0.0781088920533113,
      "loss": 0.3593,
      "num_input_tokens_seen": 49968496,
      "step": 26365
    },
    {
      "epoch": 4.301819071702423,
      "grad_norm": 0.003621526760980487,
      "learning_rate": 0.07805719884418257,
      "loss": 0.3491,
      "num_input_tokens_seen": 49977792,
      "step": 26370
    },
    {
      "epoch": 4.30263479892324,
      "grad_norm": 0.0026833582669496536,
      "learning_rate": 0.07800551672953779,
      "loss": 0.3725,
      "num_input_tokens_seen": 49987152,
      "step": 26375
    },
    {
      "epoch": 4.303450526144058,
      "grad_norm": 0.0013384618796408176,
      "learning_rate": 0.07795384571734709,
      "loss": 0.3417,
      "num_input_tokens_seen": 49997920,
      "step": 26380
    },
    {
      "epoch": 4.304266253364875,
      "grad_norm": 0.0007923850207589567,
      "learning_rate": 0.07790218581557883,
      "loss": 0.3279,
      "num_input_tokens_seen": 50007280,
      "step": 26385
    },
    {
      "epoch": 4.305081980585692,
      "grad_norm": 0.0026556418742984533,
      "learning_rate": 0.07785053703219949,
      "loss": 0.3871,
      "num_input_tokens_seen": 50015600,
      "step": 26390
    },
    {
      "epoch": 4.305897707806509,
      "grad_norm": 0.0018380613764747977,
      "learning_rate": 0.07779889937517409,
      "loss": 0.3255,
      "num_input_tokens_seen": 50025120,
      "step": 26395
    },
    {
      "epoch": 4.306713435027326,
      "grad_norm": 0.003187389113008976,
      "learning_rate": 0.0777472728524657,
      "loss": 0.3332,
      "num_input_tokens_seen": 50034848,
      "step": 26400
    },
    {
      "epoch": 4.306713435027326,
      "eval_loss": 0.3220772445201874,
      "eval_runtime": 152.9095,
      "eval_samples_per_second": 17.821,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 50034848,
      "step": 26400
    },
    {
      "epoch": 4.307529162248144,
      "grad_norm": 0.00285917310975492,
      "learning_rate": 0.07769565747203584,
      "loss": 0.3364,
      "num_input_tokens_seen": 50044288,
      "step": 26405
    },
    {
      "epoch": 4.3083448894689615,
      "grad_norm": 0.0017644979525357485,
      "learning_rate": 0.07764405324184427,
      "loss": 0.3389,
      "num_input_tokens_seen": 50054176,
      "step": 26410
    },
    {
      "epoch": 4.309160616689779,
      "grad_norm": 0.0024373228661715984,
      "learning_rate": 0.07759246016984889,
      "loss": 0.2409,
      "num_input_tokens_seen": 50063136,
      "step": 26415
    },
    {
      "epoch": 4.309976343910597,
      "grad_norm": 0.0026307785883545876,
      "learning_rate": 0.07754087826400609,
      "loss": 0.3311,
      "num_input_tokens_seen": 50072096,
      "step": 26420
    },
    {
      "epoch": 4.310792071131414,
      "grad_norm": 0.0022501545026898384,
      "learning_rate": 0.0774893075322705,
      "loss": 0.3979,
      "num_input_tokens_seen": 50081120,
      "step": 26425
    },
    {
      "epoch": 4.311607798352231,
      "grad_norm": 0.0018682373920455575,
      "learning_rate": 0.07743774798259484,
      "loss": 0.3027,
      "num_input_tokens_seen": 50090304,
      "step": 26430
    },
    {
      "epoch": 4.312423525573048,
      "grad_norm": 0.0021342833060771227,
      "learning_rate": 0.07738619962293032,
      "loss": 0.3161,
      "num_input_tokens_seen": 50100400,
      "step": 26435
    },
    {
      "epoch": 4.313239252793866,
      "grad_norm": 0.0013982071541249752,
      "learning_rate": 0.0773346624612264,
      "loss": 0.3419,
      "num_input_tokens_seen": 50109632,
      "step": 26440
    },
    {
      "epoch": 4.314054980014683,
      "grad_norm": 0.0009636518079787493,
      "learning_rate": 0.07728313650543066,
      "loss": 0.3315,
      "num_input_tokens_seen": 50119552,
      "step": 26445
    },
    {
      "epoch": 4.3148707072355,
      "grad_norm": 0.001968870870769024,
      "learning_rate": 0.07723162176348913,
      "loss": 0.3439,
      "num_input_tokens_seen": 50129008,
      "step": 26450
    },
    {
      "epoch": 4.3156864344563175,
      "grad_norm": 0.0014370238641276956,
      "learning_rate": 0.07718011824334593,
      "loss": 0.3305,
      "num_input_tokens_seen": 50137904,
      "step": 26455
    },
    {
      "epoch": 4.3165021616771355,
      "grad_norm": 0.0009763387497514486,
      "learning_rate": 0.07712862595294363,
      "loss": 0.3569,
      "num_input_tokens_seen": 50147632,
      "step": 26460
    },
    {
      "epoch": 4.317317888897953,
      "grad_norm": 0.0011420761002227664,
      "learning_rate": 0.07707714490022301,
      "loss": 0.302,
      "num_input_tokens_seen": 50157104,
      "step": 26465
    },
    {
      "epoch": 4.31813361611877,
      "grad_norm": 0.002746396465227008,
      "learning_rate": 0.07702567509312298,
      "loss": 0.3855,
      "num_input_tokens_seen": 50166640,
      "step": 26470
    },
    {
      "epoch": 4.318949343339587,
      "grad_norm": 0.0016006357036530972,
      "learning_rate": 0.07697421653958098,
      "loss": 0.3384,
      "num_input_tokens_seen": 50176848,
      "step": 26475
    },
    {
      "epoch": 4.319765070560405,
      "grad_norm": 0.0011206287890672684,
      "learning_rate": 0.07692276924753247,
      "loss": 0.3476,
      "num_input_tokens_seen": 50186064,
      "step": 26480
    },
    {
      "epoch": 4.320580797781222,
      "grad_norm": 0.0021877784747630358,
      "learning_rate": 0.07687133322491124,
      "loss": 0.3487,
      "num_input_tokens_seen": 50195088,
      "step": 26485
    },
    {
      "epoch": 4.321396525002039,
      "grad_norm": 0.0016786300111562014,
      "learning_rate": 0.07681990847964948,
      "loss": 0.3158,
      "num_input_tokens_seen": 50204640,
      "step": 26490
    },
    {
      "epoch": 4.322212252222856,
      "grad_norm": 0.0012778572272509336,
      "learning_rate": 0.0767684950196774,
      "loss": 0.316,
      "num_input_tokens_seen": 50214512,
      "step": 26495
    },
    {
      "epoch": 4.323027979443674,
      "grad_norm": 0.0010971089359372854,
      "learning_rate": 0.0767170928529237,
      "loss": 0.3309,
      "num_input_tokens_seen": 50225136,
      "step": 26500
    },
    {
      "epoch": 4.3238437066644915,
      "grad_norm": 0.0019054891308769584,
      "learning_rate": 0.07666570198731526,
      "loss": 0.3441,
      "num_input_tokens_seen": 50234880,
      "step": 26505
    },
    {
      "epoch": 4.324659433885309,
      "grad_norm": 0.0027352396864444017,
      "learning_rate": 0.07661432243077708,
      "loss": 0.3543,
      "num_input_tokens_seen": 50244208,
      "step": 26510
    },
    {
      "epoch": 4.325475161106126,
      "grad_norm": 0.000718708906788379,
      "learning_rate": 0.0765629541912326,
      "loss": 0.3116,
      "num_input_tokens_seen": 50254880,
      "step": 26515
    },
    {
      "epoch": 4.326290888326944,
      "grad_norm": 0.0010432286653667688,
      "learning_rate": 0.07651159727660352,
      "loss": 0.316,
      "num_input_tokens_seen": 50263568,
      "step": 26520
    },
    {
      "epoch": 4.327106615547761,
      "grad_norm": 0.0022367662750184536,
      "learning_rate": 0.07646025169480959,
      "loss": 0.3206,
      "num_input_tokens_seen": 50272736,
      "step": 26525
    },
    {
      "epoch": 4.327922342768578,
      "grad_norm": 0.0015025617321953177,
      "learning_rate": 0.07640891745376908,
      "loss": 0.3277,
      "num_input_tokens_seen": 50281232,
      "step": 26530
    },
    {
      "epoch": 4.328738069989395,
      "grad_norm": 0.001014519133605063,
      "learning_rate": 0.07635759456139822,
      "loss": 0.3449,
      "num_input_tokens_seen": 50290288,
      "step": 26535
    },
    {
      "epoch": 4.329553797210213,
      "grad_norm": 0.0012665885733440518,
      "learning_rate": 0.0763062830256118,
      "loss": 0.273,
      "num_input_tokens_seen": 50300048,
      "step": 26540
    },
    {
      "epoch": 4.33036952443103,
      "grad_norm": 0.000765150529332459,
      "learning_rate": 0.07625498285432258,
      "loss": 0.3188,
      "num_input_tokens_seen": 50310624,
      "step": 26545
    },
    {
      "epoch": 4.331185251651847,
      "grad_norm": 0.002724100835621357,
      "learning_rate": 0.07620369405544176,
      "loss": 0.361,
      "num_input_tokens_seen": 50319888,
      "step": 26550
    },
    {
      "epoch": 4.332000978872665,
      "grad_norm": 0.0009455408435314894,
      "learning_rate": 0.07615241663687868,
      "loss": 0.3142,
      "num_input_tokens_seen": 50329376,
      "step": 26555
    },
    {
      "epoch": 4.332816706093483,
      "grad_norm": 0.0014740474289283156,
      "learning_rate": 0.07610115060654106,
      "loss": 0.3204,
      "num_input_tokens_seen": 50338064,
      "step": 26560
    },
    {
      "epoch": 4.3336324333143,
      "grad_norm": 0.0010298321722075343,
      "learning_rate": 0.07604989597233458,
      "loss": 0.325,
      "num_input_tokens_seen": 50348128,
      "step": 26565
    },
    {
      "epoch": 4.334448160535117,
      "grad_norm": 0.001327702309936285,
      "learning_rate": 0.07599865274216352,
      "loss": 0.3234,
      "num_input_tokens_seen": 50357312,
      "step": 26570
    },
    {
      "epoch": 4.335263887755934,
      "grad_norm": 0.001213515060953796,
      "learning_rate": 0.07594742092393013,
      "loss": 0.3317,
      "num_input_tokens_seen": 50366592,
      "step": 26575
    },
    {
      "epoch": 4.336079614976752,
      "grad_norm": 0.0013760479632765055,
      "learning_rate": 0.07589620052553503,
      "loss": 0.3255,
      "num_input_tokens_seen": 50376112,
      "step": 26580
    },
    {
      "epoch": 4.336895342197569,
      "grad_norm": 0.001045191427692771,
      "learning_rate": 0.0758449915548771,
      "loss": 0.2858,
      "num_input_tokens_seen": 50386192,
      "step": 26585
    },
    {
      "epoch": 4.337711069418386,
      "grad_norm": 0.0013624151470139623,
      "learning_rate": 0.07579379401985332,
      "loss": 0.3611,
      "num_input_tokens_seen": 50394864,
      "step": 26590
    },
    {
      "epoch": 4.338526796639204,
      "grad_norm": 0.002260581124573946,
      "learning_rate": 0.07574260792835905,
      "loss": 0.4104,
      "num_input_tokens_seen": 50404960,
      "step": 26595
    },
    {
      "epoch": 4.339342523860021,
      "grad_norm": 0.004017337691038847,
      "learning_rate": 0.07569143328828784,
      "loss": 0.3795,
      "num_input_tokens_seen": 50413376,
      "step": 26600
    },
    {
      "epoch": 4.339342523860021,
      "eval_loss": 0.32693833112716675,
      "eval_runtime": 152.9244,
      "eval_samples_per_second": 17.819,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 50413376,
      "step": 26600
    },
    {
      "epoch": 4.3401582510808385,
      "grad_norm": 0.0008425320265814662,
      "learning_rate": 0.0756402701075314,
      "loss": 0.3516,
      "num_input_tokens_seen": 50423824,
      "step": 26605
    },
    {
      "epoch": 4.340973978301656,
      "grad_norm": 0.0007542672101408243,
      "learning_rate": 0.07558911839397982,
      "loss": 0.3555,
      "num_input_tokens_seen": 50433056,
      "step": 26610
    },
    {
      "epoch": 4.341789705522474,
      "grad_norm": 0.0007706525502726436,
      "learning_rate": 0.07553797815552123,
      "loss": 0.325,
      "num_input_tokens_seen": 50442176,
      "step": 26615
    },
    {
      "epoch": 4.342605432743291,
      "grad_norm": 0.0011391171719878912,
      "learning_rate": 0.07548684940004222,
      "loss": 0.2933,
      "num_input_tokens_seen": 50452032,
      "step": 26620
    },
    {
      "epoch": 4.343421159964108,
      "grad_norm": 0.002249622717499733,
      "learning_rate": 0.07543573213542744,
      "loss": 0.2618,
      "num_input_tokens_seen": 50462768,
      "step": 26625
    },
    {
      "epoch": 4.344236887184925,
      "grad_norm": 0.0027800793759524822,
      "learning_rate": 0.0753846263695597,
      "loss": 0.3121,
      "num_input_tokens_seen": 50473216,
      "step": 26630
    },
    {
      "epoch": 4.345052614405743,
      "grad_norm": 0.0015983942430466413,
      "learning_rate": 0.07533353211032029,
      "loss": 0.3749,
      "num_input_tokens_seen": 50483456,
      "step": 26635
    },
    {
      "epoch": 4.34586834162656,
      "grad_norm": 0.002249671146273613,
      "learning_rate": 0.07528244936558857,
      "loss": 0.3875,
      "num_input_tokens_seen": 50492992,
      "step": 26640
    },
    {
      "epoch": 4.346684068847377,
      "grad_norm": 0.0014234088594093919,
      "learning_rate": 0.07523137814324206,
      "loss": 0.3056,
      "num_input_tokens_seen": 50502208,
      "step": 26645
    },
    {
      "epoch": 4.3474997960681945,
      "grad_norm": 0.0018890190403908491,
      "learning_rate": 0.07518031845115672,
      "loss": 0.2817,
      "num_input_tokens_seen": 50511616,
      "step": 26650
    },
    {
      "epoch": 4.3483155232890125,
      "grad_norm": 0.0005453414050862193,
      "learning_rate": 0.07512927029720647,
      "loss": 0.3645,
      "num_input_tokens_seen": 50521952,
      "step": 26655
    },
    {
      "epoch": 4.34913125050983,
      "grad_norm": 0.001248177606612444,
      "learning_rate": 0.0750782336892636,
      "loss": 0.323,
      "num_input_tokens_seen": 50531296,
      "step": 26660
    },
    {
      "epoch": 4.349946977730647,
      "grad_norm": 0.000963873288128525,
      "learning_rate": 0.0750272086351987,
      "loss": 0.3106,
      "num_input_tokens_seen": 50540688,
      "step": 26665
    },
    {
      "epoch": 4.350762704951464,
      "grad_norm": 0.00166997779160738,
      "learning_rate": 0.07497619514288031,
      "loss": 0.3614,
      "num_input_tokens_seen": 50550320,
      "step": 26670
    },
    {
      "epoch": 4.351578432172282,
      "grad_norm": 0.0017131338827311993,
      "learning_rate": 0.07492519322017545,
      "loss": 0.3606,
      "num_input_tokens_seen": 50559424,
      "step": 26675
    },
    {
      "epoch": 4.352394159393099,
      "grad_norm": 0.0027018741238862276,
      "learning_rate": 0.0748742028749493,
      "loss": 0.3473,
      "num_input_tokens_seen": 50569536,
      "step": 26680
    },
    {
      "epoch": 4.353209886613916,
      "grad_norm": 0.0033322565723210573,
      "learning_rate": 0.0748232241150651,
      "loss": 0.3231,
      "num_input_tokens_seen": 50579648,
      "step": 26685
    },
    {
      "epoch": 4.354025613834733,
      "grad_norm": 0.001650832244195044,
      "learning_rate": 0.07477225694838453,
      "loss": 0.3443,
      "num_input_tokens_seen": 50589328,
      "step": 26690
    },
    {
      "epoch": 4.354841341055551,
      "grad_norm": 0.000628096517175436,
      "learning_rate": 0.07472130138276731,
      "loss": 0.4114,
      "num_input_tokens_seen": 50598208,
      "step": 26695
    },
    {
      "epoch": 4.3556570682763684,
      "grad_norm": 0.002825111150741577,
      "learning_rate": 0.07467035742607138,
      "loss": 0.3482,
      "num_input_tokens_seen": 50607920,
      "step": 26700
    },
    {
      "epoch": 4.356472795497186,
      "grad_norm": 0.0012615309096872807,
      "learning_rate": 0.07461942508615303,
      "loss": 0.265,
      "num_input_tokens_seen": 50617712,
      "step": 26705
    },
    {
      "epoch": 4.357288522718003,
      "grad_norm": 0.001781081547960639,
      "learning_rate": 0.07456850437086657,
      "loss": 0.2969,
      "num_input_tokens_seen": 50626640,
      "step": 26710
    },
    {
      "epoch": 4.358104249938821,
      "grad_norm": 0.0017530051991343498,
      "learning_rate": 0.07451759528806468,
      "loss": 0.3757,
      "num_input_tokens_seen": 50635984,
      "step": 26715
    },
    {
      "epoch": 4.358919977159638,
      "grad_norm": 0.0009197702747769654,
      "learning_rate": 0.0744666978455982,
      "loss": 0.2927,
      "num_input_tokens_seen": 50645504,
      "step": 26720
    },
    {
      "epoch": 4.359735704380455,
      "grad_norm": 0.0012307880679145455,
      "learning_rate": 0.07441581205131609,
      "loss": 0.3352,
      "num_input_tokens_seen": 50653824,
      "step": 26725
    },
    {
      "epoch": 4.360551431601272,
      "grad_norm": 0.001685073133558035,
      "learning_rate": 0.07436493791306566,
      "loss": 0.3273,
      "num_input_tokens_seen": 50663472,
      "step": 26730
    },
    {
      "epoch": 4.36136715882209,
      "grad_norm": 0.0016699954867362976,
      "learning_rate": 0.07431407543869223,
      "loss": 0.3222,
      "num_input_tokens_seen": 50672272,
      "step": 26735
    },
    {
      "epoch": 4.362182886042907,
      "grad_norm": 0.0023771077394485474,
      "learning_rate": 0.0742632246360395,
      "loss": 0.3467,
      "num_input_tokens_seen": 50682400,
      "step": 26740
    },
    {
      "epoch": 4.362998613263724,
      "grad_norm": 0.002152740489691496,
      "learning_rate": 0.07421238551294934,
      "loss": 0.3528,
      "num_input_tokens_seen": 50692240,
      "step": 26745
    },
    {
      "epoch": 4.3638143404845415,
      "grad_norm": 0.0010346658527851105,
      "learning_rate": 0.07416155807726171,
      "loss": 0.341,
      "num_input_tokens_seen": 50700608,
      "step": 26750
    },
    {
      "epoch": 4.3646300677053596,
      "grad_norm": 0.0022601555101573467,
      "learning_rate": 0.07411074233681492,
      "loss": 0.3037,
      "num_input_tokens_seen": 50710528,
      "step": 26755
    },
    {
      "epoch": 4.365445794926177,
      "grad_norm": 0.0006440271390601993,
      "learning_rate": 0.07405993829944528,
      "loss": 0.3829,
      "num_input_tokens_seen": 50719136,
      "step": 26760
    },
    {
      "epoch": 4.366261522146994,
      "grad_norm": 0.0021451946813613176,
      "learning_rate": 0.07400914597298755,
      "loss": 0.3408,
      "num_input_tokens_seen": 50728416,
      "step": 26765
    },
    {
      "epoch": 4.367077249367812,
      "grad_norm": 0.0018709212308749557,
      "learning_rate": 0.07395836536527445,
      "loss": 0.2861,
      "num_input_tokens_seen": 50738416,
      "step": 26770
    },
    {
      "epoch": 4.367892976588629,
      "grad_norm": 0.0006942061008885503,
      "learning_rate": 0.07390759648413696,
      "loss": 0.3421,
      "num_input_tokens_seen": 50747392,
      "step": 26775
    },
    {
      "epoch": 4.368708703809446,
      "grad_norm": 0.0008939357940107584,
      "learning_rate": 0.07385683933740435,
      "loss": 0.2845,
      "num_input_tokens_seen": 50758096,
      "step": 26780
    },
    {
      "epoch": 4.369524431030263,
      "grad_norm": 0.002244410337880254,
      "learning_rate": 0.07380609393290402,
      "loss": 0.3919,
      "num_input_tokens_seen": 50767632,
      "step": 26785
    },
    {
      "epoch": 4.370340158251081,
      "grad_norm": 0.001421002671122551,
      "learning_rate": 0.07375536027846147,
      "loss": 0.3068,
      "num_input_tokens_seen": 50776096,
      "step": 26790
    },
    {
      "epoch": 4.371155885471898,
      "grad_norm": 0.0009532374679110944,
      "learning_rate": 0.07370463838190057,
      "loss": 0.2963,
      "num_input_tokens_seen": 50784816,
      "step": 26795
    },
    {
      "epoch": 4.3719716126927155,
      "grad_norm": 0.002606383990496397,
      "learning_rate": 0.07365392825104317,
      "loss": 0.3959,
      "num_input_tokens_seen": 50793248,
      "step": 26800
    },
    {
      "epoch": 4.3719716126927155,
      "eval_loss": 0.32157963514328003,
      "eval_runtime": 152.8971,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 50793248,
      "step": 26800
    },
    {
      "epoch": 4.372787339913533,
      "grad_norm": 0.0018025123281404376,
      "learning_rate": 0.07360322989370945,
      "loss": 0.3197,
      "num_input_tokens_seen": 50802480,
      "step": 26805
    },
    {
      "epoch": 4.373603067134351,
      "grad_norm": 0.0008243407937698066,
      "learning_rate": 0.07355254331771781,
      "loss": 0.3282,
      "num_input_tokens_seen": 50812704,
      "step": 26810
    },
    {
      "epoch": 4.374418794355168,
      "grad_norm": 0.0015403610887005925,
      "learning_rate": 0.07350186853088461,
      "loss": 0.3167,
      "num_input_tokens_seen": 50821488,
      "step": 26815
    },
    {
      "epoch": 4.375234521575985,
      "grad_norm": 0.0014441510429605842,
      "learning_rate": 0.07345120554102462,
      "loss": 0.3276,
      "num_input_tokens_seen": 50831712,
      "step": 26820
    },
    {
      "epoch": 4.376050248796802,
      "grad_norm": 0.0012268169084563851,
      "learning_rate": 0.07340055435595079,
      "loss": 0.3426,
      "num_input_tokens_seen": 50840480,
      "step": 26825
    },
    {
      "epoch": 4.37686597601762,
      "grad_norm": 0.0018569394014775753,
      "learning_rate": 0.07334991498347401,
      "loss": 0.3298,
      "num_input_tokens_seen": 50849872,
      "step": 26830
    },
    {
      "epoch": 4.377681703238437,
      "grad_norm": 0.0011178329586982727,
      "learning_rate": 0.07329928743140365,
      "loss": 0.3151,
      "num_input_tokens_seen": 50860048,
      "step": 26835
    },
    {
      "epoch": 4.378497430459254,
      "grad_norm": 0.002487520920112729,
      "learning_rate": 0.07324867170754705,
      "loss": 0.3596,
      "num_input_tokens_seen": 50869360,
      "step": 26840
    },
    {
      "epoch": 4.3793131576800715,
      "grad_norm": 0.0015372629277408123,
      "learning_rate": 0.07319806781970974,
      "loss": 0.2915,
      "num_input_tokens_seen": 50879536,
      "step": 26845
    },
    {
      "epoch": 4.3801288849008895,
      "grad_norm": 0.0029275533743202686,
      "learning_rate": 0.07314747577569555,
      "loss": 0.4135,
      "num_input_tokens_seen": 50889424,
      "step": 26850
    },
    {
      "epoch": 4.380944612121707,
      "grad_norm": 0.0006352727068588138,
      "learning_rate": 0.07309689558330636,
      "loss": 0.2941,
      "num_input_tokens_seen": 50898432,
      "step": 26855
    },
    {
      "epoch": 4.381760339342524,
      "grad_norm": 0.0010997508652508259,
      "learning_rate": 0.0730463272503423,
      "loss": 0.3052,
      "num_input_tokens_seen": 50907664,
      "step": 26860
    },
    {
      "epoch": 4.382576066563341,
      "grad_norm": 0.0006830379134044051,
      "learning_rate": 0.07299577078460168,
      "loss": 0.332,
      "num_input_tokens_seen": 50917664,
      "step": 26865
    },
    {
      "epoch": 4.383391793784159,
      "grad_norm": 0.0019092425936833024,
      "learning_rate": 0.07294522619388083,
      "loss": 0.3581,
      "num_input_tokens_seen": 50927088,
      "step": 26870
    },
    {
      "epoch": 4.384207521004976,
      "grad_norm": 0.0012670016149058938,
      "learning_rate": 0.07289469348597452,
      "loss": 0.2858,
      "num_input_tokens_seen": 50935792,
      "step": 26875
    },
    {
      "epoch": 4.385023248225793,
      "grad_norm": 0.0013648353051394224,
      "learning_rate": 0.07284417266867535,
      "loss": 0.3411,
      "num_input_tokens_seen": 50944832,
      "step": 26880
    },
    {
      "epoch": 4.38583897544661,
      "grad_norm": 0.0033749183639883995,
      "learning_rate": 0.07279366374977439,
      "loss": 0.3813,
      "num_input_tokens_seen": 50955072,
      "step": 26885
    },
    {
      "epoch": 4.386654702667428,
      "grad_norm": 0.0023657376877963543,
      "learning_rate": 0.07274316673706074,
      "loss": 0.3171,
      "num_input_tokens_seen": 50964608,
      "step": 26890
    },
    {
      "epoch": 4.387470429888245,
      "grad_norm": 0.0019943888764828444,
      "learning_rate": 0.07269268163832161,
      "loss": 0.3159,
      "num_input_tokens_seen": 50974960,
      "step": 26895
    },
    {
      "epoch": 4.388286157109063,
      "grad_norm": 0.002077764365822077,
      "learning_rate": 0.07264220846134248,
      "loss": 0.3436,
      "num_input_tokens_seen": 50984128,
      "step": 26900
    },
    {
      "epoch": 4.38910188432988,
      "grad_norm": 0.0017478994559496641,
      "learning_rate": 0.07259174721390699,
      "loss": 0.3268,
      "num_input_tokens_seen": 50994320,
      "step": 26905
    },
    {
      "epoch": 4.389917611550698,
      "grad_norm": 0.0016386995557695627,
      "learning_rate": 0.07254129790379686,
      "loss": 0.2987,
      "num_input_tokens_seen": 51003552,
      "step": 26910
    },
    {
      "epoch": 4.390733338771515,
      "grad_norm": 0.0010646263835951686,
      "learning_rate": 0.072490860538792,
      "loss": 0.3405,
      "num_input_tokens_seen": 51013424,
      "step": 26915
    },
    {
      "epoch": 4.391549065992332,
      "grad_norm": 0.0026694107800722122,
      "learning_rate": 0.07244043512667042,
      "loss": 0.3768,
      "num_input_tokens_seen": 51022208,
      "step": 26920
    },
    {
      "epoch": 4.392364793213149,
      "grad_norm": 0.0008592754020355642,
      "learning_rate": 0.07239002167520843,
      "loss": 0.3024,
      "num_input_tokens_seen": 51031216,
      "step": 26925
    },
    {
      "epoch": 4.393180520433967,
      "grad_norm": 0.002179959788918495,
      "learning_rate": 0.07233962019218045,
      "loss": 0.3382,
      "num_input_tokens_seen": 51039904,
      "step": 26930
    },
    {
      "epoch": 4.393996247654784,
      "grad_norm": 0.0009548417292535305,
      "learning_rate": 0.07228923068535892,
      "loss": 0.2914,
      "num_input_tokens_seen": 51049440,
      "step": 26935
    },
    {
      "epoch": 4.394811974875601,
      "grad_norm": 0.002528941724449396,
      "learning_rate": 0.0722388531625146,
      "loss": 0.3259,
      "num_input_tokens_seen": 51059152,
      "step": 26940
    },
    {
      "epoch": 4.395627702096419,
      "grad_norm": 0.002382064936682582,
      "learning_rate": 0.07218848763141639,
      "loss": 0.3587,
      "num_input_tokens_seen": 51069184,
      "step": 26945
    },
    {
      "epoch": 4.3964434293172365,
      "grad_norm": 0.0014132001670077443,
      "learning_rate": 0.07213813409983118,
      "loss": 0.3438,
      "num_input_tokens_seen": 51077360,
      "step": 26950
    },
    {
      "epoch": 4.397259156538054,
      "grad_norm": 0.0017948714084923267,
      "learning_rate": 0.0720877925755242,
      "loss": 0.3219,
      "num_input_tokens_seen": 51086480,
      "step": 26955
    },
    {
      "epoch": 4.398074883758871,
      "grad_norm": 0.002731035929173231,
      "learning_rate": 0.07203746306625866,
      "loss": 0.3271,
      "num_input_tokens_seen": 51096656,
      "step": 26960
    },
    {
      "epoch": 4.398890610979688,
      "grad_norm": 0.002546439878642559,
      "learning_rate": 0.07198714557979606,
      "loss": 0.3903,
      "num_input_tokens_seen": 51106592,
      "step": 26965
    },
    {
      "epoch": 4.399706338200506,
      "grad_norm": 0.0034012615215033293,
      "learning_rate": 0.07193684012389602,
      "loss": 0.3668,
      "num_input_tokens_seen": 51114560,
      "step": 26970
    },
    {
      "epoch": 4.400522065421323,
      "grad_norm": 0.0004860719491261989,
      "learning_rate": 0.07188654670631621,
      "loss": 0.292,
      "num_input_tokens_seen": 51123776,
      "step": 26975
    },
    {
      "epoch": 4.40133779264214,
      "grad_norm": 0.0004288450290914625,
      "learning_rate": 0.07183626533481258,
      "loss": 0.2862,
      "num_input_tokens_seen": 51133296,
      "step": 26980
    },
    {
      "epoch": 4.402153519862958,
      "grad_norm": 0.0022586036939173937,
      "learning_rate": 0.07178599601713909,
      "loss": 0.3075,
      "num_input_tokens_seen": 51142848,
      "step": 26985
    },
    {
      "epoch": 4.402969247083775,
      "grad_norm": 0.0012113277334719896,
      "learning_rate": 0.07173573876104786,
      "loss": 0.3252,
      "num_input_tokens_seen": 51152160,
      "step": 26990
    },
    {
      "epoch": 4.4037849743045925,
      "grad_norm": 0.002796364715322852,
      "learning_rate": 0.0716854935742893,
      "loss": 0.3267,
      "num_input_tokens_seen": 51161584,
      "step": 26995
    },
    {
      "epoch": 4.40460070152541,
      "grad_norm": 0.001982889836654067,
      "learning_rate": 0.07163526046461174,
      "loss": 0.3603,
      "num_input_tokens_seen": 51170976,
      "step": 27000
    },
    {
      "epoch": 4.40460070152541,
      "eval_loss": 0.32147035002708435,
      "eval_runtime": 152.9516,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 51170976,
      "step": 27000
    },
    {
      "epoch": 4.405416428746228,
      "grad_norm": 0.0009136209264397621,
      "learning_rate": 0.07158503943976181,
      "loss": 0.3153,
      "num_input_tokens_seen": 51180560,
      "step": 27005
    },
    {
      "epoch": 4.406232155967045,
      "grad_norm": 0.0015096766874194145,
      "learning_rate": 0.07153483050748427,
      "loss": 0.3606,
      "num_input_tokens_seen": 51190784,
      "step": 27010
    },
    {
      "epoch": 4.407047883187862,
      "grad_norm": 0.0006607297691516578,
      "learning_rate": 0.07148463367552188,
      "loss": 0.3168,
      "num_input_tokens_seen": 51201424,
      "step": 27015
    },
    {
      "epoch": 4.407863610408679,
      "grad_norm": 0.0028298175893723965,
      "learning_rate": 0.07143444895161565,
      "loss": 0.2892,
      "num_input_tokens_seen": 51211536,
      "step": 27020
    },
    {
      "epoch": 4.408679337629497,
      "grad_norm": 0.001811813679523766,
      "learning_rate": 0.07138427634350476,
      "loss": 0.3103,
      "num_input_tokens_seen": 51220704,
      "step": 27025
    },
    {
      "epoch": 4.409495064850314,
      "grad_norm": 0.0009929562220349908,
      "learning_rate": 0.07133411585892636,
      "loss": 0.2978,
      "num_input_tokens_seen": 51229376,
      "step": 27030
    },
    {
      "epoch": 4.410310792071131,
      "grad_norm": 0.0013486243551597,
      "learning_rate": 0.07128396750561593,
      "loss": 0.2979,
      "num_input_tokens_seen": 51239424,
      "step": 27035
    },
    {
      "epoch": 4.411126519291948,
      "grad_norm": 0.0018396982923150063,
      "learning_rate": 0.07123383129130685,
      "loss": 0.274,
      "num_input_tokens_seen": 51248384,
      "step": 27040
    },
    {
      "epoch": 4.4119422465127665,
      "grad_norm": 0.0009974896674975753,
      "learning_rate": 0.07118370722373084,
      "loss": 0.2843,
      "num_input_tokens_seen": 51257856,
      "step": 27045
    },
    {
      "epoch": 4.412757973733584,
      "grad_norm": 0.0010316475527361035,
      "learning_rate": 0.07113359531061769,
      "loss": 0.2948,
      "num_input_tokens_seen": 51266256,
      "step": 27050
    },
    {
      "epoch": 4.413573700954401,
      "grad_norm": 0.001795531832613051,
      "learning_rate": 0.07108349555969525,
      "loss": 0.3491,
      "num_input_tokens_seen": 51276128,
      "step": 27055
    },
    {
      "epoch": 4.414389428175218,
      "grad_norm": 0.0006599706830456853,
      "learning_rate": 0.07103340797868944,
      "loss": 0.3219,
      "num_input_tokens_seen": 51285504,
      "step": 27060
    },
    {
      "epoch": 4.415205155396036,
      "grad_norm": 0.0021844315342605114,
      "learning_rate": 0.07098333257532453,
      "loss": 0.321,
      "num_input_tokens_seen": 51294320,
      "step": 27065
    },
    {
      "epoch": 4.416020882616853,
      "grad_norm": 0.003156965132802725,
      "learning_rate": 0.07093326935732269,
      "loss": 0.3138,
      "num_input_tokens_seen": 51303952,
      "step": 27070
    },
    {
      "epoch": 4.41683660983767,
      "grad_norm": 0.003882769728079438,
      "learning_rate": 0.0708832183324044,
      "loss": 0.353,
      "num_input_tokens_seen": 51314336,
      "step": 27075
    },
    {
      "epoch": 4.417652337058487,
      "grad_norm": 0.0019743586890399456,
      "learning_rate": 0.07083317950828799,
      "loss": 0.3061,
      "num_input_tokens_seen": 51324656,
      "step": 27080
    },
    {
      "epoch": 4.418468064279305,
      "grad_norm": 0.0022612805478274822,
      "learning_rate": 0.0707831528926902,
      "loss": 0.3084,
      "num_input_tokens_seen": 51333664,
      "step": 27085
    },
    {
      "epoch": 4.419283791500122,
      "grad_norm": 0.004896833095699549,
      "learning_rate": 0.07073313849332578,
      "loss": 0.3324,
      "num_input_tokens_seen": 51343824,
      "step": 27090
    },
    {
      "epoch": 4.4200995187209395,
      "grad_norm": 0.0024212163407355547,
      "learning_rate": 0.07068313631790749,
      "loss": 0.3437,
      "num_input_tokens_seen": 51352864,
      "step": 27095
    },
    {
      "epoch": 4.420915245941757,
      "grad_norm": 0.0010750459041446447,
      "learning_rate": 0.07063314637414632,
      "loss": 0.3305,
      "num_input_tokens_seen": 51362864,
      "step": 27100
    },
    {
      "epoch": 4.421730973162575,
      "grad_norm": 0.002451438456773758,
      "learning_rate": 0.07058316866975144,
      "loss": 0.3299,
      "num_input_tokens_seen": 51372928,
      "step": 27105
    },
    {
      "epoch": 4.422546700383392,
      "grad_norm": 0.0017062281258404255,
      "learning_rate": 0.0705332032124299,
      "loss": 0.2856,
      "num_input_tokens_seen": 51383680,
      "step": 27110
    },
    {
      "epoch": 4.423362427604209,
      "grad_norm": 0.0021566981449723244,
      "learning_rate": 0.0704832500098871,
      "loss": 0.3203,
      "num_input_tokens_seen": 51393088,
      "step": 27115
    },
    {
      "epoch": 4.424178154825027,
      "grad_norm": 0.0014018910005688667,
      "learning_rate": 0.07043330906982641,
      "loss": 0.2796,
      "num_input_tokens_seen": 51402672,
      "step": 27120
    },
    {
      "epoch": 4.424993882045844,
      "grad_norm": 0.0013652511406689882,
      "learning_rate": 0.07038338039994936,
      "loss": 0.2693,
      "num_input_tokens_seen": 51412448,
      "step": 27125
    },
    {
      "epoch": 4.425809609266661,
      "grad_norm": 0.002275054343044758,
      "learning_rate": 0.07033346400795562,
      "loss": 0.3662,
      "num_input_tokens_seen": 51422160,
      "step": 27130
    },
    {
      "epoch": 4.426625336487478,
      "grad_norm": 0.0011340683558955789,
      "learning_rate": 0.07028355990154282,
      "loss": 0.3246,
      "num_input_tokens_seen": 51431568,
      "step": 27135
    },
    {
      "epoch": 4.4274410637082955,
      "grad_norm": 0.0012247789418324828,
      "learning_rate": 0.07023366808840685,
      "loss": 0.3539,
      "num_input_tokens_seen": 51441504,
      "step": 27140
    },
    {
      "epoch": 4.4282567909291135,
      "grad_norm": 0.004707252606749535,
      "learning_rate": 0.07018378857624172,
      "loss": 0.3614,
      "num_input_tokens_seen": 51451376,
      "step": 27145
    },
    {
      "epoch": 4.429072518149931,
      "grad_norm": 0.0026429209392517805,
      "learning_rate": 0.0701339213727394,
      "loss": 0.3029,
      "num_input_tokens_seen": 51459728,
      "step": 27150
    },
    {
      "epoch": 4.429888245370748,
      "grad_norm": 0.0013729454949498177,
      "learning_rate": 0.07008406648559008,
      "loss": 0.3167,
      "num_input_tokens_seen": 51468816,
      "step": 27155
    },
    {
      "epoch": 4.430703972591566,
      "grad_norm": 0.0012456132099032402,
      "learning_rate": 0.07003422392248196,
      "loss": 0.2845,
      "num_input_tokens_seen": 51478512,
      "step": 27160
    },
    {
      "epoch": 4.431519699812383,
      "grad_norm": 0.0014775473391637206,
      "learning_rate": 0.06998439369110142,
      "loss": 0.2665,
      "num_input_tokens_seen": 51487456,
      "step": 27165
    },
    {
      "epoch": 4.4323354270332,
      "grad_norm": 0.002912358148023486,
      "learning_rate": 0.06993457579913295,
      "loss": 0.2575,
      "num_input_tokens_seen": 51497136,
      "step": 27170
    },
    {
      "epoch": 4.433151154254017,
      "grad_norm": 0.0012534564593806863,
      "learning_rate": 0.06988477025425903,
      "loss": 0.4361,
      "num_input_tokens_seen": 51507456,
      "step": 27175
    },
    {
      "epoch": 4.433966881474835,
      "grad_norm": 0.002817329252138734,
      "learning_rate": 0.06983497706416032,
      "loss": 0.3075,
      "num_input_tokens_seen": 51518480,
      "step": 27180
    },
    {
      "epoch": 4.434782608695652,
      "grad_norm": 0.002536369953304529,
      "learning_rate": 0.0697851962365156,
      "loss": 0.3079,
      "num_input_tokens_seen": 51529136,
      "step": 27185
    },
    {
      "epoch": 4.4355983359164695,
      "grad_norm": 0.0026176851242780685,
      "learning_rate": 0.06973542777900163,
      "loss": 0.3221,
      "num_input_tokens_seen": 51539376,
      "step": 27190
    },
    {
      "epoch": 4.436414063137287,
      "grad_norm": 0.001760531566105783,
      "learning_rate": 0.06968567169929342,
      "loss": 0.2984,
      "num_input_tokens_seen": 51549776,
      "step": 27195
    },
    {
      "epoch": 4.437229790358105,
      "grad_norm": 0.0025513360742479563,
      "learning_rate": 0.06963592800506392,
      "loss": 0.321,
      "num_input_tokens_seen": 51559504,
      "step": 27200
    },
    {
      "epoch": 4.437229790358105,
      "eval_loss": 0.3249574899673462,
      "eval_runtime": 152.8571,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 51559504,
      "step": 27200
    },
    {
      "epoch": 4.438045517578922,
      "grad_norm": 0.00313751888461411,
      "learning_rate": 0.06958619670398417,
      "loss": 0.3691,
      "num_input_tokens_seen": 51568240,
      "step": 27205
    },
    {
      "epoch": 4.438861244799739,
      "grad_norm": 0.0006007772753946483,
      "learning_rate": 0.0695364778037235,
      "loss": 0.3496,
      "num_input_tokens_seen": 51576560,
      "step": 27210
    },
    {
      "epoch": 4.439676972020556,
      "grad_norm": 0.001121395151130855,
      "learning_rate": 0.06948677131194907,
      "loss": 0.3156,
      "num_input_tokens_seen": 51585904,
      "step": 27215
    },
    {
      "epoch": 4.440492699241374,
      "grad_norm": 0.0008176675764843822,
      "learning_rate": 0.06943707723632629,
      "loss": 0.3119,
      "num_input_tokens_seen": 51594832,
      "step": 27220
    },
    {
      "epoch": 4.441308426462191,
      "grad_norm": 0.001540923723950982,
      "learning_rate": 0.06938739558451867,
      "loss": 0.3492,
      "num_input_tokens_seen": 51605152,
      "step": 27225
    },
    {
      "epoch": 4.442124153683008,
      "grad_norm": 0.0009831464849412441,
      "learning_rate": 0.06933772636418763,
      "loss": 0.3321,
      "num_input_tokens_seen": 51615120,
      "step": 27230
    },
    {
      "epoch": 4.442939880903825,
      "grad_norm": 0.002254297723993659,
      "learning_rate": 0.06928806958299293,
      "loss": 0.2751,
      "num_input_tokens_seen": 51624816,
      "step": 27235
    },
    {
      "epoch": 4.443755608124643,
      "grad_norm": 0.0017434742767363787,
      "learning_rate": 0.06923842524859211,
      "loss": 0.3598,
      "num_input_tokens_seen": 51633440,
      "step": 27240
    },
    {
      "epoch": 4.444571335345461,
      "grad_norm": 0.0019853035919368267,
      "learning_rate": 0.06918879336864105,
      "loss": 0.2993,
      "num_input_tokens_seen": 51642336,
      "step": 27245
    },
    {
      "epoch": 4.445387062566278,
      "grad_norm": 0.002268366515636444,
      "learning_rate": 0.06913917395079362,
      "loss": 0.3206,
      "num_input_tokens_seen": 51651488,
      "step": 27250
    },
    {
      "epoch": 4.446202789787095,
      "grad_norm": 0.0020257714204490185,
      "learning_rate": 0.0690895670027017,
      "loss": 0.3104,
      "num_input_tokens_seen": 51660512,
      "step": 27255
    },
    {
      "epoch": 4.447018517007913,
      "grad_norm": 0.0021064700558781624,
      "learning_rate": 0.06903997253201531,
      "loss": 0.3504,
      "num_input_tokens_seen": 51669952,
      "step": 27260
    },
    {
      "epoch": 4.44783424422873,
      "grad_norm": 0.0014586959732696414,
      "learning_rate": 0.06899039054638263,
      "loss": 0.3267,
      "num_input_tokens_seen": 51679136,
      "step": 27265
    },
    {
      "epoch": 4.448649971449547,
      "grad_norm": 0.0012536488939076662,
      "learning_rate": 0.06894082105344976,
      "loss": 0.3263,
      "num_input_tokens_seen": 51689200,
      "step": 27270
    },
    {
      "epoch": 4.449465698670364,
      "grad_norm": 0.0011570058995857835,
      "learning_rate": 0.06889126406086087,
      "loss": 0.3435,
      "num_input_tokens_seen": 51697648,
      "step": 27275
    },
    {
      "epoch": 4.450281425891182,
      "grad_norm": 0.000611045048572123,
      "learning_rate": 0.0688417195762584,
      "loss": 0.3155,
      "num_input_tokens_seen": 51707328,
      "step": 27280
    },
    {
      "epoch": 4.451097153111999,
      "grad_norm": 0.002996681025251746,
      "learning_rate": 0.06879218760728262,
      "loss": 0.3377,
      "num_input_tokens_seen": 51717312,
      "step": 27285
    },
    {
      "epoch": 4.4519128803328165,
      "grad_norm": 0.0005114722298458219,
      "learning_rate": 0.06874266816157207,
      "loss": 0.3095,
      "num_input_tokens_seen": 51726224,
      "step": 27290
    },
    {
      "epoch": 4.4527286075536345,
      "grad_norm": 0.0005909035680815578,
      "learning_rate": 0.06869316124676321,
      "loss": 0.3202,
      "num_input_tokens_seen": 51735872,
      "step": 27295
    },
    {
      "epoch": 4.453544334774452,
      "grad_norm": 0.0005879760137759149,
      "learning_rate": 0.06864366687049062,
      "loss": 0.2895,
      "num_input_tokens_seen": 51745552,
      "step": 27300
    },
    {
      "epoch": 4.454360061995269,
      "grad_norm": 0.001415942911989987,
      "learning_rate": 0.06859418504038704,
      "loss": 0.2955,
      "num_input_tokens_seen": 51754192,
      "step": 27305
    },
    {
      "epoch": 4.455175789216086,
      "grad_norm": 0.0007879819022491574,
      "learning_rate": 0.06854471576408311,
      "loss": 0.3249,
      "num_input_tokens_seen": 51763168,
      "step": 27310
    },
    {
      "epoch": 4.455991516436903,
      "grad_norm": 0.0019676212687045336,
      "learning_rate": 0.06849525904920767,
      "loss": 0.2813,
      "num_input_tokens_seen": 51773136,
      "step": 27315
    },
    {
      "epoch": 4.456807243657721,
      "grad_norm": 0.0014330765698105097,
      "learning_rate": 0.06844581490338748,
      "loss": 0.2859,
      "num_input_tokens_seen": 51781728,
      "step": 27320
    },
    {
      "epoch": 4.457622970878538,
      "grad_norm": 0.002589210867881775,
      "learning_rate": 0.06839638333424752,
      "loss": 0.3116,
      "num_input_tokens_seen": 51791184,
      "step": 27325
    },
    {
      "epoch": 4.458438698099355,
      "grad_norm": 0.0011223703622817993,
      "learning_rate": 0.06834696434941082,
      "loss": 0.31,
      "num_input_tokens_seen": 51801808,
      "step": 27330
    },
    {
      "epoch": 4.459254425320173,
      "grad_norm": 0.0009695794433355331,
      "learning_rate": 0.06829755795649824,
      "loss": 0.27,
      "num_input_tokens_seen": 51810336,
      "step": 27335
    },
    {
      "epoch": 4.4600701525409905,
      "grad_norm": 0.0019125242251902819,
      "learning_rate": 0.06824816416312904,
      "loss": 0.2353,
      "num_input_tokens_seen": 51819760,
      "step": 27340
    },
    {
      "epoch": 4.460885879761808,
      "grad_norm": 0.002795002656057477,
      "learning_rate": 0.06819878297692027,
      "loss": 0.3828,
      "num_input_tokens_seen": 51827936,
      "step": 27345
    },
    {
      "epoch": 4.461701606982625,
      "grad_norm": 0.0019032685086131096,
      "learning_rate": 0.0681494144054871,
      "loss": 0.435,
      "num_input_tokens_seen": 51836512,
      "step": 27350
    },
    {
      "epoch": 4.462517334203443,
      "grad_norm": 0.002241827081888914,
      "learning_rate": 0.06810005845644286,
      "loss": 0.3949,
      "num_input_tokens_seen": 51845312,
      "step": 27355
    },
    {
      "epoch": 4.46333306142426,
      "grad_norm": 0.0031299288384616375,
      "learning_rate": 0.06805071513739878,
      "loss": 0.3366,
      "num_input_tokens_seen": 51855168,
      "step": 27360
    },
    {
      "epoch": 4.464148788645077,
      "grad_norm": 0.0014948268653824925,
      "learning_rate": 0.06800138445596428,
      "loss": 0.3678,
      "num_input_tokens_seen": 51863712,
      "step": 27365
    },
    {
      "epoch": 4.464964515865894,
      "grad_norm": 0.0017663298640400171,
      "learning_rate": 0.06795206641974678,
      "loss": 0.2968,
      "num_input_tokens_seen": 51873088,
      "step": 27370
    },
    {
      "epoch": 4.465780243086712,
      "grad_norm": 0.0012422427535057068,
      "learning_rate": 0.06790276103635169,
      "loss": 0.2836,
      "num_input_tokens_seen": 51882128,
      "step": 27375
    },
    {
      "epoch": 4.466595970307529,
      "grad_norm": 0.0024892918299883604,
      "learning_rate": 0.0678534683133826,
      "loss": 0.3123,
      "num_input_tokens_seen": 51890048,
      "step": 27380
    },
    {
      "epoch": 4.467411697528346,
      "grad_norm": 0.0015406248858198524,
      "learning_rate": 0.06780418825844095,
      "loss": 0.3257,
      "num_input_tokens_seen": 51898592,
      "step": 27385
    },
    {
      "epoch": 4.468227424749164,
      "grad_norm": 0.0011228022631257772,
      "learning_rate": 0.0677549208791264,
      "loss": 0.3642,
      "num_input_tokens_seen": 51908624,
      "step": 27390
    },
    {
      "epoch": 4.469043151969982,
      "grad_norm": 0.0022808134090155363,
      "learning_rate": 0.06770566618303668,
      "loss": 0.3151,
      "num_input_tokens_seen": 51918752,
      "step": 27395
    },
    {
      "epoch": 4.469858879190799,
      "grad_norm": 0.00044703204184770584,
      "learning_rate": 0.06765642417776736,
      "loss": 0.3445,
      "num_input_tokens_seen": 51928704,
      "step": 27400
    },
    {
      "epoch": 4.469858879190799,
      "eval_loss": 0.321026086807251,
      "eval_runtime": 152.9168,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 51928704,
      "step": 27400
    },
    {
      "epoch": 4.470674606411616,
      "grad_norm": 0.0013484269147738814,
      "learning_rate": 0.0676071948709122,
      "loss": 0.3921,
      "num_input_tokens_seen": 51938480,
      "step": 27405
    },
    {
      "epoch": 4.471490333632433,
      "grad_norm": 0.0009617304312996566,
      "learning_rate": 0.06755797827006307,
      "loss": 0.3028,
      "num_input_tokens_seen": 51947376,
      "step": 27410
    },
    {
      "epoch": 4.472306060853251,
      "grad_norm": 0.001279977266676724,
      "learning_rate": 0.06750877438280974,
      "loss": 0.3084,
      "num_input_tokens_seen": 51957136,
      "step": 27415
    },
    {
      "epoch": 4.473121788074068,
      "grad_norm": 0.003265456296503544,
      "learning_rate": 0.06745958321673998,
      "loss": 0.3153,
      "num_input_tokens_seen": 51966752,
      "step": 27420
    },
    {
      "epoch": 4.473937515294885,
      "grad_norm": 0.0009574844734743237,
      "learning_rate": 0.0674104047794398,
      "loss": 0.3124,
      "num_input_tokens_seen": 51975520,
      "step": 27425
    },
    {
      "epoch": 4.474753242515702,
      "grad_norm": 0.00156564696226269,
      "learning_rate": 0.06736123907849303,
      "loss": 0.3659,
      "num_input_tokens_seen": 51983584,
      "step": 27430
    },
    {
      "epoch": 4.47556896973652,
      "grad_norm": 0.00043730781180784106,
      "learning_rate": 0.06731208612148178,
      "loss": 0.3257,
      "num_input_tokens_seen": 51993088,
      "step": 27435
    },
    {
      "epoch": 4.4763846969573375,
      "grad_norm": 0.003032906912267208,
      "learning_rate": 0.0672629459159859,
      "loss": 0.3622,
      "num_input_tokens_seen": 52002912,
      "step": 27440
    },
    {
      "epoch": 4.477200424178155,
      "grad_norm": 0.0007614858914166689,
      "learning_rate": 0.0672138184695835,
      "loss": 0.3571,
      "num_input_tokens_seen": 52011392,
      "step": 27445
    },
    {
      "epoch": 4.478016151398972,
      "grad_norm": 0.0011043398408219218,
      "learning_rate": 0.0671647037898507,
      "loss": 0.3498,
      "num_input_tokens_seen": 52020816,
      "step": 27450
    },
    {
      "epoch": 4.47883187861979,
      "grad_norm": 0.0012592995772138238,
      "learning_rate": 0.0671156018843615,
      "loss": 0.3456,
      "num_input_tokens_seen": 52030576,
      "step": 27455
    },
    {
      "epoch": 4.479647605840607,
      "grad_norm": 0.0009010698995552957,
      "learning_rate": 0.06706651276068812,
      "loss": 0.2832,
      "num_input_tokens_seen": 52040432,
      "step": 27460
    },
    {
      "epoch": 4.480463333061424,
      "grad_norm": 0.0009200729546137154,
      "learning_rate": 0.06701743642640064,
      "loss": 0.2542,
      "num_input_tokens_seen": 52050400,
      "step": 27465
    },
    {
      "epoch": 4.481279060282241,
      "grad_norm": 0.0014821913791820407,
      "learning_rate": 0.06696837288906729,
      "loss": 0.3034,
      "num_input_tokens_seen": 52059792,
      "step": 27470
    },
    {
      "epoch": 4.482094787503059,
      "grad_norm": 0.002011601347476244,
      "learning_rate": 0.06691932215625432,
      "loss": 0.2712,
      "num_input_tokens_seen": 52068864,
      "step": 27475
    },
    {
      "epoch": 4.482910514723876,
      "grad_norm": 0.0013288998743519187,
      "learning_rate": 0.06687028423552589,
      "loss": 0.3204,
      "num_input_tokens_seen": 52077760,
      "step": 27480
    },
    {
      "epoch": 4.4837262419446935,
      "grad_norm": 0.0013994378969073296,
      "learning_rate": 0.06682125913444435,
      "loss": 0.3599,
      "num_input_tokens_seen": 52085968,
      "step": 27485
    },
    {
      "epoch": 4.484541969165511,
      "grad_norm": 0.0012002979638054967,
      "learning_rate": 0.0667722468605699,
      "loss": 0.3567,
      "num_input_tokens_seen": 52094592,
      "step": 27490
    },
    {
      "epoch": 4.485357696386329,
      "grad_norm": 0.0007320260046981275,
      "learning_rate": 0.06672324742146094,
      "loss": 0.3285,
      "num_input_tokens_seen": 52103216,
      "step": 27495
    },
    {
      "epoch": 4.486173423607146,
      "grad_norm": 0.0012728178407996893,
      "learning_rate": 0.06667426082467373,
      "loss": 0.3141,
      "num_input_tokens_seen": 52112496,
      "step": 27500
    },
    {
      "epoch": 4.486989150827963,
      "grad_norm": 0.0012622219510376453,
      "learning_rate": 0.0666252870777626,
      "loss": 0.2968,
      "num_input_tokens_seen": 52121312,
      "step": 27505
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 0.001110896933823824,
      "learning_rate": 0.06657632618827995,
      "loss": 0.328,
      "num_input_tokens_seen": 52129920,
      "step": 27510
    },
    {
      "epoch": 4.488620605269598,
      "grad_norm": 0.0014716391451656818,
      "learning_rate": 0.06652737816377623,
      "loss": 0.3452,
      "num_input_tokens_seen": 52137904,
      "step": 27515
    },
    {
      "epoch": 4.489436332490415,
      "grad_norm": 0.0009814389050006866,
      "learning_rate": 0.06647844301179971,
      "loss": 0.2686,
      "num_input_tokens_seen": 52147856,
      "step": 27520
    },
    {
      "epoch": 4.490252059711232,
      "grad_norm": 0.001022798242047429,
      "learning_rate": 0.06642952073989689,
      "loss": 0.2793,
      "num_input_tokens_seen": 52158416,
      "step": 27525
    },
    {
      "epoch": 4.49106778693205,
      "grad_norm": 0.0009227570844814181,
      "learning_rate": 0.06638061135561223,
      "loss": 0.2896,
      "num_input_tokens_seen": 52166832,
      "step": 27530
    },
    {
      "epoch": 4.4918835141528675,
      "grad_norm": 0.00123644492123276,
      "learning_rate": 0.06633171486648808,
      "loss": 0.3222,
      "num_input_tokens_seen": 52176832,
      "step": 27535
    },
    {
      "epoch": 4.492699241373685,
      "grad_norm": 0.000819325097836554,
      "learning_rate": 0.06628283128006499,
      "loss": 0.3514,
      "num_input_tokens_seen": 52185728,
      "step": 27540
    },
    {
      "epoch": 4.493514968594502,
      "grad_norm": 0.0014561270363628864,
      "learning_rate": 0.0662339606038813,
      "loss": 0.334,
      "num_input_tokens_seen": 52194736,
      "step": 27545
    },
    {
      "epoch": 4.49433069581532,
      "grad_norm": 0.0008674662094563246,
      "learning_rate": 0.06618510284547358,
      "loss": 0.3528,
      "num_input_tokens_seen": 52203968,
      "step": 27550
    },
    {
      "epoch": 4.495146423036137,
      "grad_norm": 0.0015435609966516495,
      "learning_rate": 0.06613625801237633,
      "loss": 0.3549,
      "num_input_tokens_seen": 52212912,
      "step": 27555
    },
    {
      "epoch": 4.495962150256954,
      "grad_norm": 0.0013076908653602004,
      "learning_rate": 0.066087426112122,
      "loss": 0.3119,
      "num_input_tokens_seen": 52222128,
      "step": 27560
    },
    {
      "epoch": 4.496777877477771,
      "grad_norm": 0.001243665348738432,
      "learning_rate": 0.06603860715224101,
      "loss": 0.3128,
      "num_input_tokens_seen": 52231376,
      "step": 27565
    },
    {
      "epoch": 4.497593604698589,
      "grad_norm": 0.002242651768028736,
      "learning_rate": 0.06598980114026198,
      "loss": 0.327,
      "num_input_tokens_seen": 52241680,
      "step": 27570
    },
    {
      "epoch": 4.498409331919406,
      "grad_norm": 0.0014567564940080047,
      "learning_rate": 0.06594100808371128,
      "loss": 0.2999,
      "num_input_tokens_seen": 52251920,
      "step": 27575
    },
    {
      "epoch": 4.499225059140223,
      "grad_norm": 0.0009562760242260993,
      "learning_rate": 0.06589222799011357,
      "loss": 0.3358,
      "num_input_tokens_seen": 52261184,
      "step": 27580
    },
    {
      "epoch": 4.5000407863610405,
      "grad_norm": 0.004216828849166632,
      "learning_rate": 0.0658434608669912,
      "loss": 0.4002,
      "num_input_tokens_seen": 52271072,
      "step": 27585
    },
    {
      "epoch": 4.500856513581859,
      "grad_norm": 0.0017902182880789042,
      "learning_rate": 0.06579470672186473,
      "loss": 0.3227,
      "num_input_tokens_seen": 52280368,
      "step": 27590
    },
    {
      "epoch": 4.501672240802676,
      "grad_norm": 0.0014595513930544257,
      "learning_rate": 0.06574596556225275,
      "loss": 0.3111,
      "num_input_tokens_seen": 52289424,
      "step": 27595
    },
    {
      "epoch": 4.502487968023493,
      "grad_norm": 0.001866992679424584,
      "learning_rate": 0.06569723739567161,
      "loss": 0.3458,
      "num_input_tokens_seen": 52297776,
      "step": 27600
    },
    {
      "epoch": 4.502487968023493,
      "eval_loss": 0.3215235769748688,
      "eval_runtime": 152.8294,
      "eval_samples_per_second": 17.83,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 52297776,
      "step": 27600
    },
    {
      "epoch": 4.50330369524431,
      "grad_norm": 0.0025033506099134684,
      "learning_rate": 0.06564852222963588,
      "loss": 0.3083,
      "num_input_tokens_seen": 52307984,
      "step": 27605
    },
    {
      "epoch": 4.504119422465128,
      "grad_norm": 0.0019480439368635416,
      "learning_rate": 0.06559982007165813,
      "loss": 0.2768,
      "num_input_tokens_seen": 52317856,
      "step": 27610
    },
    {
      "epoch": 4.504935149685945,
      "grad_norm": 0.0008303248323500156,
      "learning_rate": 0.06555113092924868,
      "loss": 0.2813,
      "num_input_tokens_seen": 52327536,
      "step": 27615
    },
    {
      "epoch": 4.505750876906762,
      "grad_norm": 0.0014336431631818414,
      "learning_rate": 0.06550245480991615,
      "loss": 0.2851,
      "num_input_tokens_seen": 52336928,
      "step": 27620
    },
    {
      "epoch": 4.506566604127579,
      "grad_norm": 0.0015627926914021373,
      "learning_rate": 0.0654537917211669,
      "loss": 0.3519,
      "num_input_tokens_seen": 52345712,
      "step": 27625
    },
    {
      "epoch": 4.507382331348397,
      "grad_norm": 0.002701214049011469,
      "learning_rate": 0.0654051416705055,
      "loss": 0.2778,
      "num_input_tokens_seen": 52354480,
      "step": 27630
    },
    {
      "epoch": 4.5081980585692145,
      "grad_norm": 0.0037349537014961243,
      "learning_rate": 0.06535650466543427,
      "loss": 0.317,
      "num_input_tokens_seen": 52363088,
      "step": 27635
    },
    {
      "epoch": 4.509013785790032,
      "grad_norm": 0.002240521600469947,
      "learning_rate": 0.0653078807134538,
      "loss": 0.3431,
      "num_input_tokens_seen": 52371248,
      "step": 27640
    },
    {
      "epoch": 4.50982951301085,
      "grad_norm": 0.00030382341356016695,
      "learning_rate": 0.06525926982206236,
      "loss": 0.3028,
      "num_input_tokens_seen": 52379760,
      "step": 27645
    },
    {
      "epoch": 4.510645240231667,
      "grad_norm": 0.002225426957011223,
      "learning_rate": 0.06521067199875648,
      "loss": 0.2648,
      "num_input_tokens_seen": 52389136,
      "step": 27650
    },
    {
      "epoch": 4.511460967452484,
      "grad_norm": 0.0007474857266061008,
      "learning_rate": 0.06516208725103047,
      "loss": 0.3479,
      "num_input_tokens_seen": 52397680,
      "step": 27655
    },
    {
      "epoch": 4.512276694673301,
      "grad_norm": 0.0019260942935943604,
      "learning_rate": 0.06511351558637678,
      "loss": 0.346,
      "num_input_tokens_seen": 52406240,
      "step": 27660
    },
    {
      "epoch": 4.513092421894118,
      "grad_norm": 0.0032502319663763046,
      "learning_rate": 0.06506495701228569,
      "loss": 0.2612,
      "num_input_tokens_seen": 52416208,
      "step": 27665
    },
    {
      "epoch": 4.513908149114936,
      "grad_norm": 0.0020947910379618406,
      "learning_rate": 0.06501641153624559,
      "loss": 0.3106,
      "num_input_tokens_seen": 52425024,
      "step": 27670
    },
    {
      "epoch": 4.514723876335753,
      "grad_norm": 0.001494416268542409,
      "learning_rate": 0.06496787916574286,
      "loss": 0.356,
      "num_input_tokens_seen": 52436112,
      "step": 27675
    },
    {
      "epoch": 4.5155396035565705,
      "grad_norm": 0.0031317651737481356,
      "learning_rate": 0.06491935990826168,
      "loss": 0.3178,
      "num_input_tokens_seen": 52445184,
      "step": 27680
    },
    {
      "epoch": 4.5163553307773885,
      "grad_norm": 0.002632174175232649,
      "learning_rate": 0.0648708537712844,
      "loss": 0.267,
      "num_input_tokens_seen": 52455264,
      "step": 27685
    },
    {
      "epoch": 4.517171057998206,
      "grad_norm": 0.0003853616362903267,
      "learning_rate": 0.06482236076229132,
      "loss": 0.359,
      "num_input_tokens_seen": 52463344,
      "step": 27690
    },
    {
      "epoch": 4.517986785219023,
      "grad_norm": 0.0021617270540446043,
      "learning_rate": 0.06477388088876056,
      "loss": 0.403,
      "num_input_tokens_seen": 52471984,
      "step": 27695
    },
    {
      "epoch": 4.51880251243984,
      "grad_norm": 0.0031961218919605017,
      "learning_rate": 0.06472541415816846,
      "loss": 0.2745,
      "num_input_tokens_seen": 52481216,
      "step": 27700
    },
    {
      "epoch": 4.519618239660657,
      "grad_norm": 0.0016528066480532289,
      "learning_rate": 0.06467696057798909,
      "loss": 0.37,
      "num_input_tokens_seen": 52490384,
      "step": 27705
    },
    {
      "epoch": 4.520433966881475,
      "grad_norm": 0.00034704720019362867,
      "learning_rate": 0.0646285201556946,
      "loss": 0.353,
      "num_input_tokens_seen": 52498720,
      "step": 27710
    },
    {
      "epoch": 4.521249694102292,
      "grad_norm": 0.0025657122023403645,
      "learning_rate": 0.06458009289875521,
      "loss": 0.3135,
      "num_input_tokens_seen": 52507456,
      "step": 27715
    },
    {
      "epoch": 4.522065421323109,
      "grad_norm": 0.0010948877315968275,
      "learning_rate": 0.0645316788146389,
      "loss": 0.3028,
      "num_input_tokens_seen": 52517936,
      "step": 27720
    },
    {
      "epoch": 4.522881148543927,
      "grad_norm": 0.0017591993091627955,
      "learning_rate": 0.06448327791081175,
      "loss": 0.2872,
      "num_input_tokens_seen": 52526928,
      "step": 27725
    },
    {
      "epoch": 4.523696875764744,
      "grad_norm": 0.0028045258950442076,
      "learning_rate": 0.0644348901947379,
      "loss": 0.332,
      "num_input_tokens_seen": 52536224,
      "step": 27730
    },
    {
      "epoch": 4.524512602985562,
      "grad_norm": 0.0016917428001761436,
      "learning_rate": 0.06438651567387917,
      "loss": 0.3103,
      "num_input_tokens_seen": 52545600,
      "step": 27735
    },
    {
      "epoch": 4.525328330206379,
      "grad_norm": 0.0013992497697472572,
      "learning_rate": 0.0643381543556957,
      "loss": 0.3724,
      "num_input_tokens_seen": 52555488,
      "step": 27740
    },
    {
      "epoch": 4.526144057427197,
      "grad_norm": 0.002611274365335703,
      "learning_rate": 0.06428980624764526,
      "loss": 0.344,
      "num_input_tokens_seen": 52566096,
      "step": 27745
    },
    {
      "epoch": 4.526959784648014,
      "grad_norm": 0.0019203065894544125,
      "learning_rate": 0.06424147135718378,
      "loss": 0.3661,
      "num_input_tokens_seen": 52575072,
      "step": 27750
    },
    {
      "epoch": 4.527775511868831,
      "grad_norm": 0.0009679358336143196,
      "learning_rate": 0.06419314969176519,
      "loss": 0.3397,
      "num_input_tokens_seen": 52584016,
      "step": 27755
    },
    {
      "epoch": 4.528591239089648,
      "grad_norm": 0.001171021955087781,
      "learning_rate": 0.06414484125884118,
      "loss": 0.3203,
      "num_input_tokens_seen": 52593776,
      "step": 27760
    },
    {
      "epoch": 4.529406966310466,
      "grad_norm": 0.0009954251581802964,
      "learning_rate": 0.06409654606586157,
      "loss": 0.291,
      "num_input_tokens_seen": 52602272,
      "step": 27765
    },
    {
      "epoch": 4.530222693531283,
      "grad_norm": 0.0009651206200942397,
      "learning_rate": 0.06404826412027415,
      "loss": 0.2879,
      "num_input_tokens_seen": 52611648,
      "step": 27770
    },
    {
      "epoch": 4.5310384207521,
      "grad_norm": 0.0018367364536970854,
      "learning_rate": 0.06399999542952453,
      "loss": 0.3376,
      "num_input_tokens_seen": 52622144,
      "step": 27775
    },
    {
      "epoch": 4.5318541479729175,
      "grad_norm": 0.0015563544584438205,
      "learning_rate": 0.0639517400010563,
      "loss": 0.3262,
      "num_input_tokens_seen": 52632080,
      "step": 27780
    },
    {
      "epoch": 4.5326698751937355,
      "grad_norm": 0.0015847401227802038,
      "learning_rate": 0.06390349784231118,
      "loss": 0.2687,
      "num_input_tokens_seen": 52641168,
      "step": 27785
    },
    {
      "epoch": 4.533485602414553,
      "grad_norm": 0.0020702541805803776,
      "learning_rate": 0.06385526896072859,
      "loss": 0.3672,
      "num_input_tokens_seen": 52650528,
      "step": 27790
    },
    {
      "epoch": 4.53430132963537,
      "grad_norm": 0.002828670432791114,
      "learning_rate": 0.06380705336374613,
      "loss": 0.3241,
      "num_input_tokens_seen": 52660272,
      "step": 27795
    },
    {
      "epoch": 4.535117056856187,
      "grad_norm": 0.000835596292745322,
      "learning_rate": 0.06375885105879918,
      "loss": 0.2784,
      "num_input_tokens_seen": 52669472,
      "step": 27800
    },
    {
      "epoch": 4.535117056856187,
      "eval_loss": 0.3208317756652832,
      "eval_runtime": 152.7902,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 52669472,
      "step": 27800
    },
    {
      "epoch": 4.535932784077005,
      "grad_norm": 0.0015775677748024464,
      "learning_rate": 0.06371066205332115,
      "loss": 0.294,
      "num_input_tokens_seen": 52678480,
      "step": 27805
    },
    {
      "epoch": 4.536748511297822,
      "grad_norm": 0.000356395699782297,
      "learning_rate": 0.06366248635474347,
      "loss": 0.3495,
      "num_input_tokens_seen": 52686736,
      "step": 27810
    },
    {
      "epoch": 4.537564238518639,
      "grad_norm": 0.004666778724640608,
      "learning_rate": 0.06361432397049532,
      "loss": 0.3242,
      "num_input_tokens_seen": 52698000,
      "step": 27815
    },
    {
      "epoch": 4.538379965739456,
      "grad_norm": 0.0025657673832029104,
      "learning_rate": 0.06356617490800408,
      "loss": 0.2726,
      "num_input_tokens_seen": 52708400,
      "step": 27820
    },
    {
      "epoch": 4.539195692960274,
      "grad_norm": 0.0022835438139736652,
      "learning_rate": 0.06351803917469478,
      "loss": 0.3983,
      "num_input_tokens_seen": 52717616,
      "step": 27825
    },
    {
      "epoch": 4.5400114201810915,
      "grad_norm": 0.0016561249503865838,
      "learning_rate": 0.06346991677799067,
      "loss": 0.3999,
      "num_input_tokens_seen": 52726368,
      "step": 27830
    },
    {
      "epoch": 4.540827147401909,
      "grad_norm": 0.0020359260961413383,
      "learning_rate": 0.06342180772531283,
      "loss": 0.3035,
      "num_input_tokens_seen": 52736368,
      "step": 27835
    },
    {
      "epoch": 4.541642874622726,
      "grad_norm": 0.0012702459935098886,
      "learning_rate": 0.06337371202408021,
      "loss": 0.3587,
      "num_input_tokens_seen": 52744704,
      "step": 27840
    },
    {
      "epoch": 4.542458601843544,
      "grad_norm": 0.00132293242495507,
      "learning_rate": 0.06332562968170984,
      "loss": 0.3237,
      "num_input_tokens_seen": 52753584,
      "step": 27845
    },
    {
      "epoch": 4.543274329064361,
      "grad_norm": 0.0024572538677603006,
      "learning_rate": 0.06327756070561656,
      "loss": 0.3782,
      "num_input_tokens_seen": 52762912,
      "step": 27850
    },
    {
      "epoch": 4.544090056285178,
      "grad_norm": 0.002708615968003869,
      "learning_rate": 0.06322950510321329,
      "loss": 0.3371,
      "num_input_tokens_seen": 52772304,
      "step": 27855
    },
    {
      "epoch": 4.544905783505996,
      "grad_norm": 0.0015571475960314274,
      "learning_rate": 0.06318146288191076,
      "loss": 0.3188,
      "num_input_tokens_seen": 52782176,
      "step": 27860
    },
    {
      "epoch": 4.545721510726813,
      "grad_norm": 0.0012893201783299446,
      "learning_rate": 0.06313343404911763,
      "loss": 0.2879,
      "num_input_tokens_seen": 52791456,
      "step": 27865
    },
    {
      "epoch": 4.54653723794763,
      "grad_norm": 0.0015067077474668622,
      "learning_rate": 0.0630854186122406,
      "loss": 0.2964,
      "num_input_tokens_seen": 52800960,
      "step": 27870
    },
    {
      "epoch": 4.5473529651684474,
      "grad_norm": 0.004116957541555166,
      "learning_rate": 0.06303741657868431,
      "loss": 0.3436,
      "num_input_tokens_seen": 52812304,
      "step": 27875
    },
    {
      "epoch": 4.548168692389265,
      "grad_norm": 0.0022949737031012774,
      "learning_rate": 0.06298942795585115,
      "loss": 0.3756,
      "num_input_tokens_seen": 52820416,
      "step": 27880
    },
    {
      "epoch": 4.548984419610083,
      "grad_norm": 0.0004182314150966704,
      "learning_rate": 0.06294145275114167,
      "loss": 0.2921,
      "num_input_tokens_seen": 52830592,
      "step": 27885
    },
    {
      "epoch": 4.5498001468309,
      "grad_norm": 0.0007379347225651145,
      "learning_rate": 0.06289349097195428,
      "loss": 0.398,
      "num_input_tokens_seen": 52839840,
      "step": 27890
    },
    {
      "epoch": 4.550615874051717,
      "grad_norm": 0.0005641351453959942,
      "learning_rate": 0.06284554262568516,
      "loss": 0.3931,
      "num_input_tokens_seen": 52849216,
      "step": 27895
    },
    {
      "epoch": 4.551431601272535,
      "grad_norm": 0.0014073916245251894,
      "learning_rate": 0.06279760771972868,
      "loss": 0.3269,
      "num_input_tokens_seen": 52858496,
      "step": 27900
    },
    {
      "epoch": 4.552247328493352,
      "grad_norm": 0.0009312547044828534,
      "learning_rate": 0.06274968626147688,
      "loss": 0.3104,
      "num_input_tokens_seen": 52867392,
      "step": 27905
    },
    {
      "epoch": 4.553063055714169,
      "grad_norm": 0.004194724839180708,
      "learning_rate": 0.06270177825831993,
      "loss": 0.3399,
      "num_input_tokens_seen": 52877664,
      "step": 27910
    },
    {
      "epoch": 4.553878782934986,
      "grad_norm": 0.0010745462495833635,
      "learning_rate": 0.06265388371764587,
      "loss": 0.2779,
      "num_input_tokens_seen": 52887152,
      "step": 27915
    },
    {
      "epoch": 4.554694510155803,
      "grad_norm": 0.004452758003026247,
      "learning_rate": 0.0626060026468406,
      "loss": 0.3748,
      "num_input_tokens_seen": 52895888,
      "step": 27920
    },
    {
      "epoch": 4.555510237376621,
      "grad_norm": 0.0023089167661964893,
      "learning_rate": 0.06255813505328794,
      "loss": 0.3214,
      "num_input_tokens_seen": 52904720,
      "step": 27925
    },
    {
      "epoch": 4.5563259645974385,
      "grad_norm": 0.0015044420724734664,
      "learning_rate": 0.06251028094436978,
      "loss": 0.3263,
      "num_input_tokens_seen": 52915216,
      "step": 27930
    },
    {
      "epoch": 4.557141691818256,
      "grad_norm": 0.00045666738878935575,
      "learning_rate": 0.06246244032746568,
      "loss": 0.2901,
      "num_input_tokens_seen": 52924064,
      "step": 27935
    },
    {
      "epoch": 4.557957419039074,
      "grad_norm": 0.0016463504871353507,
      "learning_rate": 0.06241461320995342,
      "loss": 0.3132,
      "num_input_tokens_seen": 52933472,
      "step": 27940
    },
    {
      "epoch": 4.558773146259891,
      "grad_norm": 0.001758001046255231,
      "learning_rate": 0.062366799599208426,
      "loss": 0.363,
      "num_input_tokens_seen": 52940992,
      "step": 27945
    },
    {
      "epoch": 4.559588873480708,
      "grad_norm": 0.0024015887174755335,
      "learning_rate": 0.06231899950260418,
      "loss": 0.3772,
      "num_input_tokens_seen": 52950432,
      "step": 27950
    },
    {
      "epoch": 4.560404600701525,
      "grad_norm": 0.0015781964175403118,
      "learning_rate": 0.06227121292751214,
      "loss": 0.3165,
      "num_input_tokens_seen": 52959904,
      "step": 27955
    },
    {
      "epoch": 4.561220327922343,
      "grad_norm": 0.0019869289826601744,
      "learning_rate": 0.062223439881301496,
      "loss": 0.2998,
      "num_input_tokens_seen": 52970128,
      "step": 27960
    },
    {
      "epoch": 4.56203605514316,
      "grad_norm": 0.0024369496386498213,
      "learning_rate": 0.06217568037133948,
      "loss": 0.3419,
      "num_input_tokens_seen": 52979344,
      "step": 27965
    },
    {
      "epoch": 4.562851782363977,
      "grad_norm": 0.002515289466828108,
      "learning_rate": 0.06212793440499126,
      "loss": 0.3263,
      "num_input_tokens_seen": 52988464,
      "step": 27970
    },
    {
      "epoch": 4.5636675095847945,
      "grad_norm": 0.0019792523235082626,
      "learning_rate": 0.062080201989619783,
      "loss": 0.3354,
      "num_input_tokens_seen": 52998144,
      "step": 27975
    },
    {
      "epoch": 4.5644832368056125,
      "grad_norm": 0.0012815872905775905,
      "learning_rate": 0.062032483132586094,
      "loss": 0.3229,
      "num_input_tokens_seen": 53008160,
      "step": 27980
    },
    {
      "epoch": 4.56529896402643,
      "grad_norm": 0.001328352140262723,
      "learning_rate": 0.0619847778412489,
      "loss": 0.3223,
      "num_input_tokens_seen": 53016832,
      "step": 27985
    },
    {
      "epoch": 4.566114691247247,
      "grad_norm": 0.0005071923369541764,
      "learning_rate": 0.06193708612296509,
      "loss": 0.2983,
      "num_input_tokens_seen": 53026352,
      "step": 27990
    },
    {
      "epoch": 4.566930418468064,
      "grad_norm": 0.0026995486114174128,
      "learning_rate": 0.06188940798508923,
      "loss": 0.3634,
      "num_input_tokens_seen": 53035824,
      "step": 27995
    },
    {
      "epoch": 4.567746145688882,
      "grad_norm": 0.0017839474603533745,
      "learning_rate": 0.06184174343497397,
      "loss": 0.3428,
      "num_input_tokens_seen": 53045856,
      "step": 28000
    },
    {
      "epoch": 4.567746145688882,
      "eval_loss": 0.3275415301322937,
      "eval_runtime": 152.815,
      "eval_samples_per_second": 17.832,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 53045856,
      "step": 28000
    },
    {
      "epoch": 4.568561872909699,
      "grad_norm": 0.0008342295186594129,
      "learning_rate": 0.061794092479969726,
      "loss": 0.2958,
      "num_input_tokens_seen": 53054368,
      "step": 28005
    },
    {
      "epoch": 4.569377600130516,
      "grad_norm": 0.001625547418370843,
      "learning_rate": 0.06174645512742485,
      "loss": 0.2765,
      "num_input_tokens_seen": 53063888,
      "step": 28010
    },
    {
      "epoch": 4.570193327351333,
      "grad_norm": 0.0016324702883139253,
      "learning_rate": 0.06169883138468565,
      "loss": 0.3401,
      "num_input_tokens_seen": 53073968,
      "step": 28015
    },
    {
      "epoch": 4.571009054572151,
      "grad_norm": 0.0017365325475111604,
      "learning_rate": 0.06165122125909637,
      "loss": 0.2904,
      "num_input_tokens_seen": 53083440,
      "step": 28020
    },
    {
      "epoch": 4.5718247817929685,
      "grad_norm": 0.0010697057005017996,
      "learning_rate": 0.061603624757998965,
      "loss": 0.3186,
      "num_input_tokens_seen": 53093776,
      "step": 28025
    },
    {
      "epoch": 4.572640509013786,
      "grad_norm": 0.002029916737228632,
      "learning_rate": 0.0615560418887335,
      "loss": 0.3398,
      "num_input_tokens_seen": 53103776,
      "step": 28030
    },
    {
      "epoch": 4.573456236234604,
      "grad_norm": 0.0014082122361287475,
      "learning_rate": 0.06150847265863787,
      "loss": 0.2756,
      "num_input_tokens_seen": 53113776,
      "step": 28035
    },
    {
      "epoch": 4.574271963455421,
      "grad_norm": 0.0015822196146473289,
      "learning_rate": 0.061460917075047757,
      "loss": 0.3886,
      "num_input_tokens_seen": 53122704,
      "step": 28040
    },
    {
      "epoch": 4.575087690676238,
      "grad_norm": 0.0016019279137253761,
      "learning_rate": 0.06141337514529694,
      "loss": 0.3179,
      "num_input_tokens_seen": 53131632,
      "step": 28045
    },
    {
      "epoch": 4.575903417897055,
      "grad_norm": 0.001441583619453013,
      "learning_rate": 0.06136584687671687,
      "loss": 0.2624,
      "num_input_tokens_seen": 53140944,
      "step": 28050
    },
    {
      "epoch": 4.576719145117872,
      "grad_norm": 0.0021480571012943983,
      "learning_rate": 0.061318332276637064,
      "loss": 0.3422,
      "num_input_tokens_seen": 53149120,
      "step": 28055
    },
    {
      "epoch": 4.57753487233869,
      "grad_norm": 0.002391914138570428,
      "learning_rate": 0.06127083135238491,
      "loss": 0.3336,
      "num_input_tokens_seen": 53159536,
      "step": 28060
    },
    {
      "epoch": 4.578350599559507,
      "grad_norm": 0.0015329401940107346,
      "learning_rate": 0.06122334411128555,
      "loss": 0.317,
      "num_input_tokens_seen": 53169216,
      "step": 28065
    },
    {
      "epoch": 4.579166326780324,
      "grad_norm": 0.001509460504166782,
      "learning_rate": 0.06117587056066223,
      "loss": 0.347,
      "num_input_tokens_seen": 53177568,
      "step": 28070
    },
    {
      "epoch": 4.5799820540011424,
      "grad_norm": 0.0015082245226949453,
      "learning_rate": 0.06112841070783589,
      "loss": 0.3247,
      "num_input_tokens_seen": 53187200,
      "step": 28075
    },
    {
      "epoch": 4.58079778122196,
      "grad_norm": 0.0016208385350182652,
      "learning_rate": 0.061080964560125406,
      "loss": 0.3585,
      "num_input_tokens_seen": 53195424,
      "step": 28080
    },
    {
      "epoch": 4.581613508442777,
      "grad_norm": 0.0026023315731436014,
      "learning_rate": 0.06103353212484766,
      "loss": 0.3325,
      "num_input_tokens_seen": 53203760,
      "step": 28085
    },
    {
      "epoch": 4.582429235663594,
      "grad_norm": 0.0009644734091125429,
      "learning_rate": 0.06098611340931722,
      "loss": 0.3169,
      "num_input_tokens_seen": 53212560,
      "step": 28090
    },
    {
      "epoch": 4.583244962884411,
      "grad_norm": 0.003021595533937216,
      "learning_rate": 0.06093870842084672,
      "loss": 0.3205,
      "num_input_tokens_seen": 53224240,
      "step": 28095
    },
    {
      "epoch": 4.584060690105229,
      "grad_norm": 0.0030217517632991076,
      "learning_rate": 0.06089131716674666,
      "loss": 0.32,
      "num_input_tokens_seen": 53232944,
      "step": 28100
    },
    {
      "epoch": 4.584876417326046,
      "grad_norm": 0.0023267671931535006,
      "learning_rate": 0.060843939654325226,
      "loss": 0.347,
      "num_input_tokens_seen": 53242544,
      "step": 28105
    },
    {
      "epoch": 4.585692144546863,
      "grad_norm": 0.000862792890984565,
      "learning_rate": 0.06079657589088873,
      "loss": 0.2809,
      "num_input_tokens_seen": 53252416,
      "step": 28110
    },
    {
      "epoch": 4.586507871767681,
      "grad_norm": 0.001449287636205554,
      "learning_rate": 0.06074922588374126,
      "loss": 0.3493,
      "num_input_tokens_seen": 53263088,
      "step": 28115
    },
    {
      "epoch": 4.587323598988498,
      "grad_norm": 0.0018769619055092335,
      "learning_rate": 0.06070188964018472,
      "loss": 0.3158,
      "num_input_tokens_seen": 53272976,
      "step": 28120
    },
    {
      "epoch": 4.5881393262093155,
      "grad_norm": 0.0006306508439593017,
      "learning_rate": 0.06065456716751902,
      "loss": 0.3231,
      "num_input_tokens_seen": 53283408,
      "step": 28125
    },
    {
      "epoch": 4.588955053430133,
      "grad_norm": 0.0016048925463110209,
      "learning_rate": 0.06060725847304182,
      "loss": 0.3166,
      "num_input_tokens_seen": 53292832,
      "step": 28130
    },
    {
      "epoch": 4.589770780650951,
      "grad_norm": 0.002539999084547162,
      "learning_rate": 0.06055996356404877,
      "loss": 0.3014,
      "num_input_tokens_seen": 53302560,
      "step": 28135
    },
    {
      "epoch": 4.590586507871768,
      "grad_norm": 0.001113156438805163,
      "learning_rate": 0.06051268244783327,
      "loss": 0.337,
      "num_input_tokens_seen": 53312464,
      "step": 28140
    },
    {
      "epoch": 4.591402235092585,
      "grad_norm": 0.0013020987389609218,
      "learning_rate": 0.06046541513168676,
      "loss": 0.2977,
      "num_input_tokens_seen": 53321376,
      "step": 28145
    },
    {
      "epoch": 4.592217962313402,
      "grad_norm": 0.0010535609908401966,
      "learning_rate": 0.060418161622898356,
      "loss": 0.2719,
      "num_input_tokens_seen": 53331456,
      "step": 28150
    },
    {
      "epoch": 4.59303368953422,
      "grad_norm": 0.002932449337095022,
      "learning_rate": 0.06037092192875521,
      "loss": 0.329,
      "num_input_tokens_seen": 53341648,
      "step": 28155
    },
    {
      "epoch": 4.593849416755037,
      "grad_norm": 0.0013176341308280826,
      "learning_rate": 0.060323696056542225,
      "loss": 0.3187,
      "num_input_tokens_seen": 53351680,
      "step": 28160
    },
    {
      "epoch": 4.594665143975854,
      "grad_norm": 0.0025671902112662792,
      "learning_rate": 0.06027648401354229,
      "loss": 0.311,
      "num_input_tokens_seen": 53362192,
      "step": 28165
    },
    {
      "epoch": 4.5954808711966715,
      "grad_norm": 0.0009077267022803426,
      "learning_rate": 0.06022928580703601,
      "loss": 0.316,
      "num_input_tokens_seen": 53372176,
      "step": 28170
    },
    {
      "epoch": 4.5962965984174895,
      "grad_norm": 0.003310086205601692,
      "learning_rate": 0.060182101444301986,
      "loss": 0.2833,
      "num_input_tokens_seen": 53381968,
      "step": 28175
    },
    {
      "epoch": 4.597112325638307,
      "grad_norm": 0.0035185934975743294,
      "learning_rate": 0.06013493093261669,
      "loss": 0.3416,
      "num_input_tokens_seen": 53391088,
      "step": 28180
    },
    {
      "epoch": 4.597928052859124,
      "grad_norm": 0.0016373677644878626,
      "learning_rate": 0.06008777427925432,
      "loss": 0.3302,
      "num_input_tokens_seen": 53400848,
      "step": 28185
    },
    {
      "epoch": 4.598743780079941,
      "grad_norm": 0.0020989584736526012,
      "learning_rate": 0.06004063149148705,
      "loss": 0.304,
      "num_input_tokens_seen": 53410096,
      "step": 28190
    },
    {
      "epoch": 4.599559507300759,
      "grad_norm": 0.00182874477468431,
      "learning_rate": 0.05999350257658497,
      "loss": 0.3046,
      "num_input_tokens_seen": 53419584,
      "step": 28195
    },
    {
      "epoch": 4.600375234521576,
      "grad_norm": 0.00157563341781497,
      "learning_rate": 0.05994638754181582,
      "loss": 0.3063,
      "num_input_tokens_seen": 53429232,
      "step": 28200
    },
    {
      "epoch": 4.600375234521576,
      "eval_loss": 0.3215504288673401,
      "eval_runtime": 152.872,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 53429232,
      "step": 28200
    },
    {
      "epoch": 4.601190961742393,
      "grad_norm": 0.0011422978714108467,
      "learning_rate": 0.059899286394445445,
      "loss": 0.2846,
      "num_input_tokens_seen": 53438896,
      "step": 28205
    },
    {
      "epoch": 4.602006688963211,
      "grad_norm": 0.0017230132361873984,
      "learning_rate": 0.059852199141737346,
      "loss": 0.3344,
      "num_input_tokens_seen": 53448192,
      "step": 28210
    },
    {
      "epoch": 4.602822416184028,
      "grad_norm": 0.002337175188586116,
      "learning_rate": 0.05980512579095304,
      "loss": 0.3273,
      "num_input_tokens_seen": 53458096,
      "step": 28215
    },
    {
      "epoch": 4.6036381434048455,
      "grad_norm": 0.0014687948860228062,
      "learning_rate": 0.05975806634935181,
      "loss": 0.2936,
      "num_input_tokens_seen": 53466960,
      "step": 28220
    },
    {
      "epoch": 4.604453870625663,
      "grad_norm": 0.0018341204850003123,
      "learning_rate": 0.05971102082419076,
      "loss": 0.2824,
      "num_input_tokens_seen": 53477456,
      "step": 28225
    },
    {
      "epoch": 4.60526959784648,
      "grad_norm": 0.0027176423463970423,
      "learning_rate": 0.05966398922272492,
      "loss": 0.3065,
      "num_input_tokens_seen": 53486656,
      "step": 28230
    },
    {
      "epoch": 4.606085325067298,
      "grad_norm": 0.0032775660511106253,
      "learning_rate": 0.059616971552207236,
      "loss": 0.3281,
      "num_input_tokens_seen": 53494944,
      "step": 28235
    },
    {
      "epoch": 4.606901052288115,
      "grad_norm": 0.0018996550934389234,
      "learning_rate": 0.059569967819888305,
      "loss": 0.3011,
      "num_input_tokens_seen": 53505008,
      "step": 28240
    },
    {
      "epoch": 4.607716779508932,
      "grad_norm": 0.0037137544713914394,
      "learning_rate": 0.05952297803301681,
      "loss": 0.3355,
      "num_input_tokens_seen": 53515120,
      "step": 28245
    },
    {
      "epoch": 4.60853250672975,
      "grad_norm": 0.0033519500866532326,
      "learning_rate": 0.059476002198839056,
      "loss": 0.317,
      "num_input_tokens_seen": 53524672,
      "step": 28250
    },
    {
      "epoch": 4.609348233950567,
      "grad_norm": 0.000830019183922559,
      "learning_rate": 0.05942904032459935,
      "loss": 0.2918,
      "num_input_tokens_seen": 53534608,
      "step": 28255
    },
    {
      "epoch": 4.610163961171384,
      "grad_norm": 0.002064059255644679,
      "learning_rate": 0.05938209241753987,
      "loss": 0.3562,
      "num_input_tokens_seen": 53543136,
      "step": 28260
    },
    {
      "epoch": 4.610979688392201,
      "grad_norm": 0.0012270263396203518,
      "learning_rate": 0.05933515848490046,
      "loss": 0.2688,
      "num_input_tokens_seen": 53552224,
      "step": 28265
    },
    {
      "epoch": 4.6117954156130185,
      "grad_norm": 0.0026194299571216106,
      "learning_rate": 0.059288238533918985,
      "loss": 0.3136,
      "num_input_tokens_seen": 53560752,
      "step": 28270
    },
    {
      "epoch": 4.6126111428338366,
      "grad_norm": 0.002359106671065092,
      "learning_rate": 0.05924133257183113,
      "loss": 0.3402,
      "num_input_tokens_seen": 53569680,
      "step": 28275
    },
    {
      "epoch": 4.613426870054654,
      "grad_norm": 0.0028521777130663395,
      "learning_rate": 0.059194440605870285,
      "loss": 0.3286,
      "num_input_tokens_seen": 53579536,
      "step": 28280
    },
    {
      "epoch": 4.614242597275471,
      "grad_norm": 0.003169839736074209,
      "learning_rate": 0.059147562643267884,
      "loss": 0.3395,
      "num_input_tokens_seen": 53590336,
      "step": 28285
    },
    {
      "epoch": 4.615058324496289,
      "grad_norm": 0.0015720229130238295,
      "learning_rate": 0.059100698691253055,
      "loss": 0.3997,
      "num_input_tokens_seen": 53599280,
      "step": 28290
    },
    {
      "epoch": 4.615874051717106,
      "grad_norm": 0.004126183222979307,
      "learning_rate": 0.05905384875705273,
      "loss": 0.3611,
      "num_input_tokens_seen": 53608192,
      "step": 28295
    },
    {
      "epoch": 4.616689778937923,
      "grad_norm": 0.0029707399662584066,
      "learning_rate": 0.05900701284789189,
      "loss": 0.2902,
      "num_input_tokens_seen": 53617168,
      "step": 28300
    },
    {
      "epoch": 4.61750550615874,
      "grad_norm": 0.00214291806332767,
      "learning_rate": 0.058960190970993115,
      "loss": 0.3001,
      "num_input_tokens_seen": 53626912,
      "step": 28305
    },
    {
      "epoch": 4.618321233379558,
      "grad_norm": 0.0026220460422337055,
      "learning_rate": 0.058913383133576955,
      "loss": 0.3664,
      "num_input_tokens_seen": 53637264,
      "step": 28310
    },
    {
      "epoch": 4.619136960600375,
      "grad_norm": 0.002678842982277274,
      "learning_rate": 0.05886658934286185,
      "loss": 0.4056,
      "num_input_tokens_seen": 53646368,
      "step": 28315
    },
    {
      "epoch": 4.6199526878211925,
      "grad_norm": 0.002496580360457301,
      "learning_rate": 0.058819809606063846,
      "loss": 0.3389,
      "num_input_tokens_seen": 53656752,
      "step": 28320
    },
    {
      "epoch": 4.62076841504201,
      "grad_norm": 0.0013068949338048697,
      "learning_rate": 0.05877304393039711,
      "loss": 0.3058,
      "num_input_tokens_seen": 53666432,
      "step": 28325
    },
    {
      "epoch": 4.621584142262828,
      "grad_norm": 0.002581753069534898,
      "learning_rate": 0.05872629232307338,
      "loss": 0.3581,
      "num_input_tokens_seen": 53677312,
      "step": 28330
    },
    {
      "epoch": 4.622399869483645,
      "grad_norm": 0.001910469145514071,
      "learning_rate": 0.05867955479130239,
      "loss": 0.3234,
      "num_input_tokens_seen": 53686848,
      "step": 28335
    },
    {
      "epoch": 4.623215596704462,
      "grad_norm": 0.0022936714813113213,
      "learning_rate": 0.058632831342291705,
      "loss": 0.3457,
      "num_input_tokens_seen": 53696240,
      "step": 28340
    },
    {
      "epoch": 4.624031323925279,
      "grad_norm": 0.0017792769940569997,
      "learning_rate": 0.05858612198324655,
      "loss": 0.3427,
      "num_input_tokens_seen": 53705472,
      "step": 28345
    },
    {
      "epoch": 4.624847051146097,
      "grad_norm": 0.001387274358421564,
      "learning_rate": 0.05853942672137025,
      "loss": 0.3756,
      "num_input_tokens_seen": 53713968,
      "step": 28350
    },
    {
      "epoch": 4.625662778366914,
      "grad_norm": 0.0016617936780676246,
      "learning_rate": 0.05849274556386363,
      "loss": 0.2954,
      "num_input_tokens_seen": 53722864,
      "step": 28355
    },
    {
      "epoch": 4.626478505587731,
      "grad_norm": 0.0011516299564391375,
      "learning_rate": 0.05844607851792567,
      "loss": 0.3239,
      "num_input_tokens_seen": 53732624,
      "step": 28360
    },
    {
      "epoch": 4.6272942328085485,
      "grad_norm": 0.001404467853717506,
      "learning_rate": 0.058399425590752924,
      "loss": 0.3189,
      "num_input_tokens_seen": 53741840,
      "step": 28365
    },
    {
      "epoch": 4.6281099600293665,
      "grad_norm": 0.0018423845758661628,
      "learning_rate": 0.05835278678953985,
      "loss": 0.2782,
      "num_input_tokens_seen": 53751952,
      "step": 28370
    },
    {
      "epoch": 4.628925687250184,
      "grad_norm": 0.0008415600168518722,
      "learning_rate": 0.05830616212147874,
      "loss": 0.3836,
      "num_input_tokens_seen": 53761920,
      "step": 28375
    },
    {
      "epoch": 4.629741414471001,
      "grad_norm": 0.0009434923995286226,
      "learning_rate": 0.058259551593759784,
      "loss": 0.2852,
      "num_input_tokens_seen": 53771728,
      "step": 28380
    },
    {
      "epoch": 4.630557141691818,
      "grad_norm": 0.0011909666936844587,
      "learning_rate": 0.058212955213570804,
      "loss": 0.3744,
      "num_input_tokens_seen": 53781040,
      "step": 28385
    },
    {
      "epoch": 4.631372868912636,
      "grad_norm": 0.0012820173287764192,
      "learning_rate": 0.0581663729880976,
      "loss": 0.2935,
      "num_input_tokens_seen": 53791456,
      "step": 28390
    },
    {
      "epoch": 4.632188596133453,
      "grad_norm": 0.0015338872326537967,
      "learning_rate": 0.05811980492452379,
      "loss": 0.3343,
      "num_input_tokens_seen": 53801216,
      "step": 28395
    },
    {
      "epoch": 4.63300432335427,
      "grad_norm": 0.0019411692628636956,
      "learning_rate": 0.058073251030030644,
      "loss": 0.2983,
      "num_input_tokens_seen": 53810560,
      "step": 28400
    },
    {
      "epoch": 4.63300432335427,
      "eval_loss": 0.32792237401008606,
      "eval_runtime": 152.8967,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.915,
      "num_input_tokens_seen": 53810560,
      "step": 28400
    },
    {
      "epoch": 4.633820050575087,
      "grad_norm": 0.0024520934093743563,
      "learning_rate": 0.05802671131179747,
      "loss": 0.305,
      "num_input_tokens_seen": 53819264,
      "step": 28405
    },
    {
      "epoch": 4.634635777795905,
      "grad_norm": 0.0004845089861191809,
      "learning_rate": 0.057980185777001154,
      "loss": 0.3628,
      "num_input_tokens_seen": 53828480,
      "step": 28410
    },
    {
      "epoch": 4.635451505016722,
      "grad_norm": 0.0020755697041749954,
      "learning_rate": 0.057933674432816606,
      "loss": 0.3247,
      "num_input_tokens_seen": 53837616,
      "step": 28415
    },
    {
      "epoch": 4.63626723223754,
      "grad_norm": 0.0015474763931706548,
      "learning_rate": 0.05788717728641648,
      "loss": 0.326,
      "num_input_tokens_seen": 53847344,
      "step": 28420
    },
    {
      "epoch": 4.637082959458358,
      "grad_norm": 0.002783282659947872,
      "learning_rate": 0.057840694344971126,
      "loss": 0.3568,
      "num_input_tokens_seen": 53858080,
      "step": 28425
    },
    {
      "epoch": 4.637898686679175,
      "grad_norm": 0.0013171856990084052,
      "learning_rate": 0.0577942256156489,
      "loss": 0.261,
      "num_input_tokens_seen": 53867616,
      "step": 28430
    },
    {
      "epoch": 4.638714413899992,
      "grad_norm": 0.0006047168863005936,
      "learning_rate": 0.057747771105615804,
      "loss": 0.3158,
      "num_input_tokens_seen": 53877520,
      "step": 28435
    },
    {
      "epoch": 4.639530141120809,
      "grad_norm": 0.0029404398519545794,
      "learning_rate": 0.05770133082203568,
      "loss": 0.2838,
      "num_input_tokens_seen": 53887440,
      "step": 28440
    },
    {
      "epoch": 4.640345868341626,
      "grad_norm": 0.0008140044519677758,
      "learning_rate": 0.0576549047720703,
      "loss": 0.3502,
      "num_input_tokens_seen": 53896480,
      "step": 28445
    },
    {
      "epoch": 4.641161595562444,
      "grad_norm": 0.0022629722952842712,
      "learning_rate": 0.05760849296287902,
      "loss": 0.3002,
      "num_input_tokens_seen": 53905328,
      "step": 28450
    },
    {
      "epoch": 4.641977322783261,
      "grad_norm": 0.001485349959693849,
      "learning_rate": 0.05756209540161919,
      "loss": 0.3637,
      "num_input_tokens_seen": 53914784,
      "step": 28455
    },
    {
      "epoch": 4.642793050004078,
      "grad_norm": 0.0009776449296623468,
      "learning_rate": 0.05751571209544595,
      "loss": 0.3071,
      "num_input_tokens_seen": 53922720,
      "step": 28460
    },
    {
      "epoch": 4.643608777224896,
      "grad_norm": 0.001948525314219296,
      "learning_rate": 0.057469343051512085,
      "loss": 0.2973,
      "num_input_tokens_seen": 53932768,
      "step": 28465
    },
    {
      "epoch": 4.6444245044457135,
      "grad_norm": 0.0009021475561894476,
      "learning_rate": 0.057422988276968324,
      "loss": 0.3218,
      "num_input_tokens_seen": 53942640,
      "step": 28470
    },
    {
      "epoch": 4.645240231666531,
      "grad_norm": 0.002131260931491852,
      "learning_rate": 0.05737664777896323,
      "loss": 0.2715,
      "num_input_tokens_seen": 53952592,
      "step": 28475
    },
    {
      "epoch": 4.646055958887348,
      "grad_norm": 0.001723572495393455,
      "learning_rate": 0.057330321564642975,
      "loss": 0.3449,
      "num_input_tokens_seen": 53962464,
      "step": 28480
    },
    {
      "epoch": 4.646871686108166,
      "grad_norm": 0.002241353504359722,
      "learning_rate": 0.05728400964115174,
      "loss": 0.3379,
      "num_input_tokens_seen": 53970576,
      "step": 28485
    },
    {
      "epoch": 4.647687413328983,
      "grad_norm": 0.0029660691507160664,
      "learning_rate": 0.057237712015631305,
      "loss": 0.3166,
      "num_input_tokens_seen": 53979776,
      "step": 28490
    },
    {
      "epoch": 4.6485031405498,
      "grad_norm": 0.0007177501101978123,
      "learning_rate": 0.057191428695221425,
      "loss": 0.2812,
      "num_input_tokens_seen": 53989568,
      "step": 28495
    },
    {
      "epoch": 4.649318867770617,
      "grad_norm": 0.0014660404995083809,
      "learning_rate": 0.05714515968705958,
      "loss": 0.2962,
      "num_input_tokens_seen": 53999552,
      "step": 28500
    },
    {
      "epoch": 4.650134594991435,
      "grad_norm": 0.0004747921775560826,
      "learning_rate": 0.05709890499828099,
      "loss": 0.3627,
      "num_input_tokens_seen": 54008912,
      "step": 28505
    },
    {
      "epoch": 4.650950322212252,
      "grad_norm": 0.0014545985031872988,
      "learning_rate": 0.05705266463601868,
      "loss": 0.2857,
      "num_input_tokens_seen": 54019312,
      "step": 28510
    },
    {
      "epoch": 4.6517660494330695,
      "grad_norm": 0.0012118201702833176,
      "learning_rate": 0.057006438607403565,
      "loss": 0.2633,
      "num_input_tokens_seen": 54029296,
      "step": 28515
    },
    {
      "epoch": 4.652581776653887,
      "grad_norm": 0.0026942237745970488,
      "learning_rate": 0.056960226919564205,
      "loss": 0.2859,
      "num_input_tokens_seen": 54038448,
      "step": 28520
    },
    {
      "epoch": 4.653397503874705,
      "grad_norm": 0.0028779045678675175,
      "learning_rate": 0.05691402957962713,
      "loss": 0.3989,
      "num_input_tokens_seen": 54047136,
      "step": 28525
    },
    {
      "epoch": 4.654213231095522,
      "grad_norm": 0.0008911049226298928,
      "learning_rate": 0.05686784659471642,
      "loss": 0.3595,
      "num_input_tokens_seen": 54056720,
      "step": 28530
    },
    {
      "epoch": 4.655028958316339,
      "grad_norm": 0.002012908924371004,
      "learning_rate": 0.056821677971954136,
      "loss": 0.3678,
      "num_input_tokens_seen": 54066336,
      "step": 28535
    },
    {
      "epoch": 4.655844685537156,
      "grad_norm": 0.0023982145357877016,
      "learning_rate": 0.05677552371846012,
      "loss": 0.3427,
      "num_input_tokens_seen": 54075744,
      "step": 28540
    },
    {
      "epoch": 4.656660412757974,
      "grad_norm": 0.0020093265920877457,
      "learning_rate": 0.05672938384135182,
      "loss": 0.298,
      "num_input_tokens_seen": 54085712,
      "step": 28545
    },
    {
      "epoch": 4.657476139978791,
      "grad_norm": 0.001559693249873817,
      "learning_rate": 0.05668325834774465,
      "loss": 0.2758,
      "num_input_tokens_seen": 54095680,
      "step": 28550
    },
    {
      "epoch": 4.658291867199608,
      "grad_norm": 0.003086221171543002,
      "learning_rate": 0.05663714724475177,
      "loss": 0.3868,
      "num_input_tokens_seen": 54105792,
      "step": 28555
    },
    {
      "epoch": 4.659107594420425,
      "grad_norm": 0.0010948125272989273,
      "learning_rate": 0.05659105053948403,
      "loss": 0.3173,
      "num_input_tokens_seen": 54115152,
      "step": 28560
    },
    {
      "epoch": 4.6599233216412435,
      "grad_norm": 0.0025502434000372887,
      "learning_rate": 0.056544968239050176,
      "loss": 0.3735,
      "num_input_tokens_seen": 54124464,
      "step": 28565
    },
    {
      "epoch": 4.660739048862061,
      "grad_norm": 0.002446269616484642,
      "learning_rate": 0.056498900350556616,
      "loss": 0.3477,
      "num_input_tokens_seen": 54134464,
      "step": 28570
    },
    {
      "epoch": 4.661554776082878,
      "grad_norm": 0.0025404065381735563,
      "learning_rate": 0.05645284688110766,
      "loss": 0.3573,
      "num_input_tokens_seen": 54143696,
      "step": 28575
    },
    {
      "epoch": 4.662370503303695,
      "grad_norm": 0.0030508439522236586,
      "learning_rate": 0.05640680783780532,
      "loss": 0.3339,
      "num_input_tokens_seen": 54152848,
      "step": 28580
    },
    {
      "epoch": 4.663186230524513,
      "grad_norm": 0.0011195279657840729,
      "learning_rate": 0.056360783227749324,
      "loss": 0.3169,
      "num_input_tokens_seen": 54162736,
      "step": 28585
    },
    {
      "epoch": 4.66400195774533,
      "grad_norm": 0.0007254189695231616,
      "learning_rate": 0.05631477305803728,
      "loss": 0.3336,
      "num_input_tokens_seen": 54171920,
      "step": 28590
    },
    {
      "epoch": 4.664817684966147,
      "grad_norm": 0.0023461957462131977,
      "learning_rate": 0.05626877733576462,
      "loss": 0.3496,
      "num_input_tokens_seen": 54181792,
      "step": 28595
    },
    {
      "epoch": 4.665633412186965,
      "grad_norm": 0.0011349038686603308,
      "learning_rate": 0.05622279606802435,
      "loss": 0.319,
      "num_input_tokens_seen": 54191536,
      "step": 28600
    },
    {
      "epoch": 4.665633412186965,
      "eval_loss": 0.32267656922340393,
      "eval_runtime": 152.9497,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 54191536,
      "step": 28600
    },
    {
      "epoch": 4.666449139407782,
      "grad_norm": 0.0032917712815105915,
      "learning_rate": 0.05617682926190744,
      "loss": 0.3277,
      "num_input_tokens_seen": 54202224,
      "step": 28605
    },
    {
      "epoch": 4.667264866628599,
      "grad_norm": 0.0019528548000380397,
      "learning_rate": 0.05613087692450248,
      "loss": 0.3519,
      "num_input_tokens_seen": 54212240,
      "step": 28610
    },
    {
      "epoch": 4.6680805938494165,
      "grad_norm": 0.002344159409403801,
      "learning_rate": 0.05608493906289592,
      "loss": 0.3592,
      "num_input_tokens_seen": 54221840,
      "step": 28615
    },
    {
      "epoch": 4.668896321070234,
      "grad_norm": 0.0010855309665203094,
      "learning_rate": 0.05603901568417201,
      "loss": 0.2832,
      "num_input_tokens_seen": 54230864,
      "step": 28620
    },
    {
      "epoch": 4.669712048291052,
      "grad_norm": 0.00044712398084811866,
      "learning_rate": 0.055993106795412625,
      "loss": 0.3191,
      "num_input_tokens_seen": 54240544,
      "step": 28625
    },
    {
      "epoch": 4.670527775511869,
      "grad_norm": 0.004138687159866095,
      "learning_rate": 0.05594721240369759,
      "loss": 0.3288,
      "num_input_tokens_seen": 54250560,
      "step": 28630
    },
    {
      "epoch": 4.671343502732686,
      "grad_norm": 0.0010615027276799083,
      "learning_rate": 0.055901332516104296,
      "loss": 0.3323,
      "num_input_tokens_seen": 54260112,
      "step": 28635
    },
    {
      "epoch": 4.672159229953504,
      "grad_norm": 0.0019482015632092953,
      "learning_rate": 0.05585546713970804,
      "loss": 0.3363,
      "num_input_tokens_seen": 54270400,
      "step": 28640
    },
    {
      "epoch": 4.672974957174321,
      "grad_norm": 0.0020777739118784666,
      "learning_rate": 0.05580961628158189,
      "loss": 0.3272,
      "num_input_tokens_seen": 54279280,
      "step": 28645
    },
    {
      "epoch": 4.673790684395138,
      "grad_norm": 0.0021108880173414946,
      "learning_rate": 0.05576377994879659,
      "loss": 0.3636,
      "num_input_tokens_seen": 54288624,
      "step": 28650
    },
    {
      "epoch": 4.674606411615955,
      "grad_norm": 0.0007757893763482571,
      "learning_rate": 0.05571795814842063,
      "loss": 0.3164,
      "num_input_tokens_seen": 54297936,
      "step": 28655
    },
    {
      "epoch": 4.6754221388367725,
      "grad_norm": 0.002181483432650566,
      "learning_rate": 0.05567215088752037,
      "loss": 0.2739,
      "num_input_tokens_seen": 54307296,
      "step": 28660
    },
    {
      "epoch": 4.6762378660575905,
      "grad_norm": 0.0011517341481521726,
      "learning_rate": 0.05562635817315981,
      "loss": 0.2944,
      "num_input_tokens_seen": 54316848,
      "step": 28665
    },
    {
      "epoch": 4.677053593278408,
      "grad_norm": 0.0008177837007679045,
      "learning_rate": 0.05558058001240083,
      "loss": 0.3532,
      "num_input_tokens_seen": 54325696,
      "step": 28670
    },
    {
      "epoch": 4.677869320499225,
      "grad_norm": 0.0007827075314708054,
      "learning_rate": 0.055534816412302915,
      "loss": 0.2976,
      "num_input_tokens_seen": 54334992,
      "step": 28675
    },
    {
      "epoch": 4.678685047720043,
      "grad_norm": 0.0008143748855218291,
      "learning_rate": 0.055489067379923436,
      "loss": 0.3273,
      "num_input_tokens_seen": 54344256,
      "step": 28680
    },
    {
      "epoch": 4.67950077494086,
      "grad_norm": 0.002761919517070055,
      "learning_rate": 0.055443332922317505,
      "loss": 0.3808,
      "num_input_tokens_seen": 54353200,
      "step": 28685
    },
    {
      "epoch": 4.680316502161677,
      "grad_norm": 0.0019137357594445348,
      "learning_rate": 0.055397613046537876,
      "loss": 0.3792,
      "num_input_tokens_seen": 54362944,
      "step": 28690
    },
    {
      "epoch": 4.681132229382494,
      "grad_norm": 0.0013045086525380611,
      "learning_rate": 0.055351907759635145,
      "loss": 0.2821,
      "num_input_tokens_seen": 54372208,
      "step": 28695
    },
    {
      "epoch": 4.681947956603312,
      "grad_norm": 0.0005917990347370505,
      "learning_rate": 0.05530621706865772,
      "loss": 0.3292,
      "num_input_tokens_seen": 54380592,
      "step": 28700
    },
    {
      "epoch": 4.682763683824129,
      "grad_norm": 0.003227689303457737,
      "learning_rate": 0.055260540980651564,
      "loss": 0.3839,
      "num_input_tokens_seen": 54388656,
      "step": 28705
    },
    {
      "epoch": 4.6835794110449465,
      "grad_norm": 0.0005885610007680953,
      "learning_rate": 0.05521487950266062,
      "loss": 0.3561,
      "num_input_tokens_seen": 54398256,
      "step": 28710
    },
    {
      "epoch": 4.684395138265764,
      "grad_norm": 0.0012389938347041607,
      "learning_rate": 0.055169232641726344,
      "loss": 0.2702,
      "num_input_tokens_seen": 54408608,
      "step": 28715
    },
    {
      "epoch": 4.685210865486582,
      "grad_norm": 0.0005666015204042196,
      "learning_rate": 0.055123600404888166,
      "loss": 0.2987,
      "num_input_tokens_seen": 54418096,
      "step": 28720
    },
    {
      "epoch": 4.686026592707399,
      "grad_norm": 0.0035341468174010515,
      "learning_rate": 0.05507798279918309,
      "loss": 0.2921,
      "num_input_tokens_seen": 54427520,
      "step": 28725
    },
    {
      "epoch": 4.686842319928216,
      "grad_norm": 0.0023379384074360132,
      "learning_rate": 0.0550323798316459,
      "loss": 0.343,
      "num_input_tokens_seen": 54437776,
      "step": 28730
    },
    {
      "epoch": 4.687658047149033,
      "grad_norm": 0.0032614939846098423,
      "learning_rate": 0.05498679150930916,
      "loss": 0.2825,
      "num_input_tokens_seen": 54447408,
      "step": 28735
    },
    {
      "epoch": 4.688473774369851,
      "grad_norm": 0.002081557409837842,
      "learning_rate": 0.05494121783920323,
      "loss": 0.3565,
      "num_input_tokens_seen": 54455824,
      "step": 28740
    },
    {
      "epoch": 4.689289501590668,
      "grad_norm": 0.0025055173318833113,
      "learning_rate": 0.05489565882835605,
      "loss": 0.2863,
      "num_input_tokens_seen": 54465040,
      "step": 28745
    },
    {
      "epoch": 4.690105228811485,
      "grad_norm": 0.0013453119900077581,
      "learning_rate": 0.05485011448379348,
      "loss": 0.3075,
      "num_input_tokens_seen": 54475056,
      "step": 28750
    },
    {
      "epoch": 4.690920956032302,
      "grad_norm": 0.0016955782193690538,
      "learning_rate": 0.05480458481253893,
      "loss": 0.36,
      "num_input_tokens_seen": 54485072,
      "step": 28755
    },
    {
      "epoch": 4.69173668325312,
      "grad_norm": 0.001850288244895637,
      "learning_rate": 0.054759069821613715,
      "loss": 0.3565,
      "num_input_tokens_seen": 54494432,
      "step": 28760
    },
    {
      "epoch": 4.692552410473938,
      "grad_norm": 0.001441435539163649,
      "learning_rate": 0.05471356951803683,
      "loss": 0.3222,
      "num_input_tokens_seen": 54503792,
      "step": 28765
    },
    {
      "epoch": 4.693368137694755,
      "grad_norm": 0.0022498564794659615,
      "learning_rate": 0.054668083908824945,
      "loss": 0.2285,
      "num_input_tokens_seen": 54513680,
      "step": 28770
    },
    {
      "epoch": 4.694183864915573,
      "grad_norm": 0.0029427839908748865,
      "learning_rate": 0.054622613000992526,
      "loss": 0.3581,
      "num_input_tokens_seen": 54524144,
      "step": 28775
    },
    {
      "epoch": 4.69499959213639,
      "grad_norm": 0.004579160828143358,
      "learning_rate": 0.05457715680155182,
      "loss": 0.2564,
      "num_input_tokens_seen": 54533440,
      "step": 28780
    },
    {
      "epoch": 4.695815319357207,
      "grad_norm": 0.0011872806353494525,
      "learning_rate": 0.05453171531751265,
      "loss": 0.3207,
      "num_input_tokens_seen": 54543472,
      "step": 28785
    },
    {
      "epoch": 4.696631046578024,
      "grad_norm": 0.0022492643911391497,
      "learning_rate": 0.05448628855588276,
      "loss": 0.3193,
      "num_input_tokens_seen": 54553392,
      "step": 28790
    },
    {
      "epoch": 4.697446773798841,
      "grad_norm": 0.002545382361859083,
      "learning_rate": 0.05444087652366746,
      "loss": 0.2768,
      "num_input_tokens_seen": 54562336,
      "step": 28795
    },
    {
      "epoch": 4.698262501019659,
      "grad_norm": 0.0018443980952724814,
      "learning_rate": 0.05439547922786984,
      "loss": 0.3985,
      "num_input_tokens_seen": 54572176,
      "step": 28800
    },
    {
      "epoch": 4.698262501019659,
      "eval_loss": 0.3220113515853882,
      "eval_runtime": 152.8138,
      "eval_samples_per_second": 17.832,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 54572176,
      "step": 28800
    },
    {
      "epoch": 4.699078228240476,
      "grad_norm": 0.000701508775819093,
      "learning_rate": 0.0543500966754908,
      "loss": 0.3228,
      "num_input_tokens_seen": 54581232,
      "step": 28805
    },
    {
      "epoch": 4.6998939554612935,
      "grad_norm": 0.0031690143514424562,
      "learning_rate": 0.05430472887352882,
      "loss": 0.3405,
      "num_input_tokens_seen": 54591328,
      "step": 28810
    },
    {
      "epoch": 4.7007096826821115,
      "grad_norm": 0.0010469077387824655,
      "learning_rate": 0.05425937582898023,
      "loss": 0.3462,
      "num_input_tokens_seen": 54601072,
      "step": 28815
    },
    {
      "epoch": 4.701525409902929,
      "grad_norm": 0.0023045435082167387,
      "learning_rate": 0.054214037548839085,
      "loss": 0.2992,
      "num_input_tokens_seen": 54610768,
      "step": 28820
    },
    {
      "epoch": 4.702341137123746,
      "grad_norm": 0.0024883390869945288,
      "learning_rate": 0.05416871404009703,
      "loss": 0.3649,
      "num_input_tokens_seen": 54619712,
      "step": 28825
    },
    {
      "epoch": 4.703156864344563,
      "grad_norm": 0.0011091609485447407,
      "learning_rate": 0.054123405309743605,
      "loss": 0.2857,
      "num_input_tokens_seen": 54629872,
      "step": 28830
    },
    {
      "epoch": 4.70397259156538,
      "grad_norm": 0.0005103892763145268,
      "learning_rate": 0.0540781113647659,
      "loss": 0.3009,
      "num_input_tokens_seen": 54639152,
      "step": 28835
    },
    {
      "epoch": 4.704788318786198,
      "grad_norm": 0.0022414110135287046,
      "learning_rate": 0.054032832212148836,
      "loss": 0.2969,
      "num_input_tokens_seen": 54648384,
      "step": 28840
    },
    {
      "epoch": 4.705604046007015,
      "grad_norm": 0.0007156640058383346,
      "learning_rate": 0.0539875678588751,
      "loss": 0.3069,
      "num_input_tokens_seen": 54658512,
      "step": 28845
    },
    {
      "epoch": 4.706419773227832,
      "grad_norm": 0.0019549524877220392,
      "learning_rate": 0.05394231831192492,
      "loss": 0.3265,
      "num_input_tokens_seen": 54667888,
      "step": 28850
    },
    {
      "epoch": 4.70723550044865,
      "grad_norm": 0.0012059977743774652,
      "learning_rate": 0.05389708357827639,
      "loss": 0.3599,
      "num_input_tokens_seen": 54677952,
      "step": 28855
    },
    {
      "epoch": 4.7080512276694675,
      "grad_norm": 0.002384425140917301,
      "learning_rate": 0.05385186366490533,
      "loss": 0.387,
      "num_input_tokens_seen": 54687072,
      "step": 28860
    },
    {
      "epoch": 4.708866954890285,
      "grad_norm": 0.003893579589203,
      "learning_rate": 0.053806658578785166,
      "loss": 0.3721,
      "num_input_tokens_seen": 54696640,
      "step": 28865
    },
    {
      "epoch": 4.709682682111102,
      "grad_norm": 0.0013148015132173896,
      "learning_rate": 0.05376146832688705,
      "loss": 0.2901,
      "num_input_tokens_seen": 54707600,
      "step": 28870
    },
    {
      "epoch": 4.71049840933192,
      "grad_norm": 0.0023131612688302994,
      "learning_rate": 0.053716292916179964,
      "loss": 0.273,
      "num_input_tokens_seen": 54716912,
      "step": 28875
    },
    {
      "epoch": 4.711314136552737,
      "grad_norm": 0.0017959576798602939,
      "learning_rate": 0.05367113235363045,
      "loss": 0.3535,
      "num_input_tokens_seen": 54727248,
      "step": 28880
    },
    {
      "epoch": 4.712129863773554,
      "grad_norm": 0.001582021708600223,
      "learning_rate": 0.05362598664620289,
      "loss": 0.392,
      "num_input_tokens_seen": 54735776,
      "step": 28885
    },
    {
      "epoch": 4.712945590994371,
      "grad_norm": 0.0005145791219547391,
      "learning_rate": 0.053580855800859285,
      "loss": 0.3238,
      "num_input_tokens_seen": 54744272,
      "step": 28890
    },
    {
      "epoch": 4.713761318215189,
      "grad_norm": 0.0021865954622626305,
      "learning_rate": 0.05353573982455938,
      "loss": 0.3333,
      "num_input_tokens_seen": 54753312,
      "step": 28895
    },
    {
      "epoch": 4.714577045436006,
      "grad_norm": 0.0013768189819529653,
      "learning_rate": 0.053490638724260686,
      "loss": 0.3263,
      "num_input_tokens_seen": 54762576,
      "step": 28900
    },
    {
      "epoch": 4.715392772656823,
      "grad_norm": 0.002475195797160268,
      "learning_rate": 0.05344555250691827,
      "loss": 0.3072,
      "num_input_tokens_seen": 54772064,
      "step": 28905
    },
    {
      "epoch": 4.716208499877641,
      "grad_norm": 0.0019915495067834854,
      "learning_rate": 0.053400481179485086,
      "loss": 0.3587,
      "num_input_tokens_seen": 54781216,
      "step": 28910
    },
    {
      "epoch": 4.717024227098459,
      "grad_norm": 0.0012409668415784836,
      "learning_rate": 0.05335542474891159,
      "loss": 0.3143,
      "num_input_tokens_seen": 54792016,
      "step": 28915
    },
    {
      "epoch": 4.717839954319276,
      "grad_norm": 0.0017520713154226542,
      "learning_rate": 0.053310383222146124,
      "loss": 0.3108,
      "num_input_tokens_seen": 54801008,
      "step": 28920
    },
    {
      "epoch": 4.718655681540093,
      "grad_norm": 0.002269704593345523,
      "learning_rate": 0.053265356606134684,
      "loss": 0.3469,
      "num_input_tokens_seen": 54809728,
      "step": 28925
    },
    {
      "epoch": 4.71947140876091,
      "grad_norm": 0.0006255783955566585,
      "learning_rate": 0.053220344907820856,
      "loss": 0.3224,
      "num_input_tokens_seen": 54819312,
      "step": 28930
    },
    {
      "epoch": 4.720287135981728,
      "grad_norm": 0.00297523383051157,
      "learning_rate": 0.05317534813414608,
      "loss": 0.3706,
      "num_input_tokens_seen": 54827232,
      "step": 28935
    },
    {
      "epoch": 4.721102863202545,
      "grad_norm": 0.001593793393112719,
      "learning_rate": 0.05313036629204942,
      "loss": 0.3027,
      "num_input_tokens_seen": 54836560,
      "step": 28940
    },
    {
      "epoch": 4.721918590423362,
      "grad_norm": 0.001438878825865686,
      "learning_rate": 0.05308539938846756,
      "loss": 0.3638,
      "num_input_tokens_seen": 54845344,
      "step": 28945
    },
    {
      "epoch": 4.72273431764418,
      "grad_norm": 0.0014217318966984749,
      "learning_rate": 0.05304044743033507,
      "loss": 0.3722,
      "num_input_tokens_seen": 54854944,
      "step": 28950
    },
    {
      "epoch": 4.723550044864997,
      "grad_norm": 0.0018565870122984052,
      "learning_rate": 0.05299551042458401,
      "loss": 0.3151,
      "num_input_tokens_seen": 54865600,
      "step": 28955
    },
    {
      "epoch": 4.7243657720858145,
      "grad_norm": 0.0008394873002544045,
      "learning_rate": 0.052950588378144266,
      "loss": 0.3322,
      "num_input_tokens_seen": 54875200,
      "step": 28960
    },
    {
      "epoch": 4.725181499306632,
      "grad_norm": 0.0006762784905731678,
      "learning_rate": 0.052905681297943465,
      "loss": 0.3296,
      "num_input_tokens_seen": 54884816,
      "step": 28965
    },
    {
      "epoch": 4.725997226527449,
      "grad_norm": 0.001631910796277225,
      "learning_rate": 0.0528607891909067,
      "loss": 0.3133,
      "num_input_tokens_seen": 54895152,
      "step": 28970
    },
    {
      "epoch": 4.726812953748267,
      "grad_norm": 0.0004659230471588671,
      "learning_rate": 0.05281591206395697,
      "loss": 0.3332,
      "num_input_tokens_seen": 54904608,
      "step": 28975
    },
    {
      "epoch": 4.727628680969084,
      "grad_norm": 0.0014828555285930634,
      "learning_rate": 0.05277104992401496,
      "loss": 0.3308,
      "num_input_tokens_seen": 54913472,
      "step": 28980
    },
    {
      "epoch": 4.728444408189901,
      "grad_norm": 0.0019571350421756506,
      "learning_rate": 0.05272620277799884,
      "loss": 0.3836,
      "num_input_tokens_seen": 54923488,
      "step": 28985
    },
    {
      "epoch": 4.729260135410719,
      "grad_norm": 0.0012880477588623762,
      "learning_rate": 0.05268137063282473,
      "loss": 0.2876,
      "num_input_tokens_seen": 54933088,
      "step": 28990
    },
    {
      "epoch": 4.730075862631536,
      "grad_norm": 0.0013102497905492783,
      "learning_rate": 0.0526365534954062,
      "loss": 0.3255,
      "num_input_tokens_seen": 54942704,
      "step": 28995
    },
    {
      "epoch": 4.730891589852353,
      "grad_norm": 0.003078958485275507,
      "learning_rate": 0.052591751372654656,
      "loss": 0.3425,
      "num_input_tokens_seen": 54952896,
      "step": 29000
    },
    {
      "epoch": 4.730891589852353,
      "eval_loss": 0.31999385356903076,
      "eval_runtime": 152.8415,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 54952896,
      "step": 29000
    },
    {
      "epoch": 4.7317073170731705,
      "grad_norm": 0.001654192223213613,
      "learning_rate": 0.05254696427147921,
      "loss": 0.3261,
      "num_input_tokens_seen": 54962928,
      "step": 29005
    },
    {
      "epoch": 4.732523044293988,
      "grad_norm": 0.0004826868243981153,
      "learning_rate": 0.052502192198786546,
      "loss": 0.2959,
      "num_input_tokens_seen": 54972384,
      "step": 29010
    },
    {
      "epoch": 4.733338771514806,
      "grad_norm": 0.0007829679525457323,
      "learning_rate": 0.05245743516148103,
      "loss": 0.3533,
      "num_input_tokens_seen": 54980384,
      "step": 29015
    },
    {
      "epoch": 4.734154498735623,
      "grad_norm": 0.0006754975183866918,
      "learning_rate": 0.05241269316646486,
      "loss": 0.3047,
      "num_input_tokens_seen": 54989936,
      "step": 29020
    },
    {
      "epoch": 4.73497022595644,
      "grad_norm": 0.0014766304520890117,
      "learning_rate": 0.052367966220637725,
      "loss": 0.3132,
      "num_input_tokens_seen": 54999024,
      "step": 29025
    },
    {
      "epoch": 4.735785953177258,
      "grad_norm": 0.0010387294460088015,
      "learning_rate": 0.05232325433089716,
      "loss": 0.353,
      "num_input_tokens_seen": 55007968,
      "step": 29030
    },
    {
      "epoch": 4.736601680398075,
      "grad_norm": 0.002891459269449115,
      "learning_rate": 0.052278557504138214,
      "loss": 0.3686,
      "num_input_tokens_seen": 55016432,
      "step": 29035
    },
    {
      "epoch": 4.737417407618892,
      "grad_norm": 0.0035894981119781733,
      "learning_rate": 0.05223387574725372,
      "loss": 0.343,
      "num_input_tokens_seen": 55026272,
      "step": 29040
    },
    {
      "epoch": 4.738233134839709,
      "grad_norm": 0.0009724499541334808,
      "learning_rate": 0.05218920906713428,
      "loss": 0.3383,
      "num_input_tokens_seen": 55036160,
      "step": 29045
    },
    {
      "epoch": 4.739048862060527,
      "grad_norm": 0.0013854356948286295,
      "learning_rate": 0.05214455747066789,
      "loss": 0.3213,
      "num_input_tokens_seen": 55044096,
      "step": 29050
    },
    {
      "epoch": 4.7398645892813445,
      "grad_norm": 0.0035818470641970634,
      "learning_rate": 0.05209992096474048,
      "loss": 0.3851,
      "num_input_tokens_seen": 55053104,
      "step": 29055
    },
    {
      "epoch": 4.740680316502162,
      "grad_norm": 0.0016993086319416761,
      "learning_rate": 0.05205529955623559,
      "loss": 0.3291,
      "num_input_tokens_seen": 55063776,
      "step": 29060
    },
    {
      "epoch": 4.741496043722979,
      "grad_norm": 0.0012581867631524801,
      "learning_rate": 0.052010693252034314,
      "loss": 0.3325,
      "num_input_tokens_seen": 55072400,
      "step": 29065
    },
    {
      "epoch": 4.742311770943797,
      "grad_norm": 0.0005067292368039489,
      "learning_rate": 0.0519661020590156,
      "loss": 0.3476,
      "num_input_tokens_seen": 55081840,
      "step": 29070
    },
    {
      "epoch": 4.743127498164614,
      "grad_norm": 0.001852752291597426,
      "learning_rate": 0.05192152598405586,
      "loss": 0.3161,
      "num_input_tokens_seen": 55090048,
      "step": 29075
    },
    {
      "epoch": 4.743943225385431,
      "grad_norm": 0.00045433573541231453,
      "learning_rate": 0.05187696503402941,
      "loss": 0.3127,
      "num_input_tokens_seen": 55099584,
      "step": 29080
    },
    {
      "epoch": 4.744758952606248,
      "grad_norm": 0.0004897768958471715,
      "learning_rate": 0.05183241921580798,
      "loss": 0.3592,
      "num_input_tokens_seen": 55109472,
      "step": 29085
    },
    {
      "epoch": 4.745574679827066,
      "grad_norm": 0.0012283219257369637,
      "learning_rate": 0.051787888536261206,
      "loss": 0.3526,
      "num_input_tokens_seen": 55119584,
      "step": 29090
    },
    {
      "epoch": 4.746390407047883,
      "grad_norm": 0.0010926412651315331,
      "learning_rate": 0.051743373002256184,
      "loss": 0.3227,
      "num_input_tokens_seen": 55128208,
      "step": 29095
    },
    {
      "epoch": 4.7472061342687,
      "grad_norm": 0.0019251239718869328,
      "learning_rate": 0.05169887262065787,
      "loss": 0.3564,
      "num_input_tokens_seen": 55137200,
      "step": 29100
    },
    {
      "epoch": 4.7480218614895175,
      "grad_norm": 0.0009537881705909967,
      "learning_rate": 0.051654387398328665,
      "loss": 0.3025,
      "num_input_tokens_seen": 55147552,
      "step": 29105
    },
    {
      "epoch": 4.748837588710336,
      "grad_norm": 0.0011467968579381704,
      "learning_rate": 0.05160991734212888,
      "loss": 0.3366,
      "num_input_tokens_seen": 55156720,
      "step": 29110
    },
    {
      "epoch": 4.749653315931153,
      "grad_norm": 0.0007410558173432946,
      "learning_rate": 0.051565462458916224,
      "loss": 0.2988,
      "num_input_tokens_seen": 55165872,
      "step": 29115
    },
    {
      "epoch": 4.75046904315197,
      "grad_norm": 0.002092484151944518,
      "learning_rate": 0.05152102275554627,
      "loss": 0.3094,
      "num_input_tokens_seen": 55176288,
      "step": 29120
    },
    {
      "epoch": 4.751284770372787,
      "grad_norm": 0.0012315964559093118,
      "learning_rate": 0.05147659823887222,
      "loss": 0.3357,
      "num_input_tokens_seen": 55186096,
      "step": 29125
    },
    {
      "epoch": 4.752100497593605,
      "grad_norm": 0.0016554499743506312,
      "learning_rate": 0.05143218891574479,
      "loss": 0.2927,
      "num_input_tokens_seen": 55194896,
      "step": 29130
    },
    {
      "epoch": 4.752916224814422,
      "grad_norm": 0.0017122963909059763,
      "learning_rate": 0.0513877947930125,
      "loss": 0.3806,
      "num_input_tokens_seen": 55204272,
      "step": 29135
    },
    {
      "epoch": 4.753731952035239,
      "grad_norm": 0.002497059525921941,
      "learning_rate": 0.051343415877521566,
      "loss": 0.3512,
      "num_input_tokens_seen": 55213792,
      "step": 29140
    },
    {
      "epoch": 4.754547679256056,
      "grad_norm": 0.00125660072080791,
      "learning_rate": 0.051299052176115634,
      "loss": 0.3286,
      "num_input_tokens_seen": 55224560,
      "step": 29145
    },
    {
      "epoch": 4.755363406476874,
      "grad_norm": 0.0023520528338849545,
      "learning_rate": 0.051254703695636256,
      "loss": 0.3515,
      "num_input_tokens_seen": 55235328,
      "step": 29150
    },
    {
      "epoch": 4.7561791336976915,
      "grad_norm": 0.0023402287624776363,
      "learning_rate": 0.05121037044292249,
      "loss": 0.3282,
      "num_input_tokens_seen": 55244640,
      "step": 29155
    },
    {
      "epoch": 4.756994860918509,
      "grad_norm": 0.0015003436710685492,
      "learning_rate": 0.05116605242481101,
      "loss": 0.3117,
      "num_input_tokens_seen": 55253904,
      "step": 29160
    },
    {
      "epoch": 4.757810588139327,
      "grad_norm": 0.0006537421140819788,
      "learning_rate": 0.05112174964813634,
      "loss": 0.2988,
      "num_input_tokens_seen": 55261872,
      "step": 29165
    },
    {
      "epoch": 4.758626315360144,
      "grad_norm": 0.0011073012137785554,
      "learning_rate": 0.05107746211973038,
      "loss": 0.3177,
      "num_input_tokens_seen": 55271488,
      "step": 29170
    },
    {
      "epoch": 4.759442042580961,
      "grad_norm": 0.0022731339558959007,
      "learning_rate": 0.05103318984642291,
      "loss": 0.354,
      "num_input_tokens_seen": 55281808,
      "step": 29175
    },
    {
      "epoch": 4.760257769801778,
      "grad_norm": 0.0034005502238869667,
      "learning_rate": 0.05098893283504131,
      "loss": 0.3598,
      "num_input_tokens_seen": 55290896,
      "step": 29180
    },
    {
      "epoch": 4.761073497022595,
      "grad_norm": 0.0016405654605478048,
      "learning_rate": 0.050944691092410475,
      "loss": 0.3558,
      "num_input_tokens_seen": 55300544,
      "step": 29185
    },
    {
      "epoch": 4.761889224243413,
      "grad_norm": 0.0018975440179929137,
      "learning_rate": 0.05090046462535313,
      "loss": 0.2989,
      "num_input_tokens_seen": 55309264,
      "step": 29190
    },
    {
      "epoch": 4.76270495146423,
      "grad_norm": 0.0011373980669304729,
      "learning_rate": 0.050856253440689454,
      "loss": 0.3135,
      "num_input_tokens_seen": 55318976,
      "step": 29195
    },
    {
      "epoch": 4.7635206786850475,
      "grad_norm": 0.0012563364580273628,
      "learning_rate": 0.050812057545237405,
      "loss": 0.236,
      "num_input_tokens_seen": 55327776,
      "step": 29200
    },
    {
      "epoch": 4.7635206786850475,
      "eval_loss": 0.32140159606933594,
      "eval_runtime": 152.8384,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 55327776,
      "step": 29200
    },
    {
      "epoch": 4.7643364059058655,
      "grad_norm": 0.0022407874930649996,
      "learning_rate": 0.0507678769458126,
      "loss": 0.2672,
      "num_input_tokens_seen": 55337408,
      "step": 29205
    },
    {
      "epoch": 4.765152133126683,
      "grad_norm": 0.0015597735764458776,
      "learning_rate": 0.050723711649228155,
      "loss": 0.3483,
      "num_input_tokens_seen": 55346176,
      "step": 29210
    },
    {
      "epoch": 4.7659678603475,
      "grad_norm": 0.001973523525521159,
      "learning_rate": 0.05067956166229496,
      "loss": 0.3609,
      "num_input_tokens_seen": 55355168,
      "step": 29215
    },
    {
      "epoch": 4.766783587568317,
      "grad_norm": 0.0014047384029254317,
      "learning_rate": 0.05063542699182155,
      "loss": 0.3233,
      "num_input_tokens_seen": 55364736,
      "step": 29220
    },
    {
      "epoch": 4.767599314789135,
      "grad_norm": 0.0015921044396236539,
      "learning_rate": 0.050591307644613996,
      "loss": 0.349,
      "num_input_tokens_seen": 55374032,
      "step": 29225
    },
    {
      "epoch": 4.768415042009952,
      "grad_norm": 0.0010310437064617872,
      "learning_rate": 0.05054720362747599,
      "loss": 0.3213,
      "num_input_tokens_seen": 55383072,
      "step": 29230
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 0.0013386313803493977,
      "learning_rate": 0.050503114947209035,
      "loss": 0.3112,
      "num_input_tokens_seen": 55393520,
      "step": 29235
    },
    {
      "epoch": 4.770046496451586,
      "grad_norm": 0.0011356977047398686,
      "learning_rate": 0.05045904161061207,
      "loss": 0.3189,
      "num_input_tokens_seen": 55402368,
      "step": 29240
    },
    {
      "epoch": 4.770862223672404,
      "grad_norm": 0.0019708985928446054,
      "learning_rate": 0.05041498362448185,
      "loss": 0.2997,
      "num_input_tokens_seen": 55411776,
      "step": 29245
    },
    {
      "epoch": 4.771677950893221,
      "grad_norm": 0.0025338472332805395,
      "learning_rate": 0.05037094099561256,
      "loss": 0.3638,
      "num_input_tokens_seen": 55420416,
      "step": 29250
    },
    {
      "epoch": 4.772493678114039,
      "grad_norm": 0.001294090528972447,
      "learning_rate": 0.05032691373079624,
      "loss": 0.3188,
      "num_input_tokens_seen": 55429888,
      "step": 29255
    },
    {
      "epoch": 4.773309405334856,
      "grad_norm": 0.00264756684191525,
      "learning_rate": 0.05028290183682234,
      "loss": 0.3275,
      "num_input_tokens_seen": 55438480,
      "step": 29260
    },
    {
      "epoch": 4.774125132555674,
      "grad_norm": 0.0016150899464264512,
      "learning_rate": 0.050238905320478096,
      "loss": 0.3554,
      "num_input_tokens_seen": 55447312,
      "step": 29265
    },
    {
      "epoch": 4.774940859776491,
      "grad_norm": 0.0028880215249955654,
      "learning_rate": 0.05019492418854838,
      "loss": 0.3838,
      "num_input_tokens_seen": 55458128,
      "step": 29270
    },
    {
      "epoch": 4.775756586997308,
      "grad_norm": 0.0014397623017430305,
      "learning_rate": 0.05015095844781554,
      "loss": 0.3058,
      "num_input_tokens_seen": 55467952,
      "step": 29275
    },
    {
      "epoch": 4.776572314218125,
      "grad_norm": 0.0007152184844017029,
      "learning_rate": 0.05010700810505968,
      "loss": 0.311,
      "num_input_tokens_seen": 55478144,
      "step": 29280
    },
    {
      "epoch": 4.777388041438943,
      "grad_norm": 0.002025113208219409,
      "learning_rate": 0.05006307316705856,
      "loss": 0.3121,
      "num_input_tokens_seen": 55487680,
      "step": 29285
    },
    {
      "epoch": 4.77820376865976,
      "grad_norm": 0.002137542236596346,
      "learning_rate": 0.0500191536405874,
      "loss": 0.2977,
      "num_input_tokens_seen": 55496480,
      "step": 29290
    },
    {
      "epoch": 4.779019495880577,
      "grad_norm": 0.0019332042429596186,
      "learning_rate": 0.04997524953241922,
      "loss": 0.2611,
      "num_input_tokens_seen": 55505888,
      "step": 29295
    },
    {
      "epoch": 4.7798352231013945,
      "grad_norm": 0.002424317877739668,
      "learning_rate": 0.049931360849324556,
      "loss": 0.3251,
      "num_input_tokens_seen": 55514928,
      "step": 29300
    },
    {
      "epoch": 4.7806509503222125,
      "grad_norm": 0.0005014113849028945,
      "learning_rate": 0.04988748759807155,
      "loss": 0.2713,
      "num_input_tokens_seen": 55524016,
      "step": 29305
    },
    {
      "epoch": 4.78146667754303,
      "grad_norm": 0.0029785833321511745,
      "learning_rate": 0.0498436297854261,
      "loss": 0.3175,
      "num_input_tokens_seen": 55533920,
      "step": 29310
    },
    {
      "epoch": 4.782282404763847,
      "grad_norm": 0.0020021051168441772,
      "learning_rate": 0.04979978741815152,
      "loss": 0.3641,
      "num_input_tokens_seen": 55543984,
      "step": 29315
    },
    {
      "epoch": 4.783098131984664,
      "grad_norm": 0.001596230547875166,
      "learning_rate": 0.04975596050300891,
      "loss": 0.3403,
      "num_input_tokens_seen": 55552384,
      "step": 29320
    },
    {
      "epoch": 4.783913859205482,
      "grad_norm": 0.004443667829036713,
      "learning_rate": 0.049712149046757005,
      "loss": 0.3605,
      "num_input_tokens_seen": 55560768,
      "step": 29325
    },
    {
      "epoch": 4.784729586426299,
      "grad_norm": 0.002079237485304475,
      "learning_rate": 0.04966835305615194,
      "loss": 0.3464,
      "num_input_tokens_seen": 55571168,
      "step": 29330
    },
    {
      "epoch": 4.785545313647116,
      "grad_norm": 0.0011958929244428873,
      "learning_rate": 0.049624572537947755,
      "loss": 0.3475,
      "num_input_tokens_seen": 55580384,
      "step": 29335
    },
    {
      "epoch": 4.786361040867934,
      "grad_norm": 0.0007936533074826002,
      "learning_rate": 0.04958080749889582,
      "loss": 0.3391,
      "num_input_tokens_seen": 55590976,
      "step": 29340
    },
    {
      "epoch": 4.787176768088751,
      "grad_norm": 0.002405182458460331,
      "learning_rate": 0.049537057945745304,
      "loss": 0.3439,
      "num_input_tokens_seen": 55601760,
      "step": 29345
    },
    {
      "epoch": 4.7879924953095685,
      "grad_norm": 0.002224984811618924,
      "learning_rate": 0.049493323885243,
      "loss": 0.3164,
      "num_input_tokens_seen": 55612448,
      "step": 29350
    },
    {
      "epoch": 4.788808222530386,
      "grad_norm": 0.0020680022425949574,
      "learning_rate": 0.04944960532413318,
      "loss": 0.2847,
      "num_input_tokens_seen": 55622080,
      "step": 29355
    },
    {
      "epoch": 4.789623949751203,
      "grad_norm": 0.0023233687970787287,
      "learning_rate": 0.049405902269157774,
      "loss": 0.3188,
      "num_input_tokens_seen": 55632096,
      "step": 29360
    },
    {
      "epoch": 4.790439676972021,
      "grad_norm": 0.0010463300859555602,
      "learning_rate": 0.04936221472705646,
      "loss": 0.3269,
      "num_input_tokens_seen": 55641024,
      "step": 29365
    },
    {
      "epoch": 4.791255404192838,
      "grad_norm": 0.0005591387744061649,
      "learning_rate": 0.04931854270456632,
      "loss": 0.3274,
      "num_input_tokens_seen": 55651296,
      "step": 29370
    },
    {
      "epoch": 4.792071131413655,
      "grad_norm": 0.0016581842210143805,
      "learning_rate": 0.049274886208422075,
      "loss": 0.2809,
      "num_input_tokens_seen": 55660240,
      "step": 29375
    },
    {
      "epoch": 4.792886858634473,
      "grad_norm": 0.0011147318873554468,
      "learning_rate": 0.049231245245356235,
      "loss": 0.3342,
      "num_input_tokens_seen": 55669904,
      "step": 29380
    },
    {
      "epoch": 4.79370258585529,
      "grad_norm": 0.0022344617173075676,
      "learning_rate": 0.049187619822098655,
      "loss": 0.3708,
      "num_input_tokens_seen": 55679584,
      "step": 29385
    },
    {
      "epoch": 4.794518313076107,
      "grad_norm": 0.001048921956680715,
      "learning_rate": 0.04914400994537705,
      "loss": 0.2808,
      "num_input_tokens_seen": 55689552,
      "step": 29390
    },
    {
      "epoch": 4.7953340402969244,
      "grad_norm": 0.0008054478676058352,
      "learning_rate": 0.049100415621916485,
      "loss": 0.3594,
      "num_input_tokens_seen": 55700864,
      "step": 29395
    },
    {
      "epoch": 4.796149767517742,
      "grad_norm": 0.001428085844963789,
      "learning_rate": 0.04905683685843981,
      "loss": 0.3071,
      "num_input_tokens_seen": 55708896,
      "step": 29400
    },
    {
      "epoch": 4.796149767517742,
      "eval_loss": 0.32077479362487793,
      "eval_runtime": 152.8739,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 55708896,
      "step": 29400
    },
    {
      "epoch": 4.79696549473856,
      "grad_norm": 0.0026172520592808723,
      "learning_rate": 0.049013273661667495,
      "loss": 0.2813,
      "num_input_tokens_seen": 55717488,
      "step": 29405
    },
    {
      "epoch": 4.797781221959377,
      "grad_norm": 0.0011066921288147569,
      "learning_rate": 0.048969726038317396,
      "loss": 0.2864,
      "num_input_tokens_seen": 55726992,
      "step": 29410
    },
    {
      "epoch": 4.798596949180194,
      "grad_norm": 0.0007394663407467306,
      "learning_rate": 0.048926193995105206,
      "loss": 0.3638,
      "num_input_tokens_seen": 55735056,
      "step": 29415
    },
    {
      "epoch": 4.799412676401012,
      "grad_norm": 0.0011866504792124033,
      "learning_rate": 0.048882677538744035,
      "loss": 0.2478,
      "num_input_tokens_seen": 55744816,
      "step": 29420
    },
    {
      "epoch": 4.800228403621829,
      "grad_norm": 0.0015130118699744344,
      "learning_rate": 0.048839176675944715,
      "loss": 0.2926,
      "num_input_tokens_seen": 55754848,
      "step": 29425
    },
    {
      "epoch": 4.801044130842646,
      "grad_norm": 0.001141851767897606,
      "learning_rate": 0.04879569141341566,
      "loss": 0.4054,
      "num_input_tokens_seen": 55763536,
      "step": 29430
    },
    {
      "epoch": 4.801859858063463,
      "grad_norm": 0.0019417348084971309,
      "learning_rate": 0.04875222175786274,
      "loss": 0.3001,
      "num_input_tokens_seen": 55771808,
      "step": 29435
    },
    {
      "epoch": 4.802675585284281,
      "grad_norm": 0.0025957729667425156,
      "learning_rate": 0.04870876771598966,
      "loss": 0.3571,
      "num_input_tokens_seen": 55781216,
      "step": 29440
    },
    {
      "epoch": 4.803491312505098,
      "grad_norm": 0.0008673366974107921,
      "learning_rate": 0.04866532929449744,
      "loss": 0.3352,
      "num_input_tokens_seen": 55790064,
      "step": 29445
    },
    {
      "epoch": 4.8043070397259156,
      "grad_norm": 0.0011509628966450691,
      "learning_rate": 0.048621906500084945,
      "loss": 0.2285,
      "num_input_tokens_seen": 55800112,
      "step": 29450
    },
    {
      "epoch": 4.805122766946733,
      "grad_norm": 0.0015133413253352046,
      "learning_rate": 0.04857849933944845,
      "loss": 0.3568,
      "num_input_tokens_seen": 55810272,
      "step": 29455
    },
    {
      "epoch": 4.805938494167551,
      "grad_norm": 0.0030429132748395205,
      "learning_rate": 0.048535107819281866,
      "loss": 0.3429,
      "num_input_tokens_seen": 55819296,
      "step": 29460
    },
    {
      "epoch": 4.806754221388368,
      "grad_norm": 0.0006931597017683089,
      "learning_rate": 0.04849173194627675,
      "loss": 0.2685,
      "num_input_tokens_seen": 55829472,
      "step": 29465
    },
    {
      "epoch": 4.807569948609185,
      "grad_norm": 0.001956766238436103,
      "learning_rate": 0.04844837172712223,
      "loss": 0.2636,
      "num_input_tokens_seen": 55840240,
      "step": 29470
    },
    {
      "epoch": 4.808385675830002,
      "grad_norm": 0.0024972991086542606,
      "learning_rate": 0.04840502716850494,
      "loss": 0.2672,
      "num_input_tokens_seen": 55849920,
      "step": 29475
    },
    {
      "epoch": 4.80920140305082,
      "grad_norm": 0.0027244791854172945,
      "learning_rate": 0.04836169827710916,
      "loss": 0.3046,
      "num_input_tokens_seen": 55859408,
      "step": 29480
    },
    {
      "epoch": 4.810017130271637,
      "grad_norm": 0.0038387237582355738,
      "learning_rate": 0.04831838505961684,
      "loss": 0.3315,
      "num_input_tokens_seen": 55869712,
      "step": 29485
    },
    {
      "epoch": 4.810832857492454,
      "grad_norm": 0.002381819300353527,
      "learning_rate": 0.048275087522707295,
      "loss": 0.3915,
      "num_input_tokens_seen": 55878592,
      "step": 29490
    },
    {
      "epoch": 4.8116485847132715,
      "grad_norm": 0.0024553982075303793,
      "learning_rate": 0.04823180567305766,
      "loss": 0.2721,
      "num_input_tokens_seen": 55888352,
      "step": 29495
    },
    {
      "epoch": 4.8124643119340895,
      "grad_norm": 0.004052952863276005,
      "learning_rate": 0.04818853951734244,
      "loss": 0.2914,
      "num_input_tokens_seen": 55898160,
      "step": 29500
    },
    {
      "epoch": 4.813280039154907,
      "grad_norm": 0.0014071351615712047,
      "learning_rate": 0.04814528906223387,
      "loss": 0.3634,
      "num_input_tokens_seen": 55907456,
      "step": 29505
    },
    {
      "epoch": 4.814095766375724,
      "grad_norm": 0.0017432688036933541,
      "learning_rate": 0.04810205431440177,
      "loss": 0.3719,
      "num_input_tokens_seen": 55916800,
      "step": 29510
    },
    {
      "epoch": 4.814911493596542,
      "grad_norm": 0.0015203169314190745,
      "learning_rate": 0.04805883528051341,
      "loss": 0.3297,
      "num_input_tokens_seen": 55926848,
      "step": 29515
    },
    {
      "epoch": 4.815727220817359,
      "grad_norm": 0.0004657046520151198,
      "learning_rate": 0.048015631967233685,
      "loss": 0.31,
      "num_input_tokens_seen": 55935520,
      "step": 29520
    },
    {
      "epoch": 4.816542948038176,
      "grad_norm": 0.0028939677868038416,
      "learning_rate": 0.04797244438122517,
      "loss": 0.3632,
      "num_input_tokens_seen": 55944912,
      "step": 29525
    },
    {
      "epoch": 4.817358675258993,
      "grad_norm": 0.0012298537185415626,
      "learning_rate": 0.04792927252914784,
      "loss": 0.3864,
      "num_input_tokens_seen": 55954112,
      "step": 29530
    },
    {
      "epoch": 4.81817440247981,
      "grad_norm": 0.0015863211592659354,
      "learning_rate": 0.04788611641765944,
      "loss": 0.3236,
      "num_input_tokens_seen": 55964608,
      "step": 29535
    },
    {
      "epoch": 4.818990129700628,
      "grad_norm": 0.0008033155463635921,
      "learning_rate": 0.04784297605341508,
      "loss": 0.3385,
      "num_input_tokens_seen": 55974176,
      "step": 29540
    },
    {
      "epoch": 4.8198058569214455,
      "grad_norm": 0.002233153907582164,
      "learning_rate": 0.04779985144306761,
      "loss": 0.307,
      "num_input_tokens_seen": 55983120,
      "step": 29545
    },
    {
      "epoch": 4.820621584142263,
      "grad_norm": 0.001734959427267313,
      "learning_rate": 0.047756742593267405,
      "loss": 0.3302,
      "num_input_tokens_seen": 55992048,
      "step": 29550
    },
    {
      "epoch": 4.821437311363081,
      "grad_norm": 0.0016161978710442781,
      "learning_rate": 0.047713649510662315,
      "loss": 0.3318,
      "num_input_tokens_seen": 56001344,
      "step": 29555
    },
    {
      "epoch": 4.822253038583898,
      "grad_norm": 0.00043678845395334065,
      "learning_rate": 0.04767057220189789,
      "loss": 0.2524,
      "num_input_tokens_seen": 56010912,
      "step": 29560
    },
    {
      "epoch": 4.823068765804715,
      "grad_norm": 0.0008214301196858287,
      "learning_rate": 0.04762751067361722,
      "loss": 0.3793,
      "num_input_tokens_seen": 56020272,
      "step": 29565
    },
    {
      "epoch": 4.823884493025532,
      "grad_norm": 0.0032326558139175177,
      "learning_rate": 0.04758446493246086,
      "loss": 0.3491,
      "num_input_tokens_seen": 56028864,
      "step": 29570
    },
    {
      "epoch": 4.824700220246349,
      "grad_norm": 0.0011280914768576622,
      "learning_rate": 0.047541434985067084,
      "loss": 0.3271,
      "num_input_tokens_seen": 56038320,
      "step": 29575
    },
    {
      "epoch": 4.825515947467167,
      "grad_norm": 0.004570401273667812,
      "learning_rate": 0.047498420838071556,
      "loss": 0.3856,
      "num_input_tokens_seen": 56047568,
      "step": 29580
    },
    {
      "epoch": 4.826331674687984,
      "grad_norm": 0.0018829436739906669,
      "learning_rate": 0.04745542249810772,
      "loss": 0.3399,
      "num_input_tokens_seen": 56056640,
      "step": 29585
    },
    {
      "epoch": 4.827147401908801,
      "grad_norm": 0.0022903624922037125,
      "learning_rate": 0.047412439971806324,
      "loss": 0.3641,
      "num_input_tokens_seen": 56066592,
      "step": 29590
    },
    {
      "epoch": 4.8279631291296194,
      "grad_norm": 0.0016166121931746602,
      "learning_rate": 0.04736947326579592,
      "loss": 0.2941,
      "num_input_tokens_seen": 56075616,
      "step": 29595
    },
    {
      "epoch": 4.828778856350437,
      "grad_norm": 0.00048573300591669977,
      "learning_rate": 0.04732652238670245,
      "loss": 0.3007,
      "num_input_tokens_seen": 56085712,
      "step": 29600
    },
    {
      "epoch": 4.828778856350437,
      "eval_loss": 0.3212120532989502,
      "eval_runtime": 152.8919,
      "eval_samples_per_second": 17.823,
      "eval_steps_per_second": 8.915,
      "num_input_tokens_seen": 56085712,
      "step": 29600
    },
    {
      "epoch": 4.829594583571254,
      "grad_norm": 0.0018413508078083396,
      "learning_rate": 0.04728358734114952,
      "loss": 0.385,
      "num_input_tokens_seen": 56094528,
      "step": 29605
    },
    {
      "epoch": 4.830410310792071,
      "grad_norm": 0.00099561910610646,
      "learning_rate": 0.04724066813575821,
      "loss": 0.3458,
      "num_input_tokens_seen": 56104304,
      "step": 29610
    },
    {
      "epoch": 4.831226038012889,
      "grad_norm": 0.0007923799566924572,
      "learning_rate": 0.04719776477714729,
      "loss": 0.3214,
      "num_input_tokens_seen": 56113328,
      "step": 29615
    },
    {
      "epoch": 4.832041765233706,
      "grad_norm": 0.001471766154281795,
      "learning_rate": 0.047154877271932856,
      "loss": 0.3169,
      "num_input_tokens_seen": 56122880,
      "step": 29620
    },
    {
      "epoch": 4.832857492454523,
      "grad_norm": 0.0021711380686610937,
      "learning_rate": 0.0471120056267288,
      "loss": 0.3555,
      "num_input_tokens_seen": 56133248,
      "step": 29625
    },
    {
      "epoch": 4.83367321967534,
      "grad_norm": 0.002514887833967805,
      "learning_rate": 0.047069149848146495,
      "loss": 0.379,
      "num_input_tokens_seen": 56142928,
      "step": 29630
    },
    {
      "epoch": 4.834488946896158,
      "grad_norm": 0.0018594280118122697,
      "learning_rate": 0.04702630994279473,
      "loss": 0.3222,
      "num_input_tokens_seen": 56152064,
      "step": 29635
    },
    {
      "epoch": 4.835304674116975,
      "grad_norm": 0.0014293310232460499,
      "learning_rate": 0.046983485917280035,
      "loss": 0.3351,
      "num_input_tokens_seen": 56160752,
      "step": 29640
    },
    {
      "epoch": 4.8361204013377925,
      "grad_norm": 0.0009254944743588567,
      "learning_rate": 0.04694067777820644,
      "loss": 0.3149,
      "num_input_tokens_seen": 56171184,
      "step": 29645
    },
    {
      "epoch": 4.83693612855861,
      "grad_norm": 0.0017633689567446709,
      "learning_rate": 0.046897885532175415,
      "loss": 0.3147,
      "num_input_tokens_seen": 56181456,
      "step": 29650
    },
    {
      "epoch": 4.837751855779428,
      "grad_norm": 0.002166440011933446,
      "learning_rate": 0.04685510918578613,
      "loss": 0.3855,
      "num_input_tokens_seen": 56191312,
      "step": 29655
    },
    {
      "epoch": 4.838567583000245,
      "grad_norm": 0.0013859485043212771,
      "learning_rate": 0.04681234874563519,
      "loss": 0.3253,
      "num_input_tokens_seen": 56199584,
      "step": 29660
    },
    {
      "epoch": 4.839383310221062,
      "grad_norm": 0.0003547222586348653,
      "learning_rate": 0.046769604218316836,
      "loss": 0.316,
      "num_input_tokens_seen": 56209904,
      "step": 29665
    },
    {
      "epoch": 4.840199037441879,
      "grad_norm": 0.000918340461794287,
      "learning_rate": 0.04672687561042279,
      "loss": 0.3173,
      "num_input_tokens_seen": 56219536,
      "step": 29670
    },
    {
      "epoch": 4.841014764662697,
      "grad_norm": 0.0008636603015474975,
      "learning_rate": 0.046684162928542286,
      "loss": 0.3225,
      "num_input_tokens_seen": 56229456,
      "step": 29675
    },
    {
      "epoch": 4.841830491883514,
      "grad_norm": 0.0010078969644382596,
      "learning_rate": 0.04664146617926222,
      "loss": 0.313,
      "num_input_tokens_seen": 56240528,
      "step": 29680
    },
    {
      "epoch": 4.842646219104331,
      "grad_norm": 0.001730246702209115,
      "learning_rate": 0.046598785369167,
      "loss": 0.3182,
      "num_input_tokens_seen": 56250464,
      "step": 29685
    },
    {
      "epoch": 4.8434619463251485,
      "grad_norm": 0.0013483189977705479,
      "learning_rate": 0.046556120504838434,
      "loss": 0.32,
      "num_input_tokens_seen": 56260640,
      "step": 29690
    },
    {
      "epoch": 4.8442776735459665,
      "grad_norm": 0.0014520747354254127,
      "learning_rate": 0.04651347159285609,
      "loss": 0.3681,
      "num_input_tokens_seen": 56271136,
      "step": 29695
    },
    {
      "epoch": 4.845093400766784,
      "grad_norm": 0.0014100857079029083,
      "learning_rate": 0.04647083863979688,
      "loss": 0.3654,
      "num_input_tokens_seen": 56280768,
      "step": 29700
    },
    {
      "epoch": 4.845909127987601,
      "grad_norm": 0.0004604186106007546,
      "learning_rate": 0.04642822165223538,
      "loss": 0.3538,
      "num_input_tokens_seen": 56290240,
      "step": 29705
    },
    {
      "epoch": 4.846724855208418,
      "grad_norm": 0.0015669586136937141,
      "learning_rate": 0.046385620636743716,
      "loss": 0.3709,
      "num_input_tokens_seen": 56299568,
      "step": 29710
    },
    {
      "epoch": 4.847540582429236,
      "grad_norm": 0.0010561378439888358,
      "learning_rate": 0.04634303559989141,
      "loss": 0.3628,
      "num_input_tokens_seen": 56309216,
      "step": 29715
    },
    {
      "epoch": 4.848356309650053,
      "grad_norm": 0.0006981531041674316,
      "learning_rate": 0.046300466548245635,
      "loss": 0.3205,
      "num_input_tokens_seen": 56318112,
      "step": 29720
    },
    {
      "epoch": 4.84917203687087,
      "grad_norm": 0.001251096953637898,
      "learning_rate": 0.04625791348837114,
      "loss": 0.2706,
      "num_input_tokens_seen": 56327056,
      "step": 29725
    },
    {
      "epoch": 4.849987764091688,
      "grad_norm": 0.0008271997794508934,
      "learning_rate": 0.046215376426830095,
      "loss": 0.3027,
      "num_input_tokens_seen": 56337264,
      "step": 29730
    },
    {
      "epoch": 4.850803491312505,
      "grad_norm": 0.0008713201386854053,
      "learning_rate": 0.04617285537018219,
      "loss": 0.3268,
      "num_input_tokens_seen": 56345632,
      "step": 29735
    },
    {
      "epoch": 4.8516192185333225,
      "grad_norm": 0.001871545915491879,
      "learning_rate": 0.046130350324984803,
      "loss": 0.289,
      "num_input_tokens_seen": 56354592,
      "step": 29740
    },
    {
      "epoch": 4.85243494575414,
      "grad_norm": 0.0032256098929792643,
      "learning_rate": 0.046087861297792666,
      "loss": 0.349,
      "num_input_tokens_seen": 56364784,
      "step": 29745
    },
    {
      "epoch": 4.853250672974957,
      "grad_norm": 0.0021813453640788794,
      "learning_rate": 0.0460453882951582,
      "loss": 0.3186,
      "num_input_tokens_seen": 56374576,
      "step": 29750
    },
    {
      "epoch": 4.854066400195775,
      "grad_norm": 0.0014946490991860628,
      "learning_rate": 0.04600293132363119,
      "loss": 0.3497,
      "num_input_tokens_seen": 56384048,
      "step": 29755
    },
    {
      "epoch": 4.854882127416592,
      "grad_norm": 0.0016221445985138416,
      "learning_rate": 0.045960490389759086,
      "loss": 0.3008,
      "num_input_tokens_seen": 56394336,
      "step": 29760
    },
    {
      "epoch": 4.855697854637409,
      "grad_norm": 0.0017727226950228214,
      "learning_rate": 0.04591806550008685,
      "loss": 0.3162,
      "num_input_tokens_seen": 56403392,
      "step": 29765
    },
    {
      "epoch": 4.856513581858227,
      "grad_norm": 0.0008579805726185441,
      "learning_rate": 0.045875656661156825,
      "loss": 0.362,
      "num_input_tokens_seen": 56412384,
      "step": 29770
    },
    {
      "epoch": 4.857329309079044,
      "grad_norm": 0.00047703980817459524,
      "learning_rate": 0.04583326387950911,
      "loss": 0.3094,
      "num_input_tokens_seen": 56420896,
      "step": 29775
    },
    {
      "epoch": 4.858145036299861,
      "grad_norm": 0.0020895814523100853,
      "learning_rate": 0.0457908871616811,
      "loss": 0.3558,
      "num_input_tokens_seen": 56429888,
      "step": 29780
    },
    {
      "epoch": 4.858960763520678,
      "grad_norm": 0.0011943912832066417,
      "learning_rate": 0.04574852651420786,
      "loss": 0.2871,
      "num_input_tokens_seen": 56438880,
      "step": 29785
    },
    {
      "epoch": 4.859776490741496,
      "grad_norm": 0.0014997560065239668,
      "learning_rate": 0.045706181943621985,
      "loss": 0.2973,
      "num_input_tokens_seen": 56448576,
      "step": 29790
    },
    {
      "epoch": 4.8605922179623136,
      "grad_norm": 0.003532164264470339,
      "learning_rate": 0.04566385345645344,
      "loss": 0.3069,
      "num_input_tokens_seen": 56457312,
      "step": 29795
    },
    {
      "epoch": 4.861407945183131,
      "grad_norm": 0.0020333612337708473,
      "learning_rate": 0.04562154105922993,
      "loss": 0.3351,
      "num_input_tokens_seen": 56467376,
      "step": 29800
    },
    {
      "epoch": 4.861407945183131,
      "eval_loss": 0.3226664066314697,
      "eval_runtime": 152.9084,
      "eval_samples_per_second": 17.821,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 56467376,
      "step": 29800
    },
    {
      "epoch": 4.862223672403948,
      "grad_norm": 0.0035721866879612207,
      "learning_rate": 0.04557924475847642,
      "loss": 0.3435,
      "num_input_tokens_seen": 56477680,
      "step": 29805
    },
    {
      "epoch": 4.863039399624766,
      "grad_norm": 0.000670538458507508,
      "learning_rate": 0.04553696456071567,
      "loss": 0.347,
      "num_input_tokens_seen": 56486080,
      "step": 29810
    },
    {
      "epoch": 4.863855126845583,
      "grad_norm": 0.0007427093805745244,
      "learning_rate": 0.045494700472467724,
      "loss": 0.3145,
      "num_input_tokens_seen": 56496144,
      "step": 29815
    },
    {
      "epoch": 4.8646708540664,
      "grad_norm": 0.0019448620732873678,
      "learning_rate": 0.04545245250025024,
      "loss": 0.3293,
      "num_input_tokens_seen": 56505984,
      "step": 29820
    },
    {
      "epoch": 4.865486581287217,
      "grad_norm": 0.002049435628578067,
      "learning_rate": 0.045410220650578384,
      "loss": 0.3158,
      "num_input_tokens_seen": 56515408,
      "step": 29825
    },
    {
      "epoch": 4.866302308508035,
      "grad_norm": 0.0017538727261126041,
      "learning_rate": 0.04536800492996492,
      "loss": 0.3314,
      "num_input_tokens_seen": 56525824,
      "step": 29830
    },
    {
      "epoch": 4.867118035728852,
      "grad_norm": 0.002753821900114417,
      "learning_rate": 0.04532580534491994,
      "loss": 0.3229,
      "num_input_tokens_seen": 56535360,
      "step": 29835
    },
    {
      "epoch": 4.8679337629496695,
      "grad_norm": 0.0008051609620451927,
      "learning_rate": 0.045283621901951183,
      "loss": 0.3482,
      "num_input_tokens_seen": 56544688,
      "step": 29840
    },
    {
      "epoch": 4.868749490170487,
      "grad_norm": 0.0048537771217525005,
      "learning_rate": 0.04524145460756393,
      "loss": 0.3739,
      "num_input_tokens_seen": 56554720,
      "step": 29845
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.0022382002789527178,
      "learning_rate": 0.045199303468260794,
      "loss": 0.3769,
      "num_input_tokens_seen": 56563392,
      "step": 29850
    },
    {
      "epoch": 4.870380944612122,
      "grad_norm": 0.0023454835172742605,
      "learning_rate": 0.04515716849054214,
      "loss": 0.3427,
      "num_input_tokens_seen": 56572816,
      "step": 29855
    },
    {
      "epoch": 4.871196671832939,
      "grad_norm": 0.0018552739638835192,
      "learning_rate": 0.04511504968090558,
      "loss": 0.3405,
      "num_input_tokens_seen": 56582848,
      "step": 29860
    },
    {
      "epoch": 4.872012399053756,
      "grad_norm": 0.002834703540429473,
      "learning_rate": 0.04507294704584644,
      "loss": 0.2813,
      "num_input_tokens_seen": 56592640,
      "step": 29865
    },
    {
      "epoch": 4.872828126274574,
      "grad_norm": 0.0025157276540994644,
      "learning_rate": 0.04503086059185749,
      "loss": 0.2935,
      "num_input_tokens_seen": 56601120,
      "step": 29870
    },
    {
      "epoch": 4.873643853495391,
      "grad_norm": 0.0009630626300349832,
      "learning_rate": 0.04498879032542893,
      "loss": 0.3501,
      "num_input_tokens_seen": 56610240,
      "step": 29875
    },
    {
      "epoch": 4.874459580716208,
      "grad_norm": 0.0006785034784115851,
      "learning_rate": 0.0449467362530486,
      "loss": 0.3272,
      "num_input_tokens_seen": 56620048,
      "step": 29880
    },
    {
      "epoch": 4.8752753079370255,
      "grad_norm": 0.0023502849508076906,
      "learning_rate": 0.04490469838120171,
      "loss": 0.3104,
      "num_input_tokens_seen": 56629456,
      "step": 29885
    },
    {
      "epoch": 4.8760910351578435,
      "grad_norm": 0.0007051785360090435,
      "learning_rate": 0.04486267671637101,
      "loss": 0.3361,
      "num_input_tokens_seen": 56637616,
      "step": 29890
    },
    {
      "epoch": 4.876906762378661,
      "grad_norm": 0.0025608879514038563,
      "learning_rate": 0.04482067126503683,
      "loss": 0.3611,
      "num_input_tokens_seen": 56645936,
      "step": 29895
    },
    {
      "epoch": 4.877722489599478,
      "grad_norm": 0.0009188076364807785,
      "learning_rate": 0.04477868203367687,
      "loss": 0.3687,
      "num_input_tokens_seen": 56655008,
      "step": 29900
    },
    {
      "epoch": 4.878538216820296,
      "grad_norm": 0.0006307153380475938,
      "learning_rate": 0.044736709028766426,
      "loss": 0.305,
      "num_input_tokens_seen": 56664032,
      "step": 29905
    },
    {
      "epoch": 4.879353944041113,
      "grad_norm": 0.0012314143823459744,
      "learning_rate": 0.04469475225677832,
      "loss": 0.3258,
      "num_input_tokens_seen": 56673056,
      "step": 29910
    },
    {
      "epoch": 4.88016967126193,
      "grad_norm": 0.0007588387234136462,
      "learning_rate": 0.04465281172418273,
      "loss": 0.3424,
      "num_input_tokens_seen": 56682704,
      "step": 29915
    },
    {
      "epoch": 4.880985398482747,
      "grad_norm": 0.001312937238253653,
      "learning_rate": 0.044610887437447476,
      "loss": 0.323,
      "num_input_tokens_seen": 56692768,
      "step": 29920
    },
    {
      "epoch": 4.881801125703564,
      "grad_norm": 0.0014621251029893756,
      "learning_rate": 0.044568979403037744,
      "loss": 0.3283,
      "num_input_tokens_seen": 56701248,
      "step": 29925
    },
    {
      "epoch": 4.882616852924382,
      "grad_norm": 0.0013917089672759175,
      "learning_rate": 0.04452708762741631,
      "loss": 0.3573,
      "num_input_tokens_seen": 56710752,
      "step": 29930
    },
    {
      "epoch": 4.883432580145199,
      "grad_norm": 0.002209973521530628,
      "learning_rate": 0.044485212117043475,
      "loss": 0.3692,
      "num_input_tokens_seen": 56719504,
      "step": 29935
    },
    {
      "epoch": 4.884248307366017,
      "grad_norm": 0.00131662015337497,
      "learning_rate": 0.04444335287837687,
      "loss": 0.3144,
      "num_input_tokens_seen": 56728800,
      "step": 29940
    },
    {
      "epoch": 4.885064034586835,
      "grad_norm": 0.0026082354597747326,
      "learning_rate": 0.04440150991787179,
      "loss": 0.3334,
      "num_input_tokens_seen": 56738736,
      "step": 29945
    },
    {
      "epoch": 4.885879761807652,
      "grad_norm": 0.0015289733419194818,
      "learning_rate": 0.04435968324198088,
      "loss": 0.2939,
      "num_input_tokens_seen": 56748576,
      "step": 29950
    },
    {
      "epoch": 4.886695489028469,
      "grad_norm": 0.0013371832901611924,
      "learning_rate": 0.04431787285715442,
      "loss": 0.32,
      "num_input_tokens_seen": 56756896,
      "step": 29955
    },
    {
      "epoch": 4.887511216249286,
      "grad_norm": 0.0019074120791628957,
      "learning_rate": 0.04427607876984004,
      "loss": 0.299,
      "num_input_tokens_seen": 56767008,
      "step": 29960
    },
    {
      "epoch": 4.888326943470103,
      "grad_norm": 0.0013171707978472114,
      "learning_rate": 0.044234300986482886,
      "loss": 0.3302,
      "num_input_tokens_seen": 56775376,
      "step": 29965
    },
    {
      "epoch": 4.889142670690921,
      "grad_norm": 0.0007311620865948498,
      "learning_rate": 0.04419253951352566,
      "loss": 0.2615,
      "num_input_tokens_seen": 56784688,
      "step": 29970
    },
    {
      "epoch": 4.889958397911738,
      "grad_norm": 0.0026101511903107166,
      "learning_rate": 0.044150794357408533,
      "loss": 0.2947,
      "num_input_tokens_seen": 56793392,
      "step": 29975
    },
    {
      "epoch": 4.890774125132555,
      "grad_norm": 0.0020665680058300495,
      "learning_rate": 0.044109065524569065,
      "loss": 0.3711,
      "num_input_tokens_seen": 56803296,
      "step": 29980
    },
    {
      "epoch": 4.891589852353373,
      "grad_norm": 0.0024282769300043583,
      "learning_rate": 0.0440673530214424,
      "loss": 0.3365,
      "num_input_tokens_seen": 56813328,
      "step": 29985
    },
    {
      "epoch": 4.8924055795741905,
      "grad_norm": 0.0005550139467231929,
      "learning_rate": 0.04402565685446117,
      "loss": 0.3414,
      "num_input_tokens_seen": 56823136,
      "step": 29990
    },
    {
      "epoch": 4.893221306795008,
      "grad_norm": 0.0020704297348856926,
      "learning_rate": 0.04398397703005536,
      "loss": 0.324,
      "num_input_tokens_seen": 56832704,
      "step": 29995
    },
    {
      "epoch": 4.894037034015825,
      "grad_norm": 0.0014695459976792336,
      "learning_rate": 0.043942313554652626,
      "loss": 0.3147,
      "num_input_tokens_seen": 56841328,
      "step": 30000
    },
    {
      "epoch": 4.894037034015825,
      "eval_loss": 0.3313918113708496,
      "eval_runtime": 152.9554,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 56841328,
      "step": 30000
    },
    {
      "epoch": 4.894852761236643,
      "grad_norm": 0.0018231152789667249,
      "learning_rate": 0.0439006664346779,
      "loss": 0.3431,
      "num_input_tokens_seen": 56849184,
      "step": 30005
    },
    {
      "epoch": 4.89566848845746,
      "grad_norm": 0.0032845211680978537,
      "learning_rate": 0.043859035676553755,
      "loss": 0.312,
      "num_input_tokens_seen": 56859440,
      "step": 30010
    },
    {
      "epoch": 4.896484215678277,
      "grad_norm": 0.0022691888734698296,
      "learning_rate": 0.043817421286700194,
      "loss": 0.3083,
      "num_input_tokens_seen": 56869808,
      "step": 30015
    },
    {
      "epoch": 4.897299942899094,
      "grad_norm": 0.0017694453708827496,
      "learning_rate": 0.043775823271534585,
      "loss": 0.3592,
      "num_input_tokens_seen": 56880768,
      "step": 30020
    },
    {
      "epoch": 4.898115670119912,
      "grad_norm": 0.003047656500712037,
      "learning_rate": 0.04373424163747197,
      "loss": 0.3285,
      "num_input_tokens_seen": 56891360,
      "step": 30025
    },
    {
      "epoch": 4.898931397340729,
      "grad_norm": 0.001612030086107552,
      "learning_rate": 0.04369267639092473,
      "loss": 0.3229,
      "num_input_tokens_seen": 56901872,
      "step": 30030
    },
    {
      "epoch": 4.8997471245615465,
      "grad_norm": 0.0007701890426687896,
      "learning_rate": 0.04365112753830268,
      "loss": 0.2954,
      "num_input_tokens_seen": 56911504,
      "step": 30035
    },
    {
      "epoch": 4.900562851782364,
      "grad_norm": 0.0020274771377444267,
      "learning_rate": 0.04360959508601327,
      "loss": 0.291,
      "num_input_tokens_seen": 56921408,
      "step": 30040
    },
    {
      "epoch": 4.901378579003182,
      "grad_norm": 0.0023212512023746967,
      "learning_rate": 0.04356807904046123,
      "loss": 0.3285,
      "num_input_tokens_seen": 56931472,
      "step": 30045
    },
    {
      "epoch": 4.902194306223999,
      "grad_norm": 0.003494130913168192,
      "learning_rate": 0.04352657940804892,
      "loss": 0.3167,
      "num_input_tokens_seen": 56940720,
      "step": 30050
    },
    {
      "epoch": 4.903010033444816,
      "grad_norm": 0.0009091430110856891,
      "learning_rate": 0.04348509619517613,
      "loss": 0.2922,
      "num_input_tokens_seen": 56949616,
      "step": 30055
    },
    {
      "epoch": 4.903825760665633,
      "grad_norm": 0.0011071801418438554,
      "learning_rate": 0.04344362940824002,
      "loss": 0.2513,
      "num_input_tokens_seen": 56958384,
      "step": 30060
    },
    {
      "epoch": 4.904641487886451,
      "grad_norm": 0.002801492577418685,
      "learning_rate": 0.04340217905363533,
      "loss": 0.333,
      "num_input_tokens_seen": 56967952,
      "step": 30065
    },
    {
      "epoch": 4.905457215107268,
      "grad_norm": 0.0014171421062201262,
      "learning_rate": 0.04336074513775425,
      "loss": 0.312,
      "num_input_tokens_seen": 56977872,
      "step": 30070
    },
    {
      "epoch": 4.906272942328085,
      "grad_norm": 0.0006402244907803833,
      "learning_rate": 0.04331932766698636,
      "loss": 0.3351,
      "num_input_tokens_seen": 56987552,
      "step": 30075
    },
    {
      "epoch": 4.907088669548903,
      "grad_norm": 0.0009318115771748126,
      "learning_rate": 0.0432779266477188,
      "loss": 0.39,
      "num_input_tokens_seen": 56996304,
      "step": 30080
    },
    {
      "epoch": 4.9079043967697205,
      "grad_norm": 0.0011451395694166422,
      "learning_rate": 0.04323654208633607,
      "loss": 0.2902,
      "num_input_tokens_seen": 57006448,
      "step": 30085
    },
    {
      "epoch": 4.908720123990538,
      "grad_norm": 0.0013227510498836637,
      "learning_rate": 0.04319517398922024,
      "loss": 0.3362,
      "num_input_tokens_seen": 57015776,
      "step": 30090
    },
    {
      "epoch": 4.909535851211355,
      "grad_norm": 0.0014572390355169773,
      "learning_rate": 0.04315382236275079,
      "loss": 0.3245,
      "num_input_tokens_seen": 57025152,
      "step": 30095
    },
    {
      "epoch": 4.910351578432172,
      "grad_norm": 0.0027419328689575195,
      "learning_rate": 0.043112487213304664,
      "loss": 0.2929,
      "num_input_tokens_seen": 57035120,
      "step": 30100
    },
    {
      "epoch": 4.91116730565299,
      "grad_norm": 0.0023697151336818933,
      "learning_rate": 0.04307116854725618,
      "loss": 0.3353,
      "num_input_tokens_seen": 57045648,
      "step": 30105
    },
    {
      "epoch": 4.911983032873807,
      "grad_norm": 0.0022127993870526552,
      "learning_rate": 0.043029866370977325,
      "loss": 0.3803,
      "num_input_tokens_seen": 57056144,
      "step": 30110
    },
    {
      "epoch": 4.912798760094624,
      "grad_norm": 0.0019427292281761765,
      "learning_rate": 0.04298858069083728,
      "loss": 0.3249,
      "num_input_tokens_seen": 57066288,
      "step": 30115
    },
    {
      "epoch": 4.913614487315442,
      "grad_norm": 0.001829467248171568,
      "learning_rate": 0.04294731151320295,
      "loss": 0.3034,
      "num_input_tokens_seen": 57076096,
      "step": 30120
    },
    {
      "epoch": 4.914430214536259,
      "grad_norm": 0.0018330714665353298,
      "learning_rate": 0.04290605884443841,
      "loss": 0.3288,
      "num_input_tokens_seen": 57085360,
      "step": 30125
    },
    {
      "epoch": 4.915245941757076,
      "grad_norm": 0.0018693687161430717,
      "learning_rate": 0.04286482269090545,
      "loss": 0.3372,
      "num_input_tokens_seen": 57096048,
      "step": 30130
    },
    {
      "epoch": 4.9160616689778935,
      "grad_norm": 0.001590260653756559,
      "learning_rate": 0.04282360305896323,
      "loss": 0.3803,
      "num_input_tokens_seen": 57106368,
      "step": 30135
    },
    {
      "epoch": 4.916877396198711,
      "grad_norm": 0.00166442827321589,
      "learning_rate": 0.04278239995496822,
      "loss": 0.3432,
      "num_input_tokens_seen": 57114512,
      "step": 30140
    },
    {
      "epoch": 4.917693123419529,
      "grad_norm": 0.0006510148523375392,
      "learning_rate": 0.042741213385274514,
      "loss": 0.3609,
      "num_input_tokens_seen": 57124432,
      "step": 30145
    },
    {
      "epoch": 4.918508850640346,
      "grad_norm": 0.002181397983804345,
      "learning_rate": 0.04270004335623366,
      "loss": 0.3503,
      "num_input_tokens_seen": 57134320,
      "step": 30150
    },
    {
      "epoch": 4.919324577861163,
      "grad_norm": 0.0011069083120673895,
      "learning_rate": 0.04265888987419448,
      "loss": 0.3608,
      "num_input_tokens_seen": 57144352,
      "step": 30155
    },
    {
      "epoch": 4.920140305081981,
      "grad_norm": 0.003236863063648343,
      "learning_rate": 0.04261775294550346,
      "loss": 0.3346,
      "num_input_tokens_seen": 57153424,
      "step": 30160
    },
    {
      "epoch": 4.920956032302798,
      "grad_norm": 0.001751481438986957,
      "learning_rate": 0.042576632576504354,
      "loss": 0.3382,
      "num_input_tokens_seen": 57162992,
      "step": 30165
    },
    {
      "epoch": 4.921771759523615,
      "grad_norm": 0.0005625418853014708,
      "learning_rate": 0.0425355287735385,
      "loss": 0.3034,
      "num_input_tokens_seen": 57171376,
      "step": 30170
    },
    {
      "epoch": 4.922587486744432,
      "grad_norm": 0.0012198669137433171,
      "learning_rate": 0.0424944415429446,
      "loss": 0.3343,
      "num_input_tokens_seen": 57181088,
      "step": 30175
    },
    {
      "epoch": 4.92340321396525,
      "grad_norm": 0.0023965290747582912,
      "learning_rate": 0.04245337089105877,
      "loss": 0.3807,
      "num_input_tokens_seen": 57189936,
      "step": 30180
    },
    {
      "epoch": 4.9242189411860675,
      "grad_norm": 0.0013745209435001016,
      "learning_rate": 0.04241231682421467,
      "loss": 0.368,
      "num_input_tokens_seen": 57198928,
      "step": 30185
    },
    {
      "epoch": 4.925034668406885,
      "grad_norm": 0.0028258352540433407,
      "learning_rate": 0.04237127934874337,
      "loss": 0.3058,
      "num_input_tokens_seen": 57208912,
      "step": 30190
    },
    {
      "epoch": 4.925850395627702,
      "grad_norm": 0.0026148774195462465,
      "learning_rate": 0.042330258470973305,
      "loss": 0.3489,
      "num_input_tokens_seen": 57217840,
      "step": 30195
    },
    {
      "epoch": 4.92666612284852,
      "grad_norm": 0.0013535538455471396,
      "learning_rate": 0.042289254197230515,
      "loss": 0.3661,
      "num_input_tokens_seen": 57227184,
      "step": 30200
    },
    {
      "epoch": 4.92666612284852,
      "eval_loss": 0.3220495581626892,
      "eval_runtime": 152.9155,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 57227184,
      "step": 30200
    },
    {
      "epoch": 4.927481850069337,
      "grad_norm": 0.0024199390318244696,
      "learning_rate": 0.04224826653383823,
      "loss": 0.3569,
      "num_input_tokens_seen": 57236192,
      "step": 30205
    },
    {
      "epoch": 4.928297577290154,
      "grad_norm": 0.0016338691348209977,
      "learning_rate": 0.04220729548711735,
      "loss": 0.2777,
      "num_input_tokens_seen": 57245776,
      "step": 30210
    },
    {
      "epoch": 4.929113304510971,
      "grad_norm": 0.0011293491115793586,
      "learning_rate": 0.04216634106338616,
      "loss": 0.3048,
      "num_input_tokens_seen": 57255408,
      "step": 30215
    },
    {
      "epoch": 4.929929031731789,
      "grad_norm": 0.0007478304323740304,
      "learning_rate": 0.04212540326896025,
      "loss": 0.2867,
      "num_input_tokens_seen": 57265296,
      "step": 30220
    },
    {
      "epoch": 4.930744758952606,
      "grad_norm": 0.0021655485033988953,
      "learning_rate": 0.0420844821101528,
      "loss": 0.3846,
      "num_input_tokens_seen": 57275024,
      "step": 30225
    },
    {
      "epoch": 4.9315604861734235,
      "grad_norm": 0.0018632466671988368,
      "learning_rate": 0.04204357759327441,
      "loss": 0.2467,
      "num_input_tokens_seen": 57284448,
      "step": 30230
    },
    {
      "epoch": 4.932376213394241,
      "grad_norm": 0.0012355414219200611,
      "learning_rate": 0.042002689724632954,
      "loss": 0.3136,
      "num_input_tokens_seen": 57294000,
      "step": 30235
    },
    {
      "epoch": 4.933191940615059,
      "grad_norm": 0.002021114807575941,
      "learning_rate": 0.04196181851053398,
      "loss": 0.2244,
      "num_input_tokens_seen": 57303536,
      "step": 30240
    },
    {
      "epoch": 4.934007667835876,
      "grad_norm": 0.003445785492658615,
      "learning_rate": 0.041920963957280295,
      "loss": 0.4051,
      "num_input_tokens_seen": 57311664,
      "step": 30245
    },
    {
      "epoch": 4.934823395056693,
      "grad_norm": 0.0008442035177722573,
      "learning_rate": 0.04188012607117212,
      "loss": 0.3604,
      "num_input_tokens_seen": 57320032,
      "step": 30250
    },
    {
      "epoch": 4.935639122277511,
      "grad_norm": 0.002250042511150241,
      "learning_rate": 0.04183930485850725,
      "loss": 0.3664,
      "num_input_tokens_seen": 57330288,
      "step": 30255
    },
    {
      "epoch": 4.936454849498328,
      "grad_norm": 0.001605390920303762,
      "learning_rate": 0.04179850032558078,
      "loss": 0.3385,
      "num_input_tokens_seen": 57340432,
      "step": 30260
    },
    {
      "epoch": 4.937270576719145,
      "grad_norm": 0.0021841763518750668,
      "learning_rate": 0.041757712478685295,
      "loss": 0.2753,
      "num_input_tokens_seen": 57350528,
      "step": 30265
    },
    {
      "epoch": 4.938086303939962,
      "grad_norm": 0.0009385370067320764,
      "learning_rate": 0.04171694132411085,
      "loss": 0.2888,
      "num_input_tokens_seen": 57360384,
      "step": 30270
    },
    {
      "epoch": 4.938902031160779,
      "grad_norm": 0.002317699370905757,
      "learning_rate": 0.04167618686814479,
      "loss": 0.3959,
      "num_input_tokens_seen": 57369056,
      "step": 30275
    },
    {
      "epoch": 4.939717758381597,
      "grad_norm": 0.0010685427114367485,
      "learning_rate": 0.041635449117072024,
      "loss": 0.3085,
      "num_input_tokens_seen": 57378672,
      "step": 30280
    },
    {
      "epoch": 4.940533485602415,
      "grad_norm": 0.0013365901540964842,
      "learning_rate": 0.04159472807717477,
      "loss": 0.3359,
      "num_input_tokens_seen": 57388944,
      "step": 30285
    },
    {
      "epoch": 4.941349212823232,
      "grad_norm": 0.0020420588552951813,
      "learning_rate": 0.041554023754732744,
      "loss": 0.3461,
      "num_input_tokens_seen": 57399168,
      "step": 30290
    },
    {
      "epoch": 4.94216494004405,
      "grad_norm": 0.0007367910584434867,
      "learning_rate": 0.04151333615602311,
      "loss": 0.3395,
      "num_input_tokens_seen": 57409264,
      "step": 30295
    },
    {
      "epoch": 4.942980667264867,
      "grad_norm": 0.0010867308592423797,
      "learning_rate": 0.04147266528732034,
      "loss": 0.3376,
      "num_input_tokens_seen": 57418400,
      "step": 30300
    },
    {
      "epoch": 4.943796394485684,
      "grad_norm": 0.0013558503706008196,
      "learning_rate": 0.0414320111548964,
      "loss": 0.3258,
      "num_input_tokens_seen": 57427824,
      "step": 30305
    },
    {
      "epoch": 4.944612121706501,
      "grad_norm": 0.0009222966036759317,
      "learning_rate": 0.04139137376502076,
      "loss": 0.3444,
      "num_input_tokens_seen": 57438176,
      "step": 30310
    },
    {
      "epoch": 4.945427848927318,
      "grad_norm": 0.0009065445046871901,
      "learning_rate": 0.04135075312396014,
      "loss": 0.3302,
      "num_input_tokens_seen": 57446816,
      "step": 30315
    },
    {
      "epoch": 4.946243576148136,
      "grad_norm": 0.001319280592724681,
      "learning_rate": 0.04131014923797875,
      "loss": 0.3464,
      "num_input_tokens_seen": 57456288,
      "step": 30320
    },
    {
      "epoch": 4.947059303368953,
      "grad_norm": 0.003121093614026904,
      "learning_rate": 0.04126956211333819,
      "loss": 0.2974,
      "num_input_tokens_seen": 57464640,
      "step": 30325
    },
    {
      "epoch": 4.9478750305897705,
      "grad_norm": 0.0020360699854791164,
      "learning_rate": 0.041228991756297545,
      "loss": 0.356,
      "num_input_tokens_seen": 57474272,
      "step": 30330
    },
    {
      "epoch": 4.9486907578105885,
      "grad_norm": 0.0020255721174180508,
      "learning_rate": 0.04118843817311332,
      "loss": 0.3573,
      "num_input_tokens_seen": 57484336,
      "step": 30335
    },
    {
      "epoch": 4.949506485031406,
      "grad_norm": 0.0010046691168099642,
      "learning_rate": 0.0411479013700393,
      "loss": 0.3129,
      "num_input_tokens_seen": 57492688,
      "step": 30340
    },
    {
      "epoch": 4.950322212252223,
      "grad_norm": 0.0007986745913513005,
      "learning_rate": 0.0411073813533268,
      "loss": 0.2602,
      "num_input_tokens_seen": 57501680,
      "step": 30345
    },
    {
      "epoch": 4.95113793947304,
      "grad_norm": 0.0015909703215584159,
      "learning_rate": 0.04106687812922456,
      "loss": 0.2829,
      "num_input_tokens_seen": 57510704,
      "step": 30350
    },
    {
      "epoch": 4.951953666693858,
      "grad_norm": 0.0013533469755202532,
      "learning_rate": 0.041026391703978635,
      "loss": 0.317,
      "num_input_tokens_seen": 57518976,
      "step": 30355
    },
    {
      "epoch": 4.952769393914675,
      "grad_norm": 0.0012248883722350001,
      "learning_rate": 0.04098592208383259,
      "loss": 0.3461,
      "num_input_tokens_seen": 57528896,
      "step": 30360
    },
    {
      "epoch": 4.953585121135492,
      "grad_norm": 0.003076091641560197,
      "learning_rate": 0.040945469275027256,
      "loss": 0.3092,
      "num_input_tokens_seen": 57537488,
      "step": 30365
    },
    {
      "epoch": 4.954400848356309,
      "grad_norm": 0.0029748387169092894,
      "learning_rate": 0.04090503328380104,
      "loss": 0.2942,
      "num_input_tokens_seen": 57547024,
      "step": 30370
    },
    {
      "epoch": 4.955216575577127,
      "grad_norm": 0.0022291657514870167,
      "learning_rate": 0.04086461411638971,
      "loss": 0.3382,
      "num_input_tokens_seen": 57556912,
      "step": 30375
    },
    {
      "epoch": 4.9560323027979445,
      "grad_norm": 0.001966622192412615,
      "learning_rate": 0.04082421177902631,
      "loss": 0.3233,
      "num_input_tokens_seen": 57566736,
      "step": 30380
    },
    {
      "epoch": 4.956848030018762,
      "grad_norm": 0.0016050196718424559,
      "learning_rate": 0.04078382627794149,
      "loss": 0.3372,
      "num_input_tokens_seen": 57577104,
      "step": 30385
    },
    {
      "epoch": 4.957663757239579,
      "grad_norm": 0.0031830938532948494,
      "learning_rate": 0.04074345761936316,
      "loss": 0.3446,
      "num_input_tokens_seen": 57587568,
      "step": 30390
    },
    {
      "epoch": 4.958479484460397,
      "grad_norm": 0.0029145898297429085,
      "learning_rate": 0.04070310580951663,
      "loss": 0.3455,
      "num_input_tokens_seen": 57596752,
      "step": 30395
    },
    {
      "epoch": 4.959295211681214,
      "grad_norm": 0.0033546360209584236,
      "learning_rate": 0.040662770854624726,
      "loss": 0.2419,
      "num_input_tokens_seen": 57605632,
      "step": 30400
    },
    {
      "epoch": 4.959295211681214,
      "eval_loss": 0.32583919167518616,
      "eval_runtime": 152.8618,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 57605632,
      "step": 30400
    },
    {
      "epoch": 4.960110938902031,
      "grad_norm": 0.0014429864240810275,
      "learning_rate": 0.040622452760907535,
      "loss": 0.3102,
      "num_input_tokens_seen": 57615824,
      "step": 30405
    },
    {
      "epoch": 4.960926666122848,
      "grad_norm": 0.00141154567245394,
      "learning_rate": 0.04058215153458265,
      "loss": 0.3168,
      "num_input_tokens_seen": 57625648,
      "step": 30410
    },
    {
      "epoch": 4.961742393343666,
      "grad_norm": 0.0010966198751702905,
      "learning_rate": 0.04054186718186507,
      "loss": 0.364,
      "num_input_tokens_seen": 57635456,
      "step": 30415
    },
    {
      "epoch": 4.962558120564483,
      "grad_norm": 0.0028302334249019623,
      "learning_rate": 0.04050159970896708,
      "loss": 0.3366,
      "num_input_tokens_seen": 57644448,
      "step": 30420
    },
    {
      "epoch": 4.9633738477853,
      "grad_norm": 0.003406164236366749,
      "learning_rate": 0.04046134912209843,
      "loss": 0.3285,
      "num_input_tokens_seen": 57653984,
      "step": 30425
    },
    {
      "epoch": 4.964189575006118,
      "grad_norm": 0.0016575108747929335,
      "learning_rate": 0.040421115427466354,
      "loss": 0.3551,
      "num_input_tokens_seen": 57663216,
      "step": 30430
    },
    {
      "epoch": 4.965005302226936,
      "grad_norm": 0.0016229995526373386,
      "learning_rate": 0.04038089863127529,
      "loss": 0.3117,
      "num_input_tokens_seen": 57672672,
      "step": 30435
    },
    {
      "epoch": 4.965821029447753,
      "grad_norm": 0.0008388226269744337,
      "learning_rate": 0.04034069873972727,
      "loss": 0.3291,
      "num_input_tokens_seen": 57682064,
      "step": 30440
    },
    {
      "epoch": 4.96663675666857,
      "grad_norm": 0.0028515553567558527,
      "learning_rate": 0.040300515759021514,
      "loss": 0.277,
      "num_input_tokens_seen": 57690752,
      "step": 30445
    },
    {
      "epoch": 4.967452483889387,
      "grad_norm": 0.001407794770784676,
      "learning_rate": 0.04026034969535478,
      "loss": 0.286,
      "num_input_tokens_seen": 57701824,
      "step": 30450
    },
    {
      "epoch": 4.968268211110205,
      "grad_norm": 0.001907889498397708,
      "learning_rate": 0.040220200554921266,
      "loss": 0.3099,
      "num_input_tokens_seen": 57710880,
      "step": 30455
    },
    {
      "epoch": 4.969083938331022,
      "grad_norm": 0.002440419513732195,
      "learning_rate": 0.0401800683439124,
      "loss": 0.3315,
      "num_input_tokens_seen": 57720112,
      "step": 30460
    },
    {
      "epoch": 4.969899665551839,
      "grad_norm": 0.0022199114318937063,
      "learning_rate": 0.04013995306851704,
      "loss": 0.3712,
      "num_input_tokens_seen": 57730384,
      "step": 30465
    },
    {
      "epoch": 4.970715392772657,
      "grad_norm": 0.002485531847923994,
      "learning_rate": 0.040099854734921545,
      "loss": 0.3957,
      "num_input_tokens_seen": 57739824,
      "step": 30470
    },
    {
      "epoch": 4.971531119993474,
      "grad_norm": 0.0009361174888908863,
      "learning_rate": 0.0400597733493095,
      "loss": 0.3607,
      "num_input_tokens_seen": 57749376,
      "step": 30475
    },
    {
      "epoch": 4.9723468472142915,
      "grad_norm": 0.0011100920382887125,
      "learning_rate": 0.04001970891786203,
      "loss": 0.2904,
      "num_input_tokens_seen": 57760080,
      "step": 30480
    },
    {
      "epoch": 4.973162574435109,
      "grad_norm": 0.0018397306557744741,
      "learning_rate": 0.03997966144675752,
      "loss": 0.273,
      "num_input_tokens_seen": 57769360,
      "step": 30485
    },
    {
      "epoch": 4.973978301655926,
      "grad_norm": 0.0025268462486565113,
      "learning_rate": 0.039939630942171796,
      "loss": 0.3312,
      "num_input_tokens_seen": 57778064,
      "step": 30490
    },
    {
      "epoch": 4.974794028876744,
      "grad_norm": 0.002969916444271803,
      "learning_rate": 0.03989961741027815,
      "loss": 0.3059,
      "num_input_tokens_seen": 57787008,
      "step": 30495
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 0.0016251642955467105,
      "learning_rate": 0.03985962085724704,
      "loss": 0.3313,
      "num_input_tokens_seen": 57796368,
      "step": 30500
    },
    {
      "epoch": 4.976425483318378,
      "grad_norm": 0.0030059106647968292,
      "learning_rate": 0.03981964128924656,
      "loss": 0.3312,
      "num_input_tokens_seen": 57804864,
      "step": 30505
    },
    {
      "epoch": 4.977241210539196,
      "grad_norm": 0.004038023762404919,
      "learning_rate": 0.03977967871244197,
      "loss": 0.3142,
      "num_input_tokens_seen": 57814160,
      "step": 30510
    },
    {
      "epoch": 4.978056937760013,
      "grad_norm": 0.0022593450266867876,
      "learning_rate": 0.03973973313299602,
      "loss": 0.3601,
      "num_input_tokens_seen": 57823136,
      "step": 30515
    },
    {
      "epoch": 4.97887266498083,
      "grad_norm": 0.0013217614032328129,
      "learning_rate": 0.0396998045570689,
      "loss": 0.3748,
      "num_input_tokens_seen": 57832640,
      "step": 30520
    },
    {
      "epoch": 4.9796883922016475,
      "grad_norm": 0.0016832289984449744,
      "learning_rate": 0.03965989299081798,
      "loss": 0.325,
      "num_input_tokens_seen": 57841408,
      "step": 30525
    },
    {
      "epoch": 4.9805041194224655,
      "grad_norm": 0.0026139961555600166,
      "learning_rate": 0.039619998440398235,
      "loss": 0.2992,
      "num_input_tokens_seen": 57850624,
      "step": 30530
    },
    {
      "epoch": 4.981319846643283,
      "grad_norm": 0.000700674019753933,
      "learning_rate": 0.03958012091196184,
      "loss": 0.3328,
      "num_input_tokens_seen": 57861008,
      "step": 30535
    },
    {
      "epoch": 4.9821355738641,
      "grad_norm": 0.001906593213789165,
      "learning_rate": 0.039540260411658396,
      "loss": 0.356,
      "num_input_tokens_seen": 57870624,
      "step": 30540
    },
    {
      "epoch": 4.982951301084917,
      "grad_norm": 0.002532272832468152,
      "learning_rate": 0.03950041694563496,
      "loss": 0.3666,
      "num_input_tokens_seen": 57880672,
      "step": 30545
    },
    {
      "epoch": 4.983767028305735,
      "grad_norm": 0.0021596765145659447,
      "learning_rate": 0.0394605905200358,
      "loss": 0.3233,
      "num_input_tokens_seen": 57890128,
      "step": 30550
    },
    {
      "epoch": 4.984582755526552,
      "grad_norm": 0.0017593051306903362,
      "learning_rate": 0.03942078114100272,
      "loss": 0.289,
      "num_input_tokens_seen": 57899760,
      "step": 30555
    },
    {
      "epoch": 4.985398482747369,
      "grad_norm": 0.0014264917699620128,
      "learning_rate": 0.03938098881467485,
      "loss": 0.3406,
      "num_input_tokens_seen": 57908624,
      "step": 30560
    },
    {
      "epoch": 4.986214209968186,
      "grad_norm": 0.0014180128928273916,
      "learning_rate": 0.039341213547188586,
      "loss": 0.3318,
      "num_input_tokens_seen": 57918112,
      "step": 30565
    },
    {
      "epoch": 4.987029937189004,
      "grad_norm": 0.0012559364549815655,
      "learning_rate": 0.03930145534467782,
      "loss": 0.3268,
      "num_input_tokens_seen": 57928864,
      "step": 30570
    },
    {
      "epoch": 4.9878456644098215,
      "grad_norm": 0.0004725331091322005,
      "learning_rate": 0.0392617142132738,
      "loss": 0.2897,
      "num_input_tokens_seen": 57938576,
      "step": 30575
    },
    {
      "epoch": 4.988661391630639,
      "grad_norm": 0.003134719096124172,
      "learning_rate": 0.03922199015910504,
      "loss": 0.3184,
      "num_input_tokens_seen": 57948272,
      "step": 30580
    },
    {
      "epoch": 4.989477118851456,
      "grad_norm": 0.0011160514550283551,
      "learning_rate": 0.039182283188297556,
      "loss": 0.308,
      "num_input_tokens_seen": 57957456,
      "step": 30585
    },
    {
      "epoch": 4.990292846072274,
      "grad_norm": 0.0010939299827441573,
      "learning_rate": 0.039142593306974595,
      "loss": 0.2611,
      "num_input_tokens_seen": 57967600,
      "step": 30590
    },
    {
      "epoch": 4.991108573293091,
      "grad_norm": 0.0015695412876084447,
      "learning_rate": 0.039102920521256856,
      "loss": 0.3416,
      "num_input_tokens_seen": 57977424,
      "step": 30595
    },
    {
      "epoch": 4.991924300513908,
      "grad_norm": 0.002423603320494294,
      "learning_rate": 0.03906326483726243,
      "loss": 0.3353,
      "num_input_tokens_seen": 57987472,
      "step": 30600
    },
    {
      "epoch": 4.991924300513908,
      "eval_loss": 0.31987160444259644,
      "eval_runtime": 152.921,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 57987472,
      "step": 30600
    },
    {
      "epoch": 4.992740027734725,
      "grad_norm": 0.001423046924173832,
      "learning_rate": 0.039023626261106704,
      "loss": 0.3603,
      "num_input_tokens_seen": 57996720,
      "step": 30605
    },
    {
      "epoch": 4.993555754955543,
      "grad_norm": 0.001670678611844778,
      "learning_rate": 0.03898400479890237,
      "loss": 0.3257,
      "num_input_tokens_seen": 58005568,
      "step": 30610
    },
    {
      "epoch": 4.99437148217636,
      "grad_norm": 0.003900732845067978,
      "learning_rate": 0.038944400456759655,
      "loss": 0.3142,
      "num_input_tokens_seen": 58015360,
      "step": 30615
    },
    {
      "epoch": 4.995187209397177,
      "grad_norm": 0.0006567695527337492,
      "learning_rate": 0.038904813240785964,
      "loss": 0.314,
      "num_input_tokens_seen": 58024480,
      "step": 30620
    },
    {
      "epoch": 4.9960029366179945,
      "grad_norm": 0.004243538249284029,
      "learning_rate": 0.03886524315708621,
      "loss": 0.4051,
      "num_input_tokens_seen": 58033104,
      "step": 30625
    },
    {
      "epoch": 4.996818663838813,
      "grad_norm": 0.0013705299934372306,
      "learning_rate": 0.03882569021176255,
      "loss": 0.4091,
      "num_input_tokens_seen": 58043456,
      "step": 30630
    },
    {
      "epoch": 4.99763439105963,
      "grad_norm": 0.0026466180570423603,
      "learning_rate": 0.038786154410914535,
      "loss": 0.3141,
      "num_input_tokens_seen": 58052752,
      "step": 30635
    },
    {
      "epoch": 4.998450118280447,
      "grad_norm": 0.0006228492711670697,
      "learning_rate": 0.03874663576063917,
      "loss": 0.2914,
      "num_input_tokens_seen": 58061056,
      "step": 30640
    },
    {
      "epoch": 4.999265845501265,
      "grad_norm": 0.0016532677691429853,
      "learning_rate": 0.038707134267030624,
      "loss": 0.3414,
      "num_input_tokens_seen": 58071584,
      "step": 30645
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0012444964377209544,
      "learning_rate": 0.038667649936180555,
      "loss": 0.3419,
      "num_input_tokens_seen": 58080480,
      "step": 30650
    },
    {
      "epoch": 5.000815727220817,
      "grad_norm": 0.002350974828004837,
      "learning_rate": 0.038628182774178,
      "loss": 0.3417,
      "num_input_tokens_seen": 58090336,
      "step": 30655
    },
    {
      "epoch": 5.001631454441635,
      "grad_norm": 0.0018081753514707088,
      "learning_rate": 0.038588732787109226,
      "loss": 0.3858,
      "num_input_tokens_seen": 58099552,
      "step": 30660
    },
    {
      "epoch": 5.002447181662452,
      "grad_norm": 0.001532127265818417,
      "learning_rate": 0.03854929998105795,
      "loss": 0.2922,
      "num_input_tokens_seen": 58109376,
      "step": 30665
    },
    {
      "epoch": 5.003262908883269,
      "grad_norm": 0.0009223315282724798,
      "learning_rate": 0.03850988436210518,
      "loss": 0.2879,
      "num_input_tokens_seen": 58119536,
      "step": 30670
    },
    {
      "epoch": 5.0040786361040865,
      "grad_norm": 0.0013138017384335399,
      "learning_rate": 0.03847048593632933,
      "loss": 0.3206,
      "num_input_tokens_seen": 58129616,
      "step": 30675
    },
    {
      "epoch": 5.004894363324905,
      "grad_norm": 0.0022742226719856262,
      "learning_rate": 0.038431104709806096,
      "loss": 0.3286,
      "num_input_tokens_seen": 58138848,
      "step": 30680
    },
    {
      "epoch": 5.005710090545722,
      "grad_norm": 0.0006857970729470253,
      "learning_rate": 0.0383917406886086,
      "loss": 0.29,
      "num_input_tokens_seen": 58148512,
      "step": 30685
    },
    {
      "epoch": 5.006525817766539,
      "grad_norm": 0.0015075387200340629,
      "learning_rate": 0.03835239387880722,
      "loss": 0.3327,
      "num_input_tokens_seen": 58157600,
      "step": 30690
    },
    {
      "epoch": 5.007341544987356,
      "grad_norm": 0.001767343608662486,
      "learning_rate": 0.03831306428646979,
      "loss": 0.3262,
      "num_input_tokens_seen": 58168528,
      "step": 30695
    },
    {
      "epoch": 5.008157272208174,
      "grad_norm": 0.0015115808928385377,
      "learning_rate": 0.03827375191766135,
      "loss": 0.3256,
      "num_input_tokens_seen": 58177632,
      "step": 30700
    },
    {
      "epoch": 5.008972999428991,
      "grad_norm": 0.0009649125277064741,
      "learning_rate": 0.03823445677844446,
      "loss": 0.3093,
      "num_input_tokens_seen": 58187488,
      "step": 30705
    },
    {
      "epoch": 5.009788726649808,
      "grad_norm": 0.0015244755195453763,
      "learning_rate": 0.03819517887487881,
      "loss": 0.3379,
      "num_input_tokens_seen": 58196432,
      "step": 30710
    },
    {
      "epoch": 5.010604453870625,
      "grad_norm": 0.0013232468627393246,
      "learning_rate": 0.03815591821302161,
      "loss": 0.3201,
      "num_input_tokens_seen": 58205072,
      "step": 30715
    },
    {
      "epoch": 5.011420181091443,
      "grad_norm": 0.001211792347021401,
      "learning_rate": 0.03811667479892739,
      "loss": 0.32,
      "num_input_tokens_seen": 58213488,
      "step": 30720
    },
    {
      "epoch": 5.0122359083122605,
      "grad_norm": 0.0014035172061994672,
      "learning_rate": 0.03807744863864788,
      "loss": 0.3588,
      "num_input_tokens_seen": 58223456,
      "step": 30725
    },
    {
      "epoch": 5.013051635533078,
      "grad_norm": 0.0009786408627405763,
      "learning_rate": 0.03803823973823229,
      "loss": 0.2925,
      "num_input_tokens_seen": 58232768,
      "step": 30730
    },
    {
      "epoch": 5.013867362753895,
      "grad_norm": 0.0003059319860767573,
      "learning_rate": 0.03799904810372719,
      "loss": 0.3055,
      "num_input_tokens_seen": 58242496,
      "step": 30735
    },
    {
      "epoch": 5.014683089974713,
      "grad_norm": 0.0007784460322000086,
      "learning_rate": 0.03795987374117632,
      "loss": 0.3268,
      "num_input_tokens_seen": 58252064,
      "step": 30740
    },
    {
      "epoch": 5.01549881719553,
      "grad_norm": 0.0010563331888988614,
      "learning_rate": 0.03792071665662093,
      "loss": 0.3458,
      "num_input_tokens_seen": 58261008,
      "step": 30745
    },
    {
      "epoch": 5.016314544416347,
      "grad_norm": 0.002546873642131686,
      "learning_rate": 0.03788157685609952,
      "loss": 0.2763,
      "num_input_tokens_seen": 58270912,
      "step": 30750
    },
    {
      "epoch": 5.017130271637164,
      "grad_norm": 0.0016959838103502989,
      "learning_rate": 0.037842454345647876,
      "loss": 0.3135,
      "num_input_tokens_seen": 58280784,
      "step": 30755
    },
    {
      "epoch": 5.017945998857982,
      "grad_norm": 0.0014855265617370605,
      "learning_rate": 0.03780334913129929,
      "loss": 0.3028,
      "num_input_tokens_seen": 58290768,
      "step": 30760
    },
    {
      "epoch": 5.018761726078799,
      "grad_norm": 0.0014230767264962196,
      "learning_rate": 0.037764261219084175,
      "loss": 0.2974,
      "num_input_tokens_seen": 58301584,
      "step": 30765
    },
    {
      "epoch": 5.0195774532996165,
      "grad_norm": 0.002604302018880844,
      "learning_rate": 0.037725190615030414,
      "loss": 0.3096,
      "num_input_tokens_seen": 58310304,
      "step": 30770
    },
    {
      "epoch": 5.020393180520434,
      "grad_norm": 0.0010338471038267016,
      "learning_rate": 0.037686137325163224,
      "loss": 0.3065,
      "num_input_tokens_seen": 58320224,
      "step": 30775
    },
    {
      "epoch": 5.021208907741252,
      "grad_norm": 0.002173066372051835,
      "learning_rate": 0.037647101355505065,
      "loss": 0.3112,
      "num_input_tokens_seen": 58329056,
      "step": 30780
    },
    {
      "epoch": 5.022024634962069,
      "grad_norm": 0.0011965486919507384,
      "learning_rate": 0.03760808271207581,
      "loss": 0.3695,
      "num_input_tokens_seen": 58338800,
      "step": 30785
    },
    {
      "epoch": 5.022840362182886,
      "grad_norm": 0.0007387250079773366,
      "learning_rate": 0.03756908140089258,
      "loss": 0.3891,
      "num_input_tokens_seen": 58348160,
      "step": 30790
    },
    {
      "epoch": 5.023656089403703,
      "grad_norm": 0.0006848479970358312,
      "learning_rate": 0.03753009742796989,
      "loss": 0.3354,
      "num_input_tokens_seen": 58359328,
      "step": 30795
    },
    {
      "epoch": 5.024471816624521,
      "grad_norm": 0.0010884515941143036,
      "learning_rate": 0.037491130799319615,
      "loss": 0.335,
      "num_input_tokens_seen": 58367056,
      "step": 30800
    },
    {
      "epoch": 5.024471816624521,
      "eval_loss": 0.3205883800983429,
      "eval_runtime": 152.8574,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 58367056,
      "step": 30800
    },
    {
      "epoch": 5.025287543845338,
      "grad_norm": 0.0024602606426924467,
      "learning_rate": 0.03745218152095079,
      "loss": 0.3257,
      "num_input_tokens_seen": 58376928,
      "step": 30805
    },
    {
      "epoch": 5.026103271066155,
      "grad_norm": 0.0011161700822412968,
      "learning_rate": 0.037413249598869935,
      "loss": 0.3235,
      "num_input_tokens_seen": 58385504,
      "step": 30810
    },
    {
      "epoch": 5.026918998286972,
      "grad_norm": 0.0013852599076926708,
      "learning_rate": 0.037374335039080886,
      "loss": 0.3677,
      "num_input_tokens_seen": 58395568,
      "step": 30815
    },
    {
      "epoch": 5.02773472550779,
      "grad_norm": 0.002159127267077565,
      "learning_rate": 0.037335437847584724,
      "loss": 0.3222,
      "num_input_tokens_seen": 58403744,
      "step": 30820
    },
    {
      "epoch": 5.028550452728608,
      "grad_norm": 0.0020922955591231585,
      "learning_rate": 0.03729655803037983,
      "loss": 0.2904,
      "num_input_tokens_seen": 58412720,
      "step": 30825
    },
    {
      "epoch": 5.029366179949425,
      "grad_norm": 0.00041907429113052785,
      "learning_rate": 0.03725769559346207,
      "loss": 0.3234,
      "num_input_tokens_seen": 58422128,
      "step": 30830
    },
    {
      "epoch": 5.030181907170242,
      "grad_norm": 0.002906300825998187,
      "learning_rate": 0.03721885054282439,
      "loss": 0.3549,
      "num_input_tokens_seen": 58429632,
      "step": 30835
    },
    {
      "epoch": 5.03099763439106,
      "grad_norm": 0.001635336084291339,
      "learning_rate": 0.03718002288445731,
      "loss": 0.3443,
      "num_input_tokens_seen": 58439344,
      "step": 30840
    },
    {
      "epoch": 5.031813361611877,
      "grad_norm": 0.00096034852322191,
      "learning_rate": 0.03714121262434844,
      "loss": 0.2848,
      "num_input_tokens_seen": 58449488,
      "step": 30845
    },
    {
      "epoch": 5.032629088832694,
      "grad_norm": 0.0024406432639807463,
      "learning_rate": 0.037102419768482844,
      "loss": 0.3761,
      "num_input_tokens_seen": 58458496,
      "step": 30850
    },
    {
      "epoch": 5.033444816053512,
      "grad_norm": 0.0016856850124895573,
      "learning_rate": 0.03706364432284293,
      "loss": 0.3806,
      "num_input_tokens_seen": 58467600,
      "step": 30855
    },
    {
      "epoch": 5.034260543274329,
      "grad_norm": 0.0015836047241464257,
      "learning_rate": 0.03702488629340828,
      "loss": 0.3294,
      "num_input_tokens_seen": 58477056,
      "step": 30860
    },
    {
      "epoch": 5.035076270495146,
      "grad_norm": 0.0011956486850976944,
      "learning_rate": 0.036986145686155915,
      "loss": 0.3232,
      "num_input_tokens_seen": 58487232,
      "step": 30865
    },
    {
      "epoch": 5.0358919977159635,
      "grad_norm": 0.0017277245642617345,
      "learning_rate": 0.036947422507060075,
      "loss": 0.3145,
      "num_input_tokens_seen": 58496592,
      "step": 30870
    },
    {
      "epoch": 5.0367077249367815,
      "grad_norm": 0.002035402925685048,
      "learning_rate": 0.0369087167620924,
      "loss": 0.3188,
      "num_input_tokens_seen": 58506960,
      "step": 30875
    },
    {
      "epoch": 5.037523452157599,
      "grad_norm": 0.0008472889894619584,
      "learning_rate": 0.03687002845722183,
      "loss": 0.3457,
      "num_input_tokens_seen": 58515920,
      "step": 30880
    },
    {
      "epoch": 5.038339179378416,
      "grad_norm": 0.001442272448912263,
      "learning_rate": 0.03683135759841451,
      "loss": 0.2845,
      "num_input_tokens_seen": 58526176,
      "step": 30885
    },
    {
      "epoch": 5.039154906599233,
      "grad_norm": 0.0010375756537541747,
      "learning_rate": 0.03679270419163406,
      "loss": 0.3493,
      "num_input_tokens_seen": 58535264,
      "step": 30890
    },
    {
      "epoch": 5.039970633820051,
      "grad_norm": 0.0009617641917429864,
      "learning_rate": 0.03675406824284127,
      "loss": 0.3447,
      "num_input_tokens_seen": 58544560,
      "step": 30895
    },
    {
      "epoch": 5.040786361040868,
      "grad_norm": 0.001921062357723713,
      "learning_rate": 0.03671544975799425,
      "loss": 0.3421,
      "num_input_tokens_seen": 58553312,
      "step": 30900
    },
    {
      "epoch": 5.041602088261685,
      "grad_norm": 0.0007829968235455453,
      "learning_rate": 0.03667684874304854,
      "loss": 0.3231,
      "num_input_tokens_seen": 58562560,
      "step": 30905
    },
    {
      "epoch": 5.042417815482502,
      "grad_norm": 0.0005043193814344704,
      "learning_rate": 0.03663826520395683,
      "loss": 0.34,
      "num_input_tokens_seen": 58571904,
      "step": 30910
    },
    {
      "epoch": 5.04323354270332,
      "grad_norm": 0.0025148997083306313,
      "learning_rate": 0.03659969914666922,
      "loss": 0.2868,
      "num_input_tokens_seen": 58581296,
      "step": 30915
    },
    {
      "epoch": 5.0440492699241375,
      "grad_norm": 0.0011906978907063603,
      "learning_rate": 0.036561150577133106,
      "loss": 0.3832,
      "num_input_tokens_seen": 58591344,
      "step": 30920
    },
    {
      "epoch": 5.044864997144955,
      "grad_norm": 0.003174965037032962,
      "learning_rate": 0.036522619501293103,
      "loss": 0.3465,
      "num_input_tokens_seen": 58601184,
      "step": 30925
    },
    {
      "epoch": 5.045680724365772,
      "grad_norm": 0.000911901006475091,
      "learning_rate": 0.03648410592509122,
      "loss": 0.3229,
      "num_input_tokens_seen": 58610480,
      "step": 30930
    },
    {
      "epoch": 5.04649645158659,
      "grad_norm": 0.002230965532362461,
      "learning_rate": 0.03644560985446676,
      "loss": 0.3455,
      "num_input_tokens_seen": 58620400,
      "step": 30935
    },
    {
      "epoch": 5.047312178807407,
      "grad_norm": 0.0025458948221057653,
      "learning_rate": 0.036407131295356256,
      "loss": 0.3542,
      "num_input_tokens_seen": 58629744,
      "step": 30940
    },
    {
      "epoch": 5.048127906028224,
      "grad_norm": 0.0006163730868138373,
      "learning_rate": 0.03636867025369362,
      "loss": 0.3326,
      "num_input_tokens_seen": 58640608,
      "step": 30945
    },
    {
      "epoch": 5.048943633249041,
      "grad_norm": 0.0008408612920902669,
      "learning_rate": 0.03633022673540999,
      "loss": 0.3039,
      "num_input_tokens_seen": 58649984,
      "step": 30950
    },
    {
      "epoch": 5.049759360469859,
      "grad_norm": 0.000542759895324707,
      "learning_rate": 0.03629180074643385,
      "loss": 0.3498,
      "num_input_tokens_seen": 58659168,
      "step": 30955
    },
    {
      "epoch": 5.050575087690676,
      "grad_norm": 0.0013442052295431495,
      "learning_rate": 0.03625339229269102,
      "loss": 0.3212,
      "num_input_tokens_seen": 58668336,
      "step": 30960
    },
    {
      "epoch": 5.051390814911493,
      "grad_norm": 0.001091520651243627,
      "learning_rate": 0.036215001380104535,
      "loss": 0.297,
      "num_input_tokens_seen": 58679424,
      "step": 30965
    },
    {
      "epoch": 5.052206542132311,
      "grad_norm": 0.0028138505294919014,
      "learning_rate": 0.03617662801459471,
      "loss": 0.3318,
      "num_input_tokens_seen": 58688832,
      "step": 30970
    },
    {
      "epoch": 5.053022269353129,
      "grad_norm": 0.001535435556434095,
      "learning_rate": 0.036138272202079276,
      "loss": 0.2856,
      "num_input_tokens_seen": 58699200,
      "step": 30975
    },
    {
      "epoch": 5.053837996573946,
      "grad_norm": 0.0011574245290830731,
      "learning_rate": 0.036099933948473106,
      "loss": 0.3539,
      "num_input_tokens_seen": 58708832,
      "step": 30980
    },
    {
      "epoch": 5.054653723794763,
      "grad_norm": 0.0017090438632294536,
      "learning_rate": 0.03606161325968851,
      "loss": 0.2952,
      "num_input_tokens_seen": 58718160,
      "step": 30985
    },
    {
      "epoch": 5.05546945101558,
      "grad_norm": 0.0007634880021214485,
      "learning_rate": 0.03602331014163496,
      "loss": 0.2982,
      "num_input_tokens_seen": 58727504,
      "step": 30990
    },
    {
      "epoch": 5.056285178236398,
      "grad_norm": 0.0012301637325435877,
      "learning_rate": 0.035985024600219295,
      "loss": 0.3438,
      "num_input_tokens_seen": 58737664,
      "step": 30995
    },
    {
      "epoch": 5.057100905457215,
      "grad_norm": 0.0006857386324554682,
      "learning_rate": 0.03594675664134569,
      "loss": 0.3406,
      "num_input_tokens_seen": 58746720,
      "step": 31000
    },
    {
      "epoch": 5.057100905457215,
      "eval_loss": 0.32492542266845703,
      "eval_runtime": 152.8782,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 58746720,
      "step": 31000
    },
    {
      "epoch": 5.057916632678032,
      "grad_norm": 0.0012324148556217551,
      "learning_rate": 0.03590850627091545,
      "loss": 0.3337,
      "num_input_tokens_seen": 58755504,
      "step": 31005
    },
    {
      "epoch": 5.058732359898849,
      "grad_norm": 0.0017896927893161774,
      "learning_rate": 0.03587027349482731,
      "loss": 0.3098,
      "num_input_tokens_seen": 58765520,
      "step": 31010
    },
    {
      "epoch": 5.059548087119667,
      "grad_norm": 0.001330618397332728,
      "learning_rate": 0.035832058318977275,
      "loss": 0.338,
      "num_input_tokens_seen": 58775392,
      "step": 31015
    },
    {
      "epoch": 5.0603638143404845,
      "grad_norm": 0.0009494447731412947,
      "learning_rate": 0.03579386074925853,
      "loss": 0.2391,
      "num_input_tokens_seen": 58785024,
      "step": 31020
    },
    {
      "epoch": 5.061179541561302,
      "grad_norm": 0.0009083256591111422,
      "learning_rate": 0.035755680791561696,
      "loss": 0.2813,
      "num_input_tokens_seen": 58793280,
      "step": 31025
    },
    {
      "epoch": 5.06199526878212,
      "grad_norm": 0.0011369670974090695,
      "learning_rate": 0.03571751845177454,
      "loss": 0.3893,
      "num_input_tokens_seen": 58802240,
      "step": 31030
    },
    {
      "epoch": 5.062810996002937,
      "grad_norm": 0.001794694922864437,
      "learning_rate": 0.03567937373578225,
      "loss": 0.3095,
      "num_input_tokens_seen": 58812704,
      "step": 31035
    },
    {
      "epoch": 5.063626723223754,
      "grad_norm": 0.0020911821629852057,
      "learning_rate": 0.03564124664946711,
      "loss": 0.3391,
      "num_input_tokens_seen": 58820640,
      "step": 31040
    },
    {
      "epoch": 5.064442450444571,
      "grad_norm": 0.002474765758961439,
      "learning_rate": 0.035603137198708924,
      "loss": 0.2967,
      "num_input_tokens_seen": 58830176,
      "step": 31045
    },
    {
      "epoch": 5.065258177665389,
      "grad_norm": 0.0011233275290578604,
      "learning_rate": 0.035565045389384514,
      "loss": 0.2682,
      "num_input_tokens_seen": 58839632,
      "step": 31050
    },
    {
      "epoch": 5.066073904886206,
      "grad_norm": 0.0015370551263913512,
      "learning_rate": 0.03552697122736823,
      "loss": 0.2422,
      "num_input_tokens_seen": 58849376,
      "step": 31055
    },
    {
      "epoch": 5.066889632107023,
      "grad_norm": 0.001549088628962636,
      "learning_rate": 0.03548891471853153,
      "loss": 0.2982,
      "num_input_tokens_seen": 58858640,
      "step": 31060
    },
    {
      "epoch": 5.0677053593278405,
      "grad_norm": 0.0005845251143909991,
      "learning_rate": 0.03545087586874322,
      "loss": 0.3473,
      "num_input_tokens_seen": 58868416,
      "step": 31065
    },
    {
      "epoch": 5.0685210865486585,
      "grad_norm": 0.00027695545577444136,
      "learning_rate": 0.03541285468386935,
      "loss": 0.315,
      "num_input_tokens_seen": 58877888,
      "step": 31070
    },
    {
      "epoch": 5.069336813769476,
      "grad_norm": 0.0013740618014708161,
      "learning_rate": 0.03537485116977327,
      "loss": 0.3246,
      "num_input_tokens_seen": 58887808,
      "step": 31075
    },
    {
      "epoch": 5.070152540990293,
      "grad_norm": 0.0032143956050276756,
      "learning_rate": 0.03533686533231565,
      "loss": 0.252,
      "num_input_tokens_seen": 58897424,
      "step": 31080
    },
    {
      "epoch": 5.07096826821111,
      "grad_norm": 0.0015595154836773872,
      "learning_rate": 0.0352988971773543,
      "loss": 0.3453,
      "num_input_tokens_seen": 58906512,
      "step": 31085
    },
    {
      "epoch": 5.071783995431928,
      "grad_norm": 0.001678891945630312,
      "learning_rate": 0.03526094671074443,
      "loss": 0.3713,
      "num_input_tokens_seen": 58915904,
      "step": 31090
    },
    {
      "epoch": 5.072599722652745,
      "grad_norm": 0.0011614392278715968,
      "learning_rate": 0.03522301393833852,
      "loss": 0.3436,
      "num_input_tokens_seen": 58925664,
      "step": 31095
    },
    {
      "epoch": 5.073415449873562,
      "grad_norm": 0.0017210951773449779,
      "learning_rate": 0.035185098865986204,
      "loss": 0.2698,
      "num_input_tokens_seen": 58935920,
      "step": 31100
    },
    {
      "epoch": 5.074231177094379,
      "grad_norm": 0.001295556197874248,
      "learning_rate": 0.03514720149953453,
      "loss": 0.3414,
      "num_input_tokens_seen": 58946064,
      "step": 31105
    },
    {
      "epoch": 5.075046904315197,
      "grad_norm": 0.002354865660890937,
      "learning_rate": 0.03510932184482773,
      "loss": 0.3175,
      "num_input_tokens_seen": 58955712,
      "step": 31110
    },
    {
      "epoch": 5.0758626315360145,
      "grad_norm": 0.0012361324625089765,
      "learning_rate": 0.03507145990770724,
      "loss": 0.3224,
      "num_input_tokens_seen": 58964832,
      "step": 31115
    },
    {
      "epoch": 5.076678358756832,
      "grad_norm": 0.0014892376493662596,
      "learning_rate": 0.035033615694011984,
      "loss": 0.3599,
      "num_input_tokens_seen": 58973840,
      "step": 31120
    },
    {
      "epoch": 5.077494085977649,
      "grad_norm": 0.00270681269466877,
      "learning_rate": 0.03499578920957788,
      "loss": 0.3694,
      "num_input_tokens_seen": 58982896,
      "step": 31125
    },
    {
      "epoch": 5.078309813198467,
      "grad_norm": 0.0020174181554466486,
      "learning_rate": 0.034957980460238375,
      "loss": 0.3873,
      "num_input_tokens_seen": 58991472,
      "step": 31130
    },
    {
      "epoch": 5.079125540419284,
      "grad_norm": 0.0024547665379941463,
      "learning_rate": 0.03492018945182393,
      "loss": 0.33,
      "num_input_tokens_seen": 59000736,
      "step": 31135
    },
    {
      "epoch": 5.079941267640101,
      "grad_norm": 0.0025497502647340298,
      "learning_rate": 0.03488241619016247,
      "loss": 0.3029,
      "num_input_tokens_seen": 59009568,
      "step": 31140
    },
    {
      "epoch": 5.080756994860918,
      "grad_norm": 0.002003750763833523,
      "learning_rate": 0.03484466068107913,
      "loss": 0.3997,
      "num_input_tokens_seen": 59020032,
      "step": 31145
    },
    {
      "epoch": 5.081572722081736,
      "grad_norm": 0.0010174497729167342,
      "learning_rate": 0.034806922930396195,
      "loss": 0.339,
      "num_input_tokens_seen": 59029056,
      "step": 31150
    },
    {
      "epoch": 5.082388449302553,
      "grad_norm": 0.0006467317580245435,
      "learning_rate": 0.03476920294393337,
      "loss": 0.3242,
      "num_input_tokens_seen": 59038384,
      "step": 31155
    },
    {
      "epoch": 5.08320417652337,
      "grad_norm": 0.003139523323625326,
      "learning_rate": 0.03473150072750755,
      "loss": 0.375,
      "num_input_tokens_seen": 59047760,
      "step": 31160
    },
    {
      "epoch": 5.0840199037441876,
      "grad_norm": 0.0009320661774836481,
      "learning_rate": 0.03469381628693284,
      "loss": 0.3464,
      "num_input_tokens_seen": 59056080,
      "step": 31165
    },
    {
      "epoch": 5.084835630965006,
      "grad_norm": 0.001258688629604876,
      "learning_rate": 0.03465614962802072,
      "loss": 0.3738,
      "num_input_tokens_seen": 59065184,
      "step": 31170
    },
    {
      "epoch": 5.085651358185823,
      "grad_norm": 0.0015709394356235862,
      "learning_rate": 0.0346185007565798,
      "loss": 0.3001,
      "num_input_tokens_seen": 59076048,
      "step": 31175
    },
    {
      "epoch": 5.08646708540664,
      "grad_norm": 0.0017216800479218364,
      "learning_rate": 0.03458086967841609,
      "loss": 0.346,
      "num_input_tokens_seen": 59085856,
      "step": 31180
    },
    {
      "epoch": 5.087282812627457,
      "grad_norm": 0.000776449334807694,
      "learning_rate": 0.03454325639933266,
      "loss": 0.321,
      "num_input_tokens_seen": 59094464,
      "step": 31185
    },
    {
      "epoch": 5.088098539848275,
      "grad_norm": 0.0028087603859603405,
      "learning_rate": 0.03450566092513007,
      "loss": 0.302,
      "num_input_tokens_seen": 59104384,
      "step": 31190
    },
    {
      "epoch": 5.088914267069092,
      "grad_norm": 0.0014058639062568545,
      "learning_rate": 0.034468083261605914,
      "loss": 0.3353,
      "num_input_tokens_seen": 59114048,
      "step": 31195
    },
    {
      "epoch": 5.089729994289909,
      "grad_norm": 0.0007410577964037657,
      "learning_rate": 0.03443052341455522,
      "loss": 0.3194,
      "num_input_tokens_seen": 59124272,
      "step": 31200
    },
    {
      "epoch": 5.089729994289909,
      "eval_loss": 0.3243216574192047,
      "eval_runtime": 152.9503,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 59124272,
      "step": 31200
    },
    {
      "epoch": 5.090545721510727,
      "grad_norm": 0.0012107589282095432,
      "learning_rate": 0.0343929813897701,
      "loss": 0.3476,
      "num_input_tokens_seen": 59134464,
      "step": 31205
    },
    {
      "epoch": 5.091361448731544,
      "grad_norm": 0.0010378019651398063,
      "learning_rate": 0.034355457193040125,
      "loss": 0.3143,
      "num_input_tokens_seen": 59143552,
      "step": 31210
    },
    {
      "epoch": 5.0921771759523615,
      "grad_norm": 0.0012597732711583376,
      "learning_rate": 0.03431795083015186,
      "loss": 0.3213,
      "num_input_tokens_seen": 59153600,
      "step": 31215
    },
    {
      "epoch": 5.092992903173179,
      "grad_norm": 0.0021207088138908148,
      "learning_rate": 0.03428046230688936,
      "loss": 0.3403,
      "num_input_tokens_seen": 59163648,
      "step": 31220
    },
    {
      "epoch": 5.093808630393997,
      "grad_norm": 0.0023498411756008863,
      "learning_rate": 0.034242991629033805,
      "loss": 0.3,
      "num_input_tokens_seen": 59173536,
      "step": 31225
    },
    {
      "epoch": 5.094624357614814,
      "grad_norm": 0.0029051199089735746,
      "learning_rate": 0.03420553880236362,
      "loss": 0.329,
      "num_input_tokens_seen": 59183440,
      "step": 31230
    },
    {
      "epoch": 5.095440084835631,
      "grad_norm": 0.0015336869983002543,
      "learning_rate": 0.03416810383265449,
      "loss": 0.3306,
      "num_input_tokens_seen": 59192816,
      "step": 31235
    },
    {
      "epoch": 5.096255812056448,
      "grad_norm": 0.0012669643620029092,
      "learning_rate": 0.03413068672567944,
      "loss": 0.3612,
      "num_input_tokens_seen": 59201952,
      "step": 31240
    },
    {
      "epoch": 5.097071539277266,
      "grad_norm": 0.0014221257297322154,
      "learning_rate": 0.034093287487208565,
      "loss": 0.3337,
      "num_input_tokens_seen": 59211536,
      "step": 31245
    },
    {
      "epoch": 5.097887266498083,
      "grad_norm": 0.002487636636942625,
      "learning_rate": 0.03405590612300937,
      "loss": 0.3141,
      "num_input_tokens_seen": 59220752,
      "step": 31250
    },
    {
      "epoch": 5.0987029937189,
      "grad_norm": 0.001200020546093583,
      "learning_rate": 0.03401854263884646,
      "loss": 0.3139,
      "num_input_tokens_seen": 59230976,
      "step": 31255
    },
    {
      "epoch": 5.0995187209397175,
      "grad_norm": 0.0016873687272891402,
      "learning_rate": 0.033981197040481824,
      "loss": 0.3174,
      "num_input_tokens_seen": 59241008,
      "step": 31260
    },
    {
      "epoch": 5.1003344481605355,
      "grad_norm": 0.0016357930144295096,
      "learning_rate": 0.03394386933367459,
      "loss": 0.3005,
      "num_input_tokens_seen": 59250080,
      "step": 31265
    },
    {
      "epoch": 5.101150175381353,
      "grad_norm": 0.001437647850252688,
      "learning_rate": 0.033906559524181104,
      "loss": 0.3257,
      "num_input_tokens_seen": 59258800,
      "step": 31270
    },
    {
      "epoch": 5.10196590260217,
      "grad_norm": 0.0015823543071746826,
      "learning_rate": 0.033869267617755085,
      "loss": 0.2952,
      "num_input_tokens_seen": 59267792,
      "step": 31275
    },
    {
      "epoch": 5.102781629822987,
      "grad_norm": 0.0016547271516174078,
      "learning_rate": 0.0338319936201474,
      "loss": 0.3263,
      "num_input_tokens_seen": 59276960,
      "step": 31280
    },
    {
      "epoch": 5.103597357043805,
      "grad_norm": 0.0020015223417431116,
      "learning_rate": 0.033794737537106136,
      "loss": 0.339,
      "num_input_tokens_seen": 59286576,
      "step": 31285
    },
    {
      "epoch": 5.104413084264622,
      "grad_norm": 0.0016135239275172353,
      "learning_rate": 0.03375749937437671,
      "loss": 0.3513,
      "num_input_tokens_seen": 59294752,
      "step": 31290
    },
    {
      "epoch": 5.105228811485439,
      "grad_norm": 0.0016585779376327991,
      "learning_rate": 0.033720279137701634,
      "loss": 0.3095,
      "num_input_tokens_seen": 59304304,
      "step": 31295
    },
    {
      "epoch": 5.106044538706256,
      "grad_norm": 0.0010192451300099492,
      "learning_rate": 0.03368307683282078,
      "loss": 0.2921,
      "num_input_tokens_seen": 59313520,
      "step": 31300
    },
    {
      "epoch": 5.106860265927074,
      "grad_norm": 0.001835645642131567,
      "learning_rate": 0.033645892465471235,
      "loss": 0.3247,
      "num_input_tokens_seen": 59323376,
      "step": 31305
    },
    {
      "epoch": 5.1076759931478914,
      "grad_norm": 0.002634247299283743,
      "learning_rate": 0.03360872604138724,
      "loss": 0.2608,
      "num_input_tokens_seen": 59332976,
      "step": 31310
    },
    {
      "epoch": 5.108491720368709,
      "grad_norm": 0.0028998972848057747,
      "learning_rate": 0.03357157756630034,
      "loss": 0.2814,
      "num_input_tokens_seen": 59342336,
      "step": 31315
    },
    {
      "epoch": 5.109307447589526,
      "grad_norm": 0.00316795171238482,
      "learning_rate": 0.033534447045939365,
      "loss": 0.3265,
      "num_input_tokens_seen": 59352688,
      "step": 31320
    },
    {
      "epoch": 5.110123174810344,
      "grad_norm": 0.0015404880978167057,
      "learning_rate": 0.03349733448603026,
      "loss": 0.2848,
      "num_input_tokens_seen": 59363088,
      "step": 31325
    },
    {
      "epoch": 5.110938902031161,
      "grad_norm": 0.001602602656930685,
      "learning_rate": 0.03346023989229619,
      "loss": 0.3584,
      "num_input_tokens_seen": 59371904,
      "step": 31330
    },
    {
      "epoch": 5.111754629251978,
      "grad_norm": 0.004160860553383827,
      "learning_rate": 0.03342316327045769,
      "loss": 0.3342,
      "num_input_tokens_seen": 59381600,
      "step": 31335
    },
    {
      "epoch": 5.112570356472795,
      "grad_norm": 0.0027515653055161238,
      "learning_rate": 0.033386104626232385,
      "loss": 0.2842,
      "num_input_tokens_seen": 59390704,
      "step": 31340
    },
    {
      "epoch": 5.113386083693613,
      "grad_norm": 0.0031460223253816366,
      "learning_rate": 0.03334906396533525,
      "loss": 0.2642,
      "num_input_tokens_seen": 59400208,
      "step": 31345
    },
    {
      "epoch": 5.11420181091443,
      "grad_norm": 0.002203085459768772,
      "learning_rate": 0.033312041293478326,
      "loss": 0.3556,
      "num_input_tokens_seen": 59410384,
      "step": 31350
    },
    {
      "epoch": 5.115017538135247,
      "grad_norm": 0.0018868952756747603,
      "learning_rate": 0.03327503661637103,
      "loss": 0.3514,
      "num_input_tokens_seen": 59420400,
      "step": 31355
    },
    {
      "epoch": 5.1158332653560645,
      "grad_norm": 0.001954849110916257,
      "learning_rate": 0.03323804993971998,
      "loss": 0.3525,
      "num_input_tokens_seen": 59430688,
      "step": 31360
    },
    {
      "epoch": 5.1166489925768825,
      "grad_norm": 0.0017879047663882375,
      "learning_rate": 0.033201081269228924,
      "loss": 0.3284,
      "num_input_tokens_seen": 59440400,
      "step": 31365
    },
    {
      "epoch": 5.1174647197977,
      "grad_norm": 0.001598120667040348,
      "learning_rate": 0.03316413061059895,
      "loss": 0.2995,
      "num_input_tokens_seen": 59449248,
      "step": 31370
    },
    {
      "epoch": 5.118280447018517,
      "grad_norm": 0.0015675774775445461,
      "learning_rate": 0.03312719796952827,
      "loss": 0.2868,
      "num_input_tokens_seen": 59457712,
      "step": 31375
    },
    {
      "epoch": 5.119096174239334,
      "grad_norm": 0.0013840071624144912,
      "learning_rate": 0.03309028335171236,
      "loss": 0.2607,
      "num_input_tokens_seen": 59467184,
      "step": 31380
    },
    {
      "epoch": 5.119911901460152,
      "grad_norm": 0.0018925770418718457,
      "learning_rate": 0.03305338676284398,
      "loss": 0.276,
      "num_input_tokens_seen": 59476496,
      "step": 31385
    },
    {
      "epoch": 5.120727628680969,
      "grad_norm": 0.0014534727670252323,
      "learning_rate": 0.03301650820861296,
      "loss": 0.2923,
      "num_input_tokens_seen": 59486064,
      "step": 31390
    },
    {
      "epoch": 5.121543355901786,
      "grad_norm": 0.0011043132981285453,
      "learning_rate": 0.03297964769470652,
      "loss": 0.2987,
      "num_input_tokens_seen": 59495552,
      "step": 31395
    },
    {
      "epoch": 5.122359083122603,
      "grad_norm": 0.0014078343519940972,
      "learning_rate": 0.032942805226808945,
      "loss": 0.3977,
      "num_input_tokens_seen": 59504688,
      "step": 31400
    },
    {
      "epoch": 5.122359083122603,
      "eval_loss": 0.3197750151157379,
      "eval_runtime": 152.901,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 59504688,
      "step": 31400
    },
    {
      "epoch": 5.123174810343421,
      "grad_norm": 0.002928523812443018,
      "learning_rate": 0.03290598081060187,
      "loss": 0.3625,
      "num_input_tokens_seen": 59513408,
      "step": 31405
    },
    {
      "epoch": 5.1239905375642385,
      "grad_norm": 0.0006313706981018186,
      "learning_rate": 0.03286917445176407,
      "loss": 0.3512,
      "num_input_tokens_seen": 59523504,
      "step": 31410
    },
    {
      "epoch": 5.124806264785056,
      "grad_norm": 0.0035056460183113813,
      "learning_rate": 0.032832386155971456,
      "loss": 0.3527,
      "num_input_tokens_seen": 59533552,
      "step": 31415
    },
    {
      "epoch": 5.125621992005874,
      "grad_norm": 0.0029766454827040434,
      "learning_rate": 0.032795615928897334,
      "loss": 0.3496,
      "num_input_tokens_seen": 59543600,
      "step": 31420
    },
    {
      "epoch": 5.126437719226691,
      "grad_norm": 0.001550739398226142,
      "learning_rate": 0.03275886377621215,
      "loss": 0.3467,
      "num_input_tokens_seen": 59553024,
      "step": 31425
    },
    {
      "epoch": 5.127253446447508,
      "grad_norm": 0.0037686701398342848,
      "learning_rate": 0.03272212970358348,
      "loss": 0.3152,
      "num_input_tokens_seen": 59560880,
      "step": 31430
    },
    {
      "epoch": 5.128069173668325,
      "grad_norm": 0.001465566921979189,
      "learning_rate": 0.032685413716676215,
      "loss": 0.2811,
      "num_input_tokens_seen": 59570096,
      "step": 31435
    },
    {
      "epoch": 5.128884900889143,
      "grad_norm": 0.001290342304855585,
      "learning_rate": 0.032648715821152474,
      "loss": 0.3228,
      "num_input_tokens_seen": 59578672,
      "step": 31440
    },
    {
      "epoch": 5.12970062810996,
      "grad_norm": 0.00046480525634251535,
      "learning_rate": 0.03261203602267143,
      "loss": 0.2728,
      "num_input_tokens_seen": 59588112,
      "step": 31445
    },
    {
      "epoch": 5.130516355330777,
      "grad_norm": 0.0009889728389680386,
      "learning_rate": 0.03257537432688966,
      "loss": 0.3105,
      "num_input_tokens_seen": 59596896,
      "step": 31450
    },
    {
      "epoch": 5.1313320825515945,
      "grad_norm": 0.0020245183259248734,
      "learning_rate": 0.03253873073946077,
      "loss": 0.3188,
      "num_input_tokens_seen": 59604960,
      "step": 31455
    },
    {
      "epoch": 5.1321478097724125,
      "grad_norm": 0.0010508581763133407,
      "learning_rate": 0.03250210526603572,
      "loss": 0.3558,
      "num_input_tokens_seen": 59612240,
      "step": 31460
    },
    {
      "epoch": 5.13296353699323,
      "grad_norm": 0.002335284138098359,
      "learning_rate": 0.03246549791226266,
      "loss": 0.3348,
      "num_input_tokens_seen": 59622288,
      "step": 31465
    },
    {
      "epoch": 5.133779264214047,
      "grad_norm": 0.0018379620742052794,
      "learning_rate": 0.03242890868378679,
      "loss": 0.3438,
      "num_input_tokens_seen": 59631296,
      "step": 31470
    },
    {
      "epoch": 5.134594991434864,
      "grad_norm": 0.002580787520855665,
      "learning_rate": 0.03239233758625074,
      "loss": 0.3252,
      "num_input_tokens_seen": 59641664,
      "step": 31475
    },
    {
      "epoch": 5.135410718655682,
      "grad_norm": 0.0020828465931117535,
      "learning_rate": 0.032355784625294204,
      "loss": 0.2815,
      "num_input_tokens_seen": 59652416,
      "step": 31480
    },
    {
      "epoch": 5.136226445876499,
      "grad_norm": 0.0013070158893242478,
      "learning_rate": 0.03231924980655402,
      "loss": 0.3099,
      "num_input_tokens_seen": 59661840,
      "step": 31485
    },
    {
      "epoch": 5.137042173097316,
      "grad_norm": 0.001558186486363411,
      "learning_rate": 0.032282733135664446,
      "loss": 0.3199,
      "num_input_tokens_seen": 59670032,
      "step": 31490
    },
    {
      "epoch": 5.137857900318133,
      "grad_norm": 0.003670568112283945,
      "learning_rate": 0.03224623461825669,
      "loss": 0.3738,
      "num_input_tokens_seen": 59679264,
      "step": 31495
    },
    {
      "epoch": 5.138673627538951,
      "grad_norm": 0.0024284834507852793,
      "learning_rate": 0.03220975425995937,
      "loss": 0.2858,
      "num_input_tokens_seen": 59688640,
      "step": 31500
    },
    {
      "epoch": 5.139489354759768,
      "grad_norm": 0.0013465835945680737,
      "learning_rate": 0.032173292066398206,
      "loss": 0.2688,
      "num_input_tokens_seen": 59697216,
      "step": 31505
    },
    {
      "epoch": 5.1403050819805856,
      "grad_norm": 0.0015080035664141178,
      "learning_rate": 0.03213684804319606,
      "loss": 0.3375,
      "num_input_tokens_seen": 59706672,
      "step": 31510
    },
    {
      "epoch": 5.141120809201403,
      "grad_norm": 0.0019459194736555219,
      "learning_rate": 0.03210042219597312,
      "loss": 0.3414,
      "num_input_tokens_seen": 59716320,
      "step": 31515
    },
    {
      "epoch": 5.141936536422221,
      "grad_norm": 0.003314505098387599,
      "learning_rate": 0.03206401453034675,
      "loss": 0.3391,
      "num_input_tokens_seen": 59726560,
      "step": 31520
    },
    {
      "epoch": 5.142752263643038,
      "grad_norm": 0.0023547152522951365,
      "learning_rate": 0.03202762505193136,
      "loss": 0.3034,
      "num_input_tokens_seen": 59735776,
      "step": 31525
    },
    {
      "epoch": 5.143567990863855,
      "grad_norm": 0.0028389173094183207,
      "learning_rate": 0.031991253766338754,
      "loss": 0.3146,
      "num_input_tokens_seen": 59744512,
      "step": 31530
    },
    {
      "epoch": 5.144383718084672,
      "grad_norm": 0.0024134255945682526,
      "learning_rate": 0.03195490067917778,
      "loss": 0.3029,
      "num_input_tokens_seen": 59754400,
      "step": 31535
    },
    {
      "epoch": 5.14519944530549,
      "grad_norm": 0.002250097692012787,
      "learning_rate": 0.03191856579605461,
      "loss": 0.3016,
      "num_input_tokens_seen": 59763456,
      "step": 31540
    },
    {
      "epoch": 5.146015172526307,
      "grad_norm": 0.0016744532622396946,
      "learning_rate": 0.031882249122572454,
      "loss": 0.3446,
      "num_input_tokens_seen": 59773648,
      "step": 31545
    },
    {
      "epoch": 5.146830899747124,
      "grad_norm": 0.002401394071057439,
      "learning_rate": 0.03184595066433188,
      "loss": 0.3242,
      "num_input_tokens_seen": 59783504,
      "step": 31550
    },
    {
      "epoch": 5.1476466269679415,
      "grad_norm": 0.0009259481448680162,
      "learning_rate": 0.03180967042693049,
      "loss": 0.3363,
      "num_input_tokens_seen": 59793456,
      "step": 31555
    },
    {
      "epoch": 5.1484623541887595,
      "grad_norm": 0.0015987867955118418,
      "learning_rate": 0.03177340841596323,
      "loss": 0.2544,
      "num_input_tokens_seen": 59802400,
      "step": 31560
    },
    {
      "epoch": 5.149278081409577,
      "grad_norm": 0.0011051826877519488,
      "learning_rate": 0.03173716463702209,
      "loss": 0.3024,
      "num_input_tokens_seen": 59811552,
      "step": 31565
    },
    {
      "epoch": 5.150093808630394,
      "grad_norm": 0.0013008477399125695,
      "learning_rate": 0.03170093909569638,
      "loss": 0.354,
      "num_input_tokens_seen": 59820560,
      "step": 31570
    },
    {
      "epoch": 5.150909535851211,
      "grad_norm": 0.0007469529518857598,
      "learning_rate": 0.03166473179757246,
      "loss": 0.3631,
      "num_input_tokens_seen": 59830960,
      "step": 31575
    },
    {
      "epoch": 5.151725263072029,
      "grad_norm": 0.0019524815725162625,
      "learning_rate": 0.031628542748234005,
      "loss": 0.2919,
      "num_input_tokens_seen": 59838784,
      "step": 31580
    },
    {
      "epoch": 5.152540990292846,
      "grad_norm": 0.0011622081510722637,
      "learning_rate": 0.03159237195326184,
      "loss": 0.2753,
      "num_input_tokens_seen": 59848432,
      "step": 31585
    },
    {
      "epoch": 5.153356717513663,
      "grad_norm": 0.0015541811008006334,
      "learning_rate": 0.031556219418233875,
      "loss": 0.2753,
      "num_input_tokens_seen": 59857616,
      "step": 31590
    },
    {
      "epoch": 5.154172444734481,
      "grad_norm": 0.0016707415925338864,
      "learning_rate": 0.03152008514872533,
      "loss": 0.3227,
      "num_input_tokens_seen": 59867232,
      "step": 31595
    },
    {
      "epoch": 5.154988171955298,
      "grad_norm": 0.0005774523015134037,
      "learning_rate": 0.03148396915030862,
      "loss": 0.26,
      "num_input_tokens_seen": 59875840,
      "step": 31600
    },
    {
      "epoch": 5.154988171955298,
      "eval_loss": 0.3204181492328644,
      "eval_runtime": 152.9785,
      "eval_samples_per_second": 17.813,
      "eval_steps_per_second": 8.91,
      "num_input_tokens_seen": 59875840,
      "step": 31600
    },
    {
      "epoch": 5.1558038991761155,
      "grad_norm": 0.003016288857907057,
      "learning_rate": 0.03144787142855318,
      "loss": 0.3114,
      "num_input_tokens_seen": 59885616,
      "step": 31605
    },
    {
      "epoch": 5.156619626396933,
      "grad_norm": 0.003826574655249715,
      "learning_rate": 0.031411791989025835,
      "loss": 0.2394,
      "num_input_tokens_seen": 59893936,
      "step": 31610
    },
    {
      "epoch": 5.157435353617751,
      "grad_norm": 0.0008336132159456611,
      "learning_rate": 0.031375730837290394,
      "loss": 0.3177,
      "num_input_tokens_seen": 59902272,
      "step": 31615
    },
    {
      "epoch": 5.158251080838568,
      "grad_norm": 0.0034359670244157314,
      "learning_rate": 0.031339687978908015,
      "loss": 0.2828,
      "num_input_tokens_seen": 59911136,
      "step": 31620
    },
    {
      "epoch": 5.159066808059385,
      "grad_norm": 0.000926813343539834,
      "learning_rate": 0.03130366341943694,
      "loss": 0.3429,
      "num_input_tokens_seen": 59920448,
      "step": 31625
    },
    {
      "epoch": 5.159882535280202,
      "grad_norm": 0.0031521334312856197,
      "learning_rate": 0.031267657164432555,
      "loss": 0.3388,
      "num_input_tokens_seen": 59929680,
      "step": 31630
    },
    {
      "epoch": 5.16069826250102,
      "grad_norm": 0.002246500225737691,
      "learning_rate": 0.03123166921944752,
      "loss": 0.4706,
      "num_input_tokens_seen": 59939648,
      "step": 31635
    },
    {
      "epoch": 5.161513989721837,
      "grad_norm": 0.0029398573096841574,
      "learning_rate": 0.031195699590031666,
      "loss": 0.2911,
      "num_input_tokens_seen": 59949296,
      "step": 31640
    },
    {
      "epoch": 5.162329716942654,
      "grad_norm": 0.0010722336592152715,
      "learning_rate": 0.031159748281731885,
      "loss": 0.3137,
      "num_input_tokens_seen": 59959392,
      "step": 31645
    },
    {
      "epoch": 5.163145444163471,
      "grad_norm": 0.0005466934526339173,
      "learning_rate": 0.031123815300092394,
      "loss": 0.2918,
      "num_input_tokens_seen": 59969664,
      "step": 31650
    },
    {
      "epoch": 5.1639611713842895,
      "grad_norm": 0.002947635715827346,
      "learning_rate": 0.031087900650654424,
      "loss": 0.343,
      "num_input_tokens_seen": 59978960,
      "step": 31655
    },
    {
      "epoch": 5.164776898605107,
      "grad_norm": 0.0020268545486032963,
      "learning_rate": 0.031052004338956534,
      "loss": 0.2709,
      "num_input_tokens_seen": 59987280,
      "step": 31660
    },
    {
      "epoch": 5.165592625825924,
      "grad_norm": 0.0006173032452352345,
      "learning_rate": 0.031016126370534407,
      "loss": 0.3462,
      "num_input_tokens_seen": 59996880,
      "step": 31665
    },
    {
      "epoch": 5.166408353046741,
      "grad_norm": 0.0015876467805355787,
      "learning_rate": 0.030980266750920804,
      "loss": 0.3439,
      "num_input_tokens_seen": 60007056,
      "step": 31670
    },
    {
      "epoch": 5.167224080267559,
      "grad_norm": 0.0010343461763113737,
      "learning_rate": 0.030944425485645747,
      "loss": 0.3332,
      "num_input_tokens_seen": 60015392,
      "step": 31675
    },
    {
      "epoch": 5.168039807488376,
      "grad_norm": 0.0019253400387242436,
      "learning_rate": 0.03090860258023647,
      "loss": 0.3712,
      "num_input_tokens_seen": 60023488,
      "step": 31680
    },
    {
      "epoch": 5.168855534709193,
      "grad_norm": 0.0060962289571762085,
      "learning_rate": 0.030872798040217236,
      "loss": 0.3513,
      "num_input_tokens_seen": 60033440,
      "step": 31685
    },
    {
      "epoch": 5.16967126193001,
      "grad_norm": 0.00113762728869915,
      "learning_rate": 0.03083701187110964,
      "loss": 0.2916,
      "num_input_tokens_seen": 60044128,
      "step": 31690
    },
    {
      "epoch": 5.170486989150828,
      "grad_norm": 0.002720179269090295,
      "learning_rate": 0.030801244078432294,
      "loss": 0.3263,
      "num_input_tokens_seen": 60052240,
      "step": 31695
    },
    {
      "epoch": 5.171302716371645,
      "grad_norm": 0.00289507070556283,
      "learning_rate": 0.030765494667701024,
      "loss": 0.2858,
      "num_input_tokens_seen": 60061104,
      "step": 31700
    },
    {
      "epoch": 5.1721184435924625,
      "grad_norm": 0.0036297496408224106,
      "learning_rate": 0.030729763644428913,
      "loss": 0.3537,
      "num_input_tokens_seen": 60071232,
      "step": 31705
    },
    {
      "epoch": 5.17293417081328,
      "grad_norm": 0.002106573898345232,
      "learning_rate": 0.030694051014126048,
      "loss": 0.3371,
      "num_input_tokens_seen": 60080720,
      "step": 31710
    },
    {
      "epoch": 5.173749898034098,
      "grad_norm": 0.003530816175043583,
      "learning_rate": 0.030658356782299792,
      "loss": 0.3912,
      "num_input_tokens_seen": 60090528,
      "step": 31715
    },
    {
      "epoch": 5.174565625254915,
      "grad_norm": 0.0020757499150931835,
      "learning_rate": 0.030622680954454726,
      "loss": 0.4309,
      "num_input_tokens_seen": 60099552,
      "step": 31720
    },
    {
      "epoch": 5.175381352475732,
      "grad_norm": 0.0011107804020866752,
      "learning_rate": 0.030587023536092398,
      "loss": 0.3523,
      "num_input_tokens_seen": 60107968,
      "step": 31725
    },
    {
      "epoch": 5.176197079696549,
      "grad_norm": 0.0013503640657290816,
      "learning_rate": 0.03055138453271171,
      "loss": 0.2901,
      "num_input_tokens_seen": 60117568,
      "step": 31730
    },
    {
      "epoch": 5.177012806917367,
      "grad_norm": 0.002122979611158371,
      "learning_rate": 0.03051576394980858,
      "loss": 0.3606,
      "num_input_tokens_seen": 60127040,
      "step": 31735
    },
    {
      "epoch": 5.177828534138184,
      "grad_norm": 0.0009817647514864802,
      "learning_rate": 0.030480161792876187,
      "loss": 0.3213,
      "num_input_tokens_seen": 60136864,
      "step": 31740
    },
    {
      "epoch": 5.178644261359001,
      "grad_norm": 0.004039100371301174,
      "learning_rate": 0.030444578067404846,
      "loss": 0.3869,
      "num_input_tokens_seen": 60145216,
      "step": 31745
    },
    {
      "epoch": 5.1794599885798185,
      "grad_norm": 0.0004711982037406415,
      "learning_rate": 0.030409012778881975,
      "loss": 0.3419,
      "num_input_tokens_seen": 60153680,
      "step": 31750
    },
    {
      "epoch": 5.1802757158006365,
      "grad_norm": 0.0030065798200666904,
      "learning_rate": 0.030373465932792235,
      "loss": 0.3059,
      "num_input_tokens_seen": 60161392,
      "step": 31755
    },
    {
      "epoch": 5.181091443021454,
      "grad_norm": 0.002233892912045121,
      "learning_rate": 0.030337937534617342,
      "loss": 0.323,
      "num_input_tokens_seen": 60170416,
      "step": 31760
    },
    {
      "epoch": 5.181907170242271,
      "grad_norm": 0.0011227544164285064,
      "learning_rate": 0.030302427589836277,
      "loss": 0.3547,
      "num_input_tokens_seen": 60179280,
      "step": 31765
    },
    {
      "epoch": 5.182722897463089,
      "grad_norm": 0.001426712959073484,
      "learning_rate": 0.030266936103925095,
      "loss": 0.2855,
      "num_input_tokens_seen": 60187824,
      "step": 31770
    },
    {
      "epoch": 5.183538624683906,
      "grad_norm": 0.0007408634992316365,
      "learning_rate": 0.030231463082356982,
      "loss": 0.311,
      "num_input_tokens_seen": 60197104,
      "step": 31775
    },
    {
      "epoch": 5.184354351904723,
      "grad_norm": 0.0017190300859510899,
      "learning_rate": 0.030196008530602367,
      "loss": 0.2934,
      "num_input_tokens_seen": 60206256,
      "step": 31780
    },
    {
      "epoch": 5.18517007912554,
      "grad_norm": 0.0014847618294879794,
      "learning_rate": 0.030160572454128842,
      "loss": 0.2891,
      "num_input_tokens_seen": 60215968,
      "step": 31785
    },
    {
      "epoch": 5.185985806346358,
      "grad_norm": 0.0013187716249376535,
      "learning_rate": 0.03012515485840098,
      "loss": 0.2979,
      "num_input_tokens_seen": 60226240,
      "step": 31790
    },
    {
      "epoch": 5.186801533567175,
      "grad_norm": 0.0013831595424562693,
      "learning_rate": 0.030089755748880734,
      "loss": 0.2692,
      "num_input_tokens_seen": 60237312,
      "step": 31795
    },
    {
      "epoch": 5.1876172607879925,
      "grad_norm": 0.0021649578120559454,
      "learning_rate": 0.030054375131027003,
      "loss": 0.3123,
      "num_input_tokens_seen": 60247360,
      "step": 31800
    },
    {
      "epoch": 5.1876172607879925,
      "eval_loss": 0.3194426894187927,
      "eval_runtime": 152.875,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 60247360,
      "step": 31800
    },
    {
      "epoch": 5.18843298800881,
      "grad_norm": 0.004760668147355318,
      "learning_rate": 0.030019013010295942,
      "loss": 0.3183,
      "num_input_tokens_seen": 60257120,
      "step": 31805
    },
    {
      "epoch": 5.189248715229628,
      "grad_norm": 0.001081531634554267,
      "learning_rate": 0.029983669392140897,
      "loss": 0.279,
      "num_input_tokens_seen": 60266112,
      "step": 31810
    },
    {
      "epoch": 5.190064442450445,
      "grad_norm": 0.0013728041667491198,
      "learning_rate": 0.029948344282012217,
      "loss": 0.3228,
      "num_input_tokens_seen": 60274928,
      "step": 31815
    },
    {
      "epoch": 5.190880169671262,
      "grad_norm": 0.0010681941639631987,
      "learning_rate": 0.029913037685357507,
      "loss": 0.3399,
      "num_input_tokens_seen": 60284368,
      "step": 31820
    },
    {
      "epoch": 5.191695896892079,
      "grad_norm": 0.002399936318397522,
      "learning_rate": 0.029877749607621528,
      "loss": 0.3281,
      "num_input_tokens_seen": 60293360,
      "step": 31825
    },
    {
      "epoch": 5.192511624112897,
      "grad_norm": 0.0011797738261520863,
      "learning_rate": 0.029842480054246077,
      "loss": 0.2657,
      "num_input_tokens_seen": 60302992,
      "step": 31830
    },
    {
      "epoch": 5.193327351333714,
      "grad_norm": 0.0034785671159625053,
      "learning_rate": 0.02980722903067022,
      "loss": 0.3418,
      "num_input_tokens_seen": 60312512,
      "step": 31835
    },
    {
      "epoch": 5.194143078554531,
      "grad_norm": 0.001759053091518581,
      "learning_rate": 0.029771996542330113,
      "loss": 0.3273,
      "num_input_tokens_seen": 60321936,
      "step": 31840
    },
    {
      "epoch": 5.194958805775348,
      "grad_norm": 0.001769563532434404,
      "learning_rate": 0.029736782594658954,
      "loss": 0.2781,
      "num_input_tokens_seen": 60331792,
      "step": 31845
    },
    {
      "epoch": 5.195774532996166,
      "grad_norm": 0.0002978703414555639,
      "learning_rate": 0.029701587193087284,
      "loss": 0.3275,
      "num_input_tokens_seen": 60339872,
      "step": 31850
    },
    {
      "epoch": 5.196590260216984,
      "grad_norm": 0.0014400698710232973,
      "learning_rate": 0.0296664103430426,
      "loss": 0.2721,
      "num_input_tokens_seen": 60349952,
      "step": 31855
    },
    {
      "epoch": 5.197405987437801,
      "grad_norm": 0.004190001171082258,
      "learning_rate": 0.029631252049949652,
      "loss": 0.3736,
      "num_input_tokens_seen": 60358896,
      "step": 31860
    },
    {
      "epoch": 5.198221714658618,
      "grad_norm": 0.0029446466360241175,
      "learning_rate": 0.02959611231923031,
      "loss": 0.3749,
      "num_input_tokens_seen": 60367584,
      "step": 31865
    },
    {
      "epoch": 5.199037441879436,
      "grad_norm": 0.0011633125832304358,
      "learning_rate": 0.029560991156303507,
      "loss": 0.3785,
      "num_input_tokens_seen": 60377360,
      "step": 31870
    },
    {
      "epoch": 5.199853169100253,
      "grad_norm": 0.0025547498371452093,
      "learning_rate": 0.02952588856658544,
      "loss": 0.3493,
      "num_input_tokens_seen": 60386288,
      "step": 31875
    },
    {
      "epoch": 5.20066889632107,
      "grad_norm": 0.0014597981935366988,
      "learning_rate": 0.029490804555489296,
      "loss": 0.3176,
      "num_input_tokens_seen": 60396352,
      "step": 31880
    },
    {
      "epoch": 5.201484623541887,
      "grad_norm": 0.0007505755056627095,
      "learning_rate": 0.029455739128425484,
      "loss": 0.3665,
      "num_input_tokens_seen": 60404848,
      "step": 31885
    },
    {
      "epoch": 5.202300350762705,
      "grad_norm": 0.0016688596224412322,
      "learning_rate": 0.029420692290801607,
      "loss": 0.3094,
      "num_input_tokens_seen": 60414384,
      "step": 31890
    },
    {
      "epoch": 5.203116077983522,
      "grad_norm": 0.0022214872296899557,
      "learning_rate": 0.02938566404802223,
      "loss": 0.3276,
      "num_input_tokens_seen": 60423136,
      "step": 31895
    },
    {
      "epoch": 5.2039318052043395,
      "grad_norm": 0.0015398231334984303,
      "learning_rate": 0.029350654405489195,
      "loss": 0.4063,
      "num_input_tokens_seen": 60431536,
      "step": 31900
    },
    {
      "epoch": 5.204747532425157,
      "grad_norm": 0.0024989889934659004,
      "learning_rate": 0.02931566336860145,
      "loss": 0.3013,
      "num_input_tokens_seen": 60441136,
      "step": 31905
    },
    {
      "epoch": 5.205563259645975,
      "grad_norm": 0.001634044572710991,
      "learning_rate": 0.02928069094275505,
      "loss": 0.3721,
      "num_input_tokens_seen": 60451984,
      "step": 31910
    },
    {
      "epoch": 5.206378986866792,
      "grad_norm": 0.0035572336055338383,
      "learning_rate": 0.02924573713334314,
      "loss": 0.381,
      "num_input_tokens_seen": 60461872,
      "step": 31915
    },
    {
      "epoch": 5.207194714087609,
      "grad_norm": 0.0007092969608493149,
      "learning_rate": 0.02921080194575603,
      "loss": 0.3419,
      "num_input_tokens_seen": 60470720,
      "step": 31920
    },
    {
      "epoch": 5.208010441308426,
      "grad_norm": 0.0011312463320791721,
      "learning_rate": 0.029175885385381177,
      "loss": 0.3416,
      "num_input_tokens_seen": 60479776,
      "step": 31925
    },
    {
      "epoch": 5.208826168529244,
      "grad_norm": 0.0010107404086738825,
      "learning_rate": 0.029140987457603223,
      "loss": 0.3158,
      "num_input_tokens_seen": 60491088,
      "step": 31930
    },
    {
      "epoch": 5.209641895750061,
      "grad_norm": 0.0018753737676888704,
      "learning_rate": 0.029106108167803763,
      "loss": 0.3179,
      "num_input_tokens_seen": 60500288,
      "step": 31935
    },
    {
      "epoch": 5.210457622970878,
      "grad_norm": 0.001070337719283998,
      "learning_rate": 0.029071247521361674,
      "loss": 0.3777,
      "num_input_tokens_seen": 60510080,
      "step": 31940
    },
    {
      "epoch": 5.211273350191696,
      "grad_norm": 0.0010463594226166606,
      "learning_rate": 0.029036405523652945,
      "loss": 0.3443,
      "num_input_tokens_seen": 60519760,
      "step": 31945
    },
    {
      "epoch": 5.2120890774125135,
      "grad_norm": 0.002859759610146284,
      "learning_rate": 0.029001582180050577,
      "loss": 0.3741,
      "num_input_tokens_seen": 60529264,
      "step": 31950
    },
    {
      "epoch": 5.212904804633331,
      "grad_norm": 0.0020310364197939634,
      "learning_rate": 0.02896677749592482,
      "loss": 0.3076,
      "num_input_tokens_seen": 60539312,
      "step": 31955
    },
    {
      "epoch": 5.213720531854148,
      "grad_norm": 0.0005743713700212538,
      "learning_rate": 0.028931991476642938,
      "loss": 0.3089,
      "num_input_tokens_seen": 60547440,
      "step": 31960
    },
    {
      "epoch": 5.214536259074965,
      "grad_norm": 0.0010099612409248948,
      "learning_rate": 0.028897224127569412,
      "loss": 0.3303,
      "num_input_tokens_seen": 60557040,
      "step": 31965
    },
    {
      "epoch": 5.215351986295783,
      "grad_norm": 0.0009102427866309881,
      "learning_rate": 0.028862475454065832,
      "loss": 0.3084,
      "num_input_tokens_seen": 60566880,
      "step": 31970
    },
    {
      "epoch": 5.2161677135166,
      "grad_norm": 0.0007633366622030735,
      "learning_rate": 0.028827745461490806,
      "loss": 0.3521,
      "num_input_tokens_seen": 60576112,
      "step": 31975
    },
    {
      "epoch": 5.216983440737417,
      "grad_norm": 0.002863978035748005,
      "learning_rate": 0.028793034155200212,
      "loss": 0.3023,
      "num_input_tokens_seen": 60586192,
      "step": 31980
    },
    {
      "epoch": 5.217799167958235,
      "grad_norm": 0.0005537871620617807,
      "learning_rate": 0.028758341540546944,
      "loss": 0.3236,
      "num_input_tokens_seen": 60594224,
      "step": 31985
    },
    {
      "epoch": 5.218614895179052,
      "grad_norm": 0.0008793530869297683,
      "learning_rate": 0.02872366762288098,
      "loss": 0.3181,
      "num_input_tokens_seen": 60603120,
      "step": 31990
    },
    {
      "epoch": 5.219430622399869,
      "grad_norm": 0.001984832575544715,
      "learning_rate": 0.028689012407549567,
      "loss": 0.3468,
      "num_input_tokens_seen": 60612208,
      "step": 31995
    },
    {
      "epoch": 5.220246349620687,
      "grad_norm": 0.00153660390060395,
      "learning_rate": 0.028654375899896892,
      "loss": 0.3155,
      "num_input_tokens_seen": 60622464,
      "step": 32000
    },
    {
      "epoch": 5.220246349620687,
      "eval_loss": 0.3232105076313019,
      "eval_runtime": 152.9408,
      "eval_samples_per_second": 17.817,
      "eval_steps_per_second": 8.912,
      "num_input_tokens_seen": 60622464,
      "step": 32000
    },
    {
      "epoch": 5.221062076841505,
      "grad_norm": 0.001023769611492753,
      "learning_rate": 0.02861975810526437,
      "loss": 0.2801,
      "num_input_tokens_seen": 60631664,
      "step": 32005
    },
    {
      "epoch": 5.221877804062322,
      "grad_norm": 0.0017813146114349365,
      "learning_rate": 0.02858515902899056,
      "loss": 0.2748,
      "num_input_tokens_seen": 60642336,
      "step": 32010
    },
    {
      "epoch": 5.222693531283139,
      "grad_norm": 0.0018075595144182444,
      "learning_rate": 0.028550578676410976,
      "loss": 0.3307,
      "num_input_tokens_seen": 60651504,
      "step": 32015
    },
    {
      "epoch": 5.223509258503956,
      "grad_norm": 0.0016426987713202834,
      "learning_rate": 0.02851601705285837,
      "loss": 0.3109,
      "num_input_tokens_seen": 60661968,
      "step": 32020
    },
    {
      "epoch": 5.224324985724774,
      "grad_norm": 0.002435283502563834,
      "learning_rate": 0.028481474163662666,
      "loss": 0.3605,
      "num_input_tokens_seen": 60671952,
      "step": 32025
    },
    {
      "epoch": 5.225140712945591,
      "grad_norm": 0.0010240785777568817,
      "learning_rate": 0.028446950014150683,
      "loss": 0.3285,
      "num_input_tokens_seen": 60681280,
      "step": 32030
    },
    {
      "epoch": 5.225956440166408,
      "grad_norm": 0.0016611729515716434,
      "learning_rate": 0.028412444609646596,
      "loss": 0.297,
      "num_input_tokens_seen": 60690368,
      "step": 32035
    },
    {
      "epoch": 5.226772167387225,
      "grad_norm": 0.002756134606897831,
      "learning_rate": 0.028377957955471465,
      "loss": 0.3536,
      "num_input_tokens_seen": 60701152,
      "step": 32040
    },
    {
      "epoch": 5.227587894608043,
      "grad_norm": 0.0013329662615433335,
      "learning_rate": 0.0283434900569436,
      "loss": 0.3233,
      "num_input_tokens_seen": 60709904,
      "step": 32045
    },
    {
      "epoch": 5.2284036218288605,
      "grad_norm": 0.000978680676780641,
      "learning_rate": 0.028309040919378456,
      "loss": 0.2806,
      "num_input_tokens_seen": 60719664,
      "step": 32050
    },
    {
      "epoch": 5.229219349049678,
      "grad_norm": 0.0012763069244101644,
      "learning_rate": 0.02827461054808848,
      "loss": 0.3316,
      "num_input_tokens_seen": 60728976,
      "step": 32055
    },
    {
      "epoch": 5.230035076270495,
      "grad_norm": 0.0006457566632889211,
      "learning_rate": 0.028240198948383186,
      "loss": 0.3324,
      "num_input_tokens_seen": 60738880,
      "step": 32060
    },
    {
      "epoch": 5.230850803491313,
      "grad_norm": 0.003498590551316738,
      "learning_rate": 0.028205806125569402,
      "loss": 0.3037,
      "num_input_tokens_seen": 60749488,
      "step": 32065
    },
    {
      "epoch": 5.23166653071213,
      "grad_norm": 0.0014980541309341788,
      "learning_rate": 0.028171432084950834,
      "loss": 0.2986,
      "num_input_tokens_seen": 60760032,
      "step": 32070
    },
    {
      "epoch": 5.232482257932947,
      "grad_norm": 0.0021721890661865473,
      "learning_rate": 0.028137076831828478,
      "loss": 0.3158,
      "num_input_tokens_seen": 60769680,
      "step": 32075
    },
    {
      "epoch": 5.233297985153764,
      "grad_norm": 0.0006465416518040001,
      "learning_rate": 0.028102740371500238,
      "loss": 0.33,
      "num_input_tokens_seen": 60779360,
      "step": 32080
    },
    {
      "epoch": 5.234113712374582,
      "grad_norm": 0.00221125315874815,
      "learning_rate": 0.0280684227092613,
      "loss": 0.3746,
      "num_input_tokens_seen": 60789632,
      "step": 32085
    },
    {
      "epoch": 5.234929439595399,
      "grad_norm": 0.003204997396096587,
      "learning_rate": 0.02803412385040392,
      "loss": 0.3364,
      "num_input_tokens_seen": 60798656,
      "step": 32090
    },
    {
      "epoch": 5.2357451668162165,
      "grad_norm": 0.001185577712021768,
      "learning_rate": 0.027999843800217306,
      "loss": 0.3489,
      "num_input_tokens_seen": 60807408,
      "step": 32095
    },
    {
      "epoch": 5.236560894037034,
      "grad_norm": 0.001222982187755406,
      "learning_rate": 0.027965582563987932,
      "loss": 0.3746,
      "num_input_tokens_seen": 60817264,
      "step": 32100
    },
    {
      "epoch": 5.237376621257852,
      "grad_norm": 0.0014125554589554667,
      "learning_rate": 0.027931340146999346,
      "loss": 0.2908,
      "num_input_tokens_seen": 60828096,
      "step": 32105
    },
    {
      "epoch": 5.238192348478669,
      "grad_norm": 0.00064658880000934,
      "learning_rate": 0.02789711655453208,
      "loss": 0.3192,
      "num_input_tokens_seen": 60837536,
      "step": 32110
    },
    {
      "epoch": 5.239008075699486,
      "grad_norm": 0.00048540046554990113,
      "learning_rate": 0.02786291179186392,
      "loss": 0.3248,
      "num_input_tokens_seen": 60846256,
      "step": 32115
    },
    {
      "epoch": 5.239823802920303,
      "grad_norm": 0.0008340830681845546,
      "learning_rate": 0.02782872586426961,
      "loss": 0.3316,
      "num_input_tokens_seen": 60856288,
      "step": 32120
    },
    {
      "epoch": 5.240639530141121,
      "grad_norm": 0.0005782184889540076,
      "learning_rate": 0.027794558777021083,
      "loss": 0.3492,
      "num_input_tokens_seen": 60865696,
      "step": 32125
    },
    {
      "epoch": 5.241455257361938,
      "grad_norm": 0.0013642648700624704,
      "learning_rate": 0.02776041053538734,
      "loss": 0.3453,
      "num_input_tokens_seen": 60875632,
      "step": 32130
    },
    {
      "epoch": 5.242270984582755,
      "grad_norm": 0.0026639425195753574,
      "learning_rate": 0.027726281144634407,
      "loss": 0.3181,
      "num_input_tokens_seen": 60884976,
      "step": 32135
    },
    {
      "epoch": 5.243086711803572,
      "grad_norm": 0.0009003148297779262,
      "learning_rate": 0.02769217061002552,
      "loss": 0.3526,
      "num_input_tokens_seen": 60894976,
      "step": 32140
    },
    {
      "epoch": 5.2439024390243905,
      "grad_norm": 0.0017759400652721524,
      "learning_rate": 0.027658078936820967,
      "loss": 0.3285,
      "num_input_tokens_seen": 60904288,
      "step": 32145
    },
    {
      "epoch": 5.244718166245208,
      "grad_norm": 0.002728536259382963,
      "learning_rate": 0.02762400613027805,
      "loss": 0.3441,
      "num_input_tokens_seen": 60913504,
      "step": 32150
    },
    {
      "epoch": 5.245533893466025,
      "grad_norm": 0.001747969421558082,
      "learning_rate": 0.027589952195651295,
      "loss": 0.3242,
      "num_input_tokens_seen": 60922640,
      "step": 32155
    },
    {
      "epoch": 5.246349620686843,
      "grad_norm": 0.0024307200219482183,
      "learning_rate": 0.027555917138192186,
      "loss": 0.2706,
      "num_input_tokens_seen": 60931808,
      "step": 32160
    },
    {
      "epoch": 5.24716534790766,
      "grad_norm": 0.0006511944229714572,
      "learning_rate": 0.027521900963149375,
      "loss": 0.2946,
      "num_input_tokens_seen": 60942096,
      "step": 32165
    },
    {
      "epoch": 5.247981075128477,
      "grad_norm": 0.0021397091913968325,
      "learning_rate": 0.027487903675768633,
      "loss": 0.3524,
      "num_input_tokens_seen": 60951168,
      "step": 32170
    },
    {
      "epoch": 5.248796802349294,
      "grad_norm": 0.0015929745277389884,
      "learning_rate": 0.027453925281292677,
      "loss": 0.3233,
      "num_input_tokens_seen": 60959504,
      "step": 32175
    },
    {
      "epoch": 5.249612529570112,
      "grad_norm": 0.0032904683612287045,
      "learning_rate": 0.027419965784961475,
      "loss": 0.3321,
      "num_input_tokens_seen": 60968320,
      "step": 32180
    },
    {
      "epoch": 5.250428256790929,
      "grad_norm": 0.0013242494314908981,
      "learning_rate": 0.027386025192012015,
      "loss": 0.3429,
      "num_input_tokens_seen": 60978800,
      "step": 32185
    },
    {
      "epoch": 5.251243984011746,
      "grad_norm": 0.0013153899926692247,
      "learning_rate": 0.027352103507678277,
      "loss": 0.368,
      "num_input_tokens_seen": 60988496,
      "step": 32190
    },
    {
      "epoch": 5.2520597112325635,
      "grad_norm": 0.0010681886924430728,
      "learning_rate": 0.027318200737191527,
      "loss": 0.3128,
      "num_input_tokens_seen": 60997776,
      "step": 32195
    },
    {
      "epoch": 5.252875438453382,
      "grad_norm": 0.0018728295108303428,
      "learning_rate": 0.027284316885779935,
      "loss": 0.3201,
      "num_input_tokens_seen": 61006768,
      "step": 32200
    },
    {
      "epoch": 5.252875438453382,
      "eval_loss": 0.3215491473674774,
      "eval_runtime": 152.8711,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 61006768,
      "step": 32200
    },
    {
      "epoch": 5.253691165674199,
      "grad_norm": 0.0007535040494985878,
      "learning_rate": 0.027250451958668785,
      "loss": 0.3554,
      "num_input_tokens_seen": 61015360,
      "step": 32205
    },
    {
      "epoch": 5.254506892895016,
      "grad_norm": 0.00137590360827744,
      "learning_rate": 0.027216605961080536,
      "loss": 0.3376,
      "num_input_tokens_seen": 61023536,
      "step": 32210
    },
    {
      "epoch": 5.255322620115833,
      "grad_norm": 0.00032893926254473627,
      "learning_rate": 0.02718277889823461,
      "loss": 0.3123,
      "num_input_tokens_seen": 61032496,
      "step": 32215
    },
    {
      "epoch": 5.256138347336651,
      "grad_norm": 0.0017133340006694198,
      "learning_rate": 0.027148970775347604,
      "loss": 0.3445,
      "num_input_tokens_seen": 61042336,
      "step": 32220
    },
    {
      "epoch": 5.256954074557468,
      "grad_norm": 0.0015770872123539448,
      "learning_rate": 0.027115181597633174,
      "loss": 0.3419,
      "num_input_tokens_seen": 61052160,
      "step": 32225
    },
    {
      "epoch": 5.257769801778285,
      "grad_norm": 0.0022295035887509584,
      "learning_rate": 0.027081411370301976,
      "loss": 0.2988,
      "num_input_tokens_seen": 61062272,
      "step": 32230
    },
    {
      "epoch": 5.258585528999102,
      "grad_norm": 0.0012814565561711788,
      "learning_rate": 0.027047660098561875,
      "loss": 0.3028,
      "num_input_tokens_seen": 61072096,
      "step": 32235
    },
    {
      "epoch": 5.25940125621992,
      "grad_norm": 0.0019424018682911992,
      "learning_rate": 0.02701392778761766,
      "loss": 0.3477,
      "num_input_tokens_seen": 61081840,
      "step": 32240
    },
    {
      "epoch": 5.2602169834407375,
      "grad_norm": 0.0014479034580290318,
      "learning_rate": 0.02698021444267133,
      "loss": 0.3204,
      "num_input_tokens_seen": 61092224,
      "step": 32245
    },
    {
      "epoch": 5.261032710661555,
      "grad_norm": 0.0009706997079774737,
      "learning_rate": 0.026946520068921915,
      "loss": 0.313,
      "num_input_tokens_seen": 61101072,
      "step": 32250
    },
    {
      "epoch": 5.261848437882372,
      "grad_norm": 0.0013810537056997418,
      "learning_rate": 0.02691284467156547,
      "loss": 0.2831,
      "num_input_tokens_seen": 61109776,
      "step": 32255
    },
    {
      "epoch": 5.26266416510319,
      "grad_norm": 0.0005897755618207157,
      "learning_rate": 0.026879188255795182,
      "loss": 0.3717,
      "num_input_tokens_seen": 61120816,
      "step": 32260
    },
    {
      "epoch": 5.263479892324007,
      "grad_norm": 0.0008635174599476159,
      "learning_rate": 0.026845550826801328,
      "loss": 0.3309,
      "num_input_tokens_seen": 61129808,
      "step": 32265
    },
    {
      "epoch": 5.264295619544824,
      "grad_norm": 0.001258361036889255,
      "learning_rate": 0.02681193238977121,
      "loss": 0.2985,
      "num_input_tokens_seen": 61139472,
      "step": 32270
    },
    {
      "epoch": 5.265111346765641,
      "grad_norm": 0.0015109872911125422,
      "learning_rate": 0.026778332949889145,
      "loss": 0.3228,
      "num_input_tokens_seen": 61148576,
      "step": 32275
    },
    {
      "epoch": 5.265927073986459,
      "grad_norm": 0.000684373895637691,
      "learning_rate": 0.026744752512336673,
      "loss": 0.3546,
      "num_input_tokens_seen": 61158144,
      "step": 32280
    },
    {
      "epoch": 5.266742801207276,
      "grad_norm": 0.003396596759557724,
      "learning_rate": 0.02671119108229225,
      "loss": 0.2828,
      "num_input_tokens_seen": 61167408,
      "step": 32285
    },
    {
      "epoch": 5.2675585284280935,
      "grad_norm": 0.000787517405115068,
      "learning_rate": 0.026677648664931556,
      "loss": 0.3234,
      "num_input_tokens_seen": 61177632,
      "step": 32290
    },
    {
      "epoch": 5.268374255648911,
      "grad_norm": 0.0008569156634621322,
      "learning_rate": 0.026644125265427154,
      "loss": 0.3335,
      "num_input_tokens_seen": 61188048,
      "step": 32295
    },
    {
      "epoch": 5.269189982869729,
      "grad_norm": 0.0018200109479948878,
      "learning_rate": 0.026610620888948822,
      "loss": 0.3492,
      "num_input_tokens_seen": 61195424,
      "step": 32300
    },
    {
      "epoch": 5.270005710090546,
      "grad_norm": 0.0017266386421397328,
      "learning_rate": 0.026577135540663408,
      "loss": 0.2813,
      "num_input_tokens_seen": 61205376,
      "step": 32305
    },
    {
      "epoch": 5.270821437311363,
      "grad_norm": 0.0016306893667206168,
      "learning_rate": 0.026543669225734673,
      "loss": 0.2973,
      "num_input_tokens_seen": 61215488,
      "step": 32310
    },
    {
      "epoch": 5.27163716453218,
      "grad_norm": 0.0013369051739573479,
      "learning_rate": 0.02651022194932363,
      "loss": 0.3541,
      "num_input_tokens_seen": 61224736,
      "step": 32315
    },
    {
      "epoch": 5.272452891752998,
      "grad_norm": 0.0021409224718809128,
      "learning_rate": 0.026476793716588194,
      "loss": 0.3291,
      "num_input_tokens_seen": 61234032,
      "step": 32320
    },
    {
      "epoch": 5.273268618973815,
      "grad_norm": 0.0019952463917434216,
      "learning_rate": 0.026443384532683467,
      "loss": 0.3018,
      "num_input_tokens_seen": 61244336,
      "step": 32325
    },
    {
      "epoch": 5.274084346194632,
      "grad_norm": 0.0024811222683638334,
      "learning_rate": 0.026409994402761584,
      "loss": 0.3408,
      "num_input_tokens_seen": 61253744,
      "step": 32330
    },
    {
      "epoch": 5.27490007341545,
      "grad_norm": 0.0009417360415682197,
      "learning_rate": 0.026376623331971653,
      "loss": 0.311,
      "num_input_tokens_seen": 61263168,
      "step": 32335
    },
    {
      "epoch": 5.275715800636267,
      "grad_norm": 0.0020539704710245132,
      "learning_rate": 0.026343271325459997,
      "loss": 0.3792,
      "num_input_tokens_seen": 61273456,
      "step": 32340
    },
    {
      "epoch": 5.276531527857085,
      "grad_norm": 0.0012886811746284366,
      "learning_rate": 0.02630993838836987,
      "loss": 0.3786,
      "num_input_tokens_seen": 61282528,
      "step": 32345
    },
    {
      "epoch": 5.277347255077902,
      "grad_norm": 0.0010348859941586852,
      "learning_rate": 0.026276624525841584,
      "loss": 0.3112,
      "num_input_tokens_seen": 61290704,
      "step": 32350
    },
    {
      "epoch": 5.27816298229872,
      "grad_norm": 0.0007374308770522475,
      "learning_rate": 0.026243329743012637,
      "loss": 0.3155,
      "num_input_tokens_seen": 61301312,
      "step": 32355
    },
    {
      "epoch": 5.278978709519537,
      "grad_norm": 0.0009401552379131317,
      "learning_rate": 0.026210054045017438,
      "loss": 0.3099,
      "num_input_tokens_seen": 61311936,
      "step": 32360
    },
    {
      "epoch": 5.279794436740354,
      "grad_norm": 0.0020875930786132812,
      "learning_rate": 0.02617679743698755,
      "loss": 0.2942,
      "num_input_tokens_seen": 61321808,
      "step": 32365
    },
    {
      "epoch": 5.280610163961171,
      "grad_norm": 0.0012718377402052283,
      "learning_rate": 0.02614355992405158,
      "loss": 0.3176,
      "num_input_tokens_seen": 61330784,
      "step": 32370
    },
    {
      "epoch": 5.281425891181989,
      "grad_norm": 0.002505979500710964,
      "learning_rate": 0.026110341511335115,
      "loss": 0.3786,
      "num_input_tokens_seen": 61339008,
      "step": 32375
    },
    {
      "epoch": 5.282241618402806,
      "grad_norm": 0.0025689697358757257,
      "learning_rate": 0.02607714220396093,
      "loss": 0.3424,
      "num_input_tokens_seen": 61348656,
      "step": 32380
    },
    {
      "epoch": 5.283057345623623,
      "grad_norm": 0.0011666908394545317,
      "learning_rate": 0.02604396200704869,
      "loss": 0.2618,
      "num_input_tokens_seen": 61357712,
      "step": 32385
    },
    {
      "epoch": 5.2838730728444405,
      "grad_norm": 0.002526539145037532,
      "learning_rate": 0.02601080092571523,
      "loss": 0.3162,
      "num_input_tokens_seen": 61367504,
      "step": 32390
    },
    {
      "epoch": 5.2846888000652585,
      "grad_norm": 0.0011166419135406613,
      "learning_rate": 0.025977658965074455,
      "loss": 0.3009,
      "num_input_tokens_seen": 61376896,
      "step": 32395
    },
    {
      "epoch": 5.285504527286076,
      "grad_norm": 0.0010398462181910872,
      "learning_rate": 0.02594453613023719,
      "loss": 0.3427,
      "num_input_tokens_seen": 61386992,
      "step": 32400
    },
    {
      "epoch": 5.285504527286076,
      "eval_loss": 0.3199571669101715,
      "eval_runtime": 152.8275,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 61386992,
      "step": 32400
    },
    {
      "epoch": 5.286320254506893,
      "grad_norm": 0.001921450369991362,
      "learning_rate": 0.025911432426311443,
      "loss": 0.3119,
      "num_input_tokens_seen": 61397072,
      "step": 32405
    },
    {
      "epoch": 5.28713598172771,
      "grad_norm": 0.001175478333607316,
      "learning_rate": 0.025878347858402234,
      "loss": 0.2931,
      "num_input_tokens_seen": 61406992,
      "step": 32410
    },
    {
      "epoch": 5.287951708948528,
      "grad_norm": 0.002475546905770898,
      "learning_rate": 0.025845282431611598,
      "loss": 0.2994,
      "num_input_tokens_seen": 61415760,
      "step": 32415
    },
    {
      "epoch": 5.288767436169345,
      "grad_norm": 0.0008156939875334501,
      "learning_rate": 0.025812236151038608,
      "loss": 0.3471,
      "num_input_tokens_seen": 61424736,
      "step": 32420
    },
    {
      "epoch": 5.289583163390162,
      "grad_norm": 0.0015590646071359515,
      "learning_rate": 0.025779209021779468,
      "loss": 0.3176,
      "num_input_tokens_seen": 61433856,
      "step": 32425
    },
    {
      "epoch": 5.290398890610979,
      "grad_norm": 0.0023971779737621546,
      "learning_rate": 0.025746201048927324,
      "loss": 0.3231,
      "num_input_tokens_seen": 61443472,
      "step": 32430
    },
    {
      "epoch": 5.291214617831797,
      "grad_norm": 0.0021152065601199865,
      "learning_rate": 0.025713212237572485,
      "loss": 0.322,
      "num_input_tokens_seen": 61453200,
      "step": 32435
    },
    {
      "epoch": 5.2920303450526145,
      "grad_norm": 0.001839383621700108,
      "learning_rate": 0.025680242592802164,
      "loss": 0.2901,
      "num_input_tokens_seen": 61463040,
      "step": 32440
    },
    {
      "epoch": 5.292846072273432,
      "grad_norm": 0.0009293335024267435,
      "learning_rate": 0.02564729211970073,
      "loss": 0.3153,
      "num_input_tokens_seen": 61473184,
      "step": 32445
    },
    {
      "epoch": 5.293661799494249,
      "grad_norm": 0.0028728623874485493,
      "learning_rate": 0.025614360823349617,
      "loss": 0.3233,
      "num_input_tokens_seen": 61481920,
      "step": 32450
    },
    {
      "epoch": 5.294477526715067,
      "grad_norm": 0.003484237240627408,
      "learning_rate": 0.025581448708827146,
      "loss": 0.2861,
      "num_input_tokens_seen": 61492464,
      "step": 32455
    },
    {
      "epoch": 5.295293253935884,
      "grad_norm": 0.0013801477616652846,
      "learning_rate": 0.025548555781208876,
      "loss": 0.3666,
      "num_input_tokens_seen": 61501280,
      "step": 32460
    },
    {
      "epoch": 5.296108981156701,
      "grad_norm": 0.0021167900413274765,
      "learning_rate": 0.02551568204556721,
      "loss": 0.2998,
      "num_input_tokens_seen": 61510416,
      "step": 32465
    },
    {
      "epoch": 5.296924708377518,
      "grad_norm": 0.0011877391953021288,
      "learning_rate": 0.02548282750697173,
      "loss": 0.282,
      "num_input_tokens_seen": 61519616,
      "step": 32470
    },
    {
      "epoch": 5.297740435598336,
      "grad_norm": 0.0016027756500989199,
      "learning_rate": 0.02544999217048909,
      "loss": 0.3465,
      "num_input_tokens_seen": 61528560,
      "step": 32475
    },
    {
      "epoch": 5.298556162819153,
      "grad_norm": 0.00047098982031457126,
      "learning_rate": 0.025417176041182793,
      "loss": 0.3269,
      "num_input_tokens_seen": 61537008,
      "step": 32480
    },
    {
      "epoch": 5.2993718900399704,
      "grad_norm": 0.002478486392647028,
      "learning_rate": 0.025384379124113596,
      "loss": 0.3262,
      "num_input_tokens_seen": 61545344,
      "step": 32485
    },
    {
      "epoch": 5.300187617260788,
      "grad_norm": 0.001287860213778913,
      "learning_rate": 0.025351601424339124,
      "loss": 0.3027,
      "num_input_tokens_seen": 61555296,
      "step": 32490
    },
    {
      "epoch": 5.301003344481606,
      "grad_norm": 0.0021858045365661383,
      "learning_rate": 0.025318842946914184,
      "loss": 0.2969,
      "num_input_tokens_seen": 61565344,
      "step": 32495
    },
    {
      "epoch": 5.301819071702423,
      "grad_norm": 0.0037514911964535713,
      "learning_rate": 0.025286103696890494,
      "loss": 0.3485,
      "num_input_tokens_seen": 61576320,
      "step": 32500
    },
    {
      "epoch": 5.30263479892324,
      "grad_norm": 0.003307736711576581,
      "learning_rate": 0.025253383679316836,
      "loss": 0.341,
      "num_input_tokens_seen": 61586352,
      "step": 32505
    },
    {
      "epoch": 5.303450526144058,
      "grad_norm": 0.0018747571157291532,
      "learning_rate": 0.025220682899239077,
      "loss": 0.3651,
      "num_input_tokens_seen": 61595296,
      "step": 32510
    },
    {
      "epoch": 5.304266253364875,
      "grad_norm": 0.001600913004949689,
      "learning_rate": 0.02518800136170013,
      "loss": 0.3336,
      "num_input_tokens_seen": 61606000,
      "step": 32515
    },
    {
      "epoch": 5.305081980585692,
      "grad_norm": 0.004017449449747801,
      "learning_rate": 0.02515533907173981,
      "loss": 0.3245,
      "num_input_tokens_seen": 61614256,
      "step": 32520
    },
    {
      "epoch": 5.305897707806509,
      "grad_norm": 0.0015702079981565475,
      "learning_rate": 0.025122696034395115,
      "loss": 0.364,
      "num_input_tokens_seen": 61624544,
      "step": 32525
    },
    {
      "epoch": 5.306713435027326,
      "grad_norm": 0.0038244472816586494,
      "learning_rate": 0.025090072254700023,
      "loss": 0.3396,
      "num_input_tokens_seen": 61634144,
      "step": 32530
    },
    {
      "epoch": 5.307529162248144,
      "grad_norm": 0.0006504295743070543,
      "learning_rate": 0.025057467737685468,
      "loss": 0.3412,
      "num_input_tokens_seen": 61643744,
      "step": 32535
    },
    {
      "epoch": 5.3083448894689615,
      "grad_norm": 0.0016637957887724042,
      "learning_rate": 0.025024882488379557,
      "loss": 0.3051,
      "num_input_tokens_seen": 61653120,
      "step": 32540
    },
    {
      "epoch": 5.309160616689779,
      "grad_norm": 0.0024915970861911774,
      "learning_rate": 0.02499231651180727,
      "loss": 0.2991,
      "num_input_tokens_seen": 61662288,
      "step": 32545
    },
    {
      "epoch": 5.309976343910597,
      "grad_norm": 0.0021413175854831934,
      "learning_rate": 0.024959769812990713,
      "loss": 0.3534,
      "num_input_tokens_seen": 61671760,
      "step": 32550
    },
    {
      "epoch": 5.310792071131414,
      "grad_norm": 0.0008886906434781849,
      "learning_rate": 0.024927242396949045,
      "loss": 0.3567,
      "num_input_tokens_seen": 61681152,
      "step": 32555
    },
    {
      "epoch": 5.311607798352231,
      "grad_norm": 0.004376787692308426,
      "learning_rate": 0.02489473426869836,
      "loss": 0.3428,
      "num_input_tokens_seen": 61690672,
      "step": 32560
    },
    {
      "epoch": 5.312423525573048,
      "grad_norm": 0.002128252061083913,
      "learning_rate": 0.024862245433251776,
      "loss": 0.3446,
      "num_input_tokens_seen": 61700192,
      "step": 32565
    },
    {
      "epoch": 5.313239252793866,
      "grad_norm": 0.004997276235371828,
      "learning_rate": 0.024829775895619577,
      "loss": 0.3244,
      "num_input_tokens_seen": 61709872,
      "step": 32570
    },
    {
      "epoch": 5.314054980014683,
      "grad_norm": 0.0005099698319099844,
      "learning_rate": 0.024797325660808882,
      "loss": 0.2733,
      "num_input_tokens_seen": 61720304,
      "step": 32575
    },
    {
      "epoch": 5.3148707072355,
      "grad_norm": 0.0009569987305440009,
      "learning_rate": 0.02476489473382401,
      "loss": 0.3633,
      "num_input_tokens_seen": 61730064,
      "step": 32580
    },
    {
      "epoch": 5.3156864344563175,
      "grad_norm": 0.0024655177257955074,
      "learning_rate": 0.024732483119666127,
      "loss": 0.3423,
      "num_input_tokens_seen": 61739808,
      "step": 32585
    },
    {
      "epoch": 5.3165021616771355,
      "grad_norm": 0.0008153150556609035,
      "learning_rate": 0.024700090823333548,
      "loss": 0.3462,
      "num_input_tokens_seen": 61750240,
      "step": 32590
    },
    {
      "epoch": 5.317317888897953,
      "grad_norm": 0.0006163104553706944,
      "learning_rate": 0.02466771784982163,
      "loss": 0.2769,
      "num_input_tokens_seen": 61759920,
      "step": 32595
    },
    {
      "epoch": 5.31813361611877,
      "grad_norm": 0.001284297090023756,
      "learning_rate": 0.024635364204122594,
      "loss": 0.3274,
      "num_input_tokens_seen": 61770000,
      "step": 32600
    },
    {
      "epoch": 5.31813361611877,
      "eval_loss": 0.3215520679950714,
      "eval_runtime": 152.8762,
      "eval_samples_per_second": 17.825,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 61770000,
      "step": 32600
    },
    {
      "epoch": 5.318949343339587,
      "grad_norm": 0.0016329267527908087,
      "learning_rate": 0.024603029891225852,
      "loss": 0.3331,
      "num_input_tokens_seen": 61780112,
      "step": 32605
    },
    {
      "epoch": 5.319765070560405,
      "grad_norm": 0.0025296215899288654,
      "learning_rate": 0.024570714916117748,
      "loss": 0.2909,
      "num_input_tokens_seen": 61790672,
      "step": 32610
    },
    {
      "epoch": 5.320580797781222,
      "grad_norm": 0.0023691272363066673,
      "learning_rate": 0.024538419283781625,
      "loss": 0.358,
      "num_input_tokens_seen": 61801120,
      "step": 32615
    },
    {
      "epoch": 5.321396525002039,
      "grad_norm": 0.000915434502530843,
      "learning_rate": 0.024506142999197938,
      "loss": 0.2946,
      "num_input_tokens_seen": 61810896,
      "step": 32620
    },
    {
      "epoch": 5.322212252222856,
      "grad_norm": 0.001314707682467997,
      "learning_rate": 0.024473886067344002,
      "loss": 0.3166,
      "num_input_tokens_seen": 61821088,
      "step": 32625
    },
    {
      "epoch": 5.323027979443674,
      "grad_norm": 0.0031032899860292673,
      "learning_rate": 0.02444164849319434,
      "loss": 0.276,
      "num_input_tokens_seen": 61829312,
      "step": 32630
    },
    {
      "epoch": 5.3238437066644915,
      "grad_norm": 0.0023605194874107838,
      "learning_rate": 0.024409430281720306,
      "loss": 0.3136,
      "num_input_tokens_seen": 61839888,
      "step": 32635
    },
    {
      "epoch": 5.324659433885309,
      "grad_norm": 0.003679840127006173,
      "learning_rate": 0.024377231437890428,
      "loss": 0.3154,
      "num_input_tokens_seen": 61850080,
      "step": 32640
    },
    {
      "epoch": 5.325475161106126,
      "grad_norm": 0.002135790418833494,
      "learning_rate": 0.024345051966670115,
      "loss": 0.3142,
      "num_input_tokens_seen": 61859808,
      "step": 32645
    },
    {
      "epoch": 5.326290888326944,
      "grad_norm": 0.0009358113748021424,
      "learning_rate": 0.024312891873021884,
      "loss": 0.3278,
      "num_input_tokens_seen": 61869440,
      "step": 32650
    },
    {
      "epoch": 5.327106615547761,
      "grad_norm": 0.0004986993153579533,
      "learning_rate": 0.024280751161905183,
      "loss": 0.3419,
      "num_input_tokens_seen": 61878512,
      "step": 32655
    },
    {
      "epoch": 5.327922342768578,
      "grad_norm": 0.0029817004688084126,
      "learning_rate": 0.02424862983827658,
      "loss": 0.299,
      "num_input_tokens_seen": 61887328,
      "step": 32660
    },
    {
      "epoch": 5.328738069989395,
      "grad_norm": 0.0033775304909795523,
      "learning_rate": 0.024216527907089495,
      "loss": 0.3508,
      "num_input_tokens_seen": 61897328,
      "step": 32665
    },
    {
      "epoch": 5.329553797210213,
      "grad_norm": 0.0023430127184838057,
      "learning_rate": 0.024184445373294505,
      "loss": 0.3138,
      "num_input_tokens_seen": 61906064,
      "step": 32670
    },
    {
      "epoch": 5.33036952443103,
      "grad_norm": 0.001881321775726974,
      "learning_rate": 0.02415238224183918,
      "loss": 0.3448,
      "num_input_tokens_seen": 61914976,
      "step": 32675
    },
    {
      "epoch": 5.331185251651847,
      "grad_norm": 0.0032362875062972307,
      "learning_rate": 0.024120338517667973,
      "loss": 0.3162,
      "num_input_tokens_seen": 61925216,
      "step": 32680
    },
    {
      "epoch": 5.332000978872665,
      "grad_norm": 0.0003890600346494466,
      "learning_rate": 0.02408831420572247,
      "loss": 0.3276,
      "num_input_tokens_seen": 61934688,
      "step": 32685
    },
    {
      "epoch": 5.332816706093483,
      "grad_norm": 0.003301743185147643,
      "learning_rate": 0.024056309310941264,
      "loss": 0.3419,
      "num_input_tokens_seen": 61943376,
      "step": 32690
    },
    {
      "epoch": 5.3336324333143,
      "grad_norm": 0.0007642484270036221,
      "learning_rate": 0.02402432383825982,
      "loss": 0.3141,
      "num_input_tokens_seen": 61954544,
      "step": 32695
    },
    {
      "epoch": 5.334448160535117,
      "grad_norm": 0.003056643530726433,
      "learning_rate": 0.023992357792610792,
      "loss": 0.3346,
      "num_input_tokens_seen": 61965232,
      "step": 32700
    },
    {
      "epoch": 5.335263887755934,
      "grad_norm": 0.0007121065864339471,
      "learning_rate": 0.0239604111789237,
      "loss": 0.3007,
      "num_input_tokens_seen": 61975680,
      "step": 32705
    },
    {
      "epoch": 5.336079614976752,
      "grad_norm": 0.0021052842494100332,
      "learning_rate": 0.023928484002125095,
      "loss": 0.3512,
      "num_input_tokens_seen": 61985152,
      "step": 32710
    },
    {
      "epoch": 5.336895342197569,
      "grad_norm": 0.0012196252355352044,
      "learning_rate": 0.023896576267138595,
      "loss": 0.3236,
      "num_input_tokens_seen": 61994896,
      "step": 32715
    },
    {
      "epoch": 5.337711069418386,
      "grad_norm": 0.0021664691157639027,
      "learning_rate": 0.02386468797888471,
      "loss": 0.362,
      "num_input_tokens_seen": 62003952,
      "step": 32720
    },
    {
      "epoch": 5.338526796639204,
      "grad_norm": 0.002395355375483632,
      "learning_rate": 0.023832819142281057,
      "loss": 0.2963,
      "num_input_tokens_seen": 62012944,
      "step": 32725
    },
    {
      "epoch": 5.339342523860021,
      "grad_norm": 0.0014332995051518083,
      "learning_rate": 0.02380096976224225,
      "loss": 0.2981,
      "num_input_tokens_seen": 62022704,
      "step": 32730
    },
    {
      "epoch": 5.3401582510808385,
      "grad_norm": 0.0006481263553723693,
      "learning_rate": 0.023769139843679777,
      "loss": 0.2847,
      "num_input_tokens_seen": 62030976,
      "step": 32735
    },
    {
      "epoch": 5.340973978301656,
      "grad_norm": 0.0022593794856220484,
      "learning_rate": 0.023737329391502287,
      "loss": 0.3576,
      "num_input_tokens_seen": 62040816,
      "step": 32740
    },
    {
      "epoch": 5.341789705522474,
      "grad_norm": 0.002804894233122468,
      "learning_rate": 0.023705538410615293,
      "loss": 0.2966,
      "num_input_tokens_seen": 62049712,
      "step": 32745
    },
    {
      "epoch": 5.342605432743291,
      "grad_norm": 0.002120835706591606,
      "learning_rate": 0.023673766905921396,
      "loss": 0.3419,
      "num_input_tokens_seen": 62058720,
      "step": 32750
    },
    {
      "epoch": 5.343421159964108,
      "grad_norm": 0.0006848982302471995,
      "learning_rate": 0.0236420148823202,
      "loss": 0.3075,
      "num_input_tokens_seen": 62068384,
      "step": 32755
    },
    {
      "epoch": 5.344236887184925,
      "grad_norm": 0.0011645805789157748,
      "learning_rate": 0.02361028234470816,
      "loss": 0.3042,
      "num_input_tokens_seen": 62078224,
      "step": 32760
    },
    {
      "epoch": 5.345052614405743,
      "grad_norm": 0.0014373594895005226,
      "learning_rate": 0.023578569297978913,
      "loss": 0.3248,
      "num_input_tokens_seen": 62087360,
      "step": 32765
    },
    {
      "epoch": 5.34586834162656,
      "grad_norm": 0.002257698681205511,
      "learning_rate": 0.023546875747023025,
      "loss": 0.2995,
      "num_input_tokens_seen": 62097120,
      "step": 32770
    },
    {
      "epoch": 5.346684068847377,
      "grad_norm": 0.003416666528210044,
      "learning_rate": 0.02351520169672801,
      "loss": 0.3559,
      "num_input_tokens_seen": 62106784,
      "step": 32775
    },
    {
      "epoch": 5.3474997960681945,
      "grad_norm": 0.0004356729914434254,
      "learning_rate": 0.023483547151978357,
      "loss": 0.2769,
      "num_input_tokens_seen": 62117616,
      "step": 32780
    },
    {
      "epoch": 5.3483155232890125,
      "grad_norm": 0.0014008762082085013,
      "learning_rate": 0.023451912117655675,
      "loss": 0.2988,
      "num_input_tokens_seen": 62128016,
      "step": 32785
    },
    {
      "epoch": 5.34913125050983,
      "grad_norm": 0.0021778002846986055,
      "learning_rate": 0.023420296598638417,
      "loss": 0.284,
      "num_input_tokens_seen": 62136672,
      "step": 32790
    },
    {
      "epoch": 5.349946977730647,
      "grad_norm": 0.0007585683488287032,
      "learning_rate": 0.023388700599802165,
      "loss": 0.3061,
      "num_input_tokens_seen": 62146000,
      "step": 32795
    },
    {
      "epoch": 5.350762704951464,
      "grad_norm": 0.0032834443263709545,
      "learning_rate": 0.023357124126019334,
      "loss": 0.3653,
      "num_input_tokens_seen": 62154640,
      "step": 32800
    },
    {
      "epoch": 5.350762704951464,
      "eval_loss": 0.31990155577659607,
      "eval_runtime": 152.8981,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 62154640,
      "step": 32800
    },
    {
      "epoch": 5.351578432172282,
      "grad_norm": 0.002296973252668977,
      "learning_rate": 0.02332556718215945,
      "loss": 0.3297,
      "num_input_tokens_seen": 62164352,
      "step": 32805
    },
    {
      "epoch": 5.352394159393099,
      "grad_norm": 0.0026297459844499826,
      "learning_rate": 0.023294029773089035,
      "loss": 0.3754,
      "num_input_tokens_seen": 62173280,
      "step": 32810
    },
    {
      "epoch": 5.353209886613916,
      "grad_norm": 0.0011424928670749068,
      "learning_rate": 0.023262511903671484,
      "loss": 0.3122,
      "num_input_tokens_seen": 62183344,
      "step": 32815
    },
    {
      "epoch": 5.354025613834733,
      "grad_norm": 0.003208334092050791,
      "learning_rate": 0.023231013578767324,
      "loss": 0.3071,
      "num_input_tokens_seen": 62192896,
      "step": 32820
    },
    {
      "epoch": 5.354841341055551,
      "grad_norm": 0.0010592784965410829,
      "learning_rate": 0.0231995348032339,
      "loss": 0.3442,
      "num_input_tokens_seen": 62202288,
      "step": 32825
    },
    {
      "epoch": 5.3556570682763684,
      "grad_norm": 0.0028556506149470806,
      "learning_rate": 0.023168075581925685,
      "loss": 0.3587,
      "num_input_tokens_seen": 62212192,
      "step": 32830
    },
    {
      "epoch": 5.356472795497186,
      "grad_norm": 0.003120910143479705,
      "learning_rate": 0.023136635919694126,
      "loss": 0.2914,
      "num_input_tokens_seen": 62221200,
      "step": 32835
    },
    {
      "epoch": 5.357288522718003,
      "grad_norm": 0.0032464966643601656,
      "learning_rate": 0.02310521582138753,
      "loss": 0.377,
      "num_input_tokens_seen": 62231248,
      "step": 32840
    },
    {
      "epoch": 5.358104249938821,
      "grad_norm": 0.0008374928729608655,
      "learning_rate": 0.023073815291851357,
      "loss": 0.3499,
      "num_input_tokens_seen": 62240416,
      "step": 32845
    },
    {
      "epoch": 5.358919977159638,
      "grad_norm": 0.001748130307532847,
      "learning_rate": 0.02304243433592788,
      "loss": 0.3142,
      "num_input_tokens_seen": 62249696,
      "step": 32850
    },
    {
      "epoch": 5.359735704380455,
      "grad_norm": 0.002546088071539998,
      "learning_rate": 0.023011072958456513,
      "loss": 0.3575,
      "num_input_tokens_seen": 62259312,
      "step": 32855
    },
    {
      "epoch": 5.360551431601272,
      "grad_norm": 0.001651560771279037,
      "learning_rate": 0.022979731164273536,
      "loss": 0.3128,
      "num_input_tokens_seen": 62269568,
      "step": 32860
    },
    {
      "epoch": 5.36136715882209,
      "grad_norm": 0.0008642159518785775,
      "learning_rate": 0.022948408958212218,
      "loss": 0.3511,
      "num_input_tokens_seen": 62279040,
      "step": 32865
    },
    {
      "epoch": 5.362182886042907,
      "grad_norm": 0.0025717485696077347,
      "learning_rate": 0.022917106345102876,
      "loss": 0.2901,
      "num_input_tokens_seen": 62288944,
      "step": 32870
    },
    {
      "epoch": 5.362998613263724,
      "grad_norm": 0.0008072039927355945,
      "learning_rate": 0.022885823329772785,
      "loss": 0.3312,
      "num_input_tokens_seen": 62299776,
      "step": 32875
    },
    {
      "epoch": 5.3638143404845415,
      "grad_norm": 0.002060903701931238,
      "learning_rate": 0.02285455991704612,
      "loss": 0.3101,
      "num_input_tokens_seen": 62308464,
      "step": 32880
    },
    {
      "epoch": 5.3646300677053596,
      "grad_norm": 0.0014833958121016622,
      "learning_rate": 0.022823316111744117,
      "loss": 0.3307,
      "num_input_tokens_seen": 62318400,
      "step": 32885
    },
    {
      "epoch": 5.365445794926177,
      "grad_norm": 0.0008170019136741757,
      "learning_rate": 0.022792091918685014,
      "loss": 0.3093,
      "num_input_tokens_seen": 62327808,
      "step": 32890
    },
    {
      "epoch": 5.366261522146994,
      "grad_norm": 0.0008852883474901319,
      "learning_rate": 0.022760887342683906,
      "loss": 0.3111,
      "num_input_tokens_seen": 62336688,
      "step": 32895
    },
    {
      "epoch": 5.367077249367812,
      "grad_norm": 0.0017119095427915454,
      "learning_rate": 0.022729702388552975,
      "loss": 0.283,
      "num_input_tokens_seen": 62346480,
      "step": 32900
    },
    {
      "epoch": 5.367892976588629,
      "grad_norm": 0.002594106597825885,
      "learning_rate": 0.022698537061101292,
      "loss": 0.3305,
      "num_input_tokens_seen": 62357360,
      "step": 32905
    },
    {
      "epoch": 5.368708703809446,
      "grad_norm": 0.0009851730428636074,
      "learning_rate": 0.022667391365134962,
      "loss": 0.3221,
      "num_input_tokens_seen": 62366864,
      "step": 32910
    },
    {
      "epoch": 5.369524431030263,
      "grad_norm": 0.0029379259794950485,
      "learning_rate": 0.022636265305457065,
      "loss": 0.3832,
      "num_input_tokens_seen": 62375936,
      "step": 32915
    },
    {
      "epoch": 5.370340158251081,
      "grad_norm": 0.0019252412021160126,
      "learning_rate": 0.02260515888686764,
      "loss": 0.3699,
      "num_input_tokens_seen": 62385008,
      "step": 32920
    },
    {
      "epoch": 5.371155885471898,
      "grad_norm": 0.004702616948634386,
      "learning_rate": 0.022574072114163596,
      "loss": 0.3257,
      "num_input_tokens_seen": 62394528,
      "step": 32925
    },
    {
      "epoch": 5.3719716126927155,
      "grad_norm": 0.0012529926607385278,
      "learning_rate": 0.022543004992139005,
      "loss": 0.3262,
      "num_input_tokens_seen": 62403408,
      "step": 32930
    },
    {
      "epoch": 5.372787339913533,
      "grad_norm": 0.0024411214981228113,
      "learning_rate": 0.022511957525584745,
      "loss": 0.316,
      "num_input_tokens_seen": 62412624,
      "step": 32935
    },
    {
      "epoch": 5.373603067134351,
      "grad_norm": 0.0015681427903473377,
      "learning_rate": 0.022480929719288778,
      "loss": 0.3185,
      "num_input_tokens_seen": 62422416,
      "step": 32940
    },
    {
      "epoch": 5.374418794355168,
      "grad_norm": 0.0011042430996894836,
      "learning_rate": 0.02244992157803592,
      "loss": 0.2887,
      "num_input_tokens_seen": 62430816,
      "step": 32945
    },
    {
      "epoch": 5.375234521575985,
      "grad_norm": 0.000656616990454495,
      "learning_rate": 0.022418933106608047,
      "loss": 0.3039,
      "num_input_tokens_seen": 62440720,
      "step": 32950
    },
    {
      "epoch": 5.376050248796802,
      "grad_norm": 0.003786609508097172,
      "learning_rate": 0.022387964309784018,
      "loss": 0.3534,
      "num_input_tokens_seen": 62450160,
      "step": 32955
    },
    {
      "epoch": 5.37686597601762,
      "grad_norm": 0.001837985124439001,
      "learning_rate": 0.022357015192339517,
      "loss": 0.3889,
      "num_input_tokens_seen": 62461536,
      "step": 32960
    },
    {
      "epoch": 5.377681703238437,
      "grad_norm": 0.0005536022945307195,
      "learning_rate": 0.02232608575904734,
      "loss": 0.345,
      "num_input_tokens_seen": 62470464,
      "step": 32965
    },
    {
      "epoch": 5.378497430459254,
      "grad_norm": 0.0017932468326762319,
      "learning_rate": 0.022295176014677225,
      "loss": 0.3486,
      "num_input_tokens_seen": 62480368,
      "step": 32970
    },
    {
      "epoch": 5.3793131576800715,
      "grad_norm": 0.0032307696528732777,
      "learning_rate": 0.02226428596399577,
      "loss": 0.3618,
      "num_input_tokens_seen": 62491056,
      "step": 32975
    },
    {
      "epoch": 5.3801288849008895,
      "grad_norm": 0.0018786384025588632,
      "learning_rate": 0.02223341561176669,
      "loss": 0.3056,
      "num_input_tokens_seen": 62502224,
      "step": 32980
    },
    {
      "epoch": 5.380944612121707,
      "grad_norm": 0.002460668794810772,
      "learning_rate": 0.0222025649627505,
      "loss": 0.3114,
      "num_input_tokens_seen": 62512192,
      "step": 32985
    },
    {
      "epoch": 5.381760339342524,
      "grad_norm": 0.0018587153172120452,
      "learning_rate": 0.022171734021704814,
      "loss": 0.2902,
      "num_input_tokens_seen": 62521952,
      "step": 32990
    },
    {
      "epoch": 5.382576066563341,
      "grad_norm": 0.003819328499957919,
      "learning_rate": 0.022140922793384116,
      "loss": 0.3562,
      "num_input_tokens_seen": 62531968,
      "step": 32995
    },
    {
      "epoch": 5.383391793784159,
      "grad_norm": 0.0012556466972455382,
      "learning_rate": 0.022110131282539934,
      "loss": 0.2958,
      "num_input_tokens_seen": 62541664,
      "step": 33000
    },
    {
      "epoch": 5.383391793784159,
      "eval_loss": 0.3205162584781647,
      "eval_runtime": 152.7992,
      "eval_samples_per_second": 17.834,
      "eval_steps_per_second": 8.92,
      "num_input_tokens_seen": 62541664,
      "step": 33000
    },
    {
      "epoch": 5.384207521004976,
      "grad_norm": 0.0026268605142831802,
      "learning_rate": 0.022079359493920675,
      "loss": 0.3078,
      "num_input_tokens_seen": 62551008,
      "step": 33005
    },
    {
      "epoch": 5.385023248225793,
      "grad_norm": 0.0029407215770334005,
      "learning_rate": 0.02204860743227169,
      "loss": 0.3096,
      "num_input_tokens_seen": 62560384,
      "step": 33010
    },
    {
      "epoch": 5.38583897544661,
      "grad_norm": 0.001746544148772955,
      "learning_rate": 0.022017875102335365,
      "loss": 0.3304,
      "num_input_tokens_seen": 62569120,
      "step": 33015
    },
    {
      "epoch": 5.386654702667428,
      "grad_norm": 0.0009281992097385228,
      "learning_rate": 0.02198716250885108,
      "loss": 0.2729,
      "num_input_tokens_seen": 62578592,
      "step": 33020
    },
    {
      "epoch": 5.387470429888245,
      "grad_norm": 0.0026419206988066435,
      "learning_rate": 0.021956469656555,
      "loss": 0.3105,
      "num_input_tokens_seen": 62587792,
      "step": 33025
    },
    {
      "epoch": 5.388286157109063,
      "grad_norm": 0.0010453201830387115,
      "learning_rate": 0.0219257965501804,
      "loss": 0.2875,
      "num_input_tokens_seen": 62597840,
      "step": 33030
    },
    {
      "epoch": 5.38910188432988,
      "grad_norm": 0.000858170329593122,
      "learning_rate": 0.021895143194457494,
      "loss": 0.3479,
      "num_input_tokens_seen": 62608048,
      "step": 33035
    },
    {
      "epoch": 5.389917611550698,
      "grad_norm": 0.0019022709457203746,
      "learning_rate": 0.021864509594113322,
      "loss": 0.3152,
      "num_input_tokens_seen": 62617376,
      "step": 33040
    },
    {
      "epoch": 5.390733338771515,
      "grad_norm": 0.0015232628211379051,
      "learning_rate": 0.02183389575387207,
      "loss": 0.3331,
      "num_input_tokens_seen": 62627776,
      "step": 33045
    },
    {
      "epoch": 5.391549065992332,
      "grad_norm": 0.0010166423162445426,
      "learning_rate": 0.021803301678454682,
      "loss": 0.3026,
      "num_input_tokens_seen": 62636752,
      "step": 33050
    },
    {
      "epoch": 5.392364793213149,
      "grad_norm": 0.0009213088778778911,
      "learning_rate": 0.021772727372579213,
      "loss": 0.3676,
      "num_input_tokens_seen": 62646624,
      "step": 33055
    },
    {
      "epoch": 5.393180520433967,
      "grad_norm": 0.001909960643388331,
      "learning_rate": 0.02174217284096061,
      "loss": 0.2786,
      "num_input_tokens_seen": 62655952,
      "step": 33060
    },
    {
      "epoch": 5.393996247654784,
      "grad_norm": 0.0007593075861223042,
      "learning_rate": 0.0217116380883107,
      "loss": 0.3327,
      "num_input_tokens_seen": 62663872,
      "step": 33065
    },
    {
      "epoch": 5.394811974875601,
      "grad_norm": 0.0017256838036701083,
      "learning_rate": 0.021681123119338425,
      "loss": 0.3058,
      "num_input_tokens_seen": 62673600,
      "step": 33070
    },
    {
      "epoch": 5.395627702096419,
      "grad_norm": 0.0014354211743921041,
      "learning_rate": 0.02165062793874951,
      "loss": 0.2932,
      "num_input_tokens_seen": 62683552,
      "step": 33075
    },
    {
      "epoch": 5.3964434293172365,
      "grad_norm": 0.0015875579556450248,
      "learning_rate": 0.021620152551246666,
      "loss": 0.312,
      "num_input_tokens_seen": 62691968,
      "step": 33080
    },
    {
      "epoch": 5.397259156538054,
      "grad_norm": 0.0007881326600909233,
      "learning_rate": 0.02158969696152967,
      "loss": 0.2849,
      "num_input_tokens_seen": 62701872,
      "step": 33085
    },
    {
      "epoch": 5.398074883758871,
      "grad_norm": 0.0004556253843475133,
      "learning_rate": 0.021559261174295057,
      "loss": 0.2784,
      "num_input_tokens_seen": 62711776,
      "step": 33090
    },
    {
      "epoch": 5.398890610979688,
      "grad_norm": 0.004185603000223637,
      "learning_rate": 0.02152884519423646,
      "loss": 0.291,
      "num_input_tokens_seen": 62721888,
      "step": 33095
    },
    {
      "epoch": 5.399706338200506,
      "grad_norm": 0.0021950260270386934,
      "learning_rate": 0.021498449026044447,
      "loss": 0.3364,
      "num_input_tokens_seen": 62730464,
      "step": 33100
    },
    {
      "epoch": 5.400522065421323,
      "grad_norm": 0.0012010021600872278,
      "learning_rate": 0.021468072674406414,
      "loss": 0.3549,
      "num_input_tokens_seen": 62739536,
      "step": 33105
    },
    {
      "epoch": 5.40133779264214,
      "grad_norm": 0.0003309854364488274,
      "learning_rate": 0.021437716144006795,
      "loss": 0.3629,
      "num_input_tokens_seen": 62749312,
      "step": 33110
    },
    {
      "epoch": 5.402153519862958,
      "grad_norm": 0.0007674672524444759,
      "learning_rate": 0.021407379439527002,
      "loss": 0.3011,
      "num_input_tokens_seen": 62758496,
      "step": 33115
    },
    {
      "epoch": 5.402969247083775,
      "grad_norm": 0.0014838160714134574,
      "learning_rate": 0.021377062565645255,
      "loss": 0.3122,
      "num_input_tokens_seen": 62766896,
      "step": 33120
    },
    {
      "epoch": 5.4037849743045925,
      "grad_norm": 0.0027477701660245657,
      "learning_rate": 0.02134676552703688,
      "loss": 0.3291,
      "num_input_tokens_seen": 62775792,
      "step": 33125
    },
    {
      "epoch": 5.40460070152541,
      "grad_norm": 0.0008684277418069541,
      "learning_rate": 0.02131648832837398,
      "loss": 0.2977,
      "num_input_tokens_seen": 62785424,
      "step": 33130
    },
    {
      "epoch": 5.405416428746228,
      "grad_norm": 0.001499815029092133,
      "learning_rate": 0.02128623097432574,
      "loss": 0.3282,
      "num_input_tokens_seen": 62794624,
      "step": 33135
    },
    {
      "epoch": 5.406232155967045,
      "grad_norm": 0.0004524104588199407,
      "learning_rate": 0.021255993469558192,
      "loss": 0.3007,
      "num_input_tokens_seen": 62804032,
      "step": 33140
    },
    {
      "epoch": 5.407047883187862,
      "grad_norm": 0.0007982086972333491,
      "learning_rate": 0.021225775818734364,
      "loss": 0.3328,
      "num_input_tokens_seen": 62813920,
      "step": 33145
    },
    {
      "epoch": 5.407863610408679,
      "grad_norm": 0.0023513082414865494,
      "learning_rate": 0.021195578026514166,
      "loss": 0.3259,
      "num_input_tokens_seen": 62821280,
      "step": 33150
    },
    {
      "epoch": 5.408679337629497,
      "grad_norm": 0.002813727129250765,
      "learning_rate": 0.02116540009755452,
      "loss": 0.3077,
      "num_input_tokens_seen": 62830368,
      "step": 33155
    },
    {
      "epoch": 5.409495064850314,
      "grad_norm": 0.0017433767206966877,
      "learning_rate": 0.021135242036509173,
      "loss": 0.3444,
      "num_input_tokens_seen": 62839456,
      "step": 33160
    },
    {
      "epoch": 5.410310792071131,
      "grad_norm": 0.002431028988212347,
      "learning_rate": 0.021105103848028967,
      "loss": 0.3584,
      "num_input_tokens_seen": 62848608,
      "step": 33165
    },
    {
      "epoch": 5.411126519291948,
      "grad_norm": 0.0021147418301552534,
      "learning_rate": 0.021074985536761504,
      "loss": 0.3483,
      "num_input_tokens_seen": 62858128,
      "step": 33170
    },
    {
      "epoch": 5.4119422465127665,
      "grad_norm": 0.0014830545987933874,
      "learning_rate": 0.021044887107351435,
      "loss": 0.3198,
      "num_input_tokens_seen": 62866752,
      "step": 33175
    },
    {
      "epoch": 5.412757973733584,
      "grad_norm": 0.0007661718409508467,
      "learning_rate": 0.021014808564440362,
      "loss": 0.301,
      "num_input_tokens_seen": 62876576,
      "step": 33180
    },
    {
      "epoch": 5.413573700954401,
      "grad_norm": 0.0006693999166600406,
      "learning_rate": 0.02098474991266671,
      "loss": 0.2633,
      "num_input_tokens_seen": 62885760,
      "step": 33185
    },
    {
      "epoch": 5.414389428175218,
      "grad_norm": 0.002589740324765444,
      "learning_rate": 0.02095471115666592,
      "loss": 0.3599,
      "num_input_tokens_seen": 62895984,
      "step": 33190
    },
    {
      "epoch": 5.415205155396036,
      "grad_norm": 0.0024767343420535326,
      "learning_rate": 0.020924692301070406,
      "loss": 0.3403,
      "num_input_tokens_seen": 62904096,
      "step": 33195
    },
    {
      "epoch": 5.416020882616853,
      "grad_norm": 0.0018090528901666403,
      "learning_rate": 0.020894693350509346,
      "loss": 0.2897,
      "num_input_tokens_seen": 62912976,
      "step": 33200
    },
    {
      "epoch": 5.416020882616853,
      "eval_loss": 0.31929609179496765,
      "eval_runtime": 152.8402,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 62912976,
      "step": 33200
    },
    {
      "epoch": 5.41683660983767,
      "grad_norm": 0.0020029789302498102,
      "learning_rate": 0.020864714309609057,
      "loss": 0.3703,
      "num_input_tokens_seen": 62922112,
      "step": 33205
    },
    {
      "epoch": 5.417652337058487,
      "grad_norm": 0.002143639139831066,
      "learning_rate": 0.020834755182992604,
      "loss": 0.3482,
      "num_input_tokens_seen": 62931136,
      "step": 33210
    },
    {
      "epoch": 5.418468064279305,
      "grad_norm": 0.0017389323329553008,
      "learning_rate": 0.02080481597528011,
      "loss": 0.2606,
      "num_input_tokens_seen": 62940384,
      "step": 33215
    },
    {
      "epoch": 5.419283791500122,
      "grad_norm": 0.0010763995815068483,
      "learning_rate": 0.020774896691088583,
      "loss": 0.3538,
      "num_input_tokens_seen": 62948224,
      "step": 33220
    },
    {
      "epoch": 5.4200995187209395,
      "grad_norm": 0.0007019622717052698,
      "learning_rate": 0.020744997335031882,
      "loss": 0.3704,
      "num_input_tokens_seen": 62957904,
      "step": 33225
    },
    {
      "epoch": 5.420915245941757,
      "grad_norm": 0.0028017982840538025,
      "learning_rate": 0.02071511791172092,
      "loss": 0.3177,
      "num_input_tokens_seen": 62968384,
      "step": 33230
    },
    {
      "epoch": 5.421730973162575,
      "grad_norm": 0.001853015972301364,
      "learning_rate": 0.02068525842576351,
      "loss": 0.3258,
      "num_input_tokens_seen": 62977376,
      "step": 33235
    },
    {
      "epoch": 5.422546700383392,
      "grad_norm": 0.0025438552256673574,
      "learning_rate": 0.020655418881764264,
      "loss": 0.3304,
      "num_input_tokens_seen": 62986912,
      "step": 33240
    },
    {
      "epoch": 5.423362427604209,
      "grad_norm": 0.0025184431578963995,
      "learning_rate": 0.020625599284324923,
      "loss": 0.357,
      "num_input_tokens_seen": 62995968,
      "step": 33245
    },
    {
      "epoch": 5.424178154825027,
      "grad_norm": 0.0007345188641920686,
      "learning_rate": 0.02059579963804396,
      "loss": 0.3493,
      "num_input_tokens_seen": 63005696,
      "step": 33250
    },
    {
      "epoch": 5.424993882045844,
      "grad_norm": 0.0011079959804192185,
      "learning_rate": 0.02056601994751688,
      "loss": 0.348,
      "num_input_tokens_seen": 63016064,
      "step": 33255
    },
    {
      "epoch": 5.425809609266661,
      "grad_norm": 0.00149555504322052,
      "learning_rate": 0.02053626021733614,
      "loss": 0.2786,
      "num_input_tokens_seen": 63025888,
      "step": 33260
    },
    {
      "epoch": 5.426625336487478,
      "grad_norm": 0.0013933221343904734,
      "learning_rate": 0.02050652045209097,
      "loss": 0.3196,
      "num_input_tokens_seen": 63034912,
      "step": 33265
    },
    {
      "epoch": 5.4274410637082955,
      "grad_norm": 0.0022601820528507233,
      "learning_rate": 0.020476800656367672,
      "loss": 0.3474,
      "num_input_tokens_seen": 63044624,
      "step": 33270
    },
    {
      "epoch": 5.4282567909291135,
      "grad_norm": 0.0016090137651190162,
      "learning_rate": 0.020447100834749425,
      "loss": 0.2973,
      "num_input_tokens_seen": 63054304,
      "step": 33275
    },
    {
      "epoch": 5.429072518149931,
      "grad_norm": 0.0015908456407487392,
      "learning_rate": 0.02041742099181627,
      "loss": 0.2834,
      "num_input_tokens_seen": 63063536,
      "step": 33280
    },
    {
      "epoch": 5.429888245370748,
      "grad_norm": 0.0007238286198116839,
      "learning_rate": 0.02038776113214526,
      "loss": 0.3135,
      "num_input_tokens_seen": 63072688,
      "step": 33285
    },
    {
      "epoch": 5.430703972591566,
      "grad_norm": 0.0021961689926683903,
      "learning_rate": 0.0203581212603103,
      "loss": 0.2776,
      "num_input_tokens_seen": 63081136,
      "step": 33290
    },
    {
      "epoch": 5.431519699812383,
      "grad_norm": 0.0022551822476089,
      "learning_rate": 0.02032850138088219,
      "loss": 0.2416,
      "num_input_tokens_seen": 63090512,
      "step": 33295
    },
    {
      "epoch": 5.4323354270332,
      "grad_norm": 0.0019276103703305125,
      "learning_rate": 0.020298901498428754,
      "loss": 0.3227,
      "num_input_tokens_seen": 63099472,
      "step": 33300
    },
    {
      "epoch": 5.433151154254017,
      "grad_norm": 0.0019789678044617176,
      "learning_rate": 0.020269321617514595,
      "loss": 0.3179,
      "num_input_tokens_seen": 63108528,
      "step": 33305
    },
    {
      "epoch": 5.433966881474835,
      "grad_norm": 0.0020463766995817423,
      "learning_rate": 0.020239761742701343,
      "loss": 0.3363,
      "num_input_tokens_seen": 63119248,
      "step": 33310
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 0.001474916934967041,
      "learning_rate": 0.02021022187854754,
      "loss": 0.3061,
      "num_input_tokens_seen": 63128256,
      "step": 33315
    },
    {
      "epoch": 5.4355983359164695,
      "grad_norm": 0.0011478669475764036,
      "learning_rate": 0.020180702029608522,
      "loss": 0.3337,
      "num_input_tokens_seen": 63137744,
      "step": 33320
    },
    {
      "epoch": 5.436414063137287,
      "grad_norm": 0.001879100571386516,
      "learning_rate": 0.020151202200436695,
      "loss": 0.3023,
      "num_input_tokens_seen": 63146688,
      "step": 33325
    },
    {
      "epoch": 5.437229790358105,
      "grad_norm": 0.002527831355109811,
      "learning_rate": 0.020121722395581226,
      "loss": 0.2848,
      "num_input_tokens_seen": 63156528,
      "step": 33330
    },
    {
      "epoch": 5.438045517578922,
      "grad_norm": 0.0015148199163377285,
      "learning_rate": 0.020092262619588342,
      "loss": 0.2623,
      "num_input_tokens_seen": 63166032,
      "step": 33335
    },
    {
      "epoch": 5.438861244799739,
      "grad_norm": 0.0010444202926009893,
      "learning_rate": 0.02006282287700109,
      "loss": 0.3021,
      "num_input_tokens_seen": 63175392,
      "step": 33340
    },
    {
      "epoch": 5.439676972020556,
      "grad_norm": 0.0019142840756103396,
      "learning_rate": 0.020033403172359427,
      "loss": 0.4019,
      "num_input_tokens_seen": 63185312,
      "step": 33345
    },
    {
      "epoch": 5.440492699241374,
      "grad_norm": 0.0009417124092578888,
      "learning_rate": 0.020004003510200284,
      "loss": 0.3488,
      "num_input_tokens_seen": 63195584,
      "step": 33350
    },
    {
      "epoch": 5.441308426462191,
      "grad_norm": 0.0007866175146773458,
      "learning_rate": 0.019974623895057407,
      "loss": 0.3453,
      "num_input_tokens_seen": 63205184,
      "step": 33355
    },
    {
      "epoch": 5.442124153683008,
      "grad_norm": 0.002684013918042183,
      "learning_rate": 0.019945264331461553,
      "loss": 0.3359,
      "num_input_tokens_seen": 63214432,
      "step": 33360
    },
    {
      "epoch": 5.442939880903825,
      "grad_norm": 0.0036560758017003536,
      "learning_rate": 0.019915924823940317,
      "loss": 0.3189,
      "num_input_tokens_seen": 63224208,
      "step": 33365
    },
    {
      "epoch": 5.443755608124643,
      "grad_norm": 0.0015159640461206436,
      "learning_rate": 0.01988660537701816,
      "loss": 0.3211,
      "num_input_tokens_seen": 63233952,
      "step": 33370
    },
    {
      "epoch": 5.444571335345461,
      "grad_norm": 0.002205482218414545,
      "learning_rate": 0.01985730599521659,
      "loss": 0.3868,
      "num_input_tokens_seen": 63244000,
      "step": 33375
    },
    {
      "epoch": 5.445387062566278,
      "grad_norm": 0.0010615804931148887,
      "learning_rate": 0.019828026683053918,
      "loss": 0.306,
      "num_input_tokens_seen": 63252064,
      "step": 33380
    },
    {
      "epoch": 5.446202789787095,
      "grad_norm": 0.0015622612554579973,
      "learning_rate": 0.01979876744504535,
      "loss": 0.32,
      "num_input_tokens_seen": 63262416,
      "step": 33385
    },
    {
      "epoch": 5.447018517007913,
      "grad_norm": 0.0020070471800863743,
      "learning_rate": 0.019769528285703046,
      "loss": 0.3781,
      "num_input_tokens_seen": 63271440,
      "step": 33390
    },
    {
      "epoch": 5.44783424422873,
      "grad_norm": 0.0010882263304665685,
      "learning_rate": 0.019740309209536098,
      "loss": 0.3168,
      "num_input_tokens_seen": 63280976,
      "step": 33395
    },
    {
      "epoch": 5.448649971449547,
      "grad_norm": 0.0013303069863468409,
      "learning_rate": 0.019711110221050387,
      "loss": 0.3029,
      "num_input_tokens_seen": 63289520,
      "step": 33400
    },
    {
      "epoch": 5.448649971449547,
      "eval_loss": 0.32216179370880127,
      "eval_runtime": 152.8617,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 63289520,
      "step": 33400
    },
    {
      "epoch": 5.449465698670364,
      "grad_norm": 0.0025205889251083136,
      "learning_rate": 0.019681931324748825,
      "loss": 0.3745,
      "num_input_tokens_seen": 63298400,
      "step": 33405
    },
    {
      "epoch": 5.450281425891182,
      "grad_norm": 0.002351972507312894,
      "learning_rate": 0.019652772525131094,
      "loss": 0.322,
      "num_input_tokens_seen": 63307808,
      "step": 33410
    },
    {
      "epoch": 5.451097153111999,
      "grad_norm": 0.0027832123450934887,
      "learning_rate": 0.019623633826693885,
      "loss": 0.295,
      "num_input_tokens_seen": 63316560,
      "step": 33415
    },
    {
      "epoch": 5.4519128803328165,
      "grad_norm": 0.0026881173253059387,
      "learning_rate": 0.019594515233930788,
      "loss": 0.3416,
      "num_input_tokens_seen": 63326288,
      "step": 33420
    },
    {
      "epoch": 5.4527286075536345,
      "grad_norm": 0.00041886887629516423,
      "learning_rate": 0.019565416751332186,
      "loss": 0.3432,
      "num_input_tokens_seen": 63335728,
      "step": 33425
    },
    {
      "epoch": 5.453544334774452,
      "grad_norm": 0.002501665148884058,
      "learning_rate": 0.019536338383385497,
      "loss": 0.3819,
      "num_input_tokens_seen": 63345168,
      "step": 33430
    },
    {
      "epoch": 5.454360061995269,
      "grad_norm": 0.004584080073982477,
      "learning_rate": 0.019507280134574933,
      "loss": 0.2775,
      "num_input_tokens_seen": 63355440,
      "step": 33435
    },
    {
      "epoch": 5.455175789216086,
      "grad_norm": 0.0007958265487104654,
      "learning_rate": 0.019478242009381624,
      "loss": 0.3209,
      "num_input_tokens_seen": 63365088,
      "step": 33440
    },
    {
      "epoch": 5.455991516436903,
      "grad_norm": 0.0026468553114682436,
      "learning_rate": 0.01944922401228367,
      "loss": 0.321,
      "num_input_tokens_seen": 63374240,
      "step": 33445
    },
    {
      "epoch": 5.456807243657721,
      "grad_norm": 0.003140017855912447,
      "learning_rate": 0.01942022614775593,
      "loss": 0.3361,
      "num_input_tokens_seen": 63383120,
      "step": 33450
    },
    {
      "epoch": 5.457622970878538,
      "grad_norm": 0.002281015506014228,
      "learning_rate": 0.01939124842027029,
      "loss": 0.3272,
      "num_input_tokens_seen": 63392128,
      "step": 33455
    },
    {
      "epoch": 5.458438698099355,
      "grad_norm": 0.003974736202508211,
      "learning_rate": 0.01936229083429551,
      "loss": 0.3162,
      "num_input_tokens_seen": 63400864,
      "step": 33460
    },
    {
      "epoch": 5.459254425320173,
      "grad_norm": 0.00269168708473444,
      "learning_rate": 0.019333353394297148,
      "loss": 0.3213,
      "num_input_tokens_seen": 63409824,
      "step": 33465
    },
    {
      "epoch": 5.4600701525409905,
      "grad_norm": 0.00187290133908391,
      "learning_rate": 0.019304436104737754,
      "loss": 0.3469,
      "num_input_tokens_seen": 63419824,
      "step": 33470
    },
    {
      "epoch": 5.460885879761808,
      "grad_norm": 0.0011676271678879857,
      "learning_rate": 0.019275538970076778,
      "loss": 0.2734,
      "num_input_tokens_seen": 63429712,
      "step": 33475
    },
    {
      "epoch": 5.461701606982625,
      "grad_norm": 0.000812985934317112,
      "learning_rate": 0.019246661994770434,
      "loss": 0.3192,
      "num_input_tokens_seen": 63439360,
      "step": 33480
    },
    {
      "epoch": 5.462517334203443,
      "grad_norm": 0.001722289132885635,
      "learning_rate": 0.019217805183271985,
      "loss": 0.3182,
      "num_input_tokens_seen": 63449696,
      "step": 33485
    },
    {
      "epoch": 5.46333306142426,
      "grad_norm": 0.0017437952337786555,
      "learning_rate": 0.019188968540031465,
      "loss": 0.273,
      "num_input_tokens_seen": 63459856,
      "step": 33490
    },
    {
      "epoch": 5.464148788645077,
      "grad_norm": 0.0016326485201716423,
      "learning_rate": 0.019160152069495867,
      "loss": 0.3635,
      "num_input_tokens_seen": 63467824,
      "step": 33495
    },
    {
      "epoch": 5.464964515865894,
      "grad_norm": 0.0018173346761614084,
      "learning_rate": 0.019131355776109103,
      "loss": 0.2961,
      "num_input_tokens_seen": 63477664,
      "step": 33500
    },
    {
      "epoch": 5.465780243086712,
      "grad_norm": 0.0023196530528366566,
      "learning_rate": 0.019102579664311857,
      "loss": 0.4055,
      "num_input_tokens_seen": 63486544,
      "step": 33505
    },
    {
      "epoch": 5.466595970307529,
      "grad_norm": 0.0016288934275507927,
      "learning_rate": 0.019073823738541763,
      "loss": 0.3421,
      "num_input_tokens_seen": 63495616,
      "step": 33510
    },
    {
      "epoch": 5.467411697528346,
      "grad_norm": 0.0006792476051487029,
      "learning_rate": 0.0190450880032334,
      "loss": 0.3767,
      "num_input_tokens_seen": 63504608,
      "step": 33515
    },
    {
      "epoch": 5.468227424749164,
      "grad_norm": 0.002909185132011771,
      "learning_rate": 0.019016372462818114,
      "loss": 0.3236,
      "num_input_tokens_seen": 63514160,
      "step": 33520
    },
    {
      "epoch": 5.469043151969982,
      "grad_norm": 0.0007362756878137589,
      "learning_rate": 0.018987677121724278,
      "loss": 0.3409,
      "num_input_tokens_seen": 63524272,
      "step": 33525
    },
    {
      "epoch": 5.469858879190799,
      "grad_norm": 0.0017193377716466784,
      "learning_rate": 0.018959001984377,
      "loss": 0.2836,
      "num_input_tokens_seen": 63534048,
      "step": 33530
    },
    {
      "epoch": 5.470674606411616,
      "grad_norm": 0.0011426947312429547,
      "learning_rate": 0.018930347055198377,
      "loss": 0.3558,
      "num_input_tokens_seen": 63543920,
      "step": 33535
    },
    {
      "epoch": 5.471490333632433,
      "grad_norm": 0.003552586305886507,
      "learning_rate": 0.01890171233860739,
      "loss": 0.2988,
      "num_input_tokens_seen": 63553072,
      "step": 33540
    },
    {
      "epoch": 5.472306060853251,
      "grad_norm": 0.0023127635940909386,
      "learning_rate": 0.018873097839019807,
      "loss": 0.2982,
      "num_input_tokens_seen": 63561648,
      "step": 33545
    },
    {
      "epoch": 5.473121788074068,
      "grad_norm": 0.0018730119336396456,
      "learning_rate": 0.0188445035608484,
      "loss": 0.3009,
      "num_input_tokens_seen": 63570320,
      "step": 33550
    },
    {
      "epoch": 5.473937515294885,
      "grad_norm": 0.0011637313291430473,
      "learning_rate": 0.018815929508502777,
      "loss": 0.2982,
      "num_input_tokens_seen": 63580576,
      "step": 33555
    },
    {
      "epoch": 5.474753242515702,
      "grad_norm": 0.0014570386847481132,
      "learning_rate": 0.01878737568638934,
      "loss": 0.3044,
      "num_input_tokens_seen": 63591232,
      "step": 33560
    },
    {
      "epoch": 5.47556896973652,
      "grad_norm": 0.003606901504099369,
      "learning_rate": 0.01875884209891152,
      "loss": 0.3633,
      "num_input_tokens_seen": 63601232,
      "step": 33565
    },
    {
      "epoch": 5.4763846969573375,
      "grad_norm": 0.0027593292761594057,
      "learning_rate": 0.018730328750469514,
      "loss": 0.357,
      "num_input_tokens_seen": 63611264,
      "step": 33570
    },
    {
      "epoch": 5.477200424178155,
      "grad_norm": 0.0009738143999129534,
      "learning_rate": 0.018701835645460473,
      "loss": 0.359,
      "num_input_tokens_seen": 63620976,
      "step": 33575
    },
    {
      "epoch": 5.478016151398972,
      "grad_norm": 0.002645037369802594,
      "learning_rate": 0.01867336278827838,
      "loss": 0.3744,
      "num_input_tokens_seen": 63630880,
      "step": 33580
    },
    {
      "epoch": 5.47883187861979,
      "grad_norm": 0.0012000323040410876,
      "learning_rate": 0.018644910183314056,
      "loss": 0.3217,
      "num_input_tokens_seen": 63640704,
      "step": 33585
    },
    {
      "epoch": 5.479647605840607,
      "grad_norm": 0.0024214470759034157,
      "learning_rate": 0.01861647783495531,
      "loss": 0.3815,
      "num_input_tokens_seen": 63649504,
      "step": 33590
    },
    {
      "epoch": 5.480463333061424,
      "grad_norm": 0.00065617635846138,
      "learning_rate": 0.01858806574758676,
      "loss": 0.3321,
      "num_input_tokens_seen": 63658448,
      "step": 33595
    },
    {
      "epoch": 5.481279060282241,
      "grad_norm": 0.0014554043300449848,
      "learning_rate": 0.01855967392558988,
      "loss": 0.3473,
      "num_input_tokens_seen": 63668416,
      "step": 33600
    },
    {
      "epoch": 5.481279060282241,
      "eval_loss": 0.32043567299842834,
      "eval_runtime": 152.9804,
      "eval_samples_per_second": 17.813,
      "eval_steps_per_second": 8.91,
      "num_input_tokens_seen": 63668416,
      "step": 33600
    },
    {
      "epoch": 5.482094787503059,
      "grad_norm": 0.000626346911303699,
      "learning_rate": 0.018531302373343096,
      "loss": 0.2853,
      "num_input_tokens_seen": 63678304,
      "step": 33605
    },
    {
      "epoch": 5.482910514723876,
      "grad_norm": 0.0025260290130972862,
      "learning_rate": 0.018502951095221588,
      "loss": 0.3701,
      "num_input_tokens_seen": 63687888,
      "step": 33610
    },
    {
      "epoch": 5.4837262419446935,
      "grad_norm": 0.002874991623684764,
      "learning_rate": 0.01847462009559751,
      "loss": 0.3051,
      "num_input_tokens_seen": 63697872,
      "step": 33615
    },
    {
      "epoch": 5.484541969165511,
      "grad_norm": 0.0039471969939768314,
      "learning_rate": 0.01844630937883992,
      "loss": 0.4196,
      "num_input_tokens_seen": 63708304,
      "step": 33620
    },
    {
      "epoch": 5.485357696386329,
      "grad_norm": 0.0023314107675105333,
      "learning_rate": 0.018418018949314573,
      "loss": 0.3139,
      "num_input_tokens_seen": 63718048,
      "step": 33625
    },
    {
      "epoch": 5.486173423607146,
      "grad_norm": 0.00212409277446568,
      "learning_rate": 0.018389748811384315,
      "loss": 0.3025,
      "num_input_tokens_seen": 63727872,
      "step": 33630
    },
    {
      "epoch": 5.486989150827963,
      "grad_norm": 0.0018473041709512472,
      "learning_rate": 0.018361498969408658,
      "loss": 0.3299,
      "num_input_tokens_seen": 63737664,
      "step": 33635
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 0.002470931503921747,
      "learning_rate": 0.01833326942774415,
      "loss": 0.3243,
      "num_input_tokens_seen": 63747504,
      "step": 33640
    },
    {
      "epoch": 5.488620605269598,
      "grad_norm": 0.0008076968952082098,
      "learning_rate": 0.018305060190744155,
      "loss": 0.2892,
      "num_input_tokens_seen": 63757408,
      "step": 33645
    },
    {
      "epoch": 5.489436332490415,
      "grad_norm": 0.0022481742780655622,
      "learning_rate": 0.018276871262758846,
      "loss": 0.3155,
      "num_input_tokens_seen": 63766720,
      "step": 33650
    },
    {
      "epoch": 5.490252059711232,
      "grad_norm": 0.003107398981228471,
      "learning_rate": 0.0182487026481353,
      "loss": 0.3444,
      "num_input_tokens_seen": 63775824,
      "step": 33655
    },
    {
      "epoch": 5.49106778693205,
      "grad_norm": 0.000537949672434479,
      "learning_rate": 0.018220554351217538,
      "loss": 0.3505,
      "num_input_tokens_seen": 63784464,
      "step": 33660
    },
    {
      "epoch": 5.4918835141528675,
      "grad_norm": 0.0013194689527153969,
      "learning_rate": 0.01819242637634629,
      "loss": 0.3362,
      "num_input_tokens_seen": 63794496,
      "step": 33665
    },
    {
      "epoch": 5.492699241373685,
      "grad_norm": 0.001703976420685649,
      "learning_rate": 0.01816431872785933,
      "loss": 0.3284,
      "num_input_tokens_seen": 63804480,
      "step": 33670
    },
    {
      "epoch": 5.493514968594502,
      "grad_norm": 0.002849158365279436,
      "learning_rate": 0.018136231410091148,
      "loss": 0.3152,
      "num_input_tokens_seen": 63813792,
      "step": 33675
    },
    {
      "epoch": 5.49433069581532,
      "grad_norm": 0.002953516086563468,
      "learning_rate": 0.018108164427373175,
      "loss": 0.3464,
      "num_input_tokens_seen": 63823440,
      "step": 33680
    },
    {
      "epoch": 5.495146423036137,
      "grad_norm": 0.0011111171916127205,
      "learning_rate": 0.01808011778403375,
      "loss": 0.3617,
      "num_input_tokens_seen": 63832160,
      "step": 33685
    },
    {
      "epoch": 5.495962150256954,
      "grad_norm": 0.001838807249441743,
      "learning_rate": 0.01805209148439793,
      "loss": 0.2732,
      "num_input_tokens_seen": 63841104,
      "step": 33690
    },
    {
      "epoch": 5.496777877477771,
      "grad_norm": 0.0016402921173721552,
      "learning_rate": 0.018024085532787757,
      "loss": 0.2788,
      "num_input_tokens_seen": 63850192,
      "step": 33695
    },
    {
      "epoch": 5.497593604698589,
      "grad_norm": 0.0007648097234778106,
      "learning_rate": 0.017996099933522164,
      "loss": 0.3541,
      "num_input_tokens_seen": 63859968,
      "step": 33700
    },
    {
      "epoch": 5.498409331919406,
      "grad_norm": 0.0025803823955357075,
      "learning_rate": 0.017968134690916775,
      "loss": 0.3627,
      "num_input_tokens_seen": 63869088,
      "step": 33705
    },
    {
      "epoch": 5.499225059140223,
      "grad_norm": 0.0012290736194700003,
      "learning_rate": 0.017940189809284263,
      "loss": 0.3035,
      "num_input_tokens_seen": 63878432,
      "step": 33710
    },
    {
      "epoch": 5.5000407863610405,
      "grad_norm": 0.0017703230259940028,
      "learning_rate": 0.017912265292934024,
      "loss": 0.2967,
      "num_input_tokens_seen": 63887792,
      "step": 33715
    },
    {
      "epoch": 5.500856513581859,
      "grad_norm": 0.0030117621645331383,
      "learning_rate": 0.017884361146172423,
      "loss": 0.2648,
      "num_input_tokens_seen": 63897456,
      "step": 33720
    },
    {
      "epoch": 5.501672240802676,
      "grad_norm": 0.00309411296620965,
      "learning_rate": 0.01785647737330261,
      "loss": 0.3515,
      "num_input_tokens_seen": 63905184,
      "step": 33725
    },
    {
      "epoch": 5.502487968023493,
      "grad_norm": 0.002045212546363473,
      "learning_rate": 0.017828613978624563,
      "loss": 0.3498,
      "num_input_tokens_seen": 63913040,
      "step": 33730
    },
    {
      "epoch": 5.50330369524431,
      "grad_norm": 0.0007679361151531339,
      "learning_rate": 0.01780077096643523,
      "loss": 0.2748,
      "num_input_tokens_seen": 63921664,
      "step": 33735
    },
    {
      "epoch": 5.504119422465128,
      "grad_norm": 0.0011397934285923839,
      "learning_rate": 0.017772948341028345,
      "loss": 0.375,
      "num_input_tokens_seen": 63930848,
      "step": 33740
    },
    {
      "epoch": 5.504935149685945,
      "grad_norm": 0.0016165439737960696,
      "learning_rate": 0.01774514610669447,
      "loss": 0.3495,
      "num_input_tokens_seen": 63940672,
      "step": 33745
    },
    {
      "epoch": 5.505750876906762,
      "grad_norm": 0.001374760060571134,
      "learning_rate": 0.017717364267721112,
      "loss": 0.295,
      "num_input_tokens_seen": 63950704,
      "step": 33750
    },
    {
      "epoch": 5.506566604127579,
      "grad_norm": 0.002327191876247525,
      "learning_rate": 0.017689602828392513,
      "loss": 0.3298,
      "num_input_tokens_seen": 63960032,
      "step": 33755
    },
    {
      "epoch": 5.507382331348397,
      "grad_norm": 0.0014424124965444207,
      "learning_rate": 0.017661861792989897,
      "loss": 0.3105,
      "num_input_tokens_seen": 63968560,
      "step": 33760
    },
    {
      "epoch": 5.5081980585692145,
      "grad_norm": 0.003243414219468832,
      "learning_rate": 0.017634141165791272,
      "loss": 0.3248,
      "num_input_tokens_seen": 63978416,
      "step": 33765
    },
    {
      "epoch": 5.509013785790032,
      "grad_norm": 0.0014480296522378922,
      "learning_rate": 0.017606440951071455,
      "loss": 0.3184,
      "num_input_tokens_seen": 63988528,
      "step": 33770
    },
    {
      "epoch": 5.50982951301085,
      "grad_norm": 0.0037957951426506042,
      "learning_rate": 0.017578761153102213,
      "loss": 0.3613,
      "num_input_tokens_seen": 63997248,
      "step": 33775
    },
    {
      "epoch": 5.510645240231667,
      "grad_norm": 0.0009963526390492916,
      "learning_rate": 0.017551101776152146,
      "loss": 0.3341,
      "num_input_tokens_seen": 64007008,
      "step": 33780
    },
    {
      "epoch": 5.511460967452484,
      "grad_norm": 0.0008724433137103915,
      "learning_rate": 0.017523462824486608,
      "loss": 0.3359,
      "num_input_tokens_seen": 64015632,
      "step": 33785
    },
    {
      "epoch": 5.512276694673301,
      "grad_norm": 0.001980450237169862,
      "learning_rate": 0.01749584430236794,
      "loss": 0.3017,
      "num_input_tokens_seen": 64024848,
      "step": 33790
    },
    {
      "epoch": 5.513092421894118,
      "grad_norm": 0.0033417725935578346,
      "learning_rate": 0.01746824621405524,
      "loss": 0.3068,
      "num_input_tokens_seen": 64034112,
      "step": 33795
    },
    {
      "epoch": 5.513908149114936,
      "grad_norm": 0.0007279135170392692,
      "learning_rate": 0.017440668563804412,
      "loss": 0.2732,
      "num_input_tokens_seen": 64043792,
      "step": 33800
    },
    {
      "epoch": 5.513908149114936,
      "eval_loss": 0.32045435905456543,
      "eval_runtime": 152.9128,
      "eval_samples_per_second": 17.821,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 64043792,
      "step": 33800
    },
    {
      "epoch": 5.514723876335753,
      "grad_norm": 0.0019851597025990486,
      "learning_rate": 0.017413111355868392,
      "loss": 0.316,
      "num_input_tokens_seen": 64053968,
      "step": 33805
    },
    {
      "epoch": 5.5155396035565705,
      "grad_norm": 0.0009320075041614473,
      "learning_rate": 0.017385574594496748,
      "loss": 0.3771,
      "num_input_tokens_seen": 64063456,
      "step": 33810
    },
    {
      "epoch": 5.5163553307773885,
      "grad_norm": 0.0017596492543816566,
      "learning_rate": 0.01735805828393605,
      "loss": 0.327,
      "num_input_tokens_seen": 64074016,
      "step": 33815
    },
    {
      "epoch": 5.517171057998206,
      "grad_norm": 0.00043908378574997187,
      "learning_rate": 0.017330562428429667,
      "loss": 0.2241,
      "num_input_tokens_seen": 64083952,
      "step": 33820
    },
    {
      "epoch": 5.517986785219023,
      "grad_norm": 0.001633551437407732,
      "learning_rate": 0.01730308703221776,
      "loss": 0.3934,
      "num_input_tokens_seen": 64094656,
      "step": 33825
    },
    {
      "epoch": 5.51880251243984,
      "grad_norm": 0.0011358277406543493,
      "learning_rate": 0.01727563209953744,
      "loss": 0.3238,
      "num_input_tokens_seen": 64105200,
      "step": 33830
    },
    {
      "epoch": 5.519618239660657,
      "grad_norm": 0.0016133999451994896,
      "learning_rate": 0.017248197634622535,
      "loss": 0.3274,
      "num_input_tokens_seen": 64114624,
      "step": 33835
    },
    {
      "epoch": 5.520433966881475,
      "grad_norm": 0.0010182417463511229,
      "learning_rate": 0.01722078364170383,
      "loss": 0.3167,
      "num_input_tokens_seen": 64125120,
      "step": 33840
    },
    {
      "epoch": 5.521249694102292,
      "grad_norm": 0.0009198072366416454,
      "learning_rate": 0.017193390125008905,
      "loss": 0.3039,
      "num_input_tokens_seen": 64135168,
      "step": 33845
    },
    {
      "epoch": 5.522065421323109,
      "grad_norm": 0.002819631714373827,
      "learning_rate": 0.017166017088762153,
      "loss": 0.2791,
      "num_input_tokens_seen": 64145312,
      "step": 33850
    },
    {
      "epoch": 5.522881148543927,
      "grad_norm": 0.0018490583170205355,
      "learning_rate": 0.017138664537184878,
      "loss": 0.304,
      "num_input_tokens_seen": 64155376,
      "step": 33855
    },
    {
      "epoch": 5.523696875764744,
      "grad_norm": 0.0016267391620203853,
      "learning_rate": 0.017111332474495172,
      "loss": 0.3372,
      "num_input_tokens_seen": 64165232,
      "step": 33860
    },
    {
      "epoch": 5.524512602985562,
      "grad_norm": 0.0006110730464570224,
      "learning_rate": 0.017084020904907998,
      "loss": 0.3734,
      "num_input_tokens_seen": 64174096,
      "step": 33865
    },
    {
      "epoch": 5.525328330206379,
      "grad_norm": 0.0015285853296518326,
      "learning_rate": 0.017056729832635103,
      "loss": 0.3615,
      "num_input_tokens_seen": 64184528,
      "step": 33870
    },
    {
      "epoch": 5.526144057427197,
      "grad_norm": 0.0017239724984392524,
      "learning_rate": 0.017029459261885153,
      "loss": 0.3389,
      "num_input_tokens_seen": 64193776,
      "step": 33875
    },
    {
      "epoch": 5.526959784648014,
      "grad_norm": 0.0018021442228928208,
      "learning_rate": 0.01700220919686359,
      "loss": 0.3301,
      "num_input_tokens_seen": 64203456,
      "step": 33880
    },
    {
      "epoch": 5.527775511868831,
      "grad_norm": 0.0011413025204092264,
      "learning_rate": 0.016974979641772723,
      "loss": 0.2972,
      "num_input_tokens_seen": 64214240,
      "step": 33885
    },
    {
      "epoch": 5.528591239089648,
      "grad_norm": 0.0011671701213344932,
      "learning_rate": 0.01694777060081169,
      "loss": 0.2851,
      "num_input_tokens_seen": 64224080,
      "step": 33890
    },
    {
      "epoch": 5.529406966310466,
      "grad_norm": 0.0015370703767985106,
      "learning_rate": 0.016920582078176444,
      "loss": 0.3768,
      "num_input_tokens_seen": 64232832,
      "step": 33895
    },
    {
      "epoch": 5.530222693531283,
      "grad_norm": 0.002625873778015375,
      "learning_rate": 0.016893414078059863,
      "loss": 0.3419,
      "num_input_tokens_seen": 64241216,
      "step": 33900
    },
    {
      "epoch": 5.5310384207521,
      "grad_norm": 0.006980437319725752,
      "learning_rate": 0.016866266604651535,
      "loss": 0.3622,
      "num_input_tokens_seen": 64251808,
      "step": 33905
    },
    {
      "epoch": 5.5318541479729175,
      "grad_norm": 0.0008657214348204434,
      "learning_rate": 0.016839139662137976,
      "loss": 0.2999,
      "num_input_tokens_seen": 64260880,
      "step": 33910
    },
    {
      "epoch": 5.5326698751937355,
      "grad_norm": 0.001388538395985961,
      "learning_rate": 0.01681203325470245,
      "loss": 0.3082,
      "num_input_tokens_seen": 64269248,
      "step": 33915
    },
    {
      "epoch": 5.533485602414553,
      "grad_norm": 0.0033924817107617855,
      "learning_rate": 0.016784947386525157,
      "loss": 0.3165,
      "num_input_tokens_seen": 64277952,
      "step": 33920
    },
    {
      "epoch": 5.53430132963537,
      "grad_norm": 0.002050059149041772,
      "learning_rate": 0.01675788206178308,
      "loss": 0.2783,
      "num_input_tokens_seen": 64287824,
      "step": 33925
    },
    {
      "epoch": 5.535117056856187,
      "grad_norm": 0.0025425299536436796,
      "learning_rate": 0.016730837284649986,
      "loss": 0.3128,
      "num_input_tokens_seen": 64296624,
      "step": 33930
    },
    {
      "epoch": 5.535932784077005,
      "grad_norm": 0.0008960096165537834,
      "learning_rate": 0.016703813059296583,
      "loss": 0.2941,
      "num_input_tokens_seen": 64306432,
      "step": 33935
    },
    {
      "epoch": 5.536748511297822,
      "grad_norm": 0.0007782781613059342,
      "learning_rate": 0.016676809389890294,
      "loss": 0.2922,
      "num_input_tokens_seen": 64316832,
      "step": 33940
    },
    {
      "epoch": 5.537564238518639,
      "grad_norm": 0.0014058732194826007,
      "learning_rate": 0.016649826280595435,
      "loss": 0.3649,
      "num_input_tokens_seen": 64326400,
      "step": 33945
    },
    {
      "epoch": 5.538379965739456,
      "grad_norm": 0.0010429790709167719,
      "learning_rate": 0.016622863735573163,
      "loss": 0.3737,
      "num_input_tokens_seen": 64337280,
      "step": 33950
    },
    {
      "epoch": 5.539195692960274,
      "grad_norm": 0.000965687504503876,
      "learning_rate": 0.016595921758981395,
      "loss": 0.3044,
      "num_input_tokens_seen": 64345312,
      "step": 33955
    },
    {
      "epoch": 5.5400114201810915,
      "grad_norm": 0.00199980684556067,
      "learning_rate": 0.01656900035497495,
      "loss": 0.3104,
      "num_input_tokens_seen": 64353376,
      "step": 33960
    },
    {
      "epoch": 5.540827147401909,
      "grad_norm": 0.0018890942446887493,
      "learning_rate": 0.016542099527705485,
      "loss": 0.292,
      "num_input_tokens_seen": 64363760,
      "step": 33965
    },
    {
      "epoch": 5.541642874622726,
      "grad_norm": 0.0014299498870968819,
      "learning_rate": 0.01651521928132138,
      "loss": 0.2931,
      "num_input_tokens_seen": 64373376,
      "step": 33970
    },
    {
      "epoch": 5.542458601843544,
      "grad_norm": 0.002516279462724924,
      "learning_rate": 0.01648835961996794,
      "loss": 0.3409,
      "num_input_tokens_seen": 64383248,
      "step": 33975
    },
    {
      "epoch": 5.543274329064361,
      "grad_norm": 0.0006352540804073215,
      "learning_rate": 0.016461520547787285,
      "loss": 0.2732,
      "num_input_tokens_seen": 64394144,
      "step": 33980
    },
    {
      "epoch": 5.544090056285178,
      "grad_norm": 0.001602311385795474,
      "learning_rate": 0.016434702068918266,
      "loss": 0.3564,
      "num_input_tokens_seen": 64403568,
      "step": 33985
    },
    {
      "epoch": 5.544905783505996,
      "grad_norm": 0.0031198018696159124,
      "learning_rate": 0.01640790418749673,
      "loss": 0.3375,
      "num_input_tokens_seen": 64414384,
      "step": 33990
    },
    {
      "epoch": 5.545721510726813,
      "grad_norm": 0.0010687351459637284,
      "learning_rate": 0.016381126907655134,
      "loss": 0.3098,
      "num_input_tokens_seen": 64424544,
      "step": 33995
    },
    {
      "epoch": 5.54653723794763,
      "grad_norm": 0.0015962766483426094,
      "learning_rate": 0.016354370233522948,
      "loss": 0.3241,
      "num_input_tokens_seen": 64433840,
      "step": 34000
    },
    {
      "epoch": 5.54653723794763,
      "eval_loss": 0.3199886381626129,
      "eval_runtime": 152.9522,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 8.911,
      "num_input_tokens_seen": 64433840,
      "step": 34000
    },
    {
      "epoch": 5.5473529651684474,
      "grad_norm": 0.000767493387684226,
      "learning_rate": 0.016327634169226394,
      "loss": 0.3692,
      "num_input_tokens_seen": 64443344,
      "step": 34005
    },
    {
      "epoch": 5.548168692389265,
      "grad_norm": 0.002404460683465004,
      "learning_rate": 0.016300918718888485,
      "loss": 0.3565,
      "num_input_tokens_seen": 64453168,
      "step": 34010
    },
    {
      "epoch": 5.548984419610083,
      "grad_norm": 0.0027817157097160816,
      "learning_rate": 0.016274223886629052,
      "loss": 0.3426,
      "num_input_tokens_seen": 64462592,
      "step": 34015
    },
    {
      "epoch": 5.5498001468309,
      "grad_norm": 0.0016496165189892054,
      "learning_rate": 0.01624754967656482,
      "loss": 0.287,
      "num_input_tokens_seen": 64471584,
      "step": 34020
    },
    {
      "epoch": 5.550615874051717,
      "grad_norm": 0.0005398383946157992,
      "learning_rate": 0.016220896092809235,
      "loss": 0.3016,
      "num_input_tokens_seen": 64480016,
      "step": 34025
    },
    {
      "epoch": 5.551431601272535,
      "grad_norm": 0.0006807293393649161,
      "learning_rate": 0.01619426313947267,
      "loss": 0.3561,
      "num_input_tokens_seen": 64488224,
      "step": 34030
    },
    {
      "epoch": 5.552247328493352,
      "grad_norm": 0.002013934077695012,
      "learning_rate": 0.016167650820662228,
      "loss": 0.3011,
      "num_input_tokens_seen": 64497792,
      "step": 34035
    },
    {
      "epoch": 5.553063055714169,
      "grad_norm": 0.0030770921148359776,
      "learning_rate": 0.016141059140481855,
      "loss": 0.3467,
      "num_input_tokens_seen": 64507136,
      "step": 34040
    },
    {
      "epoch": 5.553878782934986,
      "grad_norm": 0.0014208321226760745,
      "learning_rate": 0.016114488103032374,
      "loss": 0.306,
      "num_input_tokens_seen": 64516496,
      "step": 34045
    },
    {
      "epoch": 5.554694510155803,
      "grad_norm": 0.0006311133620329201,
      "learning_rate": 0.016087937712411293,
      "loss": 0.2719,
      "num_input_tokens_seen": 64525504,
      "step": 34050
    },
    {
      "epoch": 5.555510237376621,
      "grad_norm": 0.0032048597931861877,
      "learning_rate": 0.01606140797271308,
      "loss": 0.3234,
      "num_input_tokens_seen": 64534320,
      "step": 34055
    },
    {
      "epoch": 5.5563259645974385,
      "grad_norm": 0.003456695703789592,
      "learning_rate": 0.01603489888802897,
      "loss": 0.3438,
      "num_input_tokens_seen": 64544384,
      "step": 34060
    },
    {
      "epoch": 5.557141691818256,
      "grad_norm": 0.000746445672120899,
      "learning_rate": 0.016008410462446918,
      "loss": 0.3071,
      "num_input_tokens_seen": 64555168,
      "step": 34065
    },
    {
      "epoch": 5.557957419039074,
      "grad_norm": 0.0013880727346986532,
      "learning_rate": 0.01598194270005185,
      "loss": 0.3451,
      "num_input_tokens_seen": 64565600,
      "step": 34070
    },
    {
      "epoch": 5.558773146259891,
      "grad_norm": 0.0012942739995196462,
      "learning_rate": 0.015955495604925356,
      "loss": 0.3361,
      "num_input_tokens_seen": 64574416,
      "step": 34075
    },
    {
      "epoch": 5.559588873480708,
      "grad_norm": 0.0023937246296554804,
      "learning_rate": 0.01592906918114598,
      "loss": 0.387,
      "num_input_tokens_seen": 64583920,
      "step": 34080
    },
    {
      "epoch": 5.560404600701525,
      "grad_norm": 0.0038155601359903812,
      "learning_rate": 0.015902663432788965,
      "loss": 0.3297,
      "num_input_tokens_seen": 64593616,
      "step": 34085
    },
    {
      "epoch": 5.561220327922343,
      "grad_norm": 0.001982623478397727,
      "learning_rate": 0.01587627836392643,
      "loss": 0.3873,
      "num_input_tokens_seen": 64602768,
      "step": 34090
    },
    {
      "epoch": 5.56203605514316,
      "grad_norm": 0.0022583198733627796,
      "learning_rate": 0.01584991397862726,
      "loss": 0.2588,
      "num_input_tokens_seen": 64612384,
      "step": 34095
    },
    {
      "epoch": 5.562851782363977,
      "grad_norm": 0.002891304437071085,
      "learning_rate": 0.015823570280957214,
      "loss": 0.3385,
      "num_input_tokens_seen": 64622384,
      "step": 34100
    },
    {
      "epoch": 5.5636675095847945,
      "grad_norm": 0.0017590092029422522,
      "learning_rate": 0.015797247274978766,
      "loss": 0.2913,
      "num_input_tokens_seen": 64630352,
      "step": 34105
    },
    {
      "epoch": 5.5644832368056125,
      "grad_norm": 0.0020493001211434603,
      "learning_rate": 0.015770944964751326,
      "loss": 0.3289,
      "num_input_tokens_seen": 64638928,
      "step": 34110
    },
    {
      "epoch": 5.56529896402643,
      "grad_norm": 0.002632771385833621,
      "learning_rate": 0.015744663354330956,
      "loss": 0.2981,
      "num_input_tokens_seen": 64647104,
      "step": 34115
    },
    {
      "epoch": 5.566114691247247,
      "grad_norm": 0.0037698072846978903,
      "learning_rate": 0.015718402447770664,
      "loss": 0.3044,
      "num_input_tokens_seen": 64657344,
      "step": 34120
    },
    {
      "epoch": 5.566930418468064,
      "grad_norm": 0.001019218354485929,
      "learning_rate": 0.015692162249120224,
      "loss": 0.3408,
      "num_input_tokens_seen": 64666768,
      "step": 34125
    },
    {
      "epoch": 5.567746145688882,
      "grad_norm": 0.000744044256862253,
      "learning_rate": 0.01566594276242615,
      "loss": 0.3471,
      "num_input_tokens_seen": 64675728,
      "step": 34130
    },
    {
      "epoch": 5.568561872909699,
      "grad_norm": 0.0016561716329306364,
      "learning_rate": 0.015639743991731857,
      "loss": 0.3418,
      "num_input_tokens_seen": 64685024,
      "step": 34135
    },
    {
      "epoch": 5.569377600130516,
      "grad_norm": 0.0005673021078109741,
      "learning_rate": 0.01561356594107755,
      "loss": 0.2652,
      "num_input_tokens_seen": 64694720,
      "step": 34140
    },
    {
      "epoch": 5.570193327351333,
      "grad_norm": 0.0011953390203416348,
      "learning_rate": 0.015587408614500147,
      "loss": 0.2643,
      "num_input_tokens_seen": 64704496,
      "step": 34145
    },
    {
      "epoch": 5.571009054572151,
      "grad_norm": 0.0007605567225255072,
      "learning_rate": 0.015561272016033505,
      "loss": 0.3061,
      "num_input_tokens_seen": 64714336,
      "step": 34150
    },
    {
      "epoch": 5.5718247817929685,
      "grad_norm": 0.000732311571482569,
      "learning_rate": 0.015535156149708167,
      "loss": 0.3475,
      "num_input_tokens_seen": 64724144,
      "step": 34155
    },
    {
      "epoch": 5.572640509013786,
      "grad_norm": 0.001888045109808445,
      "learning_rate": 0.015509061019551528,
      "loss": 0.2841,
      "num_input_tokens_seen": 64733360,
      "step": 34160
    },
    {
      "epoch": 5.573456236234604,
      "grad_norm": 0.0023130455520004034,
      "learning_rate": 0.015482986629587818,
      "loss": 0.3187,
      "num_input_tokens_seen": 64742432,
      "step": 34165
    },
    {
      "epoch": 5.574271963455421,
      "grad_norm": 0.0021992838010191917,
      "learning_rate": 0.01545693298383799,
      "loss": 0.2736,
      "num_input_tokens_seen": 64751696,
      "step": 34170
    },
    {
      "epoch": 5.575087690676238,
      "grad_norm": 0.0012572529958561063,
      "learning_rate": 0.015430900086319858,
      "loss": 0.3016,
      "num_input_tokens_seen": 64761728,
      "step": 34175
    },
    {
      "epoch": 5.575903417897055,
      "grad_norm": 0.003828279674053192,
      "learning_rate": 0.015404887941048084,
      "loss": 0.3354,
      "num_input_tokens_seen": 64770368,
      "step": 34180
    },
    {
      "epoch": 5.576719145117872,
      "grad_norm": 0.0025359035935252905,
      "learning_rate": 0.01537889655203397,
      "loss": 0.3188,
      "num_input_tokens_seen": 64780144,
      "step": 34185
    },
    {
      "epoch": 5.57753487233869,
      "grad_norm": 0.001424567075446248,
      "learning_rate": 0.015352925923285798,
      "loss": 0.304,
      "num_input_tokens_seen": 64790288,
      "step": 34190
    },
    {
      "epoch": 5.578350599559507,
      "grad_norm": 0.002859173808246851,
      "learning_rate": 0.015326976058808511,
      "loss": 0.2722,
      "num_input_tokens_seen": 64799584,
      "step": 34195
    },
    {
      "epoch": 5.579166326780324,
      "grad_norm": 0.0011581286089494824,
      "learning_rate": 0.015301046962603908,
      "loss": 0.2618,
      "num_input_tokens_seen": 64808624,
      "step": 34200
    },
    {
      "epoch": 5.579166326780324,
      "eval_loss": 0.32136669754981995,
      "eval_runtime": 152.8445,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 64808624,
      "step": 34200
    },
    {
      "epoch": 5.5799820540011424,
      "grad_norm": 0.002755186054855585,
      "learning_rate": 0.015275138638670626,
      "loss": 0.3187,
      "num_input_tokens_seen": 64817808,
      "step": 34205
    },
    {
      "epoch": 5.58079778122196,
      "grad_norm": 0.0031999978236854076,
      "learning_rate": 0.015249251091004001,
      "loss": 0.3686,
      "num_input_tokens_seen": 64827728,
      "step": 34210
    },
    {
      "epoch": 5.581613508442777,
      "grad_norm": 0.0017498983070254326,
      "learning_rate": 0.01522338432359624,
      "loss": 0.3327,
      "num_input_tokens_seen": 64836496,
      "step": 34215
    },
    {
      "epoch": 5.582429235663594,
      "grad_norm": 0.000522889313288033,
      "learning_rate": 0.01519753834043635,
      "loss": 0.3087,
      "num_input_tokens_seen": 64845728,
      "step": 34220
    },
    {
      "epoch": 5.583244962884411,
      "grad_norm": 0.0015048027271404862,
      "learning_rate": 0.015171713145510095,
      "loss": 0.3168,
      "num_input_tokens_seen": 64855936,
      "step": 34225
    },
    {
      "epoch": 5.584060690105229,
      "grad_norm": 0.0035882522352039814,
      "learning_rate": 0.01514590874279999,
      "loss": 0.3075,
      "num_input_tokens_seen": 64865424,
      "step": 34230
    },
    {
      "epoch": 5.584876417326046,
      "grad_norm": 0.0005039914976805449,
      "learning_rate": 0.015120125136285467,
      "loss": 0.3016,
      "num_input_tokens_seen": 64874304,
      "step": 34235
    },
    {
      "epoch": 5.585692144546863,
      "grad_norm": 0.00293421046808362,
      "learning_rate": 0.015094362329942629,
      "loss": 0.3688,
      "num_input_tokens_seen": 64885232,
      "step": 34240
    },
    {
      "epoch": 5.586507871767681,
      "grad_norm": 0.00150556571315974,
      "learning_rate": 0.01506862032774448,
      "loss": 0.2922,
      "num_input_tokens_seen": 64895264,
      "step": 34245
    },
    {
      "epoch": 5.587323598988498,
      "grad_norm": 0.0014506644802168012,
      "learning_rate": 0.015042899133660697,
      "loss": 0.335,
      "num_input_tokens_seen": 64904096,
      "step": 34250
    },
    {
      "epoch": 5.5881393262093155,
      "grad_norm": 0.0003209729038644582,
      "learning_rate": 0.01501719875165789,
      "loss": 0.2893,
      "num_input_tokens_seen": 64913216,
      "step": 34255
    },
    {
      "epoch": 5.588955053430133,
      "grad_norm": 0.0021669757552444935,
      "learning_rate": 0.014991519185699286,
      "loss": 0.2849,
      "num_input_tokens_seen": 64923184,
      "step": 34260
    },
    {
      "epoch": 5.589770780650951,
      "grad_norm": 0.0023893723264336586,
      "learning_rate": 0.014965860439745054,
      "loss": 0.3307,
      "num_input_tokens_seen": 64930448,
      "step": 34265
    },
    {
      "epoch": 5.590586507871768,
      "grad_norm": 0.0031401750165969133,
      "learning_rate": 0.01494022251775211,
      "loss": 0.3298,
      "num_input_tokens_seen": 64940368,
      "step": 34270
    },
    {
      "epoch": 5.591402235092585,
      "grad_norm": 0.003901425749063492,
      "learning_rate": 0.014914605423674109,
      "loss": 0.2853,
      "num_input_tokens_seen": 64949728,
      "step": 34275
    },
    {
      "epoch": 5.592217962313402,
      "grad_norm": 0.0012496761046350002,
      "learning_rate": 0.014889009161461525,
      "loss": 0.3056,
      "num_input_tokens_seen": 64958704,
      "step": 34280
    },
    {
      "epoch": 5.59303368953422,
      "grad_norm": 0.0015730189625173807,
      "learning_rate": 0.014863433735061665,
      "loss": 0.335,
      "num_input_tokens_seen": 64967568,
      "step": 34285
    },
    {
      "epoch": 5.593849416755037,
      "grad_norm": 0.0011439980007708073,
      "learning_rate": 0.014837879148418541,
      "loss": 0.3032,
      "num_input_tokens_seen": 64977760,
      "step": 34290
    },
    {
      "epoch": 5.594665143975854,
      "grad_norm": 0.0027525487821549177,
      "learning_rate": 0.01481234540547302,
      "loss": 0.3085,
      "num_input_tokens_seen": 64986800,
      "step": 34295
    },
    {
      "epoch": 5.5954808711966715,
      "grad_norm": 0.003367881290614605,
      "learning_rate": 0.014786832510162717,
      "loss": 0.3254,
      "num_input_tokens_seen": 64996208,
      "step": 34300
    },
    {
      "epoch": 5.5962965984174895,
      "grad_norm": 0.0018874717643484473,
      "learning_rate": 0.014761340466422017,
      "loss": 0.3001,
      "num_input_tokens_seen": 65005600,
      "step": 34305
    },
    {
      "epoch": 5.597112325638307,
      "grad_norm": 0.0031940292101353407,
      "learning_rate": 0.014735869278182144,
      "loss": 0.2509,
      "num_input_tokens_seen": 65014688,
      "step": 34310
    },
    {
      "epoch": 5.597928052859124,
      "grad_norm": 0.0005121819558553398,
      "learning_rate": 0.014710418949371057,
      "loss": 0.3725,
      "num_input_tokens_seen": 65023504,
      "step": 34315
    },
    {
      "epoch": 5.598743780079941,
      "grad_norm": 0.0014994373777881265,
      "learning_rate": 0.014684989483913495,
      "loss": 0.2677,
      "num_input_tokens_seen": 65033936,
      "step": 34320
    },
    {
      "epoch": 5.599559507300759,
      "grad_norm": 0.0021229067351669073,
      "learning_rate": 0.014659580885731077,
      "loss": 0.3079,
      "num_input_tokens_seen": 65042224,
      "step": 34325
    },
    {
      "epoch": 5.600375234521576,
      "grad_norm": 0.004625482484698296,
      "learning_rate": 0.014634193158742047,
      "loss": 0.3486,
      "num_input_tokens_seen": 65052368,
      "step": 34330
    },
    {
      "epoch": 5.601190961742393,
      "grad_norm": 0.0013867969391867518,
      "learning_rate": 0.014608826306861576,
      "loss": 0.3117,
      "num_input_tokens_seen": 65061952,
      "step": 34335
    },
    {
      "epoch": 5.602006688963211,
      "grad_norm": 0.0025344896130263805,
      "learning_rate": 0.014583480334001486,
      "loss": 0.349,
      "num_input_tokens_seen": 65070176,
      "step": 34340
    },
    {
      "epoch": 5.602822416184028,
      "grad_norm": 0.0024319023359566927,
      "learning_rate": 0.014558155244070496,
      "loss": 0.3405,
      "num_input_tokens_seen": 65080048,
      "step": 34345
    },
    {
      "epoch": 5.6036381434048455,
      "grad_norm": 0.002955754054710269,
      "learning_rate": 0.014532851040974036,
      "loss": 0.2793,
      "num_input_tokens_seen": 65089056,
      "step": 34350
    },
    {
      "epoch": 5.604453870625663,
      "grad_norm": 0.0029117618687450886,
      "learning_rate": 0.014507567728614335,
      "loss": 0.3355,
      "num_input_tokens_seen": 65098176,
      "step": 34355
    },
    {
      "epoch": 5.60526959784648,
      "grad_norm": 0.001394640188664198,
      "learning_rate": 0.01448230531089037,
      "loss": 0.2759,
      "num_input_tokens_seen": 65107312,
      "step": 34360
    },
    {
      "epoch": 5.606085325067298,
      "grad_norm": 0.0030497985426336527,
      "learning_rate": 0.014457063791697993,
      "loss": 0.3475,
      "num_input_tokens_seen": 65117088,
      "step": 34365
    },
    {
      "epoch": 5.606901052288115,
      "grad_norm": 0.0008151049842126667,
      "learning_rate": 0.01443184317492971,
      "loss": 0.3112,
      "num_input_tokens_seen": 65126896,
      "step": 34370
    },
    {
      "epoch": 5.607716779508932,
      "grad_norm": 0.0021662586368620396,
      "learning_rate": 0.014406643464474822,
      "loss": 0.3633,
      "num_input_tokens_seen": 65135504,
      "step": 34375
    },
    {
      "epoch": 5.60853250672975,
      "grad_norm": 0.0006494586123153567,
      "learning_rate": 0.014381464664219539,
      "loss": 0.3348,
      "num_input_tokens_seen": 65144864,
      "step": 34380
    },
    {
      "epoch": 5.609348233950567,
      "grad_norm": 0.0036434566136449575,
      "learning_rate": 0.014356306778046656,
      "loss": 0.3068,
      "num_input_tokens_seen": 65154736,
      "step": 34385
    },
    {
      "epoch": 5.610163961171384,
      "grad_norm": 0.0017371115973219275,
      "learning_rate": 0.014331169809835885,
      "loss": 0.3242,
      "num_input_tokens_seen": 65164704,
      "step": 34390
    },
    {
      "epoch": 5.610979688392201,
      "grad_norm": 0.0035178903490304947,
      "learning_rate": 0.014306053763463644,
      "loss": 0.3932,
      "num_input_tokens_seen": 65173920,
      "step": 34395
    },
    {
      "epoch": 5.6117954156130185,
      "grad_norm": 0.002305178437381983,
      "learning_rate": 0.014280958642803147,
      "loss": 0.2975,
      "num_input_tokens_seen": 65182704,
      "step": 34400
    },
    {
      "epoch": 5.6117954156130185,
      "eval_loss": 0.31985244154930115,
      "eval_runtime": 152.8496,
      "eval_samples_per_second": 17.828,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 65182704,
      "step": 34400
    },
    {
      "epoch": 5.6126111428338366,
      "grad_norm": 0.002470227424055338,
      "learning_rate": 0.014255884451724404,
      "loss": 0.3137,
      "num_input_tokens_seen": 65191888,
      "step": 34405
    },
    {
      "epoch": 5.613426870054654,
      "grad_norm": 0.0007940063369460404,
      "learning_rate": 0.014230831194094101,
      "loss": 0.2922,
      "num_input_tokens_seen": 65200176,
      "step": 34410
    },
    {
      "epoch": 5.614242597275471,
      "grad_norm": 0.0013320206198841333,
      "learning_rate": 0.014205798873775865,
      "loss": 0.2915,
      "num_input_tokens_seen": 65210464,
      "step": 34415
    },
    {
      "epoch": 5.615058324496289,
      "grad_norm": 0.0010024867951869965,
      "learning_rate": 0.014180787494629893,
      "loss": 0.3358,
      "num_input_tokens_seen": 65219696,
      "step": 34420
    },
    {
      "epoch": 5.615874051717106,
      "grad_norm": 0.0011504200519993901,
      "learning_rate": 0.014155797060513314,
      "loss": 0.344,
      "num_input_tokens_seen": 65229024,
      "step": 34425
    },
    {
      "epoch": 5.616689778937923,
      "grad_norm": 0.00350532797165215,
      "learning_rate": 0.014130827575279963,
      "loss": 0.3686,
      "num_input_tokens_seen": 65238352,
      "step": 34430
    },
    {
      "epoch": 5.61750550615874,
      "grad_norm": 0.0023824889212846756,
      "learning_rate": 0.014105879042780427,
      "loss": 0.2942,
      "num_input_tokens_seen": 65248208,
      "step": 34435
    },
    {
      "epoch": 5.618321233379558,
      "grad_norm": 0.0024101310409605503,
      "learning_rate": 0.014080951466862113,
      "loss": 0.2648,
      "num_input_tokens_seen": 65256944,
      "step": 34440
    },
    {
      "epoch": 5.619136960600375,
      "grad_norm": 0.001134027261286974,
      "learning_rate": 0.014056044851369126,
      "loss": 0.2907,
      "num_input_tokens_seen": 65267456,
      "step": 34445
    },
    {
      "epoch": 5.6199526878211925,
      "grad_norm": 0.0027524796314537525,
      "learning_rate": 0.014031159200142428,
      "loss": 0.3085,
      "num_input_tokens_seen": 65277232,
      "step": 34450
    },
    {
      "epoch": 5.62076841504201,
      "grad_norm": 0.0017261751927435398,
      "learning_rate": 0.014006294517019667,
      "loss": 0.2912,
      "num_input_tokens_seen": 65286544,
      "step": 34455
    },
    {
      "epoch": 5.621584142262828,
      "grad_norm": 0.0021411313209682703,
      "learning_rate": 0.013981450805835276,
      "loss": 0.2567,
      "num_input_tokens_seen": 65295584,
      "step": 34460
    },
    {
      "epoch": 5.622399869483645,
      "grad_norm": 0.0006149215623736382,
      "learning_rate": 0.01395662807042049,
      "loss": 0.3649,
      "num_input_tokens_seen": 65305968,
      "step": 34465
    },
    {
      "epoch": 5.623215596704462,
      "grad_norm": 0.0034930044785141945,
      "learning_rate": 0.013931826314603296,
      "loss": 0.3895,
      "num_input_tokens_seen": 65315728,
      "step": 34470
    },
    {
      "epoch": 5.624031323925279,
      "grad_norm": 0.0043055289424955845,
      "learning_rate": 0.013907045542208401,
      "loss": 0.3677,
      "num_input_tokens_seen": 65325136,
      "step": 34475
    },
    {
      "epoch": 5.624847051146097,
      "grad_norm": 0.002575572347268462,
      "learning_rate": 0.013882285757057333,
      "loss": 0.2745,
      "num_input_tokens_seen": 65335968,
      "step": 34480
    },
    {
      "epoch": 5.625662778366914,
      "grad_norm": 0.0025402302853763103,
      "learning_rate": 0.013857546962968403,
      "loss": 0.3439,
      "num_input_tokens_seen": 65345344,
      "step": 34485
    },
    {
      "epoch": 5.626478505587731,
      "grad_norm": 0.004225618205964565,
      "learning_rate": 0.013832829163756577,
      "loss": 0.3424,
      "num_input_tokens_seen": 65352832,
      "step": 34490
    },
    {
      "epoch": 5.6272942328085485,
      "grad_norm": 0.0016225890722125769,
      "learning_rate": 0.013808132363233689,
      "loss": 0.3612,
      "num_input_tokens_seen": 65362592,
      "step": 34495
    },
    {
      "epoch": 5.6281099600293665,
      "grad_norm": 0.003018541494384408,
      "learning_rate": 0.013783456565208256,
      "loss": 0.2896,
      "num_input_tokens_seen": 65371952,
      "step": 34500
    },
    {
      "epoch": 5.628925687250184,
      "grad_norm": 0.002283412730321288,
      "learning_rate": 0.01375880177348564,
      "loss": 0.3312,
      "num_input_tokens_seen": 65381584,
      "step": 34505
    },
    {
      "epoch": 5.629741414471001,
      "grad_norm": 0.0010046879760921001,
      "learning_rate": 0.013734167991867928,
      "loss": 0.2959,
      "num_input_tokens_seen": 65389504,
      "step": 34510
    },
    {
      "epoch": 5.630557141691818,
      "grad_norm": 0.0017333680298179388,
      "learning_rate": 0.013709555224153935,
      "loss": 0.2997,
      "num_input_tokens_seen": 65398384,
      "step": 34515
    },
    {
      "epoch": 5.631372868912636,
      "grad_norm": 0.0026731733232736588,
      "learning_rate": 0.013684963474139222,
      "loss": 0.3213,
      "num_input_tokens_seen": 65407872,
      "step": 34520
    },
    {
      "epoch": 5.632188596133453,
      "grad_norm": 0.002016458660364151,
      "learning_rate": 0.013660392745616224,
      "loss": 0.3115,
      "num_input_tokens_seen": 65418096,
      "step": 34525
    },
    {
      "epoch": 5.63300432335427,
      "grad_norm": 0.0016021019546315074,
      "learning_rate": 0.013635843042373974,
      "loss": 0.3332,
      "num_input_tokens_seen": 65426560,
      "step": 34530
    },
    {
      "epoch": 5.633820050575087,
      "grad_norm": 0.001393337151966989,
      "learning_rate": 0.01361131436819843,
      "loss": 0.3317,
      "num_input_tokens_seen": 65437408,
      "step": 34535
    },
    {
      "epoch": 5.634635777795905,
      "grad_norm": 0.001404671696946025,
      "learning_rate": 0.013586806726872147,
      "loss": 0.2843,
      "num_input_tokens_seen": 65447648,
      "step": 34540
    },
    {
      "epoch": 5.635451505016722,
      "grad_norm": 0.0015823515132069588,
      "learning_rate": 0.013562320122174537,
      "loss": 0.2954,
      "num_input_tokens_seen": 65457584,
      "step": 34545
    },
    {
      "epoch": 5.63626723223754,
      "grad_norm": 0.001766922534443438,
      "learning_rate": 0.013537854557881762,
      "loss": 0.3287,
      "num_input_tokens_seen": 65468272,
      "step": 34550
    },
    {
      "epoch": 5.637082959458358,
      "grad_norm": 0.00040532133425585926,
      "learning_rate": 0.013513410037766687,
      "loss": 0.3626,
      "num_input_tokens_seen": 65478000,
      "step": 34555
    },
    {
      "epoch": 5.637898686679175,
      "grad_norm": 0.0012759399833157659,
      "learning_rate": 0.013488986565598998,
      "loss": 0.287,
      "num_input_tokens_seen": 65487216,
      "step": 34560
    },
    {
      "epoch": 5.638714413899992,
      "grad_norm": 0.0009916998678818345,
      "learning_rate": 0.013464584145145097,
      "loss": 0.2923,
      "num_input_tokens_seen": 65497088,
      "step": 34565
    },
    {
      "epoch": 5.639530141120809,
      "grad_norm": 0.002442446304485202,
      "learning_rate": 0.013440202780168109,
      "loss": 0.3038,
      "num_input_tokens_seen": 65507280,
      "step": 34570
    },
    {
      "epoch": 5.640345868341626,
      "grad_norm": 0.0013310805661603808,
      "learning_rate": 0.01341584247442799,
      "loss": 0.3285,
      "num_input_tokens_seen": 65517008,
      "step": 34575
    },
    {
      "epoch": 5.641161595562444,
      "grad_norm": 0.002894950332120061,
      "learning_rate": 0.013391503231681355,
      "loss": 0.2507,
      "num_input_tokens_seen": 65526816,
      "step": 34580
    },
    {
      "epoch": 5.641977322783261,
      "grad_norm": 0.0010426546214148402,
      "learning_rate": 0.013367185055681685,
      "loss": 0.3472,
      "num_input_tokens_seen": 65535520,
      "step": 34585
    },
    {
      "epoch": 5.642793050004078,
      "grad_norm": 0.0007071575964801013,
      "learning_rate": 0.013342887950179095,
      "loss": 0.3114,
      "num_input_tokens_seen": 65544096,
      "step": 34590
    },
    {
      "epoch": 5.643608777224896,
      "grad_norm": 0.0010991350281983614,
      "learning_rate": 0.013318611918920554,
      "loss": 0.2507,
      "num_input_tokens_seen": 65552736,
      "step": 34595
    },
    {
      "epoch": 5.6444245044457135,
      "grad_norm": 0.0013326945481821895,
      "learning_rate": 0.01329435696564965,
      "loss": 0.326,
      "num_input_tokens_seen": 65562192,
      "step": 34600
    },
    {
      "epoch": 5.6444245044457135,
      "eval_loss": 0.3208852708339691,
      "eval_runtime": 152.8674,
      "eval_samples_per_second": 17.826,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 65562192,
      "step": 34600
    },
    {
      "epoch": 5.645240231666531,
      "grad_norm": 0.0018247278640046716,
      "learning_rate": 0.013270123094106894,
      "loss": 0.3489,
      "num_input_tokens_seen": 65572096,
      "step": 34605
    },
    {
      "epoch": 5.646055958887348,
      "grad_norm": 0.0011069633765146136,
      "learning_rate": 0.013245910308029395,
      "loss": 0.3724,
      "num_input_tokens_seen": 65581888,
      "step": 34610
    },
    {
      "epoch": 5.646871686108166,
      "grad_norm": 0.0024630739353597164,
      "learning_rate": 0.0132217186111511,
      "loss": 0.4318,
      "num_input_tokens_seen": 65591232,
      "step": 34615
    },
    {
      "epoch": 5.647687413328983,
      "grad_norm": 0.002114941133186221,
      "learning_rate": 0.013197548007202626,
      "loss": 0.338,
      "num_input_tokens_seen": 65600224,
      "step": 34620
    },
    {
      "epoch": 5.6485031405498,
      "grad_norm": 0.001869482104666531,
      "learning_rate": 0.01317339849991142,
      "loss": 0.335,
      "num_input_tokens_seen": 65610272,
      "step": 34625
    },
    {
      "epoch": 5.649318867770617,
      "grad_norm": 0.0023699933663010597,
      "learning_rate": 0.013149270093001675,
      "loss": 0.3298,
      "num_input_tokens_seen": 65620128,
      "step": 34630
    },
    {
      "epoch": 5.650134594991435,
      "grad_norm": 0.004367510322481394,
      "learning_rate": 0.013125162790194227,
      "loss": 0.3087,
      "num_input_tokens_seen": 65630416,
      "step": 34635
    },
    {
      "epoch": 5.650950322212252,
      "grad_norm": 0.0014971166383475065,
      "learning_rate": 0.01310107659520674,
      "loss": 0.3012,
      "num_input_tokens_seen": 65639968,
      "step": 34640
    },
    {
      "epoch": 5.6517660494330695,
      "grad_norm": 0.0031153722666203976,
      "learning_rate": 0.013077011511753655,
      "loss": 0.324,
      "num_input_tokens_seen": 65649440,
      "step": 34645
    },
    {
      "epoch": 5.652581776653887,
      "grad_norm": 0.0016347769415006042,
      "learning_rate": 0.013052967543546056,
      "loss": 0.2881,
      "num_input_tokens_seen": 65658944,
      "step": 34650
    },
    {
      "epoch": 5.653397503874705,
      "grad_norm": 0.0029591533821076155,
      "learning_rate": 0.01302894469429186,
      "loss": 0.3542,
      "num_input_tokens_seen": 65668048,
      "step": 34655
    },
    {
      "epoch": 5.654213231095522,
      "grad_norm": 0.0016772426897659898,
      "learning_rate": 0.013004942967695653,
      "loss": 0.3514,
      "num_input_tokens_seen": 65676784,
      "step": 34660
    },
    {
      "epoch": 5.655028958316339,
      "grad_norm": 0.001305589685216546,
      "learning_rate": 0.012980962367458859,
      "loss": 0.3671,
      "num_input_tokens_seen": 65686480,
      "step": 34665
    },
    {
      "epoch": 5.655844685537156,
      "grad_norm": 0.0006353477947413921,
      "learning_rate": 0.012957002897279567,
      "loss": 0.3476,
      "num_input_tokens_seen": 65695264,
      "step": 34670
    },
    {
      "epoch": 5.656660412757974,
      "grad_norm": 0.0015175852458924055,
      "learning_rate": 0.012933064560852576,
      "loss": 0.3447,
      "num_input_tokens_seen": 65704400,
      "step": 34675
    },
    {
      "epoch": 5.657476139978791,
      "grad_norm": 0.0017537373350933194,
      "learning_rate": 0.012909147361869527,
      "loss": 0.3075,
      "num_input_tokens_seen": 65713616,
      "step": 34680
    },
    {
      "epoch": 5.658291867199608,
      "grad_norm": 0.0029228501953184605,
      "learning_rate": 0.012885251304018774,
      "loss": 0.302,
      "num_input_tokens_seen": 65721776,
      "step": 34685
    },
    {
      "epoch": 5.659107594420425,
      "grad_norm": 0.002277122577652335,
      "learning_rate": 0.012861376390985335,
      "loss": 0.3277,
      "num_input_tokens_seen": 65732160,
      "step": 34690
    },
    {
      "epoch": 5.6599233216412435,
      "grad_norm": 0.0014216999989002943,
      "learning_rate": 0.012837522626451063,
      "loss": 0.3037,
      "num_input_tokens_seen": 65741776,
      "step": 34695
    },
    {
      "epoch": 5.660739048862061,
      "grad_norm": 0.0005382577073760331,
      "learning_rate": 0.01281369001409447,
      "loss": 0.326,
      "num_input_tokens_seen": 65750640,
      "step": 34700
    },
    {
      "epoch": 5.661554776082878,
      "grad_norm": 0.005187476985156536,
      "learning_rate": 0.012789878557590877,
      "loss": 0.3121,
      "num_input_tokens_seen": 65761632,
      "step": 34705
    },
    {
      "epoch": 5.662370503303695,
      "grad_norm": 0.003447171300649643,
      "learning_rate": 0.012766088260612334,
      "loss": 0.3181,
      "num_input_tokens_seen": 65771312,
      "step": 34710
    },
    {
      "epoch": 5.663186230524513,
      "grad_norm": 0.0015910064103081822,
      "learning_rate": 0.012742319126827523,
      "loss": 0.3095,
      "num_input_tokens_seen": 65780880,
      "step": 34715
    },
    {
      "epoch": 5.66400195774533,
      "grad_norm": 0.003050118451938033,
      "learning_rate": 0.012718571159902008,
      "loss": 0.3861,
      "num_input_tokens_seen": 65790256,
      "step": 34720
    },
    {
      "epoch": 5.664817684966147,
      "grad_norm": 0.0032292127143591642,
      "learning_rate": 0.01269484436349803,
      "loss": 0.2929,
      "num_input_tokens_seen": 65799664,
      "step": 34725
    },
    {
      "epoch": 5.665633412186965,
      "grad_norm": 0.0012681590160354972,
      "learning_rate": 0.012671138741274528,
      "loss": 0.3158,
      "num_input_tokens_seen": 65808624,
      "step": 34730
    },
    {
      "epoch": 5.666449139407782,
      "grad_norm": 0.001643758499994874,
      "learning_rate": 0.012647454296887194,
      "loss": 0.2824,
      "num_input_tokens_seen": 65817760,
      "step": 34735
    },
    {
      "epoch": 5.667264866628599,
      "grad_norm": 0.0005381849477998912,
      "learning_rate": 0.012623791033988507,
      "loss": 0.3745,
      "num_input_tokens_seen": 65826848,
      "step": 34740
    },
    {
      "epoch": 5.6680805938494165,
      "grad_norm": 0.0023644359316676855,
      "learning_rate": 0.012600148956227597,
      "loss": 0.2873,
      "num_input_tokens_seen": 65835808,
      "step": 34745
    },
    {
      "epoch": 5.668896321070234,
      "grad_norm": 0.0010461579076945782,
      "learning_rate": 0.012576528067250414,
      "loss": 0.3222,
      "num_input_tokens_seen": 65845520,
      "step": 34750
    },
    {
      "epoch": 5.669712048291052,
      "grad_norm": 0.005362044088542461,
      "learning_rate": 0.012552928370699561,
      "loss": 0.294,
      "num_input_tokens_seen": 65854624,
      "step": 34755
    },
    {
      "epoch": 5.670527775511869,
      "grad_norm": 0.003074767766520381,
      "learning_rate": 0.012529349870214411,
      "loss": 0.3538,
      "num_input_tokens_seen": 65863808,
      "step": 34760
    },
    {
      "epoch": 5.671343502732686,
      "grad_norm": 0.0016975216567516327,
      "learning_rate": 0.012505792569431106,
      "loss": 0.3267,
      "num_input_tokens_seen": 65873680,
      "step": 34765
    },
    {
      "epoch": 5.672159229953504,
      "grad_norm": 0.0015982955228537321,
      "learning_rate": 0.012482256471982422,
      "loss": 0.3073,
      "num_input_tokens_seen": 65883376,
      "step": 34770
    },
    {
      "epoch": 5.672974957174321,
      "grad_norm": 0.00311048305593431,
      "learning_rate": 0.012458741581497956,
      "loss": 0.3262,
      "num_input_tokens_seen": 65894048,
      "step": 34775
    },
    {
      "epoch": 5.673790684395138,
      "grad_norm": 0.0037790683563798666,
      "learning_rate": 0.012435247901603974,
      "loss": 0.3238,
      "num_input_tokens_seen": 65904320,
      "step": 34780
    },
    {
      "epoch": 5.674606411615955,
      "grad_norm": 0.0014016850618645549,
      "learning_rate": 0.012411775435923528,
      "loss": 0.3354,
      "num_input_tokens_seen": 65913024,
      "step": 34785
    },
    {
      "epoch": 5.6754221388367725,
      "grad_norm": 0.0012644530506804585,
      "learning_rate": 0.012388324188076354,
      "loss": 0.3254,
      "num_input_tokens_seen": 65922432,
      "step": 34790
    },
    {
      "epoch": 5.6762378660575905,
      "grad_norm": 0.0025970819406211376,
      "learning_rate": 0.012364894161678913,
      "loss": 0.2847,
      "num_input_tokens_seen": 65931744,
      "step": 34795
    },
    {
      "epoch": 5.677053593278408,
      "grad_norm": 0.004736014176160097,
      "learning_rate": 0.012341485360344445,
      "loss": 0.3478,
      "num_input_tokens_seen": 65940816,
      "step": 34800
    },
    {
      "epoch": 5.677053593278408,
      "eval_loss": 0.32321298122406006,
      "eval_runtime": 152.889,
      "eval_samples_per_second": 17.823,
      "eval_steps_per_second": 8.915,
      "num_input_tokens_seen": 65940816,
      "step": 34800
    },
    {
      "epoch": 5.677869320499225,
      "grad_norm": 0.0022123849485069513,
      "learning_rate": 0.01231809778768283,
      "loss": 0.2999,
      "num_input_tokens_seen": 65950176,
      "step": 34805
    },
    {
      "epoch": 5.678685047720043,
      "grad_norm": 0.00048660350148566067,
      "learning_rate": 0.012294731447300799,
      "loss": 0.3433,
      "num_input_tokens_seen": 65962032,
      "step": 34810
    },
    {
      "epoch": 5.67950077494086,
      "grad_norm": 0.002205837517976761,
      "learning_rate": 0.012271386342801671,
      "loss": 0.3396,
      "num_input_tokens_seen": 65971648,
      "step": 34815
    },
    {
      "epoch": 5.680316502161677,
      "grad_norm": 0.0012320068199187517,
      "learning_rate": 0.012248062477785565,
      "loss": 0.2616,
      "num_input_tokens_seen": 65980400,
      "step": 34820
    },
    {
      "epoch": 5.681132229382494,
      "grad_norm": 0.0014711387921124697,
      "learning_rate": 0.012224759855849305,
      "loss": 0.2527,
      "num_input_tokens_seen": 65989872,
      "step": 34825
    },
    {
      "epoch": 5.681947956603312,
      "grad_norm": 0.001311514526605606,
      "learning_rate": 0.012201478480586513,
      "loss": 0.407,
      "num_input_tokens_seen": 65997760,
      "step": 34830
    },
    {
      "epoch": 5.682763683824129,
      "grad_norm": 0.001558742020279169,
      "learning_rate": 0.012178218355587389,
      "loss": 0.3688,
      "num_input_tokens_seen": 66008032,
      "step": 34835
    },
    {
      "epoch": 5.6835794110449465,
      "grad_norm": 0.0010301382280886173,
      "learning_rate": 0.01215497948443896,
      "loss": 0.2799,
      "num_input_tokens_seen": 66017552,
      "step": 34840
    },
    {
      "epoch": 5.684395138265764,
      "grad_norm": 0.003098101355135441,
      "learning_rate": 0.012131761870724993,
      "loss": 0.3187,
      "num_input_tokens_seen": 66026560,
      "step": 34845
    },
    {
      "epoch": 5.685210865486582,
      "grad_norm": 0.001912763575091958,
      "learning_rate": 0.012108565518025893,
      "loss": 0.3083,
      "num_input_tokens_seen": 66035664,
      "step": 34850
    },
    {
      "epoch": 5.686026592707399,
      "grad_norm": 0.002857390558347106,
      "learning_rate": 0.012085390429918862,
      "loss": 0.3123,
      "num_input_tokens_seen": 66045504,
      "step": 34855
    },
    {
      "epoch": 5.686842319928216,
      "grad_norm": 0.0011659665033221245,
      "learning_rate": 0.012062236609977744,
      "loss": 0.3568,
      "num_input_tokens_seen": 66055664,
      "step": 34860
    },
    {
      "epoch": 5.687658047149033,
      "grad_norm": 0.0006549429381266236,
      "learning_rate": 0.01203910406177318,
      "loss": 0.3298,
      "num_input_tokens_seen": 66064336,
      "step": 34865
    },
    {
      "epoch": 5.688473774369851,
      "grad_norm": 0.0014156043762341142,
      "learning_rate": 0.01201599278887252,
      "loss": 0.3647,
      "num_input_tokens_seen": 66074848,
      "step": 34870
    },
    {
      "epoch": 5.689289501590668,
      "grad_norm": 0.0013830390525981784,
      "learning_rate": 0.011992902794839744,
      "loss": 0.3672,
      "num_input_tokens_seen": 66085248,
      "step": 34875
    },
    {
      "epoch": 5.690105228811485,
      "grad_norm": 0.0014083973364904523,
      "learning_rate": 0.011969834083235703,
      "loss": 0.3777,
      "num_input_tokens_seen": 66094736,
      "step": 34880
    },
    {
      "epoch": 5.690920956032302,
      "grad_norm": 0.0026175465900450945,
      "learning_rate": 0.011946786657617836,
      "loss": 0.3123,
      "num_input_tokens_seen": 66105472,
      "step": 34885
    },
    {
      "epoch": 5.69173668325312,
      "grad_norm": 0.0013038276229053736,
      "learning_rate": 0.011923760521540332,
      "loss": 0.2955,
      "num_input_tokens_seen": 66114256,
      "step": 34890
    },
    {
      "epoch": 5.692552410473938,
      "grad_norm": 0.0008919531828723848,
      "learning_rate": 0.011900755678554153,
      "loss": 0.3426,
      "num_input_tokens_seen": 66123648,
      "step": 34895
    },
    {
      "epoch": 5.693368137694755,
      "grad_norm": 0.0014751754933968186,
      "learning_rate": 0.011877772132206893,
      "loss": 0.242,
      "num_input_tokens_seen": 66133424,
      "step": 34900
    },
    {
      "epoch": 5.694183864915573,
      "grad_norm": 0.0017086692387238145,
      "learning_rate": 0.011854809886042915,
      "loss": 0.2641,
      "num_input_tokens_seen": 66144944,
      "step": 34905
    },
    {
      "epoch": 5.69499959213639,
      "grad_norm": 0.0022881831973791122,
      "learning_rate": 0.011831868943603325,
      "loss": 0.3569,
      "num_input_tokens_seen": 66155552,
      "step": 34910
    },
    {
      "epoch": 5.695815319357207,
      "grad_norm": 0.002180441515520215,
      "learning_rate": 0.011808949308425836,
      "loss": 0.3008,
      "num_input_tokens_seen": 66165040,
      "step": 34915
    },
    {
      "epoch": 5.696631046578024,
      "grad_norm": 0.0024651114363223314,
      "learning_rate": 0.01178605098404501,
      "loss": 0.2939,
      "num_input_tokens_seen": 66175120,
      "step": 34920
    },
    {
      "epoch": 5.697446773798841,
      "grad_norm": 0.0024411496706306934,
      "learning_rate": 0.011763173973992002,
      "loss": 0.2874,
      "num_input_tokens_seen": 66184496,
      "step": 34925
    },
    {
      "epoch": 5.698262501019659,
      "grad_norm": 0.0034681647084653378,
      "learning_rate": 0.011740318281794776,
      "loss": 0.4044,
      "num_input_tokens_seen": 66194384,
      "step": 34930
    },
    {
      "epoch": 5.699078228240476,
      "grad_norm": 0.0010936904000118375,
      "learning_rate": 0.01171748391097796,
      "loss": 0.2775,
      "num_input_tokens_seen": 66205072,
      "step": 34935
    },
    {
      "epoch": 5.6998939554612935,
      "grad_norm": 0.0041312179528176785,
      "learning_rate": 0.011694670865062873,
      "loss": 0.34,
      "num_input_tokens_seen": 66214560,
      "step": 34940
    },
    {
      "epoch": 5.7007096826821115,
      "grad_norm": 0.0030821135733276606,
      "learning_rate": 0.011671879147567616,
      "loss": 0.2699,
      "num_input_tokens_seen": 66224224,
      "step": 34945
    },
    {
      "epoch": 5.701525409902929,
      "grad_norm": 0.002095409668982029,
      "learning_rate": 0.011649108762006893,
      "loss": 0.302,
      "num_input_tokens_seen": 66234880,
      "step": 34950
    },
    {
      "epoch": 5.702341137123746,
      "grad_norm": 0.0025607238058000803,
      "learning_rate": 0.011626359711892265,
      "loss": 0.2777,
      "num_input_tokens_seen": 66244096,
      "step": 34955
    },
    {
      "epoch": 5.703156864344563,
      "grad_norm": 0.001555734546855092,
      "learning_rate": 0.01160363200073189,
      "loss": 0.2891,
      "num_input_tokens_seen": 66254304,
      "step": 34960
    },
    {
      "epoch": 5.70397259156538,
      "grad_norm": 0.000734093424398452,
      "learning_rate": 0.011580925632030614,
      "loss": 0.2884,
      "num_input_tokens_seen": 66263328,
      "step": 34965
    },
    {
      "epoch": 5.704788318786198,
      "grad_norm": 0.003663755487650633,
      "learning_rate": 0.011558240609290104,
      "loss": 0.3248,
      "num_input_tokens_seen": 66271456,
      "step": 34970
    },
    {
      "epoch": 5.705604046007015,
      "grad_norm": 0.001997964456677437,
      "learning_rate": 0.011535576936008679,
      "loss": 0.2797,
      "num_input_tokens_seen": 66279584,
      "step": 34975
    },
    {
      "epoch": 5.706419773227832,
      "grad_norm": 0.001256963238120079,
      "learning_rate": 0.011512934615681309,
      "loss": 0.317,
      "num_input_tokens_seen": 66290240,
      "step": 34980
    },
    {
      "epoch": 5.70723550044865,
      "grad_norm": 0.0023886533454060555,
      "learning_rate": 0.011490313651799765,
      "loss": 0.294,
      "num_input_tokens_seen": 66299056,
      "step": 34985
    },
    {
      "epoch": 5.7080512276694675,
      "grad_norm": 0.0017222712049260736,
      "learning_rate": 0.011467714047852512,
      "loss": 0.3081,
      "num_input_tokens_seen": 66308976,
      "step": 34990
    },
    {
      "epoch": 5.708866954890285,
      "grad_norm": 0.003019446274265647,
      "learning_rate": 0.011445135807324624,
      "loss": 0.3766,
      "num_input_tokens_seen": 66318336,
      "step": 34995
    },
    {
      "epoch": 5.709682682111102,
      "grad_norm": 0.001205683103762567,
      "learning_rate": 0.011422578933698002,
      "loss": 0.3652,
      "num_input_tokens_seen": 66326768,
      "step": 35000
    },
    {
      "epoch": 5.709682682111102,
      "eval_loss": 0.31983235478401184,
      "eval_runtime": 152.8577,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 66326768,
      "step": 35000
    },
    {
      "epoch": 5.71049840933192,
      "grad_norm": 0.0016108420677483082,
      "learning_rate": 0.011400043430451161,
      "loss": 0.3416,
      "num_input_tokens_seen": 66335968,
      "step": 35005
    },
    {
      "epoch": 5.711314136552737,
      "grad_norm": 0.0025751173961907625,
      "learning_rate": 0.011377529301059392,
      "loss": 0.3448,
      "num_input_tokens_seen": 66343728,
      "step": 35010
    },
    {
      "epoch": 5.712129863773554,
      "grad_norm": 0.0016489813569933176,
      "learning_rate": 0.011355036548994646,
      "loss": 0.3107,
      "num_input_tokens_seen": 66352640,
      "step": 35015
    },
    {
      "epoch": 5.712945590994371,
      "grad_norm": 0.002198026282712817,
      "learning_rate": 0.011332565177725584,
      "loss": 0.3231,
      "num_input_tokens_seen": 66361680,
      "step": 35020
    },
    {
      "epoch": 5.713761318215189,
      "grad_norm": 0.0033974216785281897,
      "learning_rate": 0.011310115190717585,
      "loss": 0.3757,
      "num_input_tokens_seen": 66370368,
      "step": 35025
    },
    {
      "epoch": 5.714577045436006,
      "grad_norm": 0.0013311387738212943,
      "learning_rate": 0.01128768659143271,
      "loss": 0.3069,
      "num_input_tokens_seen": 66381120,
      "step": 35030
    },
    {
      "epoch": 5.715392772656823,
      "grad_norm": 0.0012950845994055271,
      "learning_rate": 0.011265279383329713,
      "loss": 0.2759,
      "num_input_tokens_seen": 66390256,
      "step": 35035
    },
    {
      "epoch": 5.716208499877641,
      "grad_norm": 0.001993686892092228,
      "learning_rate": 0.01124289356986411,
      "loss": 0.3539,
      "num_input_tokens_seen": 66399152,
      "step": 35040
    },
    {
      "epoch": 5.717024227098459,
      "grad_norm": 0.0018857656978070736,
      "learning_rate": 0.011220529154488023,
      "loss": 0.3629,
      "num_input_tokens_seen": 66407488,
      "step": 35045
    },
    {
      "epoch": 5.717839954319276,
      "grad_norm": 0.0009385361918248236,
      "learning_rate": 0.011198186140650346,
      "loss": 0.377,
      "num_input_tokens_seen": 66418144,
      "step": 35050
    },
    {
      "epoch": 5.718655681540093,
      "grad_norm": 0.0021727087441831827,
      "learning_rate": 0.011175864531796685,
      "loss": 0.3354,
      "num_input_tokens_seen": 66427568,
      "step": 35055
    },
    {
      "epoch": 5.71947140876091,
      "grad_norm": 0.0012611791025847197,
      "learning_rate": 0.011153564331369258,
      "loss": 0.3583,
      "num_input_tokens_seen": 66438160,
      "step": 35060
    },
    {
      "epoch": 5.720287135981728,
      "grad_norm": 0.00296773761510849,
      "learning_rate": 0.011131285542807078,
      "loss": 0.2959,
      "num_input_tokens_seen": 66447568,
      "step": 35065
    },
    {
      "epoch": 5.721102863202545,
      "grad_norm": 0.0024041596334427595,
      "learning_rate": 0.011109028169545815,
      "loss": 0.2686,
      "num_input_tokens_seen": 66457280,
      "step": 35070
    },
    {
      "epoch": 5.721918590423362,
      "grad_norm": 0.0015700648073107004,
      "learning_rate": 0.011086792215017804,
      "loss": 0.3097,
      "num_input_tokens_seen": 66467600,
      "step": 35075
    },
    {
      "epoch": 5.72273431764418,
      "grad_norm": 0.001367663498967886,
      "learning_rate": 0.011064577682652137,
      "loss": 0.3219,
      "num_input_tokens_seen": 66476992,
      "step": 35080
    },
    {
      "epoch": 5.723550044864997,
      "grad_norm": 0.0031919574830681086,
      "learning_rate": 0.011042384575874559,
      "loss": 0.3329,
      "num_input_tokens_seen": 66485520,
      "step": 35085
    },
    {
      "epoch": 5.7243657720858145,
      "grad_norm": 0.0012609910918399692,
      "learning_rate": 0.011020212898107512,
      "loss": 0.2964,
      "num_input_tokens_seen": 66495600,
      "step": 35090
    },
    {
      "epoch": 5.725181499306632,
      "grad_norm": 0.0022148271091282368,
      "learning_rate": 0.010998062652770197,
      "loss": 0.3268,
      "num_input_tokens_seen": 66504832,
      "step": 35095
    },
    {
      "epoch": 5.725997226527449,
      "grad_norm": 0.0005461641703732312,
      "learning_rate": 0.010975933843278428,
      "loss": 0.3467,
      "num_input_tokens_seen": 66514784,
      "step": 35100
    },
    {
      "epoch": 5.726812953748267,
      "grad_norm": 0.0016297698020935059,
      "learning_rate": 0.010953826473044714,
      "loss": 0.3196,
      "num_input_tokens_seen": 66524464,
      "step": 35105
    },
    {
      "epoch": 5.727628680969084,
      "grad_norm": 0.002112000249326229,
      "learning_rate": 0.010931740545478357,
      "loss": 0.3201,
      "num_input_tokens_seen": 66534336,
      "step": 35110
    },
    {
      "epoch": 5.728444408189901,
      "grad_norm": 0.0022232176270335913,
      "learning_rate": 0.010909676063985218,
      "loss": 0.3057,
      "num_input_tokens_seen": 66543952,
      "step": 35115
    },
    {
      "epoch": 5.729260135410719,
      "grad_norm": 0.001832677866332233,
      "learning_rate": 0.010887633031967974,
      "loss": 0.3681,
      "num_input_tokens_seen": 66553760,
      "step": 35120
    },
    {
      "epoch": 5.730075862631536,
      "grad_norm": 0.003928083926439285,
      "learning_rate": 0.01086561145282589,
      "loss": 0.3239,
      "num_input_tokens_seen": 66562192,
      "step": 35125
    },
    {
      "epoch": 5.730891589852353,
      "grad_norm": 0.007415103260427713,
      "learning_rate": 0.010843611329954983,
      "loss": 0.3726,
      "num_input_tokens_seen": 66572256,
      "step": 35130
    },
    {
      "epoch": 5.7317073170731705,
      "grad_norm": 0.0018984187627211213,
      "learning_rate": 0.010821632666747988,
      "loss": 0.2889,
      "num_input_tokens_seen": 66581376,
      "step": 35135
    },
    {
      "epoch": 5.732523044293988,
      "grad_norm": 0.004056280478835106,
      "learning_rate": 0.010799675466594244,
      "loss": 0.3988,
      "num_input_tokens_seen": 66592256,
      "step": 35140
    },
    {
      "epoch": 5.733338771514806,
      "grad_norm": 0.002791491337120533,
      "learning_rate": 0.010777739732879826,
      "loss": 0.3803,
      "num_input_tokens_seen": 66602272,
      "step": 35145
    },
    {
      "epoch": 5.734154498735623,
      "grad_norm": 0.002475915476679802,
      "learning_rate": 0.010755825468987562,
      "loss": 0.3308,
      "num_input_tokens_seen": 66612400,
      "step": 35150
    },
    {
      "epoch": 5.73497022595644,
      "grad_norm": 0.0006045934860594571,
      "learning_rate": 0.010733932678296814,
      "loss": 0.277,
      "num_input_tokens_seen": 66623248,
      "step": 35155
    },
    {
      "epoch": 5.735785953177258,
      "grad_norm": 0.0013387328945100307,
      "learning_rate": 0.010712061364183817,
      "loss": 0.2867,
      "num_input_tokens_seen": 66632880,
      "step": 35160
    },
    {
      "epoch": 5.736601680398075,
      "grad_norm": 0.0020500144455581903,
      "learning_rate": 0.010690211530021337,
      "loss": 0.304,
      "num_input_tokens_seen": 66642688,
      "step": 35165
    },
    {
      "epoch": 5.737417407618892,
      "grad_norm": 0.0007903265650384128,
      "learning_rate": 0.01066838317917893,
      "loss": 0.3545,
      "num_input_tokens_seen": 66652288,
      "step": 35170
    },
    {
      "epoch": 5.738233134839709,
      "grad_norm": 0.003101607318967581,
      "learning_rate": 0.010646576315022787,
      "loss": 0.3117,
      "num_input_tokens_seen": 66660928,
      "step": 35175
    },
    {
      "epoch": 5.739048862060527,
      "grad_norm": 0.0026678815484046936,
      "learning_rate": 0.010624790940915785,
      "loss": 0.3684,
      "num_input_tokens_seen": 66670240,
      "step": 35180
    },
    {
      "epoch": 5.7398645892813445,
      "grad_norm": 0.0031840603332966566,
      "learning_rate": 0.0106030270602175,
      "loss": 0.3532,
      "num_input_tokens_seen": 66679360,
      "step": 35185
    },
    {
      "epoch": 5.740680316502162,
      "grad_norm": 0.0014451369643211365,
      "learning_rate": 0.010581284676284252,
      "loss": 0.3375,
      "num_input_tokens_seen": 66688032,
      "step": 35190
    },
    {
      "epoch": 5.741496043722979,
      "grad_norm": 0.0006510294042527676,
      "learning_rate": 0.010559563792468923,
      "loss": 0.3092,
      "num_input_tokens_seen": 66696512,
      "step": 35195
    },
    {
      "epoch": 5.742311770943797,
      "grad_norm": 0.0011894431663677096,
      "learning_rate": 0.010537864412121217,
      "loss": 0.3207,
      "num_input_tokens_seen": 66705744,
      "step": 35200
    },
    {
      "epoch": 5.742311770943797,
      "eval_loss": 0.3225876986980438,
      "eval_runtime": 152.9164,
      "eval_samples_per_second": 17.82,
      "eval_steps_per_second": 8.913,
      "num_input_tokens_seen": 66705744,
      "step": 35200
    },
    {
      "epoch": 5.743127498164614,
      "grad_norm": 0.0010645106667652726,
      "learning_rate": 0.010516186538587357,
      "loss": 0.2928,
      "num_input_tokens_seen": 66715520,
      "step": 35205
    },
    {
      "epoch": 5.743943225385431,
      "grad_norm": 0.0021473197266459465,
      "learning_rate": 0.01049453017521042,
      "loss": 0.3605,
      "num_input_tokens_seen": 66726208,
      "step": 35210
    },
    {
      "epoch": 5.744758952606248,
      "grad_norm": 0.0009215413592755795,
      "learning_rate": 0.010472895325330083,
      "loss": 0.358,
      "num_input_tokens_seen": 66736128,
      "step": 35215
    },
    {
      "epoch": 5.745574679827066,
      "grad_norm": 0.002511046826839447,
      "learning_rate": 0.010451281992282662,
      "loss": 0.324,
      "num_input_tokens_seen": 66745952,
      "step": 35220
    },
    {
      "epoch": 5.746390407047883,
      "grad_norm": 0.002306781942024827,
      "learning_rate": 0.01042969017940124,
      "loss": 0.3028,
      "num_input_tokens_seen": 66755792,
      "step": 35225
    },
    {
      "epoch": 5.7472061342687,
      "grad_norm": 0.0017944814171642065,
      "learning_rate": 0.01040811989001557,
      "loss": 0.3234,
      "num_input_tokens_seen": 66765360,
      "step": 35230
    },
    {
      "epoch": 5.7480218614895175,
      "grad_norm": 0.002012907760217786,
      "learning_rate": 0.010386571127451992,
      "loss": 0.3081,
      "num_input_tokens_seen": 66775776,
      "step": 35235
    },
    {
      "epoch": 5.748837588710336,
      "grad_norm": 0.00345303094945848,
      "learning_rate": 0.010365043895033682,
      "loss": 0.4171,
      "num_input_tokens_seen": 66785008,
      "step": 35240
    },
    {
      "epoch": 5.749653315931153,
      "grad_norm": 0.00240922044031322,
      "learning_rate": 0.010343538196080365,
      "loss": 0.3465,
      "num_input_tokens_seen": 66795040,
      "step": 35245
    },
    {
      "epoch": 5.75046904315197,
      "grad_norm": 0.0006093735573813319,
      "learning_rate": 0.010322054033908457,
      "loss": 0.3366,
      "num_input_tokens_seen": 66803504,
      "step": 35250
    },
    {
      "epoch": 5.751284770372787,
      "grad_norm": 0.0026375872548669577,
      "learning_rate": 0.010300591411831156,
      "loss": 0.3036,
      "num_input_tokens_seen": 66814064,
      "step": 35255
    },
    {
      "epoch": 5.752100497593605,
      "grad_norm": 0.002834106096997857,
      "learning_rate": 0.010279150333158198,
      "loss": 0.3516,
      "num_input_tokens_seen": 66823280,
      "step": 35260
    },
    {
      "epoch": 5.752916224814422,
      "grad_norm": 0.003360109170898795,
      "learning_rate": 0.010257730801196107,
      "loss": 0.3083,
      "num_input_tokens_seen": 66832384,
      "step": 35265
    },
    {
      "epoch": 5.753731952035239,
      "grad_norm": 0.002607869217172265,
      "learning_rate": 0.010236332819248056,
      "loss": 0.2893,
      "num_input_tokens_seen": 66840592,
      "step": 35270
    },
    {
      "epoch": 5.754547679256056,
      "grad_norm": 0.0020178041886538267,
      "learning_rate": 0.010214956390613854,
      "loss": 0.3387,
      "num_input_tokens_seen": 66850448,
      "step": 35275
    },
    {
      "epoch": 5.755363406476874,
      "grad_norm": 0.0019964382518082857,
      "learning_rate": 0.010193601518590034,
      "loss": 0.3139,
      "num_input_tokens_seen": 66859536,
      "step": 35280
    },
    {
      "epoch": 5.7561791336976915,
      "grad_norm": 0.002850423799827695,
      "learning_rate": 0.010172268206469758,
      "loss": 0.3177,
      "num_input_tokens_seen": 66869408,
      "step": 35285
    },
    {
      "epoch": 5.756994860918509,
      "grad_norm": 0.0038131000474095345,
      "learning_rate": 0.010150956457542897,
      "loss": 0.3029,
      "num_input_tokens_seen": 66879728,
      "step": 35290
    },
    {
      "epoch": 5.757810588139327,
      "grad_norm": 0.002863852074369788,
      "learning_rate": 0.010129666275096054,
      "loss": 0.2598,
      "num_input_tokens_seen": 66888704,
      "step": 35295
    },
    {
      "epoch": 5.758626315360144,
      "grad_norm": 0.002932174364104867,
      "learning_rate": 0.010108397662412338,
      "loss": 0.4081,
      "num_input_tokens_seen": 66897872,
      "step": 35300
    },
    {
      "epoch": 5.759442042580961,
      "grad_norm": 0.0029335683211684227,
      "learning_rate": 0.010087150622771707,
      "loss": 0.3164,
      "num_input_tokens_seen": 66906832,
      "step": 35305
    },
    {
      "epoch": 5.760257769801778,
      "grad_norm": 0.0007742081652395427,
      "learning_rate": 0.010065925159450739,
      "loss": 0.2991,
      "num_input_tokens_seen": 66915024,
      "step": 35310
    },
    {
      "epoch": 5.761073497022595,
      "grad_norm": 0.002197937574237585,
      "learning_rate": 0.010044721275722618,
      "loss": 0.4254,
      "num_input_tokens_seen": 66924624,
      "step": 35315
    },
    {
      "epoch": 5.761889224243413,
      "grad_norm": 0.0033643541391938925,
      "learning_rate": 0.01002353897485726,
      "loss": 0.3168,
      "num_input_tokens_seen": 66933728,
      "step": 35320
    },
    {
      "epoch": 5.76270495146423,
      "grad_norm": 0.002690626075491309,
      "learning_rate": 0.010002378260121236,
      "loss": 0.3442,
      "num_input_tokens_seen": 66943648,
      "step": 35325
    },
    {
      "epoch": 5.7635206786850475,
      "grad_norm": 0.0038697782438248396,
      "learning_rate": 0.009981239134777786,
      "loss": 0.3321,
      "num_input_tokens_seen": 66951760,
      "step": 35330
    },
    {
      "epoch": 5.7643364059058655,
      "grad_norm": 0.0012029296485707164,
      "learning_rate": 0.009960121602086884,
      "loss": 0.3189,
      "num_input_tokens_seen": 66961200,
      "step": 35335
    },
    {
      "epoch": 5.765152133126683,
      "grad_norm": 0.0014373267767950892,
      "learning_rate": 0.009939025665305062,
      "loss": 0.395,
      "num_input_tokens_seen": 66970208,
      "step": 35340
    },
    {
      "epoch": 5.7659678603475,
      "grad_norm": 0.004879266954958439,
      "learning_rate": 0.009917951327685597,
      "loss": 0.3481,
      "num_input_tokens_seen": 66980272,
      "step": 35345
    },
    {
      "epoch": 5.766783587568317,
      "grad_norm": 0.0023001611698418856,
      "learning_rate": 0.009896898592478425,
      "loss": 0.4055,
      "num_input_tokens_seen": 66989584,
      "step": 35350
    },
    {
      "epoch": 5.767599314789135,
      "grad_norm": 0.0011961179552599788,
      "learning_rate": 0.009875867462930132,
      "loss": 0.334,
      "num_input_tokens_seen": 66997840,
      "step": 35355
    },
    {
      "epoch": 5.768415042009952,
      "grad_norm": 0.0013782009482383728,
      "learning_rate": 0.009854857942284006,
      "loss": 0.3501,
      "num_input_tokens_seen": 67006432,
      "step": 35360
    },
    {
      "epoch": 5.769230769230769,
      "grad_norm": 0.0007584196864627302,
      "learning_rate": 0.009833870033779923,
      "loss": 0.3509,
      "num_input_tokens_seen": 67016640,
      "step": 35365
    },
    {
      "epoch": 5.770046496451586,
      "grad_norm": 0.00223543937318027,
      "learning_rate": 0.009812903740654527,
      "loss": 0.3354,
      "num_input_tokens_seen": 67026288,
      "step": 35370
    },
    {
      "epoch": 5.770862223672404,
      "grad_norm": 0.000619258324149996,
      "learning_rate": 0.009791959066141097,
      "loss": 0.3394,
      "num_input_tokens_seen": 67036560,
      "step": 35375
    },
    {
      "epoch": 5.771677950893221,
      "grad_norm": 0.0025257766246795654,
      "learning_rate": 0.009771036013469537,
      "loss": 0.3221,
      "num_input_tokens_seen": 67046112,
      "step": 35380
    },
    {
      "epoch": 5.772493678114039,
      "grad_norm": 0.003007137682288885,
      "learning_rate": 0.00975013458586646,
      "loss": 0.3464,
      "num_input_tokens_seen": 67056832,
      "step": 35385
    },
    {
      "epoch": 5.773309405334856,
      "grad_norm": 0.0016583199612796307,
      "learning_rate": 0.009729254786555107,
      "loss": 0.3087,
      "num_input_tokens_seen": 67064784,
      "step": 35390
    },
    {
      "epoch": 5.774125132555674,
      "grad_norm": 0.0006218899507075548,
      "learning_rate": 0.009708396618755421,
      "loss": 0.3079,
      "num_input_tokens_seen": 67074720,
      "step": 35395
    },
    {
      "epoch": 5.774940859776491,
      "grad_norm": 0.0006073175463825464,
      "learning_rate": 0.009687560085683994,
      "loss": 0.3184,
      "num_input_tokens_seen": 67084928,
      "step": 35400
    },
    {
      "epoch": 5.774940859776491,
      "eval_loss": 0.32309210300445557,
      "eval_runtime": 152.9821,
      "eval_samples_per_second": 17.813,
      "eval_steps_per_second": 8.91,
      "num_input_tokens_seen": 67084928,
      "step": 35400
    },
    {
      "epoch": 5.775756586997308,
      "grad_norm": 0.0012135819997638464,
      "learning_rate": 0.009666745190554054,
      "loss": 0.3169,
      "num_input_tokens_seen": 67094208,
      "step": 35405
    },
    {
      "epoch": 5.776572314218125,
      "grad_norm": 0.0017893329495564103,
      "learning_rate": 0.009645951936575553,
      "loss": 0.2985,
      "num_input_tokens_seen": 67103872,
      "step": 35410
    },
    {
      "epoch": 5.777388041438943,
      "grad_norm": 0.0005117548862472177,
      "learning_rate": 0.00962518032695509,
      "loss": 0.344,
      "num_input_tokens_seen": 67112816,
      "step": 35415
    },
    {
      "epoch": 5.77820376865976,
      "grad_norm": 0.0021068863570690155,
      "learning_rate": 0.009604430364895855,
      "loss": 0.3197,
      "num_input_tokens_seen": 67122480,
      "step": 35420
    },
    {
      "epoch": 5.779019495880577,
      "grad_norm": 0.001394381164573133,
      "learning_rate": 0.00958370205359777,
      "loss": 0.2757,
      "num_input_tokens_seen": 67130736,
      "step": 35425
    },
    {
      "epoch": 5.7798352231013945,
      "grad_norm": 0.0016675040824338794,
      "learning_rate": 0.009562995396257445,
      "loss": 0.2884,
      "num_input_tokens_seen": 67140480,
      "step": 35430
    },
    {
      "epoch": 5.7806509503222125,
      "grad_norm": 0.0014595211250707507,
      "learning_rate": 0.009542310396068026,
      "loss": 0.354,
      "num_input_tokens_seen": 67149792,
      "step": 35435
    },
    {
      "epoch": 5.78146667754303,
      "grad_norm": 0.0022991604637354612,
      "learning_rate": 0.009521647056219495,
      "loss": 0.4034,
      "num_input_tokens_seen": 67158688,
      "step": 35440
    },
    {
      "epoch": 5.782282404763847,
      "grad_norm": 0.0022104589734226465,
      "learning_rate": 0.00950100537989832,
      "loss": 0.3497,
      "num_input_tokens_seen": 67168224,
      "step": 35445
    },
    {
      "epoch": 5.783098131984664,
      "grad_norm": 0.0017759832553565502,
      "learning_rate": 0.00948038537028772,
      "loss": 0.2809,
      "num_input_tokens_seen": 67178208,
      "step": 35450
    },
    {
      "epoch": 5.783913859205482,
      "grad_norm": 0.0009445344912819564,
      "learning_rate": 0.009459787030567617,
      "loss": 0.3177,
      "num_input_tokens_seen": 67188064,
      "step": 35455
    },
    {
      "epoch": 5.784729586426299,
      "grad_norm": 0.001928856479935348,
      "learning_rate": 0.00943921036391449,
      "loss": 0.3336,
      "num_input_tokens_seen": 67198080,
      "step": 35460
    },
    {
      "epoch": 5.785545313647116,
      "grad_norm": 0.002028623828664422,
      "learning_rate": 0.009418655373501483,
      "loss": 0.3133,
      "num_input_tokens_seen": 67207840,
      "step": 35465
    },
    {
      "epoch": 5.786361040867934,
      "grad_norm": 0.001490298192948103,
      "learning_rate": 0.00939812206249851,
      "loss": 0.326,
      "num_input_tokens_seen": 67217216,
      "step": 35470
    },
    {
      "epoch": 5.787176768088751,
      "grad_norm": 0.0020794044248759747,
      "learning_rate": 0.009377610434072004,
      "loss": 0.3367,
      "num_input_tokens_seen": 67226448,
      "step": 35475
    },
    {
      "epoch": 5.7879924953095685,
      "grad_norm": 0.0012832889333367348,
      "learning_rate": 0.009357120491385167,
      "loss": 0.3423,
      "num_input_tokens_seen": 67235328,
      "step": 35480
    },
    {
      "epoch": 5.788808222530386,
      "grad_norm": 0.0021951189264655113,
      "learning_rate": 0.009336652237597743,
      "loss": 0.3627,
      "num_input_tokens_seen": 67244368,
      "step": 35485
    },
    {
      "epoch": 5.789623949751203,
      "grad_norm": 0.0011191468220204115,
      "learning_rate": 0.009316205675866251,
      "loss": 0.2784,
      "num_input_tokens_seen": 67253648,
      "step": 35490
    },
    {
      "epoch": 5.790439676972021,
      "grad_norm": 0.0021957417484372854,
      "learning_rate": 0.00929578080934379,
      "loss": 0.3305,
      "num_input_tokens_seen": 67264624,
      "step": 35495
    },
    {
      "epoch": 5.791255404192838,
      "grad_norm": 0.0007961058872751892,
      "learning_rate": 0.00927537764118012,
      "loss": 0.266,
      "num_input_tokens_seen": 67275216,
      "step": 35500
    },
    {
      "epoch": 5.792071131413655,
      "grad_norm": 0.0021158228628337383,
      "learning_rate": 0.009254996174521678,
      "loss": 0.3201,
      "num_input_tokens_seen": 67284352,
      "step": 35505
    },
    {
      "epoch": 5.792886858634473,
      "grad_norm": 0.0012134785065427423,
      "learning_rate": 0.009234636412511531,
      "loss": 0.3442,
      "num_input_tokens_seen": 67294000,
      "step": 35510
    },
    {
      "epoch": 5.79370258585529,
      "grad_norm": 0.0022744263987988234,
      "learning_rate": 0.009214298358289418,
      "loss": 0.3484,
      "num_input_tokens_seen": 67303808,
      "step": 35515
    },
    {
      "epoch": 5.794518313076107,
      "grad_norm": 0.0008781035430729389,
      "learning_rate": 0.00919398201499173,
      "loss": 0.302,
      "num_input_tokens_seen": 67313520,
      "step": 35520
    },
    {
      "epoch": 5.7953340402969244,
      "grad_norm": 0.004746865481138229,
      "learning_rate": 0.009173687385751495,
      "loss": 0.3466,
      "num_input_tokens_seen": 67321920,
      "step": 35525
    },
    {
      "epoch": 5.796149767517742,
      "grad_norm": 0.001623838092200458,
      "learning_rate": 0.009153414473698407,
      "loss": 0.3017,
      "num_input_tokens_seen": 67331120,
      "step": 35530
    },
    {
      "epoch": 5.79696549473856,
      "grad_norm": 0.002619013888761401,
      "learning_rate": 0.009133163281958784,
      "loss": 0.3527,
      "num_input_tokens_seen": 67341184,
      "step": 35535
    },
    {
      "epoch": 5.797781221959377,
      "grad_norm": 0.0020709678065031767,
      "learning_rate": 0.009112933813655627,
      "loss": 0.3348,
      "num_input_tokens_seen": 67350864,
      "step": 35540
    },
    {
      "epoch": 5.798596949180194,
      "grad_norm": 0.0016806324711069465,
      "learning_rate": 0.009092726071908573,
      "loss": 0.2548,
      "num_input_tokens_seen": 67359664,
      "step": 35545
    },
    {
      "epoch": 5.799412676401012,
      "grad_norm": 0.001384969917126,
      "learning_rate": 0.0090725400598339,
      "loss": 0.3531,
      "num_input_tokens_seen": 67369520,
      "step": 35550
    },
    {
      "epoch": 5.800228403621829,
      "grad_norm": 0.0015408385079354048,
      "learning_rate": 0.009052375780544563,
      "loss": 0.3557,
      "num_input_tokens_seen": 67378160,
      "step": 35555
    },
    {
      "epoch": 5.801044130842646,
      "grad_norm": 0.0029152012430131435,
      "learning_rate": 0.009032233237150144,
      "loss": 0.2767,
      "num_input_tokens_seen": 67387472,
      "step": 35560
    },
    {
      "epoch": 5.801859858063463,
      "grad_norm": 0.0023337802849709988,
      "learning_rate": 0.009012112432756875,
      "loss": 0.3456,
      "num_input_tokens_seen": 67396064,
      "step": 35565
    },
    {
      "epoch": 5.802675585284281,
      "grad_norm": 0.0033932109363377094,
      "learning_rate": 0.008992013370467605,
      "loss": 0.335,
      "num_input_tokens_seen": 67404832,
      "step": 35570
    },
    {
      "epoch": 5.803491312505098,
      "grad_norm": 0.002902162494137883,
      "learning_rate": 0.008971936053381924,
      "loss": 0.3166,
      "num_input_tokens_seen": 67414688,
      "step": 35575
    },
    {
      "epoch": 5.8043070397259156,
      "grad_norm": 0.0013233679346740246,
      "learning_rate": 0.008951880484595953,
      "loss": 0.317,
      "num_input_tokens_seen": 67424352,
      "step": 35580
    },
    {
      "epoch": 5.805122766946733,
      "grad_norm": 0.0022622346878051758,
      "learning_rate": 0.008931846667202552,
      "loss": 0.2457,
      "num_input_tokens_seen": 67434224,
      "step": 35585
    },
    {
      "epoch": 5.805938494167551,
      "grad_norm": 0.0015836091479286551,
      "learning_rate": 0.008911834604291152,
      "loss": 0.3616,
      "num_input_tokens_seen": 67443312,
      "step": 35590
    },
    {
      "epoch": 5.806754221388368,
      "grad_norm": 0.0015137253794819117,
      "learning_rate": 0.008891844298947882,
      "loss": 0.325,
      "num_input_tokens_seen": 67452224,
      "step": 35595
    },
    {
      "epoch": 5.807569948609185,
      "grad_norm": 0.001738866325467825,
      "learning_rate": 0.008871875754255508,
      "loss": 0.3183,
      "num_input_tokens_seen": 67462064,
      "step": 35600
    },
    {
      "epoch": 5.807569948609185,
      "eval_loss": 0.32366129755973816,
      "eval_runtime": 152.9805,
      "eval_samples_per_second": 17.813,
      "eval_steps_per_second": 8.91,
      "num_input_tokens_seen": 67462064,
      "step": 35600
    },
    {
      "epoch": 5.808385675830002,
      "grad_norm": 0.0018324849661439657,
      "learning_rate": 0.008851928973293422,
      "loss": 0.3458,
      "num_input_tokens_seen": 67471728,
      "step": 35605
    },
    {
      "epoch": 5.80920140305082,
      "grad_norm": 0.002434868598356843,
      "learning_rate": 0.00883200395913764,
      "loss": 0.3385,
      "num_input_tokens_seen": 67481968,
      "step": 35610
    },
    {
      "epoch": 5.810017130271637,
      "grad_norm": 0.0017218060093000531,
      "learning_rate": 0.00881210071486091,
      "loss": 0.3365,
      "num_input_tokens_seen": 67492304,
      "step": 35615
    },
    {
      "epoch": 5.810832857492454,
      "grad_norm": 0.0020260042510926723,
      "learning_rate": 0.008792219243532505,
      "loss": 0.3157,
      "num_input_tokens_seen": 67500816,
      "step": 35620
    },
    {
      "epoch": 5.8116485847132715,
      "grad_norm": 0.005043891258537769,
      "learning_rate": 0.008772359548218428,
      "loss": 0.3085,
      "num_input_tokens_seen": 67511232,
      "step": 35625
    },
    {
      "epoch": 5.8124643119340895,
      "grad_norm": 0.000980724347755313,
      "learning_rate": 0.008752521631981274,
      "loss": 0.3284,
      "num_input_tokens_seen": 67520048,
      "step": 35630
    },
    {
      "epoch": 5.813280039154907,
      "grad_norm": 0.0027372483164072037,
      "learning_rate": 0.008732705497880315,
      "loss": 0.3825,
      "num_input_tokens_seen": 67528480,
      "step": 35635
    },
    {
      "epoch": 5.814095766375724,
      "grad_norm": 0.0009225704707205296,
      "learning_rate": 0.008712911148971459,
      "loss": 0.3643,
      "num_input_tokens_seen": 67537904,
      "step": 35640
    },
    {
      "epoch": 5.814911493596542,
      "grad_norm": 0.000729429186321795,
      "learning_rate": 0.008693138588307208,
      "loss": 0.2937,
      "num_input_tokens_seen": 67546800,
      "step": 35645
    },
    {
      "epoch": 5.815727220817359,
      "grad_norm": 0.004871939308941364,
      "learning_rate": 0.008673387818936762,
      "loss": 0.3335,
      "num_input_tokens_seen": 67555904,
      "step": 35650
    },
    {
      "epoch": 5.816542948038176,
      "grad_norm": 0.0026189717464149,
      "learning_rate": 0.008653658843905948,
      "loss": 0.3793,
      "num_input_tokens_seen": 67564560,
      "step": 35655
    },
    {
      "epoch": 5.817358675258993,
      "grad_norm": 0.0048306905664503574,
      "learning_rate": 0.0086339516662572,
      "loss": 0.338,
      "num_input_tokens_seen": 67573232,
      "step": 35660
    },
    {
      "epoch": 5.81817440247981,
      "grad_norm": 0.0030875022057443857,
      "learning_rate": 0.008614266289029638,
      "loss": 0.3362,
      "num_input_tokens_seen": 67584112,
      "step": 35665
    },
    {
      "epoch": 5.818990129700628,
      "grad_norm": 0.0018501213053241372,
      "learning_rate": 0.008594602715258965,
      "loss": 0.3118,
      "num_input_tokens_seen": 67592592,
      "step": 35670
    },
    {
      "epoch": 5.8198058569214455,
      "grad_norm": 0.0008929290343075991,
      "learning_rate": 0.008574960947977573,
      "loss": 0.2791,
      "num_input_tokens_seen": 67602208,
      "step": 35675
    },
    {
      "epoch": 5.820621584142263,
      "grad_norm": 0.003328314982354641,
      "learning_rate": 0.008555340990214438,
      "loss": 0.3455,
      "num_input_tokens_seen": 67611760,
      "step": 35680
    },
    {
      "epoch": 5.821437311363081,
      "grad_norm": 0.0013410437386482954,
      "learning_rate": 0.008535742844995258,
      "loss": 0.2912,
      "num_input_tokens_seen": 67622384,
      "step": 35685
    },
    {
      "epoch": 5.822253038583898,
      "grad_norm": 0.00206751492805779,
      "learning_rate": 0.008516166515342266,
      "loss": 0.305,
      "num_input_tokens_seen": 67631456,
      "step": 35690
    },
    {
      "epoch": 5.823068765804715,
      "grad_norm": 0.002840962028130889,
      "learning_rate": 0.008496612004274411,
      "loss": 0.3245,
      "num_input_tokens_seen": 67641536,
      "step": 35695
    },
    {
      "epoch": 5.823884493025532,
      "grad_norm": 0.0009324545389972627,
      "learning_rate": 0.008477079314807201,
      "loss": 0.2975,
      "num_input_tokens_seen": 67648416,
      "step": 35700
    },
    {
      "epoch": 5.824700220246349,
      "grad_norm": 0.0020145776215940714,
      "learning_rate": 0.008457568449952874,
      "loss": 0.3892,
      "num_input_tokens_seen": 67658000,
      "step": 35705
    },
    {
      "epoch": 5.825515947467167,
      "grad_norm": 0.0011622877791523933,
      "learning_rate": 0.008438079412720189,
      "loss": 0.3968,
      "num_input_tokens_seen": 67669104,
      "step": 35710
    },
    {
      "epoch": 5.826331674687984,
      "grad_norm": 0.0045653413981199265,
      "learning_rate": 0.00841861220611466,
      "loss": 0.3603,
      "num_input_tokens_seen": 67679040,
      "step": 35715
    },
    {
      "epoch": 5.827147401908801,
      "grad_norm": 0.0024555635172873735,
      "learning_rate": 0.008399166833138355,
      "loss": 0.326,
      "num_input_tokens_seen": 67689712,
      "step": 35720
    },
    {
      "epoch": 5.8279631291296194,
      "grad_norm": 0.002454361179843545,
      "learning_rate": 0.008379743296789987,
      "loss": 0.349,
      "num_input_tokens_seen": 67700352,
      "step": 35725
    },
    {
      "epoch": 5.828778856350437,
      "grad_norm": 0.0023722320329397917,
      "learning_rate": 0.008360341600064896,
      "loss": 0.3403,
      "num_input_tokens_seen": 67708880,
      "step": 35730
    },
    {
      "epoch": 5.829594583571254,
      "grad_norm": 0.00289272447116673,
      "learning_rate": 0.008340961745955121,
      "loss": 0.3734,
      "num_input_tokens_seen": 67719744,
      "step": 35735
    },
    {
      "epoch": 5.830410310792071,
      "grad_norm": 0.0015106556238606572,
      "learning_rate": 0.008321603737449224,
      "loss": 0.3395,
      "num_input_tokens_seen": 67729696,
      "step": 35740
    },
    {
      "epoch": 5.831226038012889,
      "grad_norm": 0.002067981055006385,
      "learning_rate": 0.008302267577532479,
      "loss": 0.2844,
      "num_input_tokens_seen": 67739952,
      "step": 35745
    },
    {
      "epoch": 5.832041765233706,
      "grad_norm": 0.002914001699537039,
      "learning_rate": 0.008282953269186771,
      "loss": 0.3241,
      "num_input_tokens_seen": 67749152,
      "step": 35750
    },
    {
      "epoch": 5.832857492454523,
      "grad_norm": 0.0007420475012622774,
      "learning_rate": 0.008263660815390567,
      "loss": 0.3593,
      "num_input_tokens_seen": 67757968,
      "step": 35755
    },
    {
      "epoch": 5.83367321967534,
      "grad_norm": 0.0012256671907380223,
      "learning_rate": 0.008244390219119069,
      "loss": 0.3413,
      "num_input_tokens_seen": 67768144,
      "step": 35760
    },
    {
      "epoch": 5.834488946896158,
      "grad_norm": 0.0012324630515649915,
      "learning_rate": 0.008225141483343967,
      "loss": 0.3295,
      "num_input_tokens_seen": 67777184,
      "step": 35765
    },
    {
      "epoch": 5.835304674116975,
      "grad_norm": 0.0017109251348301768,
      "learning_rate": 0.00820591461103372,
      "loss": 0.3469,
      "num_input_tokens_seen": 67787504,
      "step": 35770
    },
    {
      "epoch": 5.8361204013377925,
      "grad_norm": 0.001196288038045168,
      "learning_rate": 0.008186709605153358,
      "loss": 0.2764,
      "num_input_tokens_seen": 67797664,
      "step": 35775
    },
    {
      "epoch": 5.83693612855861,
      "grad_norm": 0.0009331470937468112,
      "learning_rate": 0.008167526468664492,
      "loss": 0.2635,
      "num_input_tokens_seen": 67808160,
      "step": 35780
    },
    {
      "epoch": 5.837751855779428,
      "grad_norm": 0.0013197639491409063,
      "learning_rate": 0.008148365204525443,
      "loss": 0.3256,
      "num_input_tokens_seen": 67817216,
      "step": 35785
    },
    {
      "epoch": 5.838567583000245,
      "grad_norm": 0.0007223634165711701,
      "learning_rate": 0.00812922581569106,
      "loss": 0.3319,
      "num_input_tokens_seen": 67824848,
      "step": 35790
    },
    {
      "epoch": 5.839383310221062,
      "grad_norm": 0.0008754496229812503,
      "learning_rate": 0.008110108305112934,
      "loss": 0.3882,
      "num_input_tokens_seen": 67835856,
      "step": 35795
    },
    {
      "epoch": 5.840199037441879,
      "grad_norm": 0.0015924449544399977,
      "learning_rate": 0.008091012675739223,
      "loss": 0.2891,
      "num_input_tokens_seen": 67846112,
      "step": 35800
    },
    {
      "epoch": 5.840199037441879,
      "eval_loss": 0.3236691951751709,
      "eval_runtime": 153.016,
      "eval_samples_per_second": 17.809,
      "eval_steps_per_second": 8.908,
      "num_input_tokens_seen": 67846112,
      "step": 35800
    },
    {
      "epoch": 5.841014764662697,
      "grad_norm": 0.0024165890645235777,
      "learning_rate": 0.008071938930514671,
      "loss": 0.383,
      "num_input_tokens_seen": 67856032,
      "step": 35805
    },
    {
      "epoch": 5.841830491883514,
      "grad_norm": 0.002338351681828499,
      "learning_rate": 0.008052887072380726,
      "loss": 0.2943,
      "num_input_tokens_seen": 67865568,
      "step": 35810
    },
    {
      "epoch": 5.842646219104331,
      "grad_norm": 0.0008229984086938202,
      "learning_rate": 0.008033857104275437,
      "loss": 0.2702,
      "num_input_tokens_seen": 67875200,
      "step": 35815
    },
    {
      "epoch": 5.8434619463251485,
      "grad_norm": 0.003333658678457141,
      "learning_rate": 0.008014849029133424,
      "loss": 0.3809,
      "num_input_tokens_seen": 67884464,
      "step": 35820
    },
    {
      "epoch": 5.8442776735459665,
      "grad_norm": 0.0011207059724256396,
      "learning_rate": 0.007995862849885975,
      "loss": 0.3053,
      "num_input_tokens_seen": 67893248,
      "step": 35825
    },
    {
      "epoch": 5.845093400766784,
      "grad_norm": 0.0014773280126973987,
      "learning_rate": 0.007976898569461032,
      "loss": 0.2982,
      "num_input_tokens_seen": 67902384,
      "step": 35830
    },
    {
      "epoch": 5.845909127987601,
      "grad_norm": 0.0010090528521686792,
      "learning_rate": 0.007957956190783088,
      "loss": 0.3576,
      "num_input_tokens_seen": 67911600,
      "step": 35835
    },
    {
      "epoch": 5.846724855208418,
      "grad_norm": 0.002197813941165805,
      "learning_rate": 0.007939035716773324,
      "loss": 0.2995,
      "num_input_tokens_seen": 67919792,
      "step": 35840
    },
    {
      "epoch": 5.847540582429236,
      "grad_norm": 0.0025969897396862507,
      "learning_rate": 0.007920137150349487,
      "loss": 0.3341,
      "num_input_tokens_seen": 67929008,
      "step": 35845
    },
    {
      "epoch": 5.848356309650053,
      "grad_norm": 0.0010986827546730638,
      "learning_rate": 0.007901260494425981,
      "loss": 0.3044,
      "num_input_tokens_seen": 67937712,
      "step": 35850
    },
    {
      "epoch": 5.84917203687087,
      "grad_norm": 0.0011772605357691646,
      "learning_rate": 0.007882405751913861,
      "loss": 0.3304,
      "num_input_tokens_seen": 67947024,
      "step": 35855
    },
    {
      "epoch": 5.849987764091688,
      "grad_norm": 0.0007300612051039934,
      "learning_rate": 0.007863572925720702,
      "loss": 0.3296,
      "num_input_tokens_seen": 67956480,
      "step": 35860
    },
    {
      "epoch": 5.850803491312505,
      "grad_norm": 0.003962668590247631,
      "learning_rate": 0.007844762018750827,
      "loss": 0.2956,
      "num_input_tokens_seen": 67965232,
      "step": 35865
    },
    {
      "epoch": 5.8516192185333225,
      "grad_norm": 0.001333537744358182,
      "learning_rate": 0.007825973033905054,
      "loss": 0.3518,
      "num_input_tokens_seen": 67975216,
      "step": 35870
    },
    {
      "epoch": 5.85243494575414,
      "grad_norm": 0.0028923149220645428,
      "learning_rate": 0.007807205974080927,
      "loss": 0.2899,
      "num_input_tokens_seen": 67983760,
      "step": 35875
    },
    {
      "epoch": 5.853250672974957,
      "grad_norm": 0.0008450672612525523,
      "learning_rate": 0.007788460842172551,
      "loss": 0.2986,
      "num_input_tokens_seen": 67994128,
      "step": 35880
    },
    {
      "epoch": 5.854066400195775,
      "grad_norm": 0.002088720677420497,
      "learning_rate": 0.0077697376410706285,
      "loss": 0.3176,
      "num_input_tokens_seen": 68002576,
      "step": 35885
    },
    {
      "epoch": 5.854882127416592,
      "grad_norm": 0.003462370950728655,
      "learning_rate": 0.007751036373662567,
      "loss": 0.2976,
      "num_input_tokens_seen": 68011536,
      "step": 35890
    },
    {
      "epoch": 5.855697854637409,
      "grad_norm": 0.0008125149179250002,
      "learning_rate": 0.00773235704283231,
      "loss": 0.3191,
      "num_input_tokens_seen": 68021264,
      "step": 35895
    },
    {
      "epoch": 5.856513581858227,
      "grad_norm": 0.0021744980476796627,
      "learning_rate": 0.007713699651460437,
      "loss": 0.2809,
      "num_input_tokens_seen": 68031360,
      "step": 35900
    },
    {
      "epoch": 5.857329309079044,
      "grad_norm": 0.0022545834071934223,
      "learning_rate": 0.007695064202424162,
      "loss": 0.3234,
      "num_input_tokens_seen": 68039408,
      "step": 35905
    },
    {
      "epoch": 5.858145036299861,
      "grad_norm": 0.002722910139709711,
      "learning_rate": 0.007676450698597286,
      "loss": 0.321,
      "num_input_tokens_seen": 68048640,
      "step": 35910
    },
    {
      "epoch": 5.858960763520678,
      "grad_norm": 0.004553500562906265,
      "learning_rate": 0.007657859142850265,
      "loss": 0.2834,
      "num_input_tokens_seen": 68058064,
      "step": 35915
    },
    {
      "epoch": 5.859776490741496,
      "grad_norm": 0.0007566954009234905,
      "learning_rate": 0.0076392895380501535,
      "loss": 0.3522,
      "num_input_tokens_seen": 68068016,
      "step": 35920
    },
    {
      "epoch": 5.8605922179623136,
      "grad_norm": 0.0010301567381247878,
      "learning_rate": 0.007620741887060611,
      "loss": 0.2745,
      "num_input_tokens_seen": 68076160,
      "step": 35925
    },
    {
      "epoch": 5.861407945183131,
      "grad_norm": 0.0006817699759267271,
      "learning_rate": 0.007602216192741901,
      "loss": 0.3254,
      "num_input_tokens_seen": 68086288,
      "step": 35930
    },
    {
      "epoch": 5.862223672403948,
      "grad_norm": 0.000984969432465732,
      "learning_rate": 0.007583712457950969,
      "loss": 0.3062,
      "num_input_tokens_seen": 68096464,
      "step": 35935
    },
    {
      "epoch": 5.863039399624766,
      "grad_norm": 0.0012720340164378285,
      "learning_rate": 0.007565230685541269,
      "loss": 0.3992,
      "num_input_tokens_seen": 68106368,
      "step": 35940
    },
    {
      "epoch": 5.863855126845583,
      "grad_norm": 0.0025414051488041878,
      "learning_rate": 0.007546770878362968,
      "loss": 0.3214,
      "num_input_tokens_seen": 68117472,
      "step": 35945
    },
    {
      "epoch": 5.8646708540664,
      "grad_norm": 0.0018609765684232116,
      "learning_rate": 0.0075283330392627405,
      "loss": 0.3762,
      "num_input_tokens_seen": 68127696,
      "step": 35950
    },
    {
      "epoch": 5.865486581287217,
      "grad_norm": 0.0018233975861221552,
      "learning_rate": 0.007509917171083979,
      "loss": 0.3532,
      "num_input_tokens_seen": 68136688,
      "step": 35955
    },
    {
      "epoch": 5.866302308508035,
      "grad_norm": 0.0024149641394615173,
      "learning_rate": 0.007491523276666662,
      "loss": 0.3208,
      "num_input_tokens_seen": 68146096,
      "step": 35960
    },
    {
      "epoch": 5.867118035728852,
      "grad_norm": 0.0009989049285650253,
      "learning_rate": 0.007473151358847318,
      "loss": 0.3454,
      "num_input_tokens_seen": 68155408,
      "step": 35965
    },
    {
      "epoch": 5.8679337629496695,
      "grad_norm": 0.0026946314610540867,
      "learning_rate": 0.007454801420459117,
      "loss": 0.3474,
      "num_input_tokens_seen": 68165200,
      "step": 35970
    },
    {
      "epoch": 5.868749490170487,
      "grad_norm": 0.00134500942658633,
      "learning_rate": 0.0074364734643319105,
      "loss": 0.3015,
      "num_input_tokens_seen": 68174528,
      "step": 35975
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 0.000506291224155575,
      "learning_rate": 0.007418167493292022,
      "loss": 0.3231,
      "num_input_tokens_seen": 68183904,
      "step": 35980
    },
    {
      "epoch": 5.870380944612122,
      "grad_norm": 0.0012532103573903441,
      "learning_rate": 0.0073998835101625245,
      "loss": 0.2832,
      "num_input_tokens_seen": 68193104,
      "step": 35985
    },
    {
      "epoch": 5.871196671832939,
      "grad_norm": 0.0022576269693672657,
      "learning_rate": 0.007381621517762998,
      "loss": 0.2915,
      "num_input_tokens_seen": 68202640,
      "step": 35990
    },
    {
      "epoch": 5.872012399053756,
      "grad_norm": 0.001448795897886157,
      "learning_rate": 0.007363381518909689,
      "loss": 0.2949,
      "num_input_tokens_seen": 68213024,
      "step": 35995
    },
    {
      "epoch": 5.872828126274574,
      "grad_norm": 0.003417057916522026,
      "learning_rate": 0.007345163516415448,
      "loss": 0.3224,
      "num_input_tokens_seen": 68221552,
      "step": 36000
    },
    {
      "epoch": 5.872828126274574,
      "eval_loss": 0.32177165150642395,
      "eval_runtime": 152.8351,
      "eval_samples_per_second": 17.83,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 68221552,
      "step": 36000
    },
    {
      "epoch": 5.873643853495391,
      "grad_norm": 0.001275992370210588,
      "learning_rate": 0.007326967513089693,
      "loss": 0.3155,
      "num_input_tokens_seen": 68231312,
      "step": 36005
    },
    {
      "epoch": 5.874459580716208,
      "grad_norm": 0.0019965707324445248,
      "learning_rate": 0.0073087935117384815,
      "loss": 0.3204,
      "num_input_tokens_seen": 68240944,
      "step": 36010
    },
    {
      "epoch": 5.8752753079370255,
      "grad_norm": 0.0014833534369245172,
      "learning_rate": 0.007290641515164503,
      "loss": 0.3724,
      "num_input_tokens_seen": 68251968,
      "step": 36015
    },
    {
      "epoch": 5.8760910351578435,
      "grad_norm": 0.002477051690220833,
      "learning_rate": 0.007272511526166986,
      "loss": 0.2991,
      "num_input_tokens_seen": 68262576,
      "step": 36020
    },
    {
      "epoch": 5.876906762378661,
      "grad_norm": 0.0014078309759497643,
      "learning_rate": 0.0072544035475418265,
      "loss": 0.3352,
      "num_input_tokens_seen": 68272192,
      "step": 36025
    },
    {
      "epoch": 5.877722489599478,
      "grad_norm": 0.001235049800015986,
      "learning_rate": 0.007236317582081475,
      "loss": 0.3186,
      "num_input_tokens_seen": 68280896,
      "step": 36030
    },
    {
      "epoch": 5.878538216820296,
      "grad_norm": 0.0007113346364349127,
      "learning_rate": 0.007218253632575066,
      "loss": 0.3004,
      "num_input_tokens_seen": 68289568,
      "step": 36035
    },
    {
      "epoch": 5.879353944041113,
      "grad_norm": 0.0026733181439340115,
      "learning_rate": 0.007200211701808223,
      "loss": 0.3055,
      "num_input_tokens_seen": 68300544,
      "step": 36040
    },
    {
      "epoch": 5.88016967126193,
      "grad_norm": 0.0044080032967031,
      "learning_rate": 0.007182191792563286,
      "loss": 0.3347,
      "num_input_tokens_seen": 68309584,
      "step": 36045
    },
    {
      "epoch": 5.880985398482747,
      "grad_norm": 0.0027557353023439646,
      "learning_rate": 0.0071641939076191145,
      "loss": 0.3485,
      "num_input_tokens_seen": 68318240,
      "step": 36050
    },
    {
      "epoch": 5.881801125703564,
      "grad_norm": 0.0012813522480428219,
      "learning_rate": 0.007146218049751257,
      "loss": 0.273,
      "num_input_tokens_seen": 68326912,
      "step": 36055
    },
    {
      "epoch": 5.882616852924382,
      "grad_norm": 0.003538062795996666,
      "learning_rate": 0.0071282642217317775,
      "loss": 0.3274,
      "num_input_tokens_seen": 68336528,
      "step": 36060
    },
    {
      "epoch": 5.883432580145199,
      "grad_norm": 0.00085820397362113,
      "learning_rate": 0.007110332426329396,
      "loss": 0.3221,
      "num_input_tokens_seen": 68346464,
      "step": 36065
    },
    {
      "epoch": 5.884248307366017,
      "grad_norm": 0.0027843990828841925,
      "learning_rate": 0.007092422666309417,
      "loss": 0.3488,
      "num_input_tokens_seen": 68356448,
      "step": 36070
    },
    {
      "epoch": 5.885064034586835,
      "grad_norm": 0.001780004589818418,
      "learning_rate": 0.0070745349444337295,
      "loss": 0.3118,
      "num_input_tokens_seen": 68366256,
      "step": 36075
    },
    {
      "epoch": 5.885879761807652,
      "grad_norm": 0.001426928211003542,
      "learning_rate": 0.007056669263460913,
      "loss": 0.2748,
      "num_input_tokens_seen": 68375936,
      "step": 36080
    },
    {
      "epoch": 5.886695489028469,
      "grad_norm": 0.003016713075339794,
      "learning_rate": 0.007038825626145995,
      "loss": 0.3646,
      "num_input_tokens_seen": 68385792,
      "step": 36085
    },
    {
      "epoch": 5.887511216249286,
      "grad_norm": 0.0005897577502764761,
      "learning_rate": 0.007021004035240724,
      "loss": 0.3674,
      "num_input_tokens_seen": 68394720,
      "step": 36090
    },
    {
      "epoch": 5.888326943470103,
      "grad_norm": 0.0017979793483391404,
      "learning_rate": 0.007003204493493453,
      "loss": 0.304,
      "num_input_tokens_seen": 68404304,
      "step": 36095
    },
    {
      "epoch": 5.889142670690921,
      "grad_norm": 0.0011750530684366822,
      "learning_rate": 0.006985427003649036,
      "loss": 0.2747,
      "num_input_tokens_seen": 68414560,
      "step": 36100
    },
    {
      "epoch": 5.889958397911738,
      "grad_norm": 0.0011297953315079212,
      "learning_rate": 0.006967671568449013,
      "loss": 0.293,
      "num_input_tokens_seen": 68424400,
      "step": 36105
    },
    {
      "epoch": 5.890774125132555,
      "grad_norm": 0.0020327253732830286,
      "learning_rate": 0.006949938190631511,
      "loss": 0.3187,
      "num_input_tokens_seen": 68434496,
      "step": 36110
    },
    {
      "epoch": 5.891589852353373,
      "grad_norm": 0.0007085913675837219,
      "learning_rate": 0.0069322268729311905,
      "loss": 0.306,
      "num_input_tokens_seen": 68443664,
      "step": 36115
    },
    {
      "epoch": 5.8924055795741905,
      "grad_norm": 0.0011252672411501408,
      "learning_rate": 0.006914537618079403,
      "loss": 0.3493,
      "num_input_tokens_seen": 68452864,
      "step": 36120
    },
    {
      "epoch": 5.893221306795008,
      "grad_norm": 0.0018565679201856256,
      "learning_rate": 0.006896870428804031,
      "loss": 0.337,
      "num_input_tokens_seen": 68462224,
      "step": 36125
    },
    {
      "epoch": 5.894037034015825,
      "grad_norm": 0.0007667793543078005,
      "learning_rate": 0.006879225307829595,
      "loss": 0.2793,
      "num_input_tokens_seen": 68471632,
      "step": 36130
    },
    {
      "epoch": 5.894852761236643,
      "grad_norm": 0.0014972984790802002,
      "learning_rate": 0.00686160225787717,
      "loss": 0.3296,
      "num_input_tokens_seen": 68481696,
      "step": 36135
    },
    {
      "epoch": 5.89566848845746,
      "grad_norm": 0.0024485273752361536,
      "learning_rate": 0.006844001281664463,
      "loss": 0.36,
      "num_input_tokens_seen": 68491328,
      "step": 36140
    },
    {
      "epoch": 5.896484215678277,
      "grad_norm": 0.001816603704355657,
      "learning_rate": 0.006826422381905789,
      "loss": 0.3073,
      "num_input_tokens_seen": 68500096,
      "step": 36145
    },
    {
      "epoch": 5.897299942899094,
      "grad_norm": 0.0032812508288770914,
      "learning_rate": 0.006808865561311994,
      "loss": 0.3613,
      "num_input_tokens_seen": 68510992,
      "step": 36150
    },
    {
      "epoch": 5.898115670119912,
      "grad_norm": 0.0012719043297693133,
      "learning_rate": 0.00679133082259058,
      "loss": 0.3374,
      "num_input_tokens_seen": 68521328,
      "step": 36155
    },
    {
      "epoch": 5.898931397340729,
      "grad_norm": 0.001773348543792963,
      "learning_rate": 0.00677381816844565,
      "loss": 0.2661,
      "num_input_tokens_seen": 68531968,
      "step": 36160
    },
    {
      "epoch": 5.8997471245615465,
      "grad_norm": 0.002249429700896144,
      "learning_rate": 0.0067563276015778434,
      "loss": 0.3185,
      "num_input_tokens_seen": 68540864,
      "step": 36165
    },
    {
      "epoch": 5.900562851782364,
      "grad_norm": 0.003118047257885337,
      "learning_rate": 0.006738859124684437,
      "loss": 0.3242,
      "num_input_tokens_seen": 68549952,
      "step": 36170
    },
    {
      "epoch": 5.901378579003182,
      "grad_norm": 0.0010753230890259147,
      "learning_rate": 0.006721412740459259,
      "loss": 0.3087,
      "num_input_tokens_seen": 68558960,
      "step": 36175
    },
    {
      "epoch": 5.902194306223999,
      "grad_norm": 0.002628894755616784,
      "learning_rate": 0.006703988451592824,
      "loss": 0.3112,
      "num_input_tokens_seen": 68567808,
      "step": 36180
    },
    {
      "epoch": 5.903010033444816,
      "grad_norm": 0.005044491961598396,
      "learning_rate": 0.006686586260772114,
      "loss": 0.3343,
      "num_input_tokens_seen": 68578064,
      "step": 36185
    },
    {
      "epoch": 5.903825760665633,
      "grad_norm": 0.0016070317942649126,
      "learning_rate": 0.006669206170680819,
      "loss": 0.2642,
      "num_input_tokens_seen": 68587920,
      "step": 36190
    },
    {
      "epoch": 5.904641487886451,
      "grad_norm": 0.0029886336997151375,
      "learning_rate": 0.0066518481839991095,
      "loss": 0.2875,
      "num_input_tokens_seen": 68597840,
      "step": 36195
    },
    {
      "epoch": 5.905457215107268,
      "grad_norm": 0.002289747353643179,
      "learning_rate": 0.006634512303403861,
      "loss": 0.2998,
      "num_input_tokens_seen": 68606416,
      "step": 36200
    },
    {
      "epoch": 5.905457215107268,
      "eval_loss": 0.32160013914108276,
      "eval_runtime": 152.7575,
      "eval_samples_per_second": 17.839,
      "eval_steps_per_second": 8.923,
      "num_input_tokens_seen": 68606416,
      "step": 36200
    },
    {
      "epoch": 5.906272942328085,
      "grad_norm": 0.0028350900392979383,
      "learning_rate": 0.0066171985315684355,
      "loss": 0.289,
      "num_input_tokens_seen": 68616224,
      "step": 36205
    },
    {
      "epoch": 5.907088669548903,
      "grad_norm": 0.00240883557125926,
      "learning_rate": 0.0065999068711628806,
      "loss": 0.331,
      "num_input_tokens_seen": 68625344,
      "step": 36210
    },
    {
      "epoch": 5.9079043967697205,
      "grad_norm": 0.0021200336050242186,
      "learning_rate": 0.0065826373248537295,
      "loss": 0.32,
      "num_input_tokens_seen": 68633680,
      "step": 36215
    },
    {
      "epoch": 5.908720123990538,
      "grad_norm": 0.0027786477003246546,
      "learning_rate": 0.006565389895304218,
      "loss": 0.2975,
      "num_input_tokens_seen": 68642880,
      "step": 36220
    },
    {
      "epoch": 5.909535851211355,
      "grad_norm": 0.004161958582699299,
      "learning_rate": 0.006548164585174104,
      "loss": 0.3532,
      "num_input_tokens_seen": 68652512,
      "step": 36225
    },
    {
      "epoch": 5.910351578432172,
      "grad_norm": 0.0010446814121678472,
      "learning_rate": 0.006530961397119728,
      "loss": 0.3075,
      "num_input_tokens_seen": 68661776,
      "step": 36230
    },
    {
      "epoch": 5.91116730565299,
      "grad_norm": 0.001731687574647367,
      "learning_rate": 0.00651378033379405,
      "loss": 0.3823,
      "num_input_tokens_seen": 68671328,
      "step": 36235
    },
    {
      "epoch": 5.911983032873807,
      "grad_norm": 0.003629962680861354,
      "learning_rate": 0.006496621397846619,
      "loss": 0.3285,
      "num_input_tokens_seen": 68679888,
      "step": 36240
    },
    {
      "epoch": 5.912798760094624,
      "grad_norm": 0.0011498340172693133,
      "learning_rate": 0.006479484591923518,
      "loss": 0.3045,
      "num_input_tokens_seen": 68690496,
      "step": 36245
    },
    {
      "epoch": 5.913614487315442,
      "grad_norm": 0.0014503876445814967,
      "learning_rate": 0.006462369918667515,
      "loss": 0.3626,
      "num_input_tokens_seen": 68699728,
      "step": 36250
    },
    {
      "epoch": 5.914430214536259,
      "grad_norm": 0.001798443030565977,
      "learning_rate": 0.006445277380717851,
      "loss": 0.3239,
      "num_input_tokens_seen": 68708160,
      "step": 36255
    },
    {
      "epoch": 5.915245941757076,
      "grad_norm": 0.0018410645425319672,
      "learning_rate": 0.006428206980710466,
      "loss": 0.3435,
      "num_input_tokens_seen": 68717904,
      "step": 36260
    },
    {
      "epoch": 5.9160616689778935,
      "grad_norm": 0.0021912148222327232,
      "learning_rate": 0.006411158721277788,
      "loss": 0.322,
      "num_input_tokens_seen": 68728048,
      "step": 36265
    },
    {
      "epoch": 5.916877396198711,
      "grad_norm": 0.0030608519446104765,
      "learning_rate": 0.00639413260504888,
      "loss": 0.345,
      "num_input_tokens_seen": 68737984,
      "step": 36270
    },
    {
      "epoch": 5.917693123419529,
      "grad_norm": 0.0025371192023158073,
      "learning_rate": 0.006377128634649376,
      "loss": 0.2782,
      "num_input_tokens_seen": 68746096,
      "step": 36275
    },
    {
      "epoch": 5.918508850640346,
      "grad_norm": 0.0008964295848272741,
      "learning_rate": 0.006360146812701528,
      "loss": 0.3466,
      "num_input_tokens_seen": 68755168,
      "step": 36280
    },
    {
      "epoch": 5.919324577861163,
      "grad_norm": 0.0011022827820852399,
      "learning_rate": 0.006343187141824125,
      "loss": 0.3178,
      "num_input_tokens_seen": 68764320,
      "step": 36285
    },
    {
      "epoch": 5.920140305081981,
      "grad_norm": 0.0015868674963712692,
      "learning_rate": 0.00632624962463259,
      "loss": 0.2965,
      "num_input_tokens_seen": 68773024,
      "step": 36290
    },
    {
      "epoch": 5.920956032302798,
      "grad_norm": 0.0007011570851318538,
      "learning_rate": 0.006309334263738853,
      "loss": 0.3291,
      "num_input_tokens_seen": 68782912,
      "step": 36295
    },
    {
      "epoch": 5.921771759523615,
      "grad_norm": 0.0006145270308479667,
      "learning_rate": 0.006292441061751508,
      "loss": 0.2518,
      "num_input_tokens_seen": 68793440,
      "step": 36300
    },
    {
      "epoch": 5.922587486744432,
      "grad_norm": 0.0013765129260718822,
      "learning_rate": 0.0062755700212757054,
      "loss": 0.3432,
      "num_input_tokens_seen": 68802224,
      "step": 36305
    },
    {
      "epoch": 5.92340321396525,
      "grad_norm": 0.0012349217431619763,
      "learning_rate": 0.006258721144913148,
      "loss": 0.3136,
      "num_input_tokens_seen": 68812112,
      "step": 36310
    },
    {
      "epoch": 5.9242189411860675,
      "grad_norm": 0.005862027872353792,
      "learning_rate": 0.0062418944352621575,
      "loss": 0.3652,
      "num_input_tokens_seen": 68822128,
      "step": 36315
    },
    {
      "epoch": 5.925034668406885,
      "grad_norm": 0.004182231612503529,
      "learning_rate": 0.0062250898949176405,
      "loss": 0.345,
      "num_input_tokens_seen": 68831936,
      "step": 36320
    },
    {
      "epoch": 5.925850395627702,
      "grad_norm": 0.001818452263250947,
      "learning_rate": 0.006208307526471041,
      "loss": 0.3244,
      "num_input_tokens_seen": 68841872,
      "step": 36325
    },
    {
      "epoch": 5.92666612284852,
      "grad_norm": 0.004710416775196791,
      "learning_rate": 0.006191547332510405,
      "loss": 0.3823,
      "num_input_tokens_seen": 68850256,
      "step": 36330
    },
    {
      "epoch": 5.927481850069337,
      "grad_norm": 0.000925011991057545,
      "learning_rate": 0.006174809315620416,
      "loss": 0.3607,
      "num_input_tokens_seen": 68860064,
      "step": 36335
    },
    {
      "epoch": 5.928297577290154,
      "grad_norm": 0.0009852285729721189,
      "learning_rate": 0.00615809347838221,
      "loss": 0.2936,
      "num_input_tokens_seen": 68870016,
      "step": 36340
    },
    {
      "epoch": 5.929113304510971,
      "grad_norm": 0.004505709744989872,
      "learning_rate": 0.006141399823373655,
      "loss": 0.3483,
      "num_input_tokens_seen": 68879248,
      "step": 36345
    },
    {
      "epoch": 5.929929031731789,
      "grad_norm": 0.0014546765014529228,
      "learning_rate": 0.0061247283531690455,
      "loss": 0.2997,
      "num_input_tokens_seen": 68887696,
      "step": 36350
    },
    {
      "epoch": 5.930744758952606,
      "grad_norm": 0.0036848646122962236,
      "learning_rate": 0.0061080790703393895,
      "loss": 0.335,
      "num_input_tokens_seen": 68897712,
      "step": 36355
    },
    {
      "epoch": 5.9315604861734235,
      "grad_norm": 0.0012449219357222319,
      "learning_rate": 0.006091451977452217,
      "loss": 0.3196,
      "num_input_tokens_seen": 68907344,
      "step": 36360
    },
    {
      "epoch": 5.932376213394241,
      "grad_norm": 0.003135907929390669,
      "learning_rate": 0.00607484707707161,
      "loss": 0.3498,
      "num_input_tokens_seen": 68916624,
      "step": 36365
    },
    {
      "epoch": 5.933191940615059,
      "grad_norm": 0.002936444478109479,
      "learning_rate": 0.006058264371758254,
      "loss": 0.3136,
      "num_input_tokens_seen": 68925376,
      "step": 36370
    },
    {
      "epoch": 5.934007667835876,
      "grad_norm": 0.002270349068567157,
      "learning_rate": 0.00604170386406942,
      "loss": 0.3,
      "num_input_tokens_seen": 68933392,
      "step": 36375
    },
    {
      "epoch": 5.934823395056693,
      "grad_norm": 0.0014991792850196362,
      "learning_rate": 0.006025165556558931,
      "loss": 0.3502,
      "num_input_tokens_seen": 68942576,
      "step": 36380
    },
    {
      "epoch": 5.935639122277511,
      "grad_norm": 0.0011488632299005985,
      "learning_rate": 0.006008649451777248,
      "loss": 0.281,
      "num_input_tokens_seen": 68951008,
      "step": 36385
    },
    {
      "epoch": 5.936454849498328,
      "grad_norm": 0.0010755283292382956,
      "learning_rate": 0.005992155552271283,
      "loss": 0.3677,
      "num_input_tokens_seen": 68961248,
      "step": 36390
    },
    {
      "epoch": 5.937270576719145,
      "grad_norm": 0.0026128070894628763,
      "learning_rate": 0.005975683860584685,
      "loss": 0.3111,
      "num_input_tokens_seen": 68970944,
      "step": 36395
    },
    {
      "epoch": 5.938086303939962,
      "grad_norm": 0.00281304819509387,
      "learning_rate": 0.0059592343792575385,
      "loss": 0.3256,
      "num_input_tokens_seen": 68980176,
      "step": 36400
    },
    {
      "epoch": 5.938086303939962,
      "eval_loss": 0.32347872853279114,
      "eval_runtime": 152.7802,
      "eval_samples_per_second": 17.836,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 68980176,
      "step": 36400
    },
    {
      "epoch": 5.938902031160779,
      "grad_norm": 0.0006328717572614551,
      "learning_rate": 0.0059428071108265975,
      "loss": 0.312,
      "num_input_tokens_seen": 68990016,
      "step": 36405
    },
    {
      "epoch": 5.939717758381597,
      "grad_norm": 0.0009481408051215112,
      "learning_rate": 0.005926402057825136,
      "loss": 0.2804,
      "num_input_tokens_seen": 68998784,
      "step": 36410
    },
    {
      "epoch": 5.940533485602415,
      "grad_norm": 0.001857440103776753,
      "learning_rate": 0.005910019222782997,
      "loss": 0.2598,
      "num_input_tokens_seen": 69008000,
      "step": 36415
    },
    {
      "epoch": 5.941349212823232,
      "grad_norm": 0.0018149014795199037,
      "learning_rate": 0.005893658608226643,
      "loss": 0.3479,
      "num_input_tokens_seen": 69017152,
      "step": 36420
    },
    {
      "epoch": 5.94216494004405,
      "grad_norm": 0.002259655389934778,
      "learning_rate": 0.0058773202166791045,
      "loss": 0.3564,
      "num_input_tokens_seen": 69027408,
      "step": 36425
    },
    {
      "epoch": 5.942980667264867,
      "grad_norm": 0.002813684055581689,
      "learning_rate": 0.005861004050659918,
      "loss": 0.3657,
      "num_input_tokens_seen": 69034384,
      "step": 36430
    },
    {
      "epoch": 5.943796394485684,
      "grad_norm": 0.0005629782099276781,
      "learning_rate": 0.005844710112685286,
      "loss": 0.3072,
      "num_input_tokens_seen": 69042576,
      "step": 36435
    },
    {
      "epoch": 5.944612121706501,
      "grad_norm": 0.0015015410026535392,
      "learning_rate": 0.005828438405267933,
      "loss": 0.2866,
      "num_input_tokens_seen": 69051616,
      "step": 36440
    },
    {
      "epoch": 5.945427848927318,
      "grad_norm": 0.003826540894806385,
      "learning_rate": 0.00581218893091715,
      "loss": 0.3369,
      "num_input_tokens_seen": 69061824,
      "step": 36445
    },
    {
      "epoch": 5.946243576148136,
      "grad_norm": 0.0013950321590527892,
      "learning_rate": 0.005795961692138801,
      "loss": 0.27,
      "num_input_tokens_seen": 69072128,
      "step": 36450
    },
    {
      "epoch": 5.947059303368953,
      "grad_norm": 0.001138050458393991,
      "learning_rate": 0.00577975669143535,
      "loss": 0.2941,
      "num_input_tokens_seen": 69080464,
      "step": 36455
    },
    {
      "epoch": 5.9478750305897705,
      "grad_norm": 0.0014399708015844226,
      "learning_rate": 0.005763573931305782,
      "loss": 0.2883,
      "num_input_tokens_seen": 69089936,
      "step": 36460
    },
    {
      "epoch": 5.9486907578105885,
      "grad_norm": 0.001723695662803948,
      "learning_rate": 0.005747413414245733,
      "loss": 0.3147,
      "num_input_tokens_seen": 69098896,
      "step": 36465
    },
    {
      "epoch": 5.949506485031406,
      "grad_norm": 0.0008164686732925475,
      "learning_rate": 0.005731275142747294,
      "loss": 0.2751,
      "num_input_tokens_seen": 69107584,
      "step": 36470
    },
    {
      "epoch": 5.950322212252223,
      "grad_norm": 0.0018885484896600246,
      "learning_rate": 0.005715159119299256,
      "loss": 0.2868,
      "num_input_tokens_seen": 69117136,
      "step": 36475
    },
    {
      "epoch": 5.95113793947304,
      "grad_norm": 0.0007102516246959567,
      "learning_rate": 0.005699065346386867,
      "loss": 0.3135,
      "num_input_tokens_seen": 69126224,
      "step": 36480
    },
    {
      "epoch": 5.951953666693858,
      "grad_norm": 0.001269806525669992,
      "learning_rate": 0.0056829938264919885,
      "loss": 0.3253,
      "num_input_tokens_seen": 69135984,
      "step": 36485
    },
    {
      "epoch": 5.952769393914675,
      "grad_norm": 0.0019651842303574085,
      "learning_rate": 0.005666944562093074,
      "loss": 0.3219,
      "num_input_tokens_seen": 69145712,
      "step": 36490
    },
    {
      "epoch": 5.953585121135492,
      "grad_norm": 0.0047486028634011745,
      "learning_rate": 0.005650917555665108,
      "loss": 0.3397,
      "num_input_tokens_seen": 69154848,
      "step": 36495
    },
    {
      "epoch": 5.954400848356309,
      "grad_norm": 0.0027935958933085203,
      "learning_rate": 0.005634912809679632,
      "loss": 0.3003,
      "num_input_tokens_seen": 69163152,
      "step": 36500
    },
    {
      "epoch": 5.955216575577127,
      "grad_norm": 0.0022325555328279734,
      "learning_rate": 0.005618930326604854,
      "loss": 0.331,
      "num_input_tokens_seen": 69172592,
      "step": 36505
    },
    {
      "epoch": 5.9560323027979445,
      "grad_norm": 0.0014533469220623374,
      "learning_rate": 0.005602970108905386,
      "loss": 0.2774,
      "num_input_tokens_seen": 69181936,
      "step": 36510
    },
    {
      "epoch": 5.956848030018762,
      "grad_norm": 0.0018214336596429348,
      "learning_rate": 0.005587032159042543,
      "loss": 0.3105,
      "num_input_tokens_seen": 69190032,
      "step": 36515
    },
    {
      "epoch": 5.957663757239579,
      "grad_norm": 0.0014410392614081502,
      "learning_rate": 0.005571116479474158,
      "loss": 0.3318,
      "num_input_tokens_seen": 69200048,
      "step": 36520
    },
    {
      "epoch": 5.958479484460397,
      "grad_norm": 0.002865551970899105,
      "learning_rate": 0.005555223072654619,
      "loss": 0.3052,
      "num_input_tokens_seen": 69208144,
      "step": 36525
    },
    {
      "epoch": 5.959295211681214,
      "grad_norm": 0.0009320135577581823,
      "learning_rate": 0.005539351941034881,
      "loss": 0.2707,
      "num_input_tokens_seen": 69217840,
      "step": 36530
    },
    {
      "epoch": 5.960110938902031,
      "grad_norm": 0.000971559202298522,
      "learning_rate": 0.0055235030870624865,
      "loss": 0.2981,
      "num_input_tokens_seen": 69228544,
      "step": 36535
    },
    {
      "epoch": 5.960926666122848,
      "grad_norm": 0.004829777870327234,
      "learning_rate": 0.005507676513181514,
      "loss": 0.3638,
      "num_input_tokens_seen": 69238016,
      "step": 36540
    },
    {
      "epoch": 5.961742393343666,
      "grad_norm": 0.0006686008418910205,
      "learning_rate": 0.005491872221832628,
      "loss": 0.2538,
      "num_input_tokens_seen": 69247552,
      "step": 36545
    },
    {
      "epoch": 5.962558120564483,
      "grad_norm": 0.0019314502133056521,
      "learning_rate": 0.005476090215453061,
      "loss": 0.3312,
      "num_input_tokens_seen": 69255824,
      "step": 36550
    },
    {
      "epoch": 5.9633738477853,
      "grad_norm": 0.0013359369477257133,
      "learning_rate": 0.0054603304964765675,
      "loss": 0.2618,
      "num_input_tokens_seen": 69266032,
      "step": 36555
    },
    {
      "epoch": 5.964189575006118,
      "grad_norm": 0.005133813712745905,
      "learning_rate": 0.005444593067333519,
      "loss": 0.4145,
      "num_input_tokens_seen": 69274080,
      "step": 36560
    },
    {
      "epoch": 5.965005302226936,
      "grad_norm": 0.0017163446173071861,
      "learning_rate": 0.00542887793045081,
      "loss": 0.3276,
      "num_input_tokens_seen": 69282880,
      "step": 36565
    },
    {
      "epoch": 5.965821029447753,
      "grad_norm": 0.001535366871394217,
      "learning_rate": 0.005413185088251932,
      "loss": 0.2649,
      "num_input_tokens_seen": 69292832,
      "step": 36570
    },
    {
      "epoch": 5.96663675666857,
      "grad_norm": 0.0015097834402695298,
      "learning_rate": 0.005397514543156884,
      "loss": 0.3207,
      "num_input_tokens_seen": 69303024,
      "step": 36575
    },
    {
      "epoch": 5.967452483889387,
      "grad_norm": 0.0014438594225794077,
      "learning_rate": 0.0053818662975822825,
      "loss": 0.3373,
      "num_input_tokens_seen": 69312544,
      "step": 36580
    },
    {
      "epoch": 5.968268211110205,
      "grad_norm": 0.0058621070347726345,
      "learning_rate": 0.005366240353941315,
      "loss": 0.3442,
      "num_input_tokens_seen": 69321664,
      "step": 36585
    },
    {
      "epoch": 5.969083938331022,
      "grad_norm": 0.001319178263656795,
      "learning_rate": 0.005350636714643636,
      "loss": 0.3361,
      "num_input_tokens_seen": 69330640,
      "step": 36590
    },
    {
      "epoch": 5.969899665551839,
      "grad_norm": 0.002089734422042966,
      "learning_rate": 0.005335055382095555,
      "loss": 0.2752,
      "num_input_tokens_seen": 69340176,
      "step": 36595
    },
    {
      "epoch": 5.970715392772657,
      "grad_norm": 0.001048204256221652,
      "learning_rate": 0.005319496358699915,
      "loss": 0.3077,
      "num_input_tokens_seen": 69349984,
      "step": 36600
    },
    {
      "epoch": 5.970715392772657,
      "eval_loss": 0.3208942115306854,
      "eval_runtime": 152.7726,
      "eval_samples_per_second": 17.837,
      "eval_steps_per_second": 8.922,
      "num_input_tokens_seen": 69349984,
      "step": 36600
    },
    {
      "epoch": 5.971531119993474,
      "grad_norm": 0.002177451504394412,
      "learning_rate": 0.005303959646856099,
      "loss": 0.2811,
      "num_input_tokens_seen": 69359760,
      "step": 36605
    },
    {
      "epoch": 5.9723468472142915,
      "grad_norm": 0.0019664622377604246,
      "learning_rate": 0.005288445248960089,
      "loss": 0.33,
      "num_input_tokens_seen": 69368864,
      "step": 36610
    },
    {
      "epoch": 5.973162574435109,
      "grad_norm": 0.0014356713509187102,
      "learning_rate": 0.005272953167404354,
      "loss": 0.3026,
      "num_input_tokens_seen": 69378224,
      "step": 36615
    },
    {
      "epoch": 5.973978301655926,
      "grad_norm": 0.0019565189722925425,
      "learning_rate": 0.005257483404578017,
      "loss": 0.3866,
      "num_input_tokens_seen": 69387984,
      "step": 36620
    },
    {
      "epoch": 5.974794028876744,
      "grad_norm": 0.002530914731323719,
      "learning_rate": 0.0052420359628666865,
      "loss": 0.3778,
      "num_input_tokens_seen": 69396480,
      "step": 36625
    },
    {
      "epoch": 5.975609756097561,
      "grad_norm": 0.0009308635490015149,
      "learning_rate": 0.00522661084465254,
      "loss": 0.3331,
      "num_input_tokens_seen": 69406688,
      "step": 36630
    },
    {
      "epoch": 5.976425483318378,
      "grad_norm": 0.004673460032790899,
      "learning_rate": 0.005211208052314326,
      "loss": 0.3005,
      "num_input_tokens_seen": 69416896,
      "step": 36635
    },
    {
      "epoch": 5.977241210539196,
      "grad_norm": 0.0017568992916494608,
      "learning_rate": 0.005195827588227391,
      "loss": 0.2738,
      "num_input_tokens_seen": 69426400,
      "step": 36640
    },
    {
      "epoch": 5.978056937760013,
      "grad_norm": 0.005150570999830961,
      "learning_rate": 0.0051804694547635255,
      "loss": 0.3697,
      "num_input_tokens_seen": 69435776,
      "step": 36645
    },
    {
      "epoch": 5.97887266498083,
      "grad_norm": 0.002884986810386181,
      "learning_rate": 0.005165133654291232,
      "loss": 0.3254,
      "num_input_tokens_seen": 69445152,
      "step": 36650
    },
    {
      "epoch": 5.9796883922016475,
      "grad_norm": 0.002431033179163933,
      "learning_rate": 0.005149820189175402,
      "loss": 0.3496,
      "num_input_tokens_seen": 69455232,
      "step": 36655
    },
    {
      "epoch": 5.9805041194224655,
      "grad_norm": 0.0012531185057014227,
      "learning_rate": 0.005134529061777598,
      "loss": 0.333,
      "num_input_tokens_seen": 69465280,
      "step": 36660
    },
    {
      "epoch": 5.981319846643283,
      "grad_norm": 0.0017377377953380346,
      "learning_rate": 0.005119260274455933,
      "loss": 0.3481,
      "num_input_tokens_seen": 69474704,
      "step": 36665
    },
    {
      "epoch": 5.9821355738641,
      "grad_norm": 0.0024568038061261177,
      "learning_rate": 0.005104013829565007,
      "loss": 0.3207,
      "num_input_tokens_seen": 69482848,
      "step": 36670
    },
    {
      "epoch": 5.982951301084917,
      "grad_norm": 0.002359677106142044,
      "learning_rate": 0.005088789729456006,
      "loss": 0.3858,
      "num_input_tokens_seen": 69493360,
      "step": 36675
    },
    {
      "epoch": 5.983767028305735,
      "grad_norm": 0.0006322651170194149,
      "learning_rate": 0.005073587976476735,
      "loss": 0.2918,
      "num_input_tokens_seen": 69501872,
      "step": 36680
    },
    {
      "epoch": 5.984582755526552,
      "grad_norm": 0.001955850748345256,
      "learning_rate": 0.005058408572971418,
      "loss": 0.3484,
      "num_input_tokens_seen": 69511648,
      "step": 36685
    },
    {
      "epoch": 5.985398482747369,
      "grad_norm": 0.0014606390614062548,
      "learning_rate": 0.005043251521280983,
      "loss": 0.333,
      "num_input_tokens_seen": 69521456,
      "step": 36690
    },
    {
      "epoch": 5.986214209968186,
      "grad_norm": 0.002223406918346882,
      "learning_rate": 0.005028116823742795,
      "loss": 0.3841,
      "num_input_tokens_seen": 69531648,
      "step": 36695
    },
    {
      "epoch": 5.987029937189004,
      "grad_norm": 0.0018392113270238042,
      "learning_rate": 0.005013004482690819,
      "loss": 0.2922,
      "num_input_tokens_seen": 69541120,
      "step": 36700
    },
    {
      "epoch": 5.9878456644098215,
      "grad_norm": 0.0022938563488423824,
      "learning_rate": 0.0049979145004555746,
      "loss": 0.2824,
      "num_input_tokens_seen": 69551088,
      "step": 36705
    },
    {
      "epoch": 5.988661391630639,
      "grad_norm": 0.00196544686332345,
      "learning_rate": 0.004982846879364116,
      "loss": 0.3274,
      "num_input_tokens_seen": 69560784,
      "step": 36710
    },
    {
      "epoch": 5.989477118851456,
      "grad_norm": 0.0015832973876968026,
      "learning_rate": 0.0049678016217400535,
      "loss": 0.2904,
      "num_input_tokens_seen": 69569664,
      "step": 36715
    },
    {
      "epoch": 5.990292846072274,
      "grad_norm": 0.0017534239450469613,
      "learning_rate": 0.004952778729903595,
      "loss": 0.3612,
      "num_input_tokens_seen": 69578224,
      "step": 36720
    },
    {
      "epoch": 5.991108573293091,
      "grad_norm": 0.0011364768724888563,
      "learning_rate": 0.004937778206171422,
      "loss": 0.3067,
      "num_input_tokens_seen": 69587696,
      "step": 36725
    },
    {
      "epoch": 5.991924300513908,
      "grad_norm": 0.0016240488039329648,
      "learning_rate": 0.004922800052856835,
      "loss": 0.3619,
      "num_input_tokens_seen": 69597904,
      "step": 36730
    },
    {
      "epoch": 5.992740027734725,
      "grad_norm": 0.001970442710444331,
      "learning_rate": 0.004907844272269602,
      "loss": 0.3145,
      "num_input_tokens_seen": 69607072,
      "step": 36735
    },
    {
      "epoch": 5.993555754955543,
      "grad_norm": 0.0014834405155852437,
      "learning_rate": 0.004892910866716144,
      "loss": 0.3029,
      "num_input_tokens_seen": 69617728,
      "step": 36740
    },
    {
      "epoch": 5.99437148217636,
      "grad_norm": 0.0011359237832948565,
      "learning_rate": 0.004877999838499369,
      "loss": 0.2762,
      "num_input_tokens_seen": 69626864,
      "step": 36745
    },
    {
      "epoch": 5.995187209397177,
      "grad_norm": 0.0016272232169285417,
      "learning_rate": 0.0048631111899187065,
      "loss": 0.3293,
      "num_input_tokens_seen": 69635024,
      "step": 36750
    },
    {
      "epoch": 5.9960029366179945,
      "grad_norm": 0.0009276380878873169,
      "learning_rate": 0.0048482449232702335,
      "loss": 0.313,
      "num_input_tokens_seen": 69644976,
      "step": 36755
    },
    {
      "epoch": 5.996818663838813,
      "grad_norm": 0.0027066892944276333,
      "learning_rate": 0.004833401040846469,
      "loss": 0.2798,
      "num_input_tokens_seen": 69654608,
      "step": 36760
    },
    {
      "epoch": 5.99763439105963,
      "grad_norm": 0.001406483119353652,
      "learning_rate": 0.004818579544936546,
      "loss": 0.3166,
      "num_input_tokens_seen": 69663904,
      "step": 36765
    },
    {
      "epoch": 5.998450118280447,
      "grad_norm": 0.0009534774580970407,
      "learning_rate": 0.004803780437826121,
      "loss": 0.277,
      "num_input_tokens_seen": 69672608,
      "step": 36770
    },
    {
      "epoch": 5.999265845501265,
      "grad_norm": 0.0031396180856972933,
      "learning_rate": 0.004789003721797402,
      "loss": 0.3788,
      "num_input_tokens_seen": 69681056,
      "step": 36775
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.000848126714117825,
      "learning_rate": 0.004774249399129132,
      "loss": 0.3106,
      "num_input_tokens_seen": 69689456,
      "step": 36780
    },
    {
      "epoch": 6.000815727220817,
      "grad_norm": 0.0013479781337082386,
      "learning_rate": 0.004759517472096642,
      "loss": 0.2721,
      "num_input_tokens_seen": 69699216,
      "step": 36785
    },
    {
      "epoch": 6.001631454441635,
      "grad_norm": 0.002030003583058715,
      "learning_rate": 0.004744807942971746,
      "loss": 0.342,
      "num_input_tokens_seen": 69709744,
      "step": 36790
    },
    {
      "epoch": 6.002447181662452,
      "grad_norm": 0.0019435803405940533,
      "learning_rate": 0.004730120814022881,
      "loss": 0.287,
      "num_input_tokens_seen": 69720256,
      "step": 36795
    },
    {
      "epoch": 6.003262908883269,
      "grad_norm": 0.001328380312770605,
      "learning_rate": 0.004715456087514935,
      "loss": 0.2706,
      "num_input_tokens_seen": 69729984,
      "step": 36800
    },
    {
      "epoch": 6.003262908883269,
      "eval_loss": 0.3216017186641693,
      "eval_runtime": 152.7864,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 69729984,
      "step": 36800
    },
    {
      "epoch": 6.0040786361040865,
      "grad_norm": 0.0014937480445951223,
      "learning_rate": 0.004700813765709432,
      "loss": 0.2734,
      "num_input_tokens_seen": 69739184,
      "step": 36805
    },
    {
      "epoch": 6.004894363324905,
      "grad_norm": 0.0017362643266096711,
      "learning_rate": 0.004686193850864401,
      "loss": 0.3149,
      "num_input_tokens_seen": 69748496,
      "step": 36810
    },
    {
      "epoch": 6.005710090545722,
      "grad_norm": 0.0011455676285549998,
      "learning_rate": 0.004671596345234385,
      "loss": 0.374,
      "num_input_tokens_seen": 69757344,
      "step": 36815
    },
    {
      "epoch": 6.006525817766539,
      "grad_norm": 0.0038479852955788374,
      "learning_rate": 0.00465702125107052,
      "loss": 0.2791,
      "num_input_tokens_seen": 69765760,
      "step": 36820
    },
    {
      "epoch": 6.007341544987356,
      "grad_norm": 0.001752715092152357,
      "learning_rate": 0.004642468570620506,
      "loss": 0.3519,
      "num_input_tokens_seen": 69776000,
      "step": 36825
    },
    {
      "epoch": 6.008157272208174,
      "grad_norm": 0.0021301968954503536,
      "learning_rate": 0.004627938306128482,
      "loss": 0.2976,
      "num_input_tokens_seen": 69784688,
      "step": 36830
    },
    {
      "epoch": 6.008972999428991,
      "grad_norm": 0.0027003344148397446,
      "learning_rate": 0.004613430459835255,
      "loss": 0.3482,
      "num_input_tokens_seen": 69792656,
      "step": 36835
    },
    {
      "epoch": 6.009788726649808,
      "grad_norm": 0.0012463605962693691,
      "learning_rate": 0.004598945033978085,
      "loss": 0.3576,
      "num_input_tokens_seen": 69802544,
      "step": 36840
    },
    {
      "epoch": 6.010604453870625,
      "grad_norm": 0.002561811124905944,
      "learning_rate": 0.004584482030790804,
      "loss": 0.3937,
      "num_input_tokens_seen": 69811200,
      "step": 36845
    },
    {
      "epoch": 6.011420181091443,
      "grad_norm": 0.0010823049815371633,
      "learning_rate": 0.004570041452503826,
      "loss": 0.337,
      "num_input_tokens_seen": 69821088,
      "step": 36850
    },
    {
      "epoch": 6.0122359083122605,
      "grad_norm": 0.001182981999590993,
      "learning_rate": 0.004555623301344003,
      "loss": 0.3441,
      "num_input_tokens_seen": 69830768,
      "step": 36855
    },
    {
      "epoch": 6.013051635533078,
      "grad_norm": 0.002419973723590374,
      "learning_rate": 0.004541227579534857,
      "loss": 0.3162,
      "num_input_tokens_seen": 69840864,
      "step": 36860
    },
    {
      "epoch": 6.013867362753895,
      "grad_norm": 0.0007844761130400002,
      "learning_rate": 0.004526854289296378,
      "loss": 0.3253,
      "num_input_tokens_seen": 69851008,
      "step": 36865
    },
    {
      "epoch": 6.014683089974713,
      "grad_norm": 0.002694777213037014,
      "learning_rate": 0.004512503432845078,
      "loss": 0.3558,
      "num_input_tokens_seen": 69861984,
      "step": 36870
    },
    {
      "epoch": 6.01549881719553,
      "grad_norm": 0.0019089477136731148,
      "learning_rate": 0.004498175012394068,
      "loss": 0.3332,
      "num_input_tokens_seen": 69872320,
      "step": 36875
    },
    {
      "epoch": 6.016314544416347,
      "grad_norm": 0.002124573104083538,
      "learning_rate": 0.004483869030152965,
      "loss": 0.2967,
      "num_input_tokens_seen": 69882624,
      "step": 36880
    },
    {
      "epoch": 6.017130271637164,
      "grad_norm": 0.0019857166334986687,
      "learning_rate": 0.004469585488327904,
      "loss": 0.2846,
      "num_input_tokens_seen": 69893488,
      "step": 36885
    },
    {
      "epoch": 6.017945998857982,
      "grad_norm": 0.0020216871052980423,
      "learning_rate": 0.0044553243891216395,
      "loss": 0.3475,
      "num_input_tokens_seen": 69901280,
      "step": 36890
    },
    {
      "epoch": 6.018761726078799,
      "grad_norm": 0.0005139761487953365,
      "learning_rate": 0.004441085734733363,
      "loss": 0.3293,
      "num_input_tokens_seen": 69910544,
      "step": 36895
    },
    {
      "epoch": 6.0195774532996165,
      "grad_norm": 0.003966699354350567,
      "learning_rate": 0.004426869527358884,
      "loss": 0.3132,
      "num_input_tokens_seen": 69919664,
      "step": 36900
    },
    {
      "epoch": 6.020393180520434,
      "grad_norm": 0.0015177637105807662,
      "learning_rate": 0.0044126757691905156,
      "loss": 0.2513,
      "num_input_tokens_seen": 69928256,
      "step": 36905
    },
    {
      "epoch": 6.021208907741252,
      "grad_norm": 0.003837206866592169,
      "learning_rate": 0.004398504462417107,
      "loss": 0.3117,
      "num_input_tokens_seen": 69937488,
      "step": 36910
    },
    {
      "epoch": 6.022024634962069,
      "grad_norm": 0.002263112226501107,
      "learning_rate": 0.0043843556092240605,
      "loss": 0.2794,
      "num_input_tokens_seen": 69945200,
      "step": 36915
    },
    {
      "epoch": 6.022840362182886,
      "grad_norm": 0.0010880599729716778,
      "learning_rate": 0.004370229211793281,
      "loss": 0.3317,
      "num_input_tokens_seen": 69953984,
      "step": 36920
    },
    {
      "epoch": 6.023656089403703,
      "grad_norm": 0.004655789118260145,
      "learning_rate": 0.0043561252723032405,
      "loss": 0.3674,
      "num_input_tokens_seen": 69963984,
      "step": 36925
    },
    {
      "epoch": 6.024471816624521,
      "grad_norm": 0.001689214026555419,
      "learning_rate": 0.004342043792929001,
      "loss": 0.2761,
      "num_input_tokens_seen": 69973200,
      "step": 36930
    },
    {
      "epoch": 6.025287543845338,
      "grad_norm": 0.0011300289770588279,
      "learning_rate": 0.004327984775842025,
      "loss": 0.3059,
      "num_input_tokens_seen": 69983664,
      "step": 36935
    },
    {
      "epoch": 6.026103271066155,
      "grad_norm": 0.0013052672147750854,
      "learning_rate": 0.004313948223210428,
      "loss": 0.3222,
      "num_input_tokens_seen": 69992560,
      "step": 36940
    },
    {
      "epoch": 6.026918998286972,
      "grad_norm": 0.0014881177339702845,
      "learning_rate": 0.004299934137198846,
      "loss": 0.3438,
      "num_input_tokens_seen": 70002528,
      "step": 36945
    },
    {
      "epoch": 6.02773472550779,
      "grad_norm": 0.001071102567948401,
      "learning_rate": 0.004285942519968383,
      "loss": 0.331,
      "num_input_tokens_seen": 70011680,
      "step": 36950
    },
    {
      "epoch": 6.028550452728608,
      "grad_norm": 0.000614615622907877,
      "learning_rate": 0.004271973373676746,
      "loss": 0.318,
      "num_input_tokens_seen": 70021104,
      "step": 36955
    },
    {
      "epoch": 6.029366179949425,
      "grad_norm": 0.0024823364801704884,
      "learning_rate": 0.004258026700478146,
      "loss": 0.346,
      "num_input_tokens_seen": 70031088,
      "step": 36960
    },
    {
      "epoch": 6.030181907170242,
      "grad_norm": 0.0028381282463669777,
      "learning_rate": 0.004244102502523328,
      "loss": 0.3257,
      "num_input_tokens_seen": 70041376,
      "step": 36965
    },
    {
      "epoch": 6.03099763439106,
      "grad_norm": 0.002437689108774066,
      "learning_rate": 0.004230200781959592,
      "loss": 0.3058,
      "num_input_tokens_seen": 70049808,
      "step": 36970
    },
    {
      "epoch": 6.031813361611877,
      "grad_norm": 0.0008033307385630906,
      "learning_rate": 0.004216321540930756,
      "loss": 0.2354,
      "num_input_tokens_seen": 70059968,
      "step": 36975
    },
    {
      "epoch": 6.032629088832694,
      "grad_norm": 0.0020870056468993425,
      "learning_rate": 0.004202464781577175,
      "loss": 0.3388,
      "num_input_tokens_seen": 70068848,
      "step": 36980
    },
    {
      "epoch": 6.033444816053512,
      "grad_norm": 0.0027867956086993217,
      "learning_rate": 0.00418863050603574,
      "loss": 0.3702,
      "num_input_tokens_seen": 70079216,
      "step": 36985
    },
    {
      "epoch": 6.034260543274329,
      "grad_norm": 0.0017229137010872364,
      "learning_rate": 0.004174818716439843,
      "loss": 0.2835,
      "num_input_tokens_seen": 70088208,
      "step": 36990
    },
    {
      "epoch": 6.035076270495146,
      "grad_norm": 0.0008188274223357439,
      "learning_rate": 0.004161029414919464,
      "loss": 0.3175,
      "num_input_tokens_seen": 70098480,
      "step": 36995
    },
    {
      "epoch": 6.0358919977159635,
      "grad_norm": 0.0029379609040915966,
      "learning_rate": 0.004147262603601071,
      "loss": 0.3618,
      "num_input_tokens_seen": 70107936,
      "step": 37000
    },
    {
      "epoch": 6.0358919977159635,
      "eval_loss": 0.3230544626712799,
      "eval_runtime": 152.8428,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.918,
      "num_input_tokens_seen": 70107936,
      "step": 37000
    },
    {
      "epoch": 6.0367077249367815,
      "grad_norm": 0.0014851799933239818,
      "learning_rate": 0.004133518284607679,
      "loss": 0.3826,
      "num_input_tokens_seen": 70115760,
      "step": 37005
    },
    {
      "epoch": 6.037523452157599,
      "grad_norm": 0.0020663924515247345,
      "learning_rate": 0.004119796460058861,
      "loss": 0.3069,
      "num_input_tokens_seen": 70124656,
      "step": 37010
    },
    {
      "epoch": 6.038339179378416,
      "grad_norm": 0.0024787066504359245,
      "learning_rate": 0.00410609713207064,
      "loss": 0.3341,
      "num_input_tokens_seen": 70134848,
      "step": 37015
    },
    {
      "epoch": 6.039154906599233,
      "grad_norm": 0.0011291233822703362,
      "learning_rate": 0.004092420302755678,
      "loss": 0.3061,
      "num_input_tokens_seen": 70144880,
      "step": 37020
    },
    {
      "epoch": 6.039970633820051,
      "grad_norm": 0.0023450125008821487,
      "learning_rate": 0.004078765974223103,
      "loss": 0.3069,
      "num_input_tokens_seen": 70154672,
      "step": 37025
    },
    {
      "epoch": 6.040786361040868,
      "grad_norm": 0.004601416178047657,
      "learning_rate": 0.004065134148578564,
      "loss": 0.3982,
      "num_input_tokens_seen": 70166000,
      "step": 37030
    },
    {
      "epoch": 6.041602088261685,
      "grad_norm": 0.0012020092690363526,
      "learning_rate": 0.004051524827924279,
      "loss": 0.3006,
      "num_input_tokens_seen": 70176096,
      "step": 37035
    },
    {
      "epoch": 6.042417815482502,
      "grad_norm": 0.0009067885694094002,
      "learning_rate": 0.004037938014358955,
      "loss": 0.3365,
      "num_input_tokens_seen": 70184656,
      "step": 37040
    },
    {
      "epoch": 6.04323354270332,
      "grad_norm": 0.0016814242117106915,
      "learning_rate": 0.004024373709977863,
      "loss": 0.2855,
      "num_input_tokens_seen": 70194112,
      "step": 37045
    },
    {
      "epoch": 6.0440492699241375,
      "grad_norm": 0.0016679654363542795,
      "learning_rate": 0.004010831916872814,
      "loss": 0.3391,
      "num_input_tokens_seen": 70204592,
      "step": 37050
    },
    {
      "epoch": 6.044864997144955,
      "grad_norm": 0.0016586249694228172,
      "learning_rate": 0.003997312637132089,
      "loss": 0.2935,
      "num_input_tokens_seen": 70213376,
      "step": 37055
    },
    {
      "epoch": 6.045680724365772,
      "grad_norm": 0.0021948337089270353,
      "learning_rate": 0.003983815872840535,
      "loss": 0.2925,
      "num_input_tokens_seen": 70224240,
      "step": 37060
    },
    {
      "epoch": 6.04649645158659,
      "grad_norm": 0.0013724714517593384,
      "learning_rate": 0.003970341626079521,
      "loss": 0.3655,
      "num_input_tokens_seen": 70234448,
      "step": 37065
    },
    {
      "epoch": 6.047312178807407,
      "grad_norm": 0.0036675818264484406,
      "learning_rate": 0.003956889898926952,
      "loss": 0.3454,
      "num_input_tokens_seen": 70243552,
      "step": 37070
    },
    {
      "epoch": 6.048127906028224,
      "grad_norm": 0.0023295492865145206,
      "learning_rate": 0.0039434606934572675,
      "loss": 0.4169,
      "num_input_tokens_seen": 70253392,
      "step": 37075
    },
    {
      "epoch": 6.048943633249041,
      "grad_norm": 0.0026207289192825556,
      "learning_rate": 0.003930054011741396,
      "loss": 0.3067,
      "num_input_tokens_seen": 70263248,
      "step": 37080
    },
    {
      "epoch": 6.049759360469859,
      "grad_norm": 0.0014123731525614858,
      "learning_rate": 0.0039166698558468155,
      "loss": 0.2977,
      "num_input_tokens_seen": 70273584,
      "step": 37085
    },
    {
      "epoch": 6.050575087690676,
      "grad_norm": 0.001430746866390109,
      "learning_rate": 0.0039033082278375594,
      "loss": 0.3311,
      "num_input_tokens_seen": 70283024,
      "step": 37090
    },
    {
      "epoch": 6.051390814911493,
      "grad_norm": 0.0021513928659260273,
      "learning_rate": 0.003889969129774112,
      "loss": 0.2904,
      "num_input_tokens_seen": 70293456,
      "step": 37095
    },
    {
      "epoch": 6.052206542132311,
      "grad_norm": 0.002103640465065837,
      "learning_rate": 0.0038766525637135784,
      "loss": 0.3342,
      "num_input_tokens_seen": 70301968,
      "step": 37100
    },
    {
      "epoch": 6.053022269353129,
      "grad_norm": 0.001540351309813559,
      "learning_rate": 0.0038633585317095318,
      "loss": 0.3451,
      "num_input_tokens_seen": 70310560,
      "step": 37105
    },
    {
      "epoch": 6.053837996573946,
      "grad_norm": 0.002049251925200224,
      "learning_rate": 0.00385008703581205,
      "loss": 0.3364,
      "num_input_tokens_seen": 70319696,
      "step": 37110
    },
    {
      "epoch": 6.054653723794763,
      "grad_norm": 0.004743368364870548,
      "learning_rate": 0.0038368380780677944,
      "loss": 0.3139,
      "num_input_tokens_seen": 70328592,
      "step": 37115
    },
    {
      "epoch": 6.05546945101558,
      "grad_norm": 0.0008423774852417409,
      "learning_rate": 0.003823611660519882,
      "loss": 0.3152,
      "num_input_tokens_seen": 70337712,
      "step": 37120
    },
    {
      "epoch": 6.056285178236398,
      "grad_norm": 0.0015996701549738646,
      "learning_rate": 0.0038104077852080475,
      "loss": 0.3096,
      "num_input_tokens_seen": 70347200,
      "step": 37125
    },
    {
      "epoch": 6.057100905457215,
      "grad_norm": 0.0005443843547254801,
      "learning_rate": 0.003797226454168462,
      "loss": 0.3388,
      "num_input_tokens_seen": 70356672,
      "step": 37130
    },
    {
      "epoch": 6.057916632678032,
      "grad_norm": 0.0010822676122188568,
      "learning_rate": 0.003784067669433849,
      "loss": 0.262,
      "num_input_tokens_seen": 70366192,
      "step": 37135
    },
    {
      "epoch": 6.058732359898849,
      "grad_norm": 0.001766352099366486,
      "learning_rate": 0.0037709314330334528,
      "loss": 0.3007,
      "num_input_tokens_seen": 70376832,
      "step": 37140
    },
    {
      "epoch": 6.059548087119667,
      "grad_norm": 0.0009362235432490706,
      "learning_rate": 0.003757817746993086,
      "loss": 0.25,
      "num_input_tokens_seen": 70387344,
      "step": 37145
    },
    {
      "epoch": 6.0603638143404845,
      "grad_norm": 0.002928632777184248,
      "learning_rate": 0.0037447266133349977,
      "loss": 0.3018,
      "num_input_tokens_seen": 70397280,
      "step": 37150
    },
    {
      "epoch": 6.061179541561302,
      "grad_norm": 0.0016282317228615284,
      "learning_rate": 0.003731658034078039,
      "loss": 0.3034,
      "num_input_tokens_seen": 70406352,
      "step": 37155
    },
    {
      "epoch": 6.06199526878212,
      "grad_norm": 0.0020361843053251505,
      "learning_rate": 0.0037186120112375153,
      "loss": 0.2549,
      "num_input_tokens_seen": 70416368,
      "step": 37160
    },
    {
      "epoch": 6.062810996002937,
      "grad_norm": 0.0008230025414377451,
      "learning_rate": 0.003705588546825317,
      "loss": 0.329,
      "num_input_tokens_seen": 70425984,
      "step": 37165
    },
    {
      "epoch": 6.063626723223754,
      "grad_norm": 0.0012843869626522064,
      "learning_rate": 0.0036925876428498205,
      "loss": 0.3015,
      "num_input_tokens_seen": 70435632,
      "step": 37170
    },
    {
      "epoch": 6.064442450444571,
      "grad_norm": 0.0021373655181378126,
      "learning_rate": 0.0036796093013159057,
      "loss": 0.3508,
      "num_input_tokens_seen": 70443952,
      "step": 37175
    },
    {
      "epoch": 6.065258177665389,
      "grad_norm": 0.002228453988209367,
      "learning_rate": 0.0036666535242250217,
      "loss": 0.3088,
      "num_input_tokens_seen": 70452640,
      "step": 37180
    },
    {
      "epoch": 6.066073904886206,
      "grad_norm": 0.0036982193123549223,
      "learning_rate": 0.003653720313575104,
      "loss": 0.2999,
      "num_input_tokens_seen": 70461648,
      "step": 37185
    },
    {
      "epoch": 6.066889632107023,
      "grad_norm": 0.0016251314664259553,
      "learning_rate": 0.003640809671360623,
      "loss": 0.318,
      "num_input_tokens_seen": 70469696,
      "step": 37190
    },
    {
      "epoch": 6.0677053593278405,
      "grad_norm": 0.000807520467787981,
      "learning_rate": 0.003627921599572553,
      "loss": 0.3021,
      "num_input_tokens_seen": 70478736,
      "step": 37195
    },
    {
      "epoch": 6.0685210865486585,
      "grad_norm": 0.0021313580218702555,
      "learning_rate": 0.003615056100198405,
      "loss": 0.3143,
      "num_input_tokens_seen": 70487856,
      "step": 37200
    },
    {
      "epoch": 6.0685210865486585,
      "eval_loss": 0.3211679756641388,
      "eval_runtime": 152.9029,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 70487856,
      "step": 37200
    },
    {
      "epoch": 6.069336813769476,
      "grad_norm": 0.0016425742069259286,
      "learning_rate": 0.003602213175222174,
      "loss": 0.3238,
      "num_input_tokens_seen": 70496848,
      "step": 37205
    },
    {
      "epoch": 6.070152540990293,
      "grad_norm": 0.0028308716136962175,
      "learning_rate": 0.0035893928266244432,
      "loss": 0.2843,
      "num_input_tokens_seen": 70506176,
      "step": 37210
    },
    {
      "epoch": 6.07096826821111,
      "grad_norm": 0.0018861532444134355,
      "learning_rate": 0.003576595056382248,
      "loss": 0.3603,
      "num_input_tokens_seen": 70515120,
      "step": 37215
    },
    {
      "epoch": 6.071783995431928,
      "grad_norm": 0.002998755080625415,
      "learning_rate": 0.0035638198664691423,
      "loss": 0.3326,
      "num_input_tokens_seen": 70525376,
      "step": 37220
    },
    {
      "epoch": 6.072599722652745,
      "grad_norm": 0.002986147766932845,
      "learning_rate": 0.003551067258855267,
      "loss": 0.306,
      "num_input_tokens_seen": 70533904,
      "step": 37225
    },
    {
      "epoch": 6.073415449873562,
      "grad_norm": 0.0028598960489034653,
      "learning_rate": 0.0035383372355071996,
      "loss": 0.3024,
      "num_input_tokens_seen": 70543472,
      "step": 37230
    },
    {
      "epoch": 6.074231177094379,
      "grad_norm": 0.002683512633666396,
      "learning_rate": 0.0035256297983881023,
      "loss": 0.3285,
      "num_input_tokens_seen": 70551856,
      "step": 37235
    },
    {
      "epoch": 6.075046904315197,
      "grad_norm": 0.001218859339132905,
      "learning_rate": 0.0035129449494575747,
      "loss": 0.3534,
      "num_input_tokens_seen": 70560384,
      "step": 37240
    },
    {
      "epoch": 6.0758626315360145,
      "grad_norm": 0.0010032746940851212,
      "learning_rate": 0.0035002826906718187,
      "loss": 0.3287,
      "num_input_tokens_seen": 70569488,
      "step": 37245
    },
    {
      "epoch": 6.076678358756832,
      "grad_norm": 0.0023290759418159723,
      "learning_rate": 0.003487643023983522,
      "loss": 0.3293,
      "num_input_tokens_seen": 70579328,
      "step": 37250
    },
    {
      "epoch": 6.077494085977649,
      "grad_norm": 0.0034720776602625847,
      "learning_rate": 0.003475025951341842,
      "loss": 0.3672,
      "num_input_tokens_seen": 70588576,
      "step": 37255
    },
    {
      "epoch": 6.078309813198467,
      "grad_norm": 0.002152068307623267,
      "learning_rate": 0.00346243147469249,
      "loss": 0.2394,
      "num_input_tokens_seen": 70598496,
      "step": 37260
    },
    {
      "epoch": 6.079125540419284,
      "grad_norm": 0.002617692109197378,
      "learning_rate": 0.0034498595959777446,
      "loss": 0.2824,
      "num_input_tokens_seen": 70607840,
      "step": 37265
    },
    {
      "epoch": 6.079941267640101,
      "grad_norm": 0.0024361396208405495,
      "learning_rate": 0.003437310317136305,
      "loss": 0.4293,
      "num_input_tokens_seen": 70618016,
      "step": 37270
    },
    {
      "epoch": 6.080756994860918,
      "grad_norm": 0.0016509495908394456,
      "learning_rate": 0.0034247836401034236,
      "loss": 0.3205,
      "num_input_tokens_seen": 70627600,
      "step": 37275
    },
    {
      "epoch": 6.081572722081736,
      "grad_norm": 0.0023082562256604433,
      "learning_rate": 0.003412279566810905,
      "loss": 0.3376,
      "num_input_tokens_seen": 70637072,
      "step": 37280
    },
    {
      "epoch": 6.082388449302553,
      "grad_norm": 0.0008401415543630719,
      "learning_rate": 0.00339979809918699,
      "loss": 0.3011,
      "num_input_tokens_seen": 70646640,
      "step": 37285
    },
    {
      "epoch": 6.08320417652337,
      "grad_norm": 0.0033978621941059828,
      "learning_rate": 0.0033873392391565228,
      "loss": 0.3447,
      "num_input_tokens_seen": 70656336,
      "step": 37290
    },
    {
      "epoch": 6.0840199037441876,
      "grad_norm": 0.001552489004097879,
      "learning_rate": 0.003374902988640782,
      "loss": 0.2918,
      "num_input_tokens_seen": 70664864,
      "step": 37295
    },
    {
      "epoch": 6.084835630965006,
      "grad_norm": 0.0018962833564728498,
      "learning_rate": 0.0033624893495576014,
      "loss": 0.2618,
      "num_input_tokens_seen": 70674768,
      "step": 37300
    },
    {
      "epoch": 6.085651358185823,
      "grad_norm": 0.001018001465126872,
      "learning_rate": 0.0033500983238213323,
      "loss": 0.3059,
      "num_input_tokens_seen": 70684032,
      "step": 37305
    },
    {
      "epoch": 6.08646708540664,
      "grad_norm": 0.0014714570716023445,
      "learning_rate": 0.0033377299133428126,
      "loss": 0.2958,
      "num_input_tokens_seen": 70694352,
      "step": 37310
    },
    {
      "epoch": 6.087282812627457,
      "grad_norm": 0.0024709179997444153,
      "learning_rate": 0.003325384120029434,
      "loss": 0.3103,
      "num_input_tokens_seen": 70703936,
      "step": 37315
    },
    {
      "epoch": 6.088098539848275,
      "grad_norm": 0.001273163128644228,
      "learning_rate": 0.0033130609457850233,
      "loss": 0.3704,
      "num_input_tokens_seen": 70714048,
      "step": 37320
    },
    {
      "epoch": 6.088914267069092,
      "grad_norm": 0.0014851948944851756,
      "learning_rate": 0.0033007603925100104,
      "loss": 0.2703,
      "num_input_tokens_seen": 70722736,
      "step": 37325
    },
    {
      "epoch": 6.089729994289909,
      "grad_norm": 0.002011277247220278,
      "learning_rate": 0.003288482462101294,
      "loss": 0.3445,
      "num_input_tokens_seen": 70732928,
      "step": 37330
    },
    {
      "epoch": 6.090545721510727,
      "grad_norm": 0.001421552151441574,
      "learning_rate": 0.0032762271564522605,
      "loss": 0.2889,
      "num_input_tokens_seen": 70741984,
      "step": 37335
    },
    {
      "epoch": 6.091361448731544,
      "grad_norm": 0.0021303179673850536,
      "learning_rate": 0.003263994477452864,
      "loss": 0.3223,
      "num_input_tokens_seen": 70750496,
      "step": 37340
    },
    {
      "epoch": 6.0921771759523615,
      "grad_norm": 0.002617036923766136,
      "learning_rate": 0.0032517844269895125,
      "loss": 0.3162,
      "num_input_tokens_seen": 70759744,
      "step": 37345
    },
    {
      "epoch": 6.092992903173179,
      "grad_norm": 0.0018384073628112674,
      "learning_rate": 0.0032395970069451496,
      "loss": 0.3109,
      "num_input_tokens_seen": 70769312,
      "step": 37350
    },
    {
      "epoch": 6.093808630393997,
      "grad_norm": 0.0030745542608201504,
      "learning_rate": 0.0032274322191992388,
      "loss": 0.3434,
      "num_input_tokens_seen": 70778720,
      "step": 37355
    },
    {
      "epoch": 6.094624357614814,
      "grad_norm": 0.003534164046868682,
      "learning_rate": 0.0032152900656277294,
      "loss": 0.3474,
      "num_input_tokens_seen": 70788480,
      "step": 37360
    },
    {
      "epoch": 6.095440084835631,
      "grad_norm": 0.0011740551562979817,
      "learning_rate": 0.0032031705481030902,
      "loss": 0.3322,
      "num_input_tokens_seen": 70798752,
      "step": 37365
    },
    {
      "epoch": 6.096255812056448,
      "grad_norm": 0.0018116015708073974,
      "learning_rate": 0.0031910736684943428,
      "loss": 0.2975,
      "num_input_tokens_seen": 70807600,
      "step": 37370
    },
    {
      "epoch": 6.097071539277266,
      "grad_norm": 0.0024199741892516613,
      "learning_rate": 0.0031789994286669453,
      "loss": 0.3339,
      "num_input_tokens_seen": 70817376,
      "step": 37375
    },
    {
      "epoch": 6.097887266498083,
      "grad_norm": 0.000747151963878423,
      "learning_rate": 0.003166947830482908,
      "loss": 0.3027,
      "num_input_tokens_seen": 70827856,
      "step": 37380
    },
    {
      "epoch": 6.0987029937189,
      "grad_norm": 0.0024868580512702465,
      "learning_rate": 0.003154918875800727,
      "loss": 0.2848,
      "num_input_tokens_seen": 70837056,
      "step": 37385
    },
    {
      "epoch": 6.0995187209397175,
      "grad_norm": 0.003146779490634799,
      "learning_rate": 0.00314291256647542,
      "loss": 0.3148,
      "num_input_tokens_seen": 70846992,
      "step": 37390
    },
    {
      "epoch": 6.1003344481605355,
      "grad_norm": 0.00107724464032799,
      "learning_rate": 0.0031309289043585375,
      "loss": 0.3799,
      "num_input_tokens_seen": 70856160,
      "step": 37395
    },
    {
      "epoch": 6.101150175381353,
      "grad_norm": 0.0036676847375929356,
      "learning_rate": 0.003118967891298069,
      "loss": 0.3113,
      "num_input_tokens_seen": 70865792,
      "step": 37400
    },
    {
      "epoch": 6.101150175381353,
      "eval_loss": 0.32034045457839966,
      "eval_runtime": 152.7909,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 8.921,
      "num_input_tokens_seen": 70865792,
      "step": 37400
    },
    {
      "epoch": 6.10196590260217,
      "grad_norm": 0.0018271435983479023,
      "learning_rate": 0.003107029529138572,
      "loss": 0.2997,
      "num_input_tokens_seen": 70874144,
      "step": 37405
    },
    {
      "epoch": 6.102781629822987,
      "grad_norm": 0.0008094010408967733,
      "learning_rate": 0.0030951138197211235,
      "loss": 0.3318,
      "num_input_tokens_seen": 70884144,
      "step": 37410
    },
    {
      "epoch": 6.103597357043805,
      "grad_norm": 0.0035508587025105953,
      "learning_rate": 0.0030832207648832377,
      "loss": 0.3189,
      "num_input_tokens_seen": 70894416,
      "step": 37415
    },
    {
      "epoch": 6.104413084264622,
      "grad_norm": 0.002675724448636174,
      "learning_rate": 0.0030713503664589635,
      "loss": 0.3426,
      "num_input_tokens_seen": 70903152,
      "step": 37420
    },
    {
      "epoch": 6.105228811485439,
      "grad_norm": 0.0038085635751485825,
      "learning_rate": 0.0030595026262788872,
      "loss": 0.2689,
      "num_input_tokens_seen": 70911888,
      "step": 37425
    },
    {
      "epoch": 6.106044538706256,
      "grad_norm": 0.003936633002012968,
      "learning_rate": 0.00304767754617008,
      "loss": 0.3559,
      "num_input_tokens_seen": 70920544,
      "step": 37430
    },
    {
      "epoch": 6.106860265927074,
      "grad_norm": 0.0033885736484080553,
      "learning_rate": 0.003035875127956117,
      "loss": 0.3272,
      "num_input_tokens_seen": 70930064,
      "step": 37435
    },
    {
      "epoch": 6.1076759931478914,
      "grad_norm": 0.0016600145027041435,
      "learning_rate": 0.0030240953734570752,
      "loss": 0.3162,
      "num_input_tokens_seen": 70939808,
      "step": 37440
    },
    {
      "epoch": 6.108491720368709,
      "grad_norm": 0.003830394707620144,
      "learning_rate": 0.003012338284489535,
      "loss": 0.3549,
      "num_input_tokens_seen": 70948032,
      "step": 37445
    },
    {
      "epoch": 6.109307447589526,
      "grad_norm": 0.004239397589117289,
      "learning_rate": 0.0030006038628665964,
      "loss": 0.2787,
      "num_input_tokens_seen": 70956560,
      "step": 37450
    },
    {
      "epoch": 6.110123174810344,
      "grad_norm": 0.0016355464467778802,
      "learning_rate": 0.002988892110397845,
      "loss": 0.3485,
      "num_input_tokens_seen": 70966064,
      "step": 37455
    },
    {
      "epoch": 6.110938902031161,
      "grad_norm": 0.0020542657002806664,
      "learning_rate": 0.0029772030288894025,
      "loss": 0.398,
      "num_input_tokens_seen": 70976016,
      "step": 37460
    },
    {
      "epoch": 6.111754629251978,
      "grad_norm": 0.0017438719514757395,
      "learning_rate": 0.0029655366201438438,
      "loss": 0.3081,
      "num_input_tokens_seen": 70985280,
      "step": 37465
    },
    {
      "epoch": 6.112570356472795,
      "grad_norm": 0.0038316601421684027,
      "learning_rate": 0.0029538928859602965,
      "loss": 0.264,
      "num_input_tokens_seen": 70993616,
      "step": 37470
    },
    {
      "epoch": 6.113386083693613,
      "grad_norm": 0.0019369609653949738,
      "learning_rate": 0.002942271828134374,
      "loss": 0.3773,
      "num_input_tokens_seen": 71002624,
      "step": 37475
    },
    {
      "epoch": 6.11420181091443,
      "grad_norm": 0.0024431210476905107,
      "learning_rate": 0.00293067344845816,
      "loss": 0.2841,
      "num_input_tokens_seen": 71012880,
      "step": 37480
    },
    {
      "epoch": 6.115017538135247,
      "grad_norm": 0.0020790458656847477,
      "learning_rate": 0.0029190977487202896,
      "loss": 0.3268,
      "num_input_tokens_seen": 71023536,
      "step": 37485
    },
    {
      "epoch": 6.1158332653560645,
      "grad_norm": 0.0015303997788578272,
      "learning_rate": 0.0029075447307058853,
      "loss": 0.2902,
      "num_input_tokens_seen": 71032992,
      "step": 37490
    },
    {
      "epoch": 6.1166489925768825,
      "grad_norm": 0.001539100194349885,
      "learning_rate": 0.0028960143961965722,
      "loss": 0.2945,
      "num_input_tokens_seen": 71042032,
      "step": 37495
    },
    {
      "epoch": 6.1174647197977,
      "grad_norm": 0.003174177836626768,
      "learning_rate": 0.002884506746970461,
      "loss": 0.314,
      "num_input_tokens_seen": 71051824,
      "step": 37500
    },
    {
      "epoch": 6.118280447018517,
      "grad_norm": 0.0016735766548663378,
      "learning_rate": 0.0028730217848021654,
      "loss": 0.3042,
      "num_input_tokens_seen": 71061328,
      "step": 37505
    },
    {
      "epoch": 6.119096174239334,
      "grad_norm": 0.001990838209167123,
      "learning_rate": 0.0028615595114628188,
      "loss": 0.3195,
      "num_input_tokens_seen": 71070880,
      "step": 37510
    },
    {
      "epoch": 6.119911901460152,
      "grad_norm": 0.0017612352967262268,
      "learning_rate": 0.002850119928720074,
      "loss": 0.2737,
      "num_input_tokens_seen": 71080736,
      "step": 37515
    },
    {
      "epoch": 6.120727628680969,
      "grad_norm": 0.0014011181192472577,
      "learning_rate": 0.0028387030383380195,
      "loss": 0.3412,
      "num_input_tokens_seen": 71089920,
      "step": 37520
    },
    {
      "epoch": 6.121543355901786,
      "grad_norm": 0.0009508139919489622,
      "learning_rate": 0.0028273088420772974,
      "loss": 0.3174,
      "num_input_tokens_seen": 71100368,
      "step": 37525
    },
    {
      "epoch": 6.122359083122603,
      "grad_norm": 0.0009068315848708153,
      "learning_rate": 0.002815937341695068,
      "loss": 0.3161,
      "num_input_tokens_seen": 71110288,
      "step": 37530
    },
    {
      "epoch": 6.123174810343421,
      "grad_norm": 0.0008728067623451352,
      "learning_rate": 0.0028045885389448963,
      "loss": 0.2969,
      "num_input_tokens_seen": 71120880,
      "step": 37535
    },
    {
      "epoch": 6.1239905375642385,
      "grad_norm": 0.0030727877747267485,
      "learning_rate": 0.002793262435576965,
      "loss": 0.3723,
      "num_input_tokens_seen": 71129744,
      "step": 37540
    },
    {
      "epoch": 6.124806264785056,
      "grad_norm": 0.0019146481063216925,
      "learning_rate": 0.0027819590333378772,
      "loss": 0.3044,
      "num_input_tokens_seen": 71139472,
      "step": 37545
    },
    {
      "epoch": 6.125621992005874,
      "grad_norm": 0.0010535692563280463,
      "learning_rate": 0.002770678333970755,
      "loss": 0.3285,
      "num_input_tokens_seen": 71148896,
      "step": 37550
    },
    {
      "epoch": 6.126437719226691,
      "grad_norm": 0.001542605459690094,
      "learning_rate": 0.0027594203392152573,
      "loss": 0.3255,
      "num_input_tokens_seen": 71158928,
      "step": 37555
    },
    {
      "epoch": 6.127253446447508,
      "grad_norm": 0.0005278187454678118,
      "learning_rate": 0.002748185050807478,
      "loss": 0.329,
      "num_input_tokens_seen": 71168240,
      "step": 37560
    },
    {
      "epoch": 6.128069173668325,
      "grad_norm": 0.005026157014071941,
      "learning_rate": 0.002736972470480031,
      "loss": 0.2288,
      "num_input_tokens_seen": 71178832,
      "step": 37565
    },
    {
      "epoch": 6.128884900889143,
      "grad_norm": 0.00131056341342628,
      "learning_rate": 0.002725782599962068,
      "loss": 0.3294,
      "num_input_tokens_seen": 71189600,
      "step": 37570
    },
    {
      "epoch": 6.12970062810996,
      "grad_norm": 0.0007943210657685995,
      "learning_rate": 0.0027146154409791734,
      "loss": 0.3208,
      "num_input_tokens_seen": 71198192,
      "step": 37575
    },
    {
      "epoch": 6.130516355330777,
      "grad_norm": 0.0008969755144789815,
      "learning_rate": 0.002703470995253504,
      "loss": 0.3746,
      "num_input_tokens_seen": 71207184,
      "step": 37580
    },
    {
      "epoch": 6.1313320825515945,
      "grad_norm": 0.0027566400822252035,
      "learning_rate": 0.0026923492645036184,
      "loss": 0.3377,
      "num_input_tokens_seen": 71216896,
      "step": 37585
    },
    {
      "epoch": 6.1321478097724125,
      "grad_norm": 0.0011300782207399607,
      "learning_rate": 0.0026812502504446776,
      "loss": 0.3219,
      "num_input_tokens_seen": 71226400,
      "step": 37590
    },
    {
      "epoch": 6.13296353699323,
      "grad_norm": 0.002311454154551029,
      "learning_rate": 0.0026701739547882798,
      "loss": 0.3869,
      "num_input_tokens_seen": 71235424,
      "step": 37595
    },
    {
      "epoch": 6.133779264214047,
      "grad_norm": 0.0010509100975468755,
      "learning_rate": 0.0026591203792425077,
      "loss": 0.2601,
      "num_input_tokens_seen": 71244784,
      "step": 37600
    },
    {
      "epoch": 6.133779264214047,
      "eval_loss": 0.3201003968715668,
      "eval_runtime": 152.8165,
      "eval_samples_per_second": 17.832,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 71244784,
      "step": 37600
    },
    {
      "epoch": 6.134594991434864,
      "grad_norm": 0.0022167256101965904,
      "learning_rate": 0.0026480895255119818,
      "loss": 0.3806,
      "num_input_tokens_seen": 71253712,
      "step": 37605
    },
    {
      "epoch": 6.135410718655682,
      "grad_norm": 0.0023937332443892956,
      "learning_rate": 0.002637081395297791,
      "loss": 0.3303,
      "num_input_tokens_seen": 71262624,
      "step": 37610
    },
    {
      "epoch": 6.136226445876499,
      "grad_norm": 0.0016066607786342502,
      "learning_rate": 0.0026260959902975113,
      "loss": 0.3345,
      "num_input_tokens_seen": 71272928,
      "step": 37615
    },
    {
      "epoch": 6.137042173097316,
      "grad_norm": 0.0020957100205123425,
      "learning_rate": 0.00261513331220527,
      "loss": 0.2913,
      "num_input_tokens_seen": 71281616,
      "step": 37620
    },
    {
      "epoch": 6.137857900318133,
      "grad_norm": 0.002604886656627059,
      "learning_rate": 0.0026041933627116154,
      "loss": 0.3008,
      "num_input_tokens_seen": 71291392,
      "step": 37625
    },
    {
      "epoch": 6.138673627538951,
      "grad_norm": 0.0010630403412505984,
      "learning_rate": 0.0025932761435036476,
      "loss": 0.2726,
      "num_input_tokens_seen": 71302160,
      "step": 37630
    },
    {
      "epoch": 6.139489354759768,
      "grad_norm": 0.0022588889114558697,
      "learning_rate": 0.002582381656264904,
      "loss": 0.3275,
      "num_input_tokens_seen": 71312640,
      "step": 37635
    },
    {
      "epoch": 6.1403050819805856,
      "grad_norm": 0.0011159846326336265,
      "learning_rate": 0.0025715099026754895,
      "loss": 0.3362,
      "num_input_tokens_seen": 71323424,
      "step": 37640
    },
    {
      "epoch": 6.141120809201403,
      "grad_norm": 0.0011831162264570594,
      "learning_rate": 0.002560660884411947,
      "loss": 0.332,
      "num_input_tokens_seen": 71332880,
      "step": 37645
    },
    {
      "epoch": 6.141936536422221,
      "grad_norm": 0.0004474413290154189,
      "learning_rate": 0.0025498346031473385,
      "loss": 0.3561,
      "num_input_tokens_seen": 71342080,
      "step": 37650
    },
    {
      "epoch": 6.142752263643038,
      "grad_norm": 0.0035099901724606752,
      "learning_rate": 0.0025390310605511945,
      "loss": 0.3483,
      "num_input_tokens_seen": 71352592,
      "step": 37655
    },
    {
      "epoch": 6.143567990863855,
      "grad_norm": 0.0019246669253334403,
      "learning_rate": 0.0025282502582895995,
      "loss": 0.2556,
      "num_input_tokens_seen": 71361664,
      "step": 37660
    },
    {
      "epoch": 6.144383718084672,
      "grad_norm": 0.003769670147448778,
      "learning_rate": 0.002517492198025023,
      "loss": 0.3413,
      "num_input_tokens_seen": 71370976,
      "step": 37665
    },
    {
      "epoch": 6.14519944530549,
      "grad_norm": 0.003307637758553028,
      "learning_rate": 0.0025067568814165554,
      "loss": 0.3635,
      "num_input_tokens_seen": 71381456,
      "step": 37670
    },
    {
      "epoch": 6.146015172526307,
      "grad_norm": 0.004180039744824171,
      "learning_rate": 0.0024960443101196884,
      "loss": 0.3153,
      "num_input_tokens_seen": 71391264,
      "step": 37675
    },
    {
      "epoch": 6.146830899747124,
      "grad_norm": 0.0009876969270408154,
      "learning_rate": 0.002485354485786434,
      "loss": 0.3242,
      "num_input_tokens_seen": 71400816,
      "step": 37680
    },
    {
      "epoch": 6.1476466269679415,
      "grad_norm": 0.0012113165576010942,
      "learning_rate": 0.002474687410065307,
      "loss": 0.3218,
      "num_input_tokens_seen": 71410816,
      "step": 37685
    },
    {
      "epoch": 6.1484623541887595,
      "grad_norm": 0.0023414466995745897,
      "learning_rate": 0.002464043084601308,
      "loss": 0.328,
      "num_input_tokens_seen": 71420416,
      "step": 37690
    },
    {
      "epoch": 6.149278081409577,
      "grad_norm": 0.001825594692490995,
      "learning_rate": 0.0024534215110358915,
      "loss": 0.3329,
      "num_input_tokens_seen": 71429376,
      "step": 37695
    },
    {
      "epoch": 6.150093808630394,
      "grad_norm": 0.0020274464040994644,
      "learning_rate": 0.002442822691007096,
      "loss": 0.3054,
      "num_input_tokens_seen": 71439584,
      "step": 37700
    },
    {
      "epoch": 6.150909535851211,
      "grad_norm": 0.001153371063992381,
      "learning_rate": 0.002432246626149348,
      "loss": 0.3129,
      "num_input_tokens_seen": 71448784,
      "step": 37705
    },
    {
      "epoch": 6.151725263072029,
      "grad_norm": 0.0011513970093801618,
      "learning_rate": 0.002421693318093626,
      "loss": 0.2863,
      "num_input_tokens_seen": 71458400,
      "step": 37710
    },
    {
      "epoch": 6.152540990292846,
      "grad_norm": 0.0011033518239855766,
      "learning_rate": 0.0024111627684673784,
      "loss": 0.3478,
      "num_input_tokens_seen": 71468352,
      "step": 37715
    },
    {
      "epoch": 6.153356717513663,
      "grad_norm": 0.0017762770876288414,
      "learning_rate": 0.0024006549788945395,
      "loss": 0.3576,
      "num_input_tokens_seen": 71478896,
      "step": 37720
    },
    {
      "epoch": 6.154172444734481,
      "grad_norm": 0.002389810513705015,
      "learning_rate": 0.0023901699509955463,
      "loss": 0.3072,
      "num_input_tokens_seen": 71488752,
      "step": 37725
    },
    {
      "epoch": 6.154988171955298,
      "grad_norm": 0.00047698101843707263,
      "learning_rate": 0.0023797076863873554,
      "loss": 0.3012,
      "num_input_tokens_seen": 71498112,
      "step": 37730
    },
    {
      "epoch": 6.1558038991761155,
      "grad_norm": 0.0018312250031158328,
      "learning_rate": 0.0023692681866833262,
      "loss": 0.295,
      "num_input_tokens_seen": 71507760,
      "step": 37735
    },
    {
      "epoch": 6.156619626396933,
      "grad_norm": 0.0015738030197098851,
      "learning_rate": 0.0023588514534934046,
      "loss": 0.2856,
      "num_input_tokens_seen": 71516560,
      "step": 37740
    },
    {
      "epoch": 6.157435353617751,
      "grad_norm": 0.0029347476083785295,
      "learning_rate": 0.002348457488423955,
      "loss": 0.2847,
      "num_input_tokens_seen": 71525984,
      "step": 37745
    },
    {
      "epoch": 6.158251080838568,
      "grad_norm": 0.0013244150904938579,
      "learning_rate": 0.0023380862930778624,
      "loss": 0.3172,
      "num_input_tokens_seen": 71536112,
      "step": 37750
    },
    {
      "epoch": 6.159066808059385,
      "grad_norm": 0.0016705758171156049,
      "learning_rate": 0.0023277378690545135,
      "loss": 0.3693,
      "num_input_tokens_seen": 71546112,
      "step": 37755
    },
    {
      "epoch": 6.159882535280202,
      "grad_norm": 0.0024621174670755863,
      "learning_rate": 0.0023174122179497325,
      "loss": 0.34,
      "num_input_tokens_seen": 71555040,
      "step": 37760
    },
    {
      "epoch": 6.16069826250102,
      "grad_norm": 0.00391779700294137,
      "learning_rate": 0.0023071093413558784,
      "loss": 0.3303,
      "num_input_tokens_seen": 71564560,
      "step": 37765
    },
    {
      "epoch": 6.161513989721837,
      "grad_norm": 0.00157329929061234,
      "learning_rate": 0.002296829240861814,
      "loss": 0.3069,
      "num_input_tokens_seen": 71574928,
      "step": 37770
    },
    {
      "epoch": 6.162329716942654,
      "grad_norm": 0.0015198574401438236,
      "learning_rate": 0.002286571918052821,
      "loss": 0.3288,
      "num_input_tokens_seen": 71583872,
      "step": 37775
    },
    {
      "epoch": 6.163145444163471,
      "grad_norm": 0.0010809338418766856,
      "learning_rate": 0.0022763373745107174,
      "loss": 0.2636,
      "num_input_tokens_seen": 71593440,
      "step": 37780
    },
    {
      "epoch": 6.1639611713842895,
      "grad_norm": 0.001885344972833991,
      "learning_rate": 0.0022661256118138074,
      "loss": 0.3054,
      "num_input_tokens_seen": 71603248,
      "step": 37785
    },
    {
      "epoch": 6.164776898605107,
      "grad_norm": 0.0019382197642698884,
      "learning_rate": 0.0022559366315368645,
      "loss": 0.3349,
      "num_input_tokens_seen": 71611904,
      "step": 37790
    },
    {
      "epoch": 6.165592625825924,
      "grad_norm": 0.0006074662669561803,
      "learning_rate": 0.002245770435251182,
      "loss": 0.2638,
      "num_input_tokens_seen": 71621584,
      "step": 37795
    },
    {
      "epoch": 6.166408353046741,
      "grad_norm": 0.002457108348608017,
      "learning_rate": 0.002235627024524456,
      "loss": 0.3489,
      "num_input_tokens_seen": 71630704,
      "step": 37800
    },
    {
      "epoch": 6.166408353046741,
      "eval_loss": 0.31930267810821533,
      "eval_runtime": 152.8227,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 71630704,
      "step": 37800
    },
    {
      "epoch": 6.167224080267559,
      "grad_norm": 0.0008410944719798863,
      "learning_rate": 0.0022255064009209847,
      "loss": 0.3359,
      "num_input_tokens_seen": 71640240,
      "step": 37805
    },
    {
      "epoch": 6.168039807488376,
      "grad_norm": 0.002890391508117318,
      "learning_rate": 0.0022154085660014864,
      "loss": 0.3277,
      "num_input_tokens_seen": 71649056,
      "step": 37810
    },
    {
      "epoch": 6.168855534709193,
      "grad_norm": 0.0016460586339235306,
      "learning_rate": 0.0022053335213231494,
      "loss": 0.3032,
      "num_input_tokens_seen": 71657552,
      "step": 37815
    },
    {
      "epoch": 6.16967126193001,
      "grad_norm": 0.0005356020992621779,
      "learning_rate": 0.002195281268439697,
      "loss": 0.3502,
      "num_input_tokens_seen": 71667888,
      "step": 37820
    },
    {
      "epoch": 6.170486989150828,
      "grad_norm": 0.0017607009503990412,
      "learning_rate": 0.002185251808901306,
      "loss": 0.3534,
      "num_input_tokens_seen": 71677456,
      "step": 37825
    },
    {
      "epoch": 6.171302716371645,
      "grad_norm": 0.0016239582328125834,
      "learning_rate": 0.0021752451442546227,
      "loss": 0.4012,
      "num_input_tokens_seen": 71687056,
      "step": 37830
    },
    {
      "epoch": 6.1721184435924625,
      "grad_norm": 0.0011997605906799436,
      "learning_rate": 0.0021652612760428456,
      "loss": 0.3184,
      "num_input_tokens_seen": 71696112,
      "step": 37835
    },
    {
      "epoch": 6.17293417081328,
      "grad_norm": 0.00533064641058445,
      "learning_rate": 0.0021553002058055603,
      "loss": 0.3032,
      "num_input_tokens_seen": 71705408,
      "step": 37840
    },
    {
      "epoch": 6.173749898034098,
      "grad_norm": 0.002131685148924589,
      "learning_rate": 0.0021453619350789376,
      "loss": 0.3711,
      "num_input_tokens_seen": 71714624,
      "step": 37845
    },
    {
      "epoch": 6.174565625254915,
      "grad_norm": 0.0009031324298121035,
      "learning_rate": 0.0021354464653955516,
      "loss": 0.3231,
      "num_input_tokens_seen": 71724384,
      "step": 37850
    },
    {
      "epoch": 6.175381352475732,
      "grad_norm": 0.0017614385578781366,
      "learning_rate": 0.002125553798284513,
      "loss": 0.3486,
      "num_input_tokens_seen": 71733328,
      "step": 37855
    },
    {
      "epoch": 6.176197079696549,
      "grad_norm": 0.0008782005170360208,
      "learning_rate": 0.002115683935271384,
      "loss": 0.2347,
      "num_input_tokens_seen": 71741856,
      "step": 37860
    },
    {
      "epoch": 6.177012806917367,
      "grad_norm": 0.0022890795953571796,
      "learning_rate": 0.0021058368778782144,
      "loss": 0.3069,
      "num_input_tokens_seen": 71751584,
      "step": 37865
    },
    {
      "epoch": 6.177828534138184,
      "grad_norm": 0.004456327762454748,
      "learning_rate": 0.002096012627623539,
      "loss": 0.3005,
      "num_input_tokens_seen": 71761104,
      "step": 37870
    },
    {
      "epoch": 6.178644261359001,
      "grad_norm": 0.002266980241984129,
      "learning_rate": 0.00208621118602243,
      "loss": 0.333,
      "num_input_tokens_seen": 71770880,
      "step": 37875
    },
    {
      "epoch": 6.1794599885798185,
      "grad_norm": 0.0037828730419278145,
      "learning_rate": 0.002076432554586327,
      "loss": 0.312,
      "num_input_tokens_seen": 71780384,
      "step": 37880
    },
    {
      "epoch": 6.1802757158006365,
      "grad_norm": 0.0011429786682128906,
      "learning_rate": 0.002066676734823258,
      "loss": 0.3028,
      "num_input_tokens_seen": 71790016,
      "step": 37885
    },
    {
      "epoch": 6.181091443021454,
      "grad_norm": 0.0019628724548965693,
      "learning_rate": 0.0020569437282376866,
      "loss": 0.3383,
      "num_input_tokens_seen": 71798704,
      "step": 37890
    },
    {
      "epoch": 6.181907170242271,
      "grad_norm": 0.001702384906820953,
      "learning_rate": 0.002047233536330545,
      "loss": 0.3009,
      "num_input_tokens_seen": 71808448,
      "step": 37895
    },
    {
      "epoch": 6.182722897463089,
      "grad_norm": 0.001356380758807063,
      "learning_rate": 0.0020375461605993015,
      "loss": 0.2713,
      "num_input_tokens_seen": 71818080,
      "step": 37900
    },
    {
      "epoch": 6.183538624683906,
      "grad_norm": 0.003148730145767331,
      "learning_rate": 0.002027881602537845,
      "loss": 0.2582,
      "num_input_tokens_seen": 71827504,
      "step": 37905
    },
    {
      "epoch": 6.184354351904723,
      "grad_norm": 0.002151461783796549,
      "learning_rate": 0.002018239863636567,
      "loss": 0.3311,
      "num_input_tokens_seen": 71837776,
      "step": 37910
    },
    {
      "epoch": 6.18517007912554,
      "grad_norm": 0.0018762820400297642,
      "learning_rate": 0.002008620945382378,
      "loss": 0.2852,
      "num_input_tokens_seen": 71846096,
      "step": 37915
    },
    {
      "epoch": 6.185985806346358,
      "grad_norm": 0.0023566456511616707,
      "learning_rate": 0.001999024849258607,
      "loss": 0.3425,
      "num_input_tokens_seen": 71855040,
      "step": 37920
    },
    {
      "epoch": 6.186801533567175,
      "grad_norm": 0.0017739776521921158,
      "learning_rate": 0.001989451576745105,
      "loss": 0.3569,
      "num_input_tokens_seen": 71864336,
      "step": 37925
    },
    {
      "epoch": 6.1876172607879925,
      "grad_norm": 0.0017439982620999217,
      "learning_rate": 0.00197990112931819,
      "loss": 0.3119,
      "num_input_tokens_seen": 71873472,
      "step": 37930
    },
    {
      "epoch": 6.18843298800881,
      "grad_norm": 0.0012776933144778013,
      "learning_rate": 0.0019703735084506345,
      "loss": 0.2686,
      "num_input_tokens_seen": 71883040,
      "step": 37935
    },
    {
      "epoch": 6.189248715229628,
      "grad_norm": 0.003307298058643937,
      "learning_rate": 0.001960868715611763,
      "loss": 0.3641,
      "num_input_tokens_seen": 71891808,
      "step": 37940
    },
    {
      "epoch": 6.190064442450445,
      "grad_norm": 0.0017425590194761753,
      "learning_rate": 0.0019513867522673034,
      "loss": 0.3533,
      "num_input_tokens_seen": 71902288,
      "step": 37945
    },
    {
      "epoch": 6.190880169671262,
      "grad_norm": 0.0014600880676880479,
      "learning_rate": 0.001941927619879502,
      "loss": 0.2672,
      "num_input_tokens_seen": 71912544,
      "step": 37950
    },
    {
      "epoch": 6.191695896892079,
      "grad_norm": 0.0022105607204139233,
      "learning_rate": 0.0019324913199070758,
      "loss": 0.2945,
      "num_input_tokens_seen": 71921840,
      "step": 37955
    },
    {
      "epoch": 6.192511624112897,
      "grad_norm": 0.000815852137748152,
      "learning_rate": 0.0019230778538052106,
      "loss": 0.3114,
      "num_input_tokens_seen": 71929792,
      "step": 37960
    },
    {
      "epoch": 6.193327351333714,
      "grad_norm": 0.0010555819608271122,
      "learning_rate": 0.0019136872230255952,
      "loss": 0.3356,
      "num_input_tokens_seen": 71937264,
      "step": 37965
    },
    {
      "epoch": 6.194143078554531,
      "grad_norm": 0.002114382106810808,
      "learning_rate": 0.0019043194290164045,
      "loss": 0.3179,
      "num_input_tokens_seen": 71946096,
      "step": 37970
    },
    {
      "epoch": 6.194958805775348,
      "grad_norm": 0.0008782107033766806,
      "learning_rate": 0.0018949744732222162,
      "loss": 0.3265,
      "num_input_tokens_seen": 71954672,
      "step": 37975
    },
    {
      "epoch": 6.195774532996166,
      "grad_norm": 0.0027377561200410128,
      "learning_rate": 0.0018856523570841776,
      "loss": 0.3107,
      "num_input_tokens_seen": 71965056,
      "step": 37980
    },
    {
      "epoch": 6.196590260216984,
      "grad_norm": 0.0010504473466426134,
      "learning_rate": 0.0018763530820398555,
      "loss": 0.3826,
      "num_input_tokens_seen": 71973936,
      "step": 37985
    },
    {
      "epoch": 6.197405987437801,
      "grad_norm": 0.001337041612714529,
      "learning_rate": 0.0018670766495233525,
      "loss": 0.289,
      "num_input_tokens_seen": 71983936,
      "step": 37990
    },
    {
      "epoch": 6.198221714658618,
      "grad_norm": 0.0033221219200640917,
      "learning_rate": 0.001857823060965158,
      "loss": 0.3627,
      "num_input_tokens_seen": 71993264,
      "step": 37995
    },
    {
      "epoch": 6.199037441879436,
      "grad_norm": 0.0029895161278545856,
      "learning_rate": 0.0018485923177923467,
      "loss": 0.2823,
      "num_input_tokens_seen": 72002688,
      "step": 38000
    },
    {
      "epoch": 6.199037441879436,
      "eval_loss": 0.3194601833820343,
      "eval_runtime": 152.8523,
      "eval_samples_per_second": 17.828,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 72002688,
      "step": 38000
    },
    {
      "epoch": 6.199853169100253,
      "grad_norm": 0.0027577762957662344,
      "learning_rate": 0.001839384421428364,
      "loss": 0.3558,
      "num_input_tokens_seen": 72012272,
      "step": 38005
    },
    {
      "epoch": 6.20066889632107,
      "grad_norm": 0.002552238292992115,
      "learning_rate": 0.0018301993732932065,
      "loss": 0.322,
      "num_input_tokens_seen": 72021328,
      "step": 38010
    },
    {
      "epoch": 6.201484623541887,
      "grad_norm": 0.0031407426577061415,
      "learning_rate": 0.0018210371748033248,
      "loss": 0.3611,
      "num_input_tokens_seen": 72032048,
      "step": 38015
    },
    {
      "epoch": 6.202300350762705,
      "grad_norm": 0.002270448487251997,
      "learning_rate": 0.0018118978273716556,
      "loss": 0.3343,
      "num_input_tokens_seen": 72042560,
      "step": 38020
    },
    {
      "epoch": 6.203116077983522,
      "grad_norm": 0.004125128034502268,
      "learning_rate": 0.001802781332407588,
      "loss": 0.3499,
      "num_input_tokens_seen": 72051712,
      "step": 38025
    },
    {
      "epoch": 6.2039318052043395,
      "grad_norm": 0.0009718389483168721,
      "learning_rate": 0.0017936876913169806,
      "loss": 0.2961,
      "num_input_tokens_seen": 72061936,
      "step": 38030
    },
    {
      "epoch": 6.204747532425157,
      "grad_norm": 0.0019898337777704,
      "learning_rate": 0.0017846169055022287,
      "loss": 0.3124,
      "num_input_tokens_seen": 72070752,
      "step": 38035
    },
    {
      "epoch": 6.205563259645975,
      "grad_norm": 0.0016438941238448024,
      "learning_rate": 0.0017755689763621295,
      "loss": 0.2915,
      "num_input_tokens_seen": 72079472,
      "step": 38040
    },
    {
      "epoch": 6.206378986866792,
      "grad_norm": 0.0019819061271846294,
      "learning_rate": 0.0017665439052920173,
      "loss": 0.3358,
      "num_input_tokens_seen": 72089120,
      "step": 38045
    },
    {
      "epoch": 6.207194714087609,
      "grad_norm": 0.0017527735326439142,
      "learning_rate": 0.0017575416936836286,
      "loss": 0.408,
      "num_input_tokens_seen": 72098064,
      "step": 38050
    },
    {
      "epoch": 6.208010441308426,
      "grad_norm": 0.0030370946042239666,
      "learning_rate": 0.0017485623429252528,
      "loss": 0.3786,
      "num_input_tokens_seen": 72107584,
      "step": 38055
    },
    {
      "epoch": 6.208826168529244,
      "grad_norm": 0.0009747457224875689,
      "learning_rate": 0.0017396058544016156,
      "loss": 0.3093,
      "num_input_tokens_seen": 72117392,
      "step": 38060
    },
    {
      "epoch": 6.209641895750061,
      "grad_norm": 0.002751583233475685,
      "learning_rate": 0.0017306722294938958,
      "loss": 0.3699,
      "num_input_tokens_seen": 72126480,
      "step": 38065
    },
    {
      "epoch": 6.210457622970878,
      "grad_norm": 0.0018538573058322072,
      "learning_rate": 0.0017217614695798078,
      "loss": 0.3633,
      "num_input_tokens_seen": 72135792,
      "step": 38070
    },
    {
      "epoch": 6.211273350191696,
      "grad_norm": 0.0013031938578933477,
      "learning_rate": 0.001712873576033469,
      "loss": 0.3658,
      "num_input_tokens_seen": 72145104,
      "step": 38075
    },
    {
      "epoch": 6.2120890774125135,
      "grad_norm": 0.0039500892162323,
      "learning_rate": 0.0017040085502255163,
      "loss": 0.3556,
      "num_input_tokens_seen": 72154640,
      "step": 38080
    },
    {
      "epoch": 6.212904804633331,
      "grad_norm": 0.002327348804101348,
      "learning_rate": 0.0016951663935230565,
      "loss": 0.3281,
      "num_input_tokens_seen": 72164768,
      "step": 38085
    },
    {
      "epoch": 6.213720531854148,
      "grad_norm": 0.0016450902912765741,
      "learning_rate": 0.0016863471072896485,
      "loss": 0.3208,
      "num_input_tokens_seen": 72175200,
      "step": 38090
    },
    {
      "epoch": 6.214536259074965,
      "grad_norm": 0.0007271571084856987,
      "learning_rate": 0.0016775506928853377,
      "loss": 0.3361,
      "num_input_tokens_seen": 72184512,
      "step": 38095
    },
    {
      "epoch": 6.215351986295783,
      "grad_norm": 0.0038236118853092194,
      "learning_rate": 0.001668777151666656,
      "loss": 0.3513,
      "num_input_tokens_seen": 72194160,
      "step": 38100
    },
    {
      "epoch": 6.2161677135166,
      "grad_norm": 0.0014640884473919868,
      "learning_rate": 0.0016600264849865709,
      "loss": 0.2718,
      "num_input_tokens_seen": 72203568,
      "step": 38105
    },
    {
      "epoch": 6.216983440737417,
      "grad_norm": 0.0006229642312973738,
      "learning_rate": 0.0016512986941945695,
      "loss": 0.3106,
      "num_input_tokens_seen": 72213072,
      "step": 38110
    },
    {
      "epoch": 6.217799167958235,
      "grad_norm": 0.0034583185333758593,
      "learning_rate": 0.0016425937806365753,
      "loss": 0.3639,
      "num_input_tokens_seen": 72221792,
      "step": 38115
    },
    {
      "epoch": 6.218614895179052,
      "grad_norm": 0.001745323184877634,
      "learning_rate": 0.0016339117456549979,
      "loss": 0.3343,
      "num_input_tokens_seen": 72231216,
      "step": 38120
    },
    {
      "epoch": 6.219430622399869,
      "grad_norm": 0.0017458245856687427,
      "learning_rate": 0.0016252525905886995,
      "loss": 0.3583,
      "num_input_tokens_seen": 72241200,
      "step": 38125
    },
    {
      "epoch": 6.220246349620687,
      "grad_norm": 0.0008056694641709328,
      "learning_rate": 0.0016166163167730617,
      "loss": 0.2731,
      "num_input_tokens_seen": 72250096,
      "step": 38130
    },
    {
      "epoch": 6.221062076841505,
      "grad_norm": 0.001272661262191832,
      "learning_rate": 0.0016080029255398864,
      "loss": 0.2884,
      "num_input_tokens_seen": 72260176,
      "step": 38135
    },
    {
      "epoch": 6.221877804062322,
      "grad_norm": 0.0019962573423981667,
      "learning_rate": 0.0015994124182174606,
      "loss": 0.321,
      "num_input_tokens_seen": 72269200,
      "step": 38140
    },
    {
      "epoch": 6.222693531283139,
      "grad_norm": 0.002067466266453266,
      "learning_rate": 0.001590844796130575,
      "loss": 0.2966,
      "num_input_tokens_seen": 72278752,
      "step": 38145
    },
    {
      "epoch": 6.223509258503956,
      "grad_norm": 0.002456369809806347,
      "learning_rate": 0.001582300060600439,
      "loss": 0.3342,
      "num_input_tokens_seen": 72288464,
      "step": 38150
    },
    {
      "epoch": 6.224324985724774,
      "grad_norm": 0.0014897382352501154,
      "learning_rate": 0.0015737782129447652,
      "loss": 0.359,
      "num_input_tokens_seen": 72298144,
      "step": 38155
    },
    {
      "epoch": 6.225140712945591,
      "grad_norm": 0.0017224818002432585,
      "learning_rate": 0.0015652792544777361,
      "loss": 0.3142,
      "num_input_tokens_seen": 72307824,
      "step": 38160
    },
    {
      "epoch": 6.225956440166408,
      "grad_norm": 0.0019137116614729166,
      "learning_rate": 0.0015568031865099863,
      "loss": 0.3213,
      "num_input_tokens_seen": 72316880,
      "step": 38165
    },
    {
      "epoch": 6.226772167387225,
      "grad_norm": 0.0017881602980196476,
      "learning_rate": 0.0015483500103486369,
      "loss": 0.3451,
      "num_input_tokens_seen": 72327056,
      "step": 38170
    },
    {
      "epoch": 6.227587894608043,
      "grad_norm": 0.0005175760597921908,
      "learning_rate": 0.0015399197272972787,
      "loss": 0.2533,
      "num_input_tokens_seen": 72337296,
      "step": 38175
    },
    {
      "epoch": 6.2284036218288605,
      "grad_norm": 0.0016352282837033272,
      "learning_rate": 0.0015315123386559714,
      "loss": 0.3207,
      "num_input_tokens_seen": 72347568,
      "step": 38180
    },
    {
      "epoch": 6.229219349049678,
      "grad_norm": 0.001206320128403604,
      "learning_rate": 0.0015231278457212283,
      "loss": 0.3052,
      "num_input_tokens_seen": 72356608,
      "step": 38185
    },
    {
      "epoch": 6.230035076270495,
      "grad_norm": 0.0013669721083715558,
      "learning_rate": 0.001514766249786048,
      "loss": 0.3959,
      "num_input_tokens_seen": 72366048,
      "step": 38190
    },
    {
      "epoch": 6.230850803491313,
      "grad_norm": 0.0006700943922623992,
      "learning_rate": 0.0015064275521398994,
      "loss": 0.336,
      "num_input_tokens_seen": 72376112,
      "step": 38195
    },
    {
      "epoch": 6.23166653071213,
      "grad_norm": 0.0019084290834143758,
      "learning_rate": 0.0014981117540686872,
      "loss": 0.3585,
      "num_input_tokens_seen": 72385776,
      "step": 38200
    },
    {
      "epoch": 6.23166653071213,
      "eval_loss": 0.31980130076408386,
      "eval_runtime": 152.7974,
      "eval_samples_per_second": 17.834,
      "eval_steps_per_second": 8.92,
      "num_input_tokens_seen": 72385776,
      "step": 38200
    },
    {
      "epoch": 6.232482257932947,
      "grad_norm": 0.002572412136942148,
      "learning_rate": 0.0014898188568548687,
      "loss": 0.2972,
      "num_input_tokens_seen": 72395168,
      "step": 38205
    },
    {
      "epoch": 6.233297985153764,
      "grad_norm": 0.0023618361447006464,
      "learning_rate": 0.0014815488617772542,
      "loss": 0.3215,
      "num_input_tokens_seen": 72405088,
      "step": 38210
    },
    {
      "epoch": 6.234113712374582,
      "grad_norm": 0.001589964609593153,
      "learning_rate": 0.0014733017701112072,
      "loss": 0.3253,
      "num_input_tokens_seen": 72414816,
      "step": 38215
    },
    {
      "epoch": 6.234929439595399,
      "grad_norm": 0.0030608191154897213,
      "learning_rate": 0.0014650775831285435,
      "loss": 0.2921,
      "num_input_tokens_seen": 72424528,
      "step": 38220
    },
    {
      "epoch": 6.2357451668162165,
      "grad_norm": 0.0012767496518790722,
      "learning_rate": 0.001456876302097515,
      "loss": 0.3346,
      "num_input_tokens_seen": 72434208,
      "step": 38225
    },
    {
      "epoch": 6.236560894037034,
      "grad_norm": 0.001815503346733749,
      "learning_rate": 0.0014486979282828604,
      "loss": 0.3956,
      "num_input_tokens_seen": 72444320,
      "step": 38230
    },
    {
      "epoch": 6.237376621257852,
      "grad_norm": 0.0009547151275910437,
      "learning_rate": 0.001440542462945804,
      "loss": 0.348,
      "num_input_tokens_seen": 72453984,
      "step": 38235
    },
    {
      "epoch": 6.238192348478669,
      "grad_norm": 0.001365954871289432,
      "learning_rate": 0.0014324099073440232,
      "loss": 0.3152,
      "num_input_tokens_seen": 72463328,
      "step": 38240
    },
    {
      "epoch": 6.239008075699486,
      "grad_norm": 0.0046540675684809685,
      "learning_rate": 0.0014243002627316482,
      "loss": 0.2676,
      "num_input_tokens_seen": 72472560,
      "step": 38245
    },
    {
      "epoch": 6.239823802920303,
      "grad_norm": 0.0013422436313703656,
      "learning_rate": 0.0014162135303592781,
      "loss": 0.3311,
      "num_input_tokens_seen": 72482448,
      "step": 38250
    },
    {
      "epoch": 6.240639530141121,
      "grad_norm": 0.0026065236888825893,
      "learning_rate": 0.001408149711474016,
      "loss": 0.34,
      "num_input_tokens_seen": 72491920,
      "step": 38255
    },
    {
      "epoch": 6.241455257361938,
      "grad_norm": 0.001679827575571835,
      "learning_rate": 0.0014001088073193834,
      "loss": 0.3583,
      "num_input_tokens_seen": 72502064,
      "step": 38260
    },
    {
      "epoch": 6.242270984582755,
      "grad_norm": 0.0008394925971515477,
      "learning_rate": 0.0013920908191354052,
      "loss": 0.3273,
      "num_input_tokens_seen": 72511328,
      "step": 38265
    },
    {
      "epoch": 6.243086711803572,
      "grad_norm": 0.004262062255293131,
      "learning_rate": 0.001384095748158526,
      "loss": 0.3319,
      "num_input_tokens_seen": 72519856,
      "step": 38270
    },
    {
      "epoch": 6.2439024390243905,
      "grad_norm": 0.0015599738107994199,
      "learning_rate": 0.0013761235956217255,
      "loss": 0.2746,
      "num_input_tokens_seen": 72529648,
      "step": 38275
    },
    {
      "epoch": 6.244718166245208,
      "grad_norm": 0.0012568613747134805,
      "learning_rate": 0.0013681743627543873,
      "loss": 0.3149,
      "num_input_tokens_seen": 72538384,
      "step": 38280
    },
    {
      "epoch": 6.245533893466025,
      "grad_norm": 0.0028234589844942093,
      "learning_rate": 0.001360248050782381,
      "loss": 0.4068,
      "num_input_tokens_seen": 72546512,
      "step": 38285
    },
    {
      "epoch": 6.246349620686843,
      "grad_norm": 0.0035179906990379095,
      "learning_rate": 0.001352344660928062,
      "loss": 0.2995,
      "num_input_tokens_seen": 72556816,
      "step": 38290
    },
    {
      "epoch": 6.24716534790766,
      "grad_norm": 0.0013723662123084068,
      "learning_rate": 0.0013444641944102052,
      "loss": 0.2448,
      "num_input_tokens_seen": 72566880,
      "step": 38295
    },
    {
      "epoch": 6.247981075128477,
      "grad_norm": 0.0006786816520616412,
      "learning_rate": 0.0013366066524441056,
      "loss": 0.379,
      "num_input_tokens_seen": 72577376,
      "step": 38300
    },
    {
      "epoch": 6.248796802349294,
      "grad_norm": 0.0021208275575190783,
      "learning_rate": 0.0013287720362414768,
      "loss": 0.3369,
      "num_input_tokens_seen": 72587664,
      "step": 38305
    },
    {
      "epoch": 6.249612529570112,
      "grad_norm": 0.0023749859537929296,
      "learning_rate": 0.0013209603470105025,
      "loss": 0.3814,
      "num_input_tokens_seen": 72596784,
      "step": 38310
    },
    {
      "epoch": 6.250428256790929,
      "grad_norm": 0.004224834498018026,
      "learning_rate": 0.0013131715859558857,
      "loss": 0.2785,
      "num_input_tokens_seen": 72605920,
      "step": 38315
    },
    {
      "epoch": 6.251243984011746,
      "grad_norm": 0.001621593488380313,
      "learning_rate": 0.001305405754278699,
      "loss": 0.303,
      "num_input_tokens_seen": 72616672,
      "step": 38320
    },
    {
      "epoch": 6.2520597112325635,
      "grad_norm": 0.0017031541792675853,
      "learning_rate": 0.0012976628531765843,
      "loss": 0.3567,
      "num_input_tokens_seen": 72626176,
      "step": 38325
    },
    {
      "epoch": 6.252875438453382,
      "grad_norm": 0.0024869407061487436,
      "learning_rate": 0.0012899428838435533,
      "loss": 0.3272,
      "num_input_tokens_seen": 72635792,
      "step": 38330
    },
    {
      "epoch": 6.253691165674199,
      "grad_norm": 0.002315002726390958,
      "learning_rate": 0.001282245847470137,
      "loss": 0.3581,
      "num_input_tokens_seen": 72643616,
      "step": 38335
    },
    {
      "epoch": 6.254506892895016,
      "grad_norm": 0.0030739621724933386,
      "learning_rate": 0.001274571745243319,
      "loss": 0.3043,
      "num_input_tokens_seen": 72653952,
      "step": 38340
    },
    {
      "epoch": 6.255322620115833,
      "grad_norm": 0.0016884560463950038,
      "learning_rate": 0.0012669205783465364,
      "loss": 0.3556,
      "num_input_tokens_seen": 72665152,
      "step": 38345
    },
    {
      "epoch": 6.256138347336651,
      "grad_norm": 0.0007421983173117042,
      "learning_rate": 0.001259292347959695,
      "loss": 0.2954,
      "num_input_tokens_seen": 72674208,
      "step": 38350
    },
    {
      "epoch": 6.256954074557468,
      "grad_norm": 0.001583400764502585,
      "learning_rate": 0.0012516870552591707,
      "loss": 0.3209,
      "num_input_tokens_seen": 72684112,
      "step": 38355
    },
    {
      "epoch": 6.257769801778285,
      "grad_norm": 0.0013129517901688814,
      "learning_rate": 0.001244104701417792,
      "loss": 0.3235,
      "num_input_tokens_seen": 72694208,
      "step": 38360
    },
    {
      "epoch": 6.258585528999102,
      "grad_norm": 0.0010081237414851785,
      "learning_rate": 0.0012365452876048565,
      "loss": 0.3344,
      "num_input_tokens_seen": 72703888,
      "step": 38365
    },
    {
      "epoch": 6.25940125621992,
      "grad_norm": 0.0008888337179087102,
      "learning_rate": 0.001229008814986099,
      "loss": 0.2918,
      "num_input_tokens_seen": 72713424,
      "step": 38370
    },
    {
      "epoch": 6.2602169834407375,
      "grad_norm": 0.0006710238521918654,
      "learning_rate": 0.0012214952847237725,
      "loss": 0.3366,
      "num_input_tokens_seen": 72723520,
      "step": 38375
    },
    {
      "epoch": 6.261032710661555,
      "grad_norm": 0.0024274890311062336,
      "learning_rate": 0.0012140046979765339,
      "loss": 0.3099,
      "num_input_tokens_seen": 72732304,
      "step": 38380
    },
    {
      "epoch": 6.261848437882372,
      "grad_norm": 0.002002850640565157,
      "learning_rate": 0.0012065370558995258,
      "loss": 0.3588,
      "num_input_tokens_seen": 72743184,
      "step": 38385
    },
    {
      "epoch": 6.26266416510319,
      "grad_norm": 0.0014317660825327039,
      "learning_rate": 0.0011990923596443602,
      "loss": 0.3115,
      "num_input_tokens_seen": 72752496,
      "step": 38390
    },
    {
      "epoch": 6.263479892324007,
      "grad_norm": 0.0028118437621742487,
      "learning_rate": 0.001191670610359119,
      "loss": 0.3291,
      "num_input_tokens_seen": 72762608,
      "step": 38395
    },
    {
      "epoch": 6.264295619544824,
      "grad_norm": 0.0010350155644118786,
      "learning_rate": 0.0011842718091882865,
      "loss": 0.2892,
      "num_input_tokens_seen": 72773152,
      "step": 38400
    },
    {
      "epoch": 6.264295619544824,
      "eval_loss": 0.32068324089050293,
      "eval_runtime": 152.7726,
      "eval_samples_per_second": 17.837,
      "eval_steps_per_second": 8.922,
      "num_input_tokens_seen": 72773152,
      "step": 38400
    },
    {
      "epoch": 6.265111346765641,
      "grad_norm": 0.0011209434596821666,
      "learning_rate": 0.0011768959572729,
      "loss": 0.3266,
      "num_input_tokens_seen": 72781648,
      "step": 38405
    },
    {
      "epoch": 6.265927073986459,
      "grad_norm": 0.0026391344144940376,
      "learning_rate": 0.001169543055750366,
      "loss": 0.2733,
      "num_input_tokens_seen": 72790384,
      "step": 38410
    },
    {
      "epoch": 6.266742801207276,
      "grad_norm": 0.0013270131312310696,
      "learning_rate": 0.0011622131057546115,
      "loss": 0.2837,
      "num_input_tokens_seen": 72800256,
      "step": 38415
    },
    {
      "epoch": 6.2675585284280935,
      "grad_norm": 0.0037906807847321033,
      "learning_rate": 0.0011549061084160316,
      "loss": 0.3795,
      "num_input_tokens_seen": 72810016,
      "step": 38420
    },
    {
      "epoch": 6.268374255648911,
      "grad_norm": 0.0012703953543677926,
      "learning_rate": 0.0011476220648614088,
      "loss": 0.2882,
      "num_input_tokens_seen": 72819008,
      "step": 38425
    },
    {
      "epoch": 6.269189982869729,
      "grad_norm": 0.0007730748620815575,
      "learning_rate": 0.0011403609762140777,
      "loss": 0.3412,
      "num_input_tokens_seen": 72827888,
      "step": 38430
    },
    {
      "epoch": 6.270005710090546,
      "grad_norm": 0.0020105033181607723,
      "learning_rate": 0.0011331228435937756,
      "loss": 0.3127,
      "num_input_tokens_seen": 72838752,
      "step": 38435
    },
    {
      "epoch": 6.270821437311363,
      "grad_norm": 0.0020962818525731564,
      "learning_rate": 0.0011259076681166935,
      "loss": 0.335,
      "num_input_tokens_seen": 72847648,
      "step": 38440
    },
    {
      "epoch": 6.27163716453218,
      "grad_norm": 0.002350050723180175,
      "learning_rate": 0.0011187154508955244,
      "loss": 0.3065,
      "num_input_tokens_seen": 72857344,
      "step": 38445
    },
    {
      "epoch": 6.272452891752998,
      "grad_norm": 0.0042540584690868855,
      "learning_rate": 0.001111546193039381,
      "loss": 0.3598,
      "num_input_tokens_seen": 72867232,
      "step": 38450
    },
    {
      "epoch": 6.273268618973815,
      "grad_norm": 0.0016596788773313165,
      "learning_rate": 0.0011043998956538792,
      "loss": 0.2887,
      "num_input_tokens_seen": 72875664,
      "step": 38455
    },
    {
      "epoch": 6.274084346194632,
      "grad_norm": 0.0023404944222420454,
      "learning_rate": 0.0010972765598410538,
      "loss": 0.3053,
      "num_input_tokens_seen": 72884992,
      "step": 38460
    },
    {
      "epoch": 6.27490007341545,
      "grad_norm": 0.0006727639702148736,
      "learning_rate": 0.0010901761866993931,
      "loss": 0.3152,
      "num_input_tokens_seen": 72893536,
      "step": 38465
    },
    {
      "epoch": 6.275715800636267,
      "grad_norm": 0.0011607770575210452,
      "learning_rate": 0.0010830987773238876,
      "loss": 0.3674,
      "num_input_tokens_seen": 72902768,
      "step": 38470
    },
    {
      "epoch": 6.276531527857085,
      "grad_norm": 0.0024963796604424715,
      "learning_rate": 0.0010760443328059644,
      "loss": 0.3482,
      "num_input_tokens_seen": 72912240,
      "step": 38475
    },
    {
      "epoch": 6.277347255077902,
      "grad_norm": 0.0005359582719393075,
      "learning_rate": 0.001069012854233503,
      "loss": 0.2874,
      "num_input_tokens_seen": 72920368,
      "step": 38480
    },
    {
      "epoch": 6.27816298229872,
      "grad_norm": 0.0012353757629171014,
      "learning_rate": 0.0010620043426908365,
      "loss": 0.3052,
      "num_input_tokens_seen": 72930720,
      "step": 38485
    },
    {
      "epoch": 6.278978709519537,
      "grad_norm": 0.0035808677785098553,
      "learning_rate": 0.0010550187992587833,
      "loss": 0.336,
      "num_input_tokens_seen": 72940048,
      "step": 38490
    },
    {
      "epoch": 6.279794436740354,
      "grad_norm": 0.004977675620466471,
      "learning_rate": 0.0010480562250145653,
      "loss": 0.2856,
      "num_input_tokens_seen": 72949552,
      "step": 38495
    },
    {
      "epoch": 6.280610163961171,
      "grad_norm": 0.001240203040651977,
      "learning_rate": 0.0010411166210319567,
      "loss": 0.3331,
      "num_input_tokens_seen": 72960256,
      "step": 38500
    },
    {
      "epoch": 6.281425891181989,
      "grad_norm": 0.0010262406431138515,
      "learning_rate": 0.0010341999883810848,
      "loss": 0.2842,
      "num_input_tokens_seen": 72969680,
      "step": 38505
    },
    {
      "epoch": 6.282241618402806,
      "grad_norm": 0.0018067241180688143,
      "learning_rate": 0.0010273063281285965,
      "loss": 0.3364,
      "num_input_tokens_seen": 72978272,
      "step": 38510
    },
    {
      "epoch": 6.283057345623623,
      "grad_norm": 0.0036734449677169323,
      "learning_rate": 0.0010204356413375747,
      "loss": 0.325,
      "num_input_tokens_seen": 72988112,
      "step": 38515
    },
    {
      "epoch": 6.2838730728444405,
      "grad_norm": 0.0028986644465476274,
      "learning_rate": 0.001013587929067572,
      "loss": 0.3292,
      "num_input_tokens_seen": 72997104,
      "step": 38520
    },
    {
      "epoch": 6.2846888000652585,
      "grad_norm": 0.0014701812760904431,
      "learning_rate": 0.00100676319237461,
      "loss": 0.2735,
      "num_input_tokens_seen": 73005984,
      "step": 38525
    },
    {
      "epoch": 6.285504527286076,
      "grad_norm": 0.0039693862199783325,
      "learning_rate": 0.0009999614323110972,
      "loss": 0.3099,
      "num_input_tokens_seen": 73015168,
      "step": 38530
    },
    {
      "epoch": 6.286320254506893,
      "grad_norm": 0.002387966960668564,
      "learning_rate": 0.000993182649926011,
      "loss": 0.3321,
      "num_input_tokens_seen": 73023952,
      "step": 38535
    },
    {
      "epoch": 6.28713598172771,
      "grad_norm": 0.0009635997703298926,
      "learning_rate": 0.000986426846264682,
      "loss": 0.289,
      "num_input_tokens_seen": 73033296,
      "step": 38540
    },
    {
      "epoch": 6.287951708948528,
      "grad_norm": 0.0009143439237959683,
      "learning_rate": 0.00097969402236896,
      "loss": 0.3468,
      "num_input_tokens_seen": 73043280,
      "step": 38545
    },
    {
      "epoch": 6.288767436169345,
      "grad_norm": 0.0009939688025042415,
      "learning_rate": 0.0009729841792771143,
      "loss": 0.3322,
      "num_input_tokens_seen": 73052192,
      "step": 38550
    },
    {
      "epoch": 6.289583163390162,
      "grad_norm": 0.002103452105075121,
      "learning_rate": 0.0009662973180239176,
      "loss": 0.3615,
      "num_input_tokens_seen": 73062128,
      "step": 38555
    },
    {
      "epoch": 6.290398890610979,
      "grad_norm": 0.00041987464646808803,
      "learning_rate": 0.0009596334396405448,
      "loss": 0.3408,
      "num_input_tokens_seen": 73071792,
      "step": 38560
    },
    {
      "epoch": 6.291214617831797,
      "grad_norm": 0.004596478305757046,
      "learning_rate": 0.0009529925451546406,
      "loss": 0.2947,
      "num_input_tokens_seen": 73081760,
      "step": 38565
    },
    {
      "epoch": 6.2920303450526145,
      "grad_norm": 0.000890263938345015,
      "learning_rate": 0.0009463746355903357,
      "loss": 0.3292,
      "num_input_tokens_seen": 73091120,
      "step": 38570
    },
    {
      "epoch": 6.292846072273432,
      "grad_norm": 0.002123358892276883,
      "learning_rate": 0.0009397797119681971,
      "loss": 0.3777,
      "num_input_tokens_seen": 73100640,
      "step": 38575
    },
    {
      "epoch": 6.293661799494249,
      "grad_norm": 0.0028430235106498003,
      "learning_rate": 0.0009332077753052281,
      "loss": 0.2727,
      "num_input_tokens_seen": 73109392,
      "step": 38580
    },
    {
      "epoch": 6.294477526715067,
      "grad_norm": 0.0019213664345443249,
      "learning_rate": 0.0009266588266149011,
      "loss": 0.3021,
      "num_input_tokens_seen": 73118320,
      "step": 38585
    },
    {
      "epoch": 6.295293253935884,
      "grad_norm": 0.0020397668704390526,
      "learning_rate": 0.0009201328669071584,
      "loss": 0.4056,
      "num_input_tokens_seen": 73128000,
      "step": 38590
    },
    {
      "epoch": 6.296108981156701,
      "grad_norm": 0.0007886707317084074,
      "learning_rate": 0.0009136298971883949,
      "loss": 0.3674,
      "num_input_tokens_seen": 73139600,
      "step": 38595
    },
    {
      "epoch": 6.296924708377518,
      "grad_norm": 0.0019085564417764544,
      "learning_rate": 0.0009071499184614251,
      "loss": 0.3261,
      "num_input_tokens_seen": 73149584,
      "step": 38600
    },
    {
      "epoch": 6.296924708377518,
      "eval_loss": 0.3204604387283325,
      "eval_runtime": 152.8216,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 73149584,
      "step": 38600
    },
    {
      "epoch": 6.297740435598336,
      "grad_norm": 0.0009717802167870104,
      "learning_rate": 0.0009006929317255663,
      "loss": 0.2858,
      "num_input_tokens_seen": 73158400,
      "step": 38605
    },
    {
      "epoch": 6.298556162819153,
      "grad_norm": 0.0011126046301797032,
      "learning_rate": 0.0008942589379765387,
      "loss": 0.3323,
      "num_input_tokens_seen": 73168528,
      "step": 38610
    },
    {
      "epoch": 6.2993718900399704,
      "grad_norm": 0.0018056281842291355,
      "learning_rate": 0.0008878479382065817,
      "loss": 0.3114,
      "num_input_tokens_seen": 73178320,
      "step": 38615
    },
    {
      "epoch": 6.300187617260788,
      "grad_norm": 0.0011667088838294148,
      "learning_rate": 0.0008814599334043215,
      "loss": 0.3131,
      "num_input_tokens_seen": 73188480,
      "step": 38620
    },
    {
      "epoch": 6.301003344481606,
      "grad_norm": 0.0009592322749085724,
      "learning_rate": 0.0008750949245548866,
      "loss": 0.3056,
      "num_input_tokens_seen": 73197360,
      "step": 38625
    },
    {
      "epoch": 6.301819071702423,
      "grad_norm": 0.0016433766577392817,
      "learning_rate": 0.0008687529126398252,
      "loss": 0.2856,
      "num_input_tokens_seen": 73207184,
      "step": 38630
    },
    {
      "epoch": 6.30263479892324,
      "grad_norm": 0.003264442551881075,
      "learning_rate": 0.0008624338986371715,
      "loss": 0.2904,
      "num_input_tokens_seen": 73215952,
      "step": 38635
    },
    {
      "epoch": 6.303450526144058,
      "grad_norm": 0.0011416554916650057,
      "learning_rate": 0.0008561378835213962,
      "loss": 0.2811,
      "num_input_tokens_seen": 73225552,
      "step": 38640
    },
    {
      "epoch": 6.304266253364875,
      "grad_norm": 0.0016338349087163806,
      "learning_rate": 0.0008498648682634058,
      "loss": 0.3247,
      "num_input_tokens_seen": 73234048,
      "step": 38645
    },
    {
      "epoch": 6.305081980585692,
      "grad_norm": 0.003990654833614826,
      "learning_rate": 0.0008436148538306099,
      "loss": 0.2834,
      "num_input_tokens_seen": 73243600,
      "step": 38650
    },
    {
      "epoch": 6.305897707806509,
      "grad_norm": 0.0032979887910187244,
      "learning_rate": 0.0008373878411868041,
      "loss": 0.4355,
      "num_input_tokens_seen": 73252624,
      "step": 38655
    },
    {
      "epoch": 6.306713435027326,
      "grad_norm": 0.0007436767918989062,
      "learning_rate": 0.000831183831292287,
      "loss": 0.2891,
      "num_input_tokens_seen": 73261280,
      "step": 38660
    },
    {
      "epoch": 6.307529162248144,
      "grad_norm": 0.0012564858188852668,
      "learning_rate": 0.0008250028251037933,
      "loss": 0.3531,
      "num_input_tokens_seen": 73270912,
      "step": 38665
    },
    {
      "epoch": 6.3083448894689615,
      "grad_norm": 0.0018921529408544302,
      "learning_rate": 0.0008188448235745271,
      "loss": 0.288,
      "num_input_tokens_seen": 73280000,
      "step": 38670
    },
    {
      "epoch": 6.309160616689779,
      "grad_norm": 0.003252727910876274,
      "learning_rate": 0.0008127098276541122,
      "loss": 0.2924,
      "num_input_tokens_seen": 73289936,
      "step": 38675
    },
    {
      "epoch": 6.309976343910597,
      "grad_norm": 0.001769128954038024,
      "learning_rate": 0.0008065978382886418,
      "loss": 0.2828,
      "num_input_tokens_seen": 73299360,
      "step": 38680
    },
    {
      "epoch": 6.310792071131414,
      "grad_norm": 0.0017232729587703943,
      "learning_rate": 0.0008005088564206785,
      "loss": 0.385,
      "num_input_tokens_seen": 73307696,
      "step": 38685
    },
    {
      "epoch": 6.311607798352231,
      "grad_norm": 0.0024186435621231794,
      "learning_rate": 0.0007944428829891881,
      "loss": 0.3017,
      "num_input_tokens_seen": 73317088,
      "step": 38690
    },
    {
      "epoch": 6.312423525573048,
      "grad_norm": 0.00308241811580956,
      "learning_rate": 0.0007883999189296386,
      "loss": 0.3269,
      "num_input_tokens_seen": 73326592,
      "step": 38695
    },
    {
      "epoch": 6.313239252793866,
      "grad_norm": 0.0028511378914117813,
      "learning_rate": 0.0007823799651739515,
      "loss": 0.3225,
      "num_input_tokens_seen": 73337040,
      "step": 38700
    },
    {
      "epoch": 6.314054980014683,
      "grad_norm": 0.0022514506708830595,
      "learning_rate": 0.0007763830226504509,
      "loss": 0.3146,
      "num_input_tokens_seen": 73345600,
      "step": 38705
    },
    {
      "epoch": 6.3148707072355,
      "grad_norm": 0.002052406081929803,
      "learning_rate": 0.0007704090922839468,
      "loss": 0.3194,
      "num_input_tokens_seen": 73352912,
      "step": 38710
    },
    {
      "epoch": 6.3156864344563175,
      "grad_norm": 0.001390534802339971,
      "learning_rate": 0.0007644581749957025,
      "loss": 0.3306,
      "num_input_tokens_seen": 73362384,
      "step": 38715
    },
    {
      "epoch": 6.3165021616771355,
      "grad_norm": 0.0036401008255779743,
      "learning_rate": 0.000758530271703417,
      "loss": 0.277,
      "num_input_tokens_seen": 73370848,
      "step": 38720
    },
    {
      "epoch": 6.317317888897953,
      "grad_norm": 0.0006811514613218606,
      "learning_rate": 0.0007526253833212426,
      "loss": 0.3644,
      "num_input_tokens_seen": 73380192,
      "step": 38725
    },
    {
      "epoch": 6.31813361611877,
      "grad_norm": 0.0021109809167683125,
      "learning_rate": 0.0007467435107598008,
      "loss": 0.3188,
      "num_input_tokens_seen": 73389936,
      "step": 38730
    },
    {
      "epoch": 6.318949343339587,
      "grad_norm": 0.003178810700774193,
      "learning_rate": 0.0007408846549261328,
      "loss": 0.3374,
      "num_input_tokens_seen": 73399536,
      "step": 38735
    },
    {
      "epoch": 6.319765070560405,
      "grad_norm": 0.0011896300129592419,
      "learning_rate": 0.0007350488167237656,
      "loss": 0.3167,
      "num_input_tokens_seen": 73409168,
      "step": 38740
    },
    {
      "epoch": 6.320580797781222,
      "grad_norm": 0.003415964310988784,
      "learning_rate": 0.0007292359970526629,
      "loss": 0.2801,
      "num_input_tokens_seen": 73418272,
      "step": 38745
    },
    {
      "epoch": 6.321396525002039,
      "grad_norm": 0.0007351295207627118,
      "learning_rate": 0.0007234461968092076,
      "loss": 0.2582,
      "num_input_tokens_seen": 73426992,
      "step": 38750
    },
    {
      "epoch": 6.322212252222856,
      "grad_norm": 0.00223365006968379,
      "learning_rate": 0.0007176794168862854,
      "loss": 0.3852,
      "num_input_tokens_seen": 73436784,
      "step": 38755
    },
    {
      "epoch": 6.323027979443674,
      "grad_norm": 0.0015495769912377,
      "learning_rate": 0.000711935658173185,
      "loss": 0.2939,
      "num_input_tokens_seen": 73446096,
      "step": 38760
    },
    {
      "epoch": 6.3238437066644915,
      "grad_norm": 0.0024447282776236534,
      "learning_rate": 0.0007062149215556812,
      "loss": 0.3145,
      "num_input_tokens_seen": 73454304,
      "step": 38765
    },
    {
      "epoch": 6.324659433885309,
      "grad_norm": 0.002974661299958825,
      "learning_rate": 0.0007005172079159849,
      "loss": 0.3617,
      "num_input_tokens_seen": 73463264,
      "step": 38770
    },
    {
      "epoch": 6.325475161106126,
      "grad_norm": 0.0017324216896668077,
      "learning_rate": 0.0006948425181327267,
      "loss": 0.2862,
      "num_input_tokens_seen": 73472688,
      "step": 38775
    },
    {
      "epoch": 6.326290888326944,
      "grad_norm": 0.0019144206307828426,
      "learning_rate": 0.000689190853081073,
      "loss": 0.3317,
      "num_input_tokens_seen": 73482592,
      "step": 38780
    },
    {
      "epoch": 6.327106615547761,
      "grad_norm": 0.001904892036691308,
      "learning_rate": 0.000683562213632527,
      "loss": 0.3287,
      "num_input_tokens_seen": 73492112,
      "step": 38785
    },
    {
      "epoch": 6.327922342768578,
      "grad_norm": 0.0009513043914921582,
      "learning_rate": 0.0006779566006551108,
      "loss": 0.3313,
      "num_input_tokens_seen": 73501568,
      "step": 38790
    },
    {
      "epoch": 6.328738069989395,
      "grad_norm": 0.0024101058952510357,
      "learning_rate": 0.0006723740150132995,
      "loss": 0.305,
      "num_input_tokens_seen": 73511136,
      "step": 38795
    },
    {
      "epoch": 6.329553797210213,
      "grad_norm": 0.0013817078433930874,
      "learning_rate": 0.0006668144575679713,
      "loss": 0.274,
      "num_input_tokens_seen": 73519536,
      "step": 38800
    },
    {
      "epoch": 6.329553797210213,
      "eval_loss": 0.3197188377380371,
      "eval_runtime": 152.8796,
      "eval_samples_per_second": 17.824,
      "eval_steps_per_second": 8.916,
      "num_input_tokens_seen": 73519536,
      "step": 38800
    },
    {
      "epoch": 6.33036952443103,
      "grad_norm": 0.0011757015017792583,
      "learning_rate": 0.0006612779291765069,
      "loss": 0.3258,
      "num_input_tokens_seen": 73529360,
      "step": 38805
    },
    {
      "epoch": 6.331185251651847,
      "grad_norm": 0.0019522526999935508,
      "learning_rate": 0.0006557644306926736,
      "loss": 0.3674,
      "num_input_tokens_seen": 73538288,
      "step": 38810
    },
    {
      "epoch": 6.332000978872665,
      "grad_norm": 0.0023019220679998398,
      "learning_rate": 0.0006502739629667575,
      "loss": 0.3638,
      "num_input_tokens_seen": 73547840,
      "step": 38815
    },
    {
      "epoch": 6.332816706093483,
      "grad_norm": 0.0034187857527285814,
      "learning_rate": 0.0006448065268454317,
      "loss": 0.3982,
      "num_input_tokens_seen": 73556848,
      "step": 38820
    },
    {
      "epoch": 6.3336324333143,
      "grad_norm": 0.0009594497387297451,
      "learning_rate": 0.0006393621231718549,
      "loss": 0.3262,
      "num_input_tokens_seen": 73567696,
      "step": 38825
    },
    {
      "epoch": 6.334448160535117,
      "grad_norm": 0.0019858693704009056,
      "learning_rate": 0.0006339407527856389,
      "loss": 0.332,
      "num_input_tokens_seen": 73577344,
      "step": 38830
    },
    {
      "epoch": 6.335263887755934,
      "grad_norm": 0.0025444552302360535,
      "learning_rate": 0.0006285424165227982,
      "loss": 0.3142,
      "num_input_tokens_seen": 73588320,
      "step": 38835
    },
    {
      "epoch": 6.336079614976752,
      "grad_norm": 0.0016733146039769053,
      "learning_rate": 0.0006231671152158169,
      "loss": 0.3272,
      "num_input_tokens_seen": 73597664,
      "step": 38840
    },
    {
      "epoch": 6.336895342197569,
      "grad_norm": 0.0038555969949811697,
      "learning_rate": 0.0006178148496936819,
      "loss": 0.2983,
      "num_input_tokens_seen": 73608768,
      "step": 38845
    },
    {
      "epoch": 6.337711069418386,
      "grad_norm": 0.0019750171341001987,
      "learning_rate": 0.000612485620781733,
      "loss": 0.3063,
      "num_input_tokens_seen": 73619392,
      "step": 38850
    },
    {
      "epoch": 6.338526796639204,
      "grad_norm": 0.003218967467546463,
      "learning_rate": 0.0006071794293018296,
      "loss": 0.3345,
      "num_input_tokens_seen": 73629312,
      "step": 38855
    },
    {
      "epoch": 6.339342523860021,
      "grad_norm": 0.0016976416809484363,
      "learning_rate": 0.0006018962760722501,
      "loss": 0.3123,
      "num_input_tokens_seen": 73639488,
      "step": 38860
    },
    {
      "epoch": 6.3401582510808385,
      "grad_norm": 0.004105562809854746,
      "learning_rate": 0.0005966361619077098,
      "loss": 0.2885,
      "num_input_tokens_seen": 73649808,
      "step": 38865
    },
    {
      "epoch": 6.340973978301656,
      "grad_norm": 0.003234853269532323,
      "learning_rate": 0.000591399087619393,
      "loss": 0.3678,
      "num_input_tokens_seen": 73659680,
      "step": 38870
    },
    {
      "epoch": 6.341789705522474,
      "grad_norm": 0.001324811833910644,
      "learning_rate": 0.0005861850540149371,
      "loss": 0.3309,
      "num_input_tokens_seen": 73668928,
      "step": 38875
    },
    {
      "epoch": 6.342605432743291,
      "grad_norm": 0.0012902783928439021,
      "learning_rate": 0.0005809940618983822,
      "loss": 0.3493,
      "num_input_tokens_seen": 73676640,
      "step": 38880
    },
    {
      "epoch": 6.343421159964108,
      "grad_norm": 0.0016557401977479458,
      "learning_rate": 0.0005758261120702712,
      "loss": 0.2821,
      "num_input_tokens_seen": 73686784,
      "step": 38885
    },
    {
      "epoch": 6.344236887184925,
      "grad_norm": 0.0008956636302173138,
      "learning_rate": 0.0005706812053275501,
      "loss": 0.3629,
      "num_input_tokens_seen": 73695952,
      "step": 38890
    },
    {
      "epoch": 6.345052614405743,
      "grad_norm": 0.0008551676874049008,
      "learning_rate": 0.0005655593424636173,
      "loss": 0.2872,
      "num_input_tokens_seen": 73706640,
      "step": 38895
    },
    {
      "epoch": 6.34586834162656,
      "grad_norm": 0.0009471682133153081,
      "learning_rate": 0.0005604605242683746,
      "loss": 0.3286,
      "num_input_tokens_seen": 73716016,
      "step": 38900
    },
    {
      "epoch": 6.346684068847377,
      "grad_norm": 0.0015132263069972396,
      "learning_rate": 0.0005553847515280596,
      "loss": 0.2988,
      "num_input_tokens_seen": 73725856,
      "step": 38905
    },
    {
      "epoch": 6.3474997960681945,
      "grad_norm": 0.001575183356180787,
      "learning_rate": 0.0005503320250254795,
      "loss": 0.3003,
      "num_input_tokens_seen": 73735984,
      "step": 38910
    },
    {
      "epoch": 6.3483155232890125,
      "grad_norm": 0.0016644925344735384,
      "learning_rate": 0.0005453023455397943,
      "loss": 0.3206,
      "num_input_tokens_seen": 73744464,
      "step": 38915
    },
    {
      "epoch": 6.34913125050983,
      "grad_norm": 0.001433936064131558,
      "learning_rate": 0.0005402957138466502,
      "loss": 0.2803,
      "num_input_tokens_seen": 73753600,
      "step": 38920
    },
    {
      "epoch": 6.349946977730647,
      "grad_norm": 0.0010350528173148632,
      "learning_rate": 0.0005353121307181463,
      "loss": 0.299,
      "num_input_tokens_seen": 73762736,
      "step": 38925
    },
    {
      "epoch": 6.350762704951464,
      "grad_norm": 0.0014093503123149276,
      "learning_rate": 0.0005303515969227845,
      "loss": 0.3045,
      "num_input_tokens_seen": 73772816,
      "step": 38930
    },
    {
      "epoch": 6.351578432172282,
      "grad_norm": 0.0029767195228487253,
      "learning_rate": 0.0005254141132255862,
      "loss": 0.307,
      "num_input_tokens_seen": 73782896,
      "step": 38935
    },
    {
      "epoch": 6.352394159393099,
      "grad_norm": 0.0038222908042371273,
      "learning_rate": 0.0005204996803879258,
      "loss": 0.2897,
      "num_input_tokens_seen": 73793216,
      "step": 38940
    },
    {
      "epoch": 6.353209886613916,
      "grad_norm": 0.0030802092514932156,
      "learning_rate": 0.0005156082991676969,
      "loss": 0.3287,
      "num_input_tokens_seen": 73802896,
      "step": 38945
    },
    {
      "epoch": 6.354025613834733,
      "grad_norm": 0.0015404997393488884,
      "learning_rate": 0.0005107399703192127,
      "loss": 0.3634,
      "num_input_tokens_seen": 73811440,
      "step": 38950
    },
    {
      "epoch": 6.354841341055551,
      "grad_norm": 0.00047965417616069317,
      "learning_rate": 0.0005058946945932063,
      "loss": 0.3051,
      "num_input_tokens_seen": 73819904,
      "step": 38955
    },
    {
      "epoch": 6.3556570682763684,
      "grad_norm": 0.001131887431256473,
      "learning_rate": 0.0005010724727369131,
      "loss": 0.331,
      "num_input_tokens_seen": 73828640,
      "step": 38960
    },
    {
      "epoch": 6.356472795497186,
      "grad_norm": 0.0039001458790153265,
      "learning_rate": 0.000496273305493955,
      "loss": 0.3274,
      "num_input_tokens_seen": 73838320,
      "step": 38965
    },
    {
      "epoch": 6.357288522718003,
      "grad_norm": 0.0021436598617583513,
      "learning_rate": 0.0004914971936044399,
      "loss": 0.3393,
      "num_input_tokens_seen": 73848608,
      "step": 38970
    },
    {
      "epoch": 6.358104249938821,
      "grad_norm": 0.0017551303608343005,
      "learning_rate": 0.00048674413780491196,
      "loss": 0.3187,
      "num_input_tokens_seen": 73857488,
      "step": 38975
    },
    {
      "epoch": 6.358919977159638,
      "grad_norm": 0.003119376488029957,
      "learning_rate": 0.0004820141388283183,
      "loss": 0.4015,
      "num_input_tokens_seen": 73865232,
      "step": 38980
    },
    {
      "epoch": 6.359735704380455,
      "grad_norm": 0.0015334243653342128,
      "learning_rate": 0.00047730719740410874,
      "loss": 0.325,
      "num_input_tokens_seen": 73875200,
      "step": 38985
    },
    {
      "epoch": 6.360551431601272,
      "grad_norm": 0.0033209132961928844,
      "learning_rate": 0.00047262331425816927,
      "loss": 0.3248,
      "num_input_tokens_seen": 73885312,
      "step": 38990
    },
    {
      "epoch": 6.36136715882209,
      "grad_norm": 0.0011587445624172688,
      "learning_rate": 0.00046796249011277213,
      "loss": 0.3311,
      "num_input_tokens_seen": 73894768,
      "step": 38995
    },
    {
      "epoch": 6.362182886042907,
      "grad_norm": 0.0013318936107680202,
      "learning_rate": 0.00046332472568669236,
      "loss": 0.2749,
      "num_input_tokens_seen": 73902896,
      "step": 39000
    },
    {
      "epoch": 6.362182886042907,
      "eval_loss": 0.320065438747406,
      "eval_runtime": 152.9005,
      "eval_samples_per_second": 17.822,
      "eval_steps_per_second": 8.914,
      "num_input_tokens_seen": 73902896,
      "step": 39000
    },
    {
      "epoch": 6.362998613263724,
      "grad_norm": 0.0034150155261158943,
      "learning_rate": 0.0004587100216951578,
      "loss": 0.3349,
      "num_input_tokens_seen": 73912688,
      "step": 39005
    },
    {
      "epoch": 6.3638143404845415,
      "grad_norm": 0.0021602206397801638,
      "learning_rate": 0.00045411837884978265,
      "loss": 0.2985,
      "num_input_tokens_seen": 73921872,
      "step": 39010
    },
    {
      "epoch": 6.3646300677053596,
      "grad_norm": 0.0028129969723522663,
      "learning_rate": 0.00044954979785865045,
      "loss": 0.2902,
      "num_input_tokens_seen": 73931664,
      "step": 39015
    },
    {
      "epoch": 6.365445794926177,
      "grad_norm": 0.0009558869060128927,
      "learning_rate": 0.00044500427942631426,
      "loss": 0.248,
      "num_input_tokens_seen": 73941344,
      "step": 39020
    },
    {
      "epoch": 6.366261522146994,
      "grad_norm": 0.001408429117873311,
      "learning_rate": 0.0004404818242537467,
      "loss": 0.3333,
      "num_input_tokens_seen": 73950032,
      "step": 39025
    },
    {
      "epoch": 6.367077249367812,
      "grad_norm": 0.0037910162936896086,
      "learning_rate": 0.00043598243303837324,
      "loss": 0.3165,
      "num_input_tokens_seen": 73959312,
      "step": 39030
    },
    {
      "epoch": 6.367892976588629,
      "grad_norm": 0.0013486341340467334,
      "learning_rate": 0.00043150610647403885,
      "loss": 0.2978,
      "num_input_tokens_seen": 73969152,
      "step": 39035
    },
    {
      "epoch": 6.368708703809446,
      "grad_norm": 0.002617456251755357,
      "learning_rate": 0.00042705284525104134,
      "loss": 0.3246,
      "num_input_tokens_seen": 73977472,
      "step": 39040
    },
    {
      "epoch": 6.369524431030263,
      "grad_norm": 0.0037207237910479307,
      "learning_rate": 0.0004226226500561647,
      "loss": 0.3679,
      "num_input_tokens_seen": 73987104,
      "step": 39045
    },
    {
      "epoch": 6.370340158251081,
      "grad_norm": 0.0023994192015379667,
      "learning_rate": 0.0004182155215725791,
      "loss": 0.3508,
      "num_input_tokens_seen": 73995248,
      "step": 39050
    },
    {
      "epoch": 6.371155885471898,
      "grad_norm": 0.005517663434147835,
      "learning_rate": 0.00041383146047992424,
      "loss": 0.3764,
      "num_input_tokens_seen": 74004992,
      "step": 39055
    },
    {
      "epoch": 6.3719716126927155,
      "grad_norm": 0.0014918470988050103,
      "learning_rate": 0.00040947046745427597,
      "loss": 0.3184,
      "num_input_tokens_seen": 74014768,
      "step": 39060
    },
    {
      "epoch": 6.372787339913533,
      "grad_norm": 0.003617997746914625,
      "learning_rate": 0.00040513254316814625,
      "loss": 0.3178,
      "num_input_tokens_seen": 74024720,
      "step": 39065
    },
    {
      "epoch": 6.373603067134351,
      "grad_norm": 0.0021347845904529095,
      "learning_rate": 0.0004008176882905168,
      "loss": 0.3324,
      "num_input_tokens_seen": 74033920,
      "step": 39070
    },
    {
      "epoch": 6.374418794355168,
      "grad_norm": 0.0034748136531561613,
      "learning_rate": 0.00039652590348677184,
      "loss": 0.3284,
      "num_input_tokens_seen": 74042960,
      "step": 39075
    },
    {
      "epoch": 6.375234521575985,
      "grad_norm": 0.0006238435162231326,
      "learning_rate": 0.00039225718941878206,
      "loss": 0.2855,
      "num_input_tokens_seen": 74052928,
      "step": 39080
    },
    {
      "epoch": 6.376050248796802,
      "grad_norm": 0.0018617762252688408,
      "learning_rate": 0.00038801154674480417,
      "loss": 0.2941,
      "num_input_tokens_seen": 74062000,
      "step": 39085
    },
    {
      "epoch": 6.37686597601762,
      "grad_norm": 0.0010334108956158161,
      "learning_rate": 0.00038378897611959784,
      "loss": 0.2769,
      "num_input_tokens_seen": 74071456,
      "step": 39090
    },
    {
      "epoch": 6.377681703238437,
      "grad_norm": 0.000750953215174377,
      "learning_rate": 0.00037958947819430875,
      "loss": 0.327,
      "num_input_tokens_seen": 74079728,
      "step": 39095
    },
    {
      "epoch": 6.378497430459254,
      "grad_norm": 0.0034784905146807432,
      "learning_rate": 0.0003754130536165856,
      "loss": 0.2781,
      "num_input_tokens_seen": 74088048,
      "step": 39100
    },
    {
      "epoch": 6.3793131576800715,
      "grad_norm": 0.002874995581805706,
      "learning_rate": 0.0003712597030304632,
      "loss": 0.3271,
      "num_input_tokens_seen": 74098256,
      "step": 39105
    },
    {
      "epoch": 6.3801288849008895,
      "grad_norm": 0.0007100168731994927,
      "learning_rate": 0.00036712942707646247,
      "loss": 0.3319,
      "num_input_tokens_seen": 74106704,
      "step": 39110
    },
    {
      "epoch": 6.380944612121707,
      "grad_norm": 0.002477079164236784,
      "learning_rate": 0.00036302222639149063,
      "loss": 0.3011,
      "num_input_tokens_seen": 74115712,
      "step": 39115
    },
    {
      "epoch": 6.381760339342524,
      "grad_norm": 0.0006141957128420472,
      "learning_rate": 0.000358938101608941,
      "loss": 0.3153,
      "num_input_tokens_seen": 74125008,
      "step": 39120
    },
    {
      "epoch": 6.382576066563341,
      "grad_norm": 0.001956216525286436,
      "learning_rate": 0.0003548770533586598,
      "loss": 0.2916,
      "num_input_tokens_seen": 74134672,
      "step": 39125
    },
    {
      "epoch": 6.383391793784159,
      "grad_norm": 0.0012452469673007727,
      "learning_rate": 0.0003508390822668961,
      "loss": 0.3135,
      "num_input_tokens_seen": 74144912,
      "step": 39130
    },
    {
      "epoch": 6.384207521004976,
      "grad_norm": 0.0028516179881989956,
      "learning_rate": 0.00034682418895633503,
      "loss": 0.3226,
      "num_input_tokens_seen": 74155456,
      "step": 39135
    },
    {
      "epoch": 6.385023248225793,
      "grad_norm": 0.0024479629937559366,
      "learning_rate": 0.0003428323740461647,
      "loss": 0.3244,
      "num_input_tokens_seen": 74165424,
      "step": 39140
    },
    {
      "epoch": 6.38583897544661,
      "grad_norm": 0.0012812053319066763,
      "learning_rate": 0.00033886363815194276,
      "loss": 0.2866,
      "num_input_tokens_seen": 74175552,
      "step": 39145
    },
    {
      "epoch": 6.386654702667428,
      "grad_norm": 0.003069325815886259,
      "learning_rate": 0.0003349179818857129,
      "loss": 0.3163,
      "num_input_tokens_seen": 74184960,
      "step": 39150
    },
    {
      "epoch": 6.387470429888245,
      "grad_norm": 0.0007291949586942792,
      "learning_rate": 0.0003309954058559383,
      "loss": 0.3276,
      "num_input_tokens_seen": 74194560,
      "step": 39155
    },
    {
      "epoch": 6.388286157109063,
      "grad_norm": 0.0020765597000718117,
      "learning_rate": 0.0003270959106675186,
      "loss": 0.3429,
      "num_input_tokens_seen": 74203568,
      "step": 39160
    },
    {
      "epoch": 6.38910188432988,
      "grad_norm": 0.0011187749914824963,
      "learning_rate": 0.0003232194969218227,
      "loss": 0.3205,
      "num_input_tokens_seen": 74213680,
      "step": 39165
    },
    {
      "epoch": 6.389917611550698,
      "grad_norm": 0.0027573436964303255,
      "learning_rate": 0.00031936616521663905,
      "loss": 0.332,
      "num_input_tokens_seen": 74223376,
      "step": 39170
    },
    {
      "epoch": 6.390733338771515,
      "grad_norm": 0.0006428036140277982,
      "learning_rate": 0.00031553591614619236,
      "loss": 0.3277,
      "num_input_tokens_seen": 74233056,
      "step": 39175
    },
    {
      "epoch": 6.391549065992332,
      "grad_norm": 0.0026072957552969456,
      "learning_rate": 0.00031172875030117676,
      "loss": 0.3353,
      "num_input_tokens_seen": 74242352,
      "step": 39180
    },
    {
      "epoch": 6.392364793213149,
      "grad_norm": 0.0013484803494066,
      "learning_rate": 0.0003079446682686726,
      "loss": 0.3312,
      "num_input_tokens_seen": 74252096,
      "step": 39185
    },
    {
      "epoch": 6.393180520433967,
      "grad_norm": 0.0024792219046503305,
      "learning_rate": 0.0003041836706322465,
      "loss": 0.3945,
      "num_input_tokens_seen": 74260400,
      "step": 39190
    },
    {
      "epoch": 6.393996247654784,
      "grad_norm": 0.002784034702926874,
      "learning_rate": 0.0003004457579719011,
      "loss": 0.327,
      "num_input_tokens_seen": 74269968,
      "step": 39195
    },
    {
      "epoch": 6.394811974875601,
      "grad_norm": 0.004006591625511646,
      "learning_rate": 0.00029673093086405867,
      "loss": 0.3728,
      "num_input_tokens_seen": 74278960,
      "step": 39200
    },
    {
      "epoch": 6.394811974875601,
      "eval_loss": 0.3205803632736206,
      "eval_runtime": 152.8212,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 74278960,
      "step": 39200
    },
    {
      "epoch": 6.395627702096419,
      "grad_norm": 0.0027470788918435574,
      "learning_rate": 0.00029303918988159426,
      "loss": 0.3151,
      "num_input_tokens_seen": 74288016,
      "step": 39205
    },
    {
      "epoch": 6.3964434293172365,
      "grad_norm": 0.0006707454449497163,
      "learning_rate": 0.0002893705355938192,
      "loss": 0.3002,
      "num_input_tokens_seen": 74297856,
      "step": 39210
    },
    {
      "epoch": 6.397259156538054,
      "grad_norm": 0.00047636826639063656,
      "learning_rate": 0.0002857249685664975,
      "loss": 0.3085,
      "num_input_tokens_seen": 74307744,
      "step": 39215
    },
    {
      "epoch": 6.398074883758871,
      "grad_norm": 0.0028473848942667246,
      "learning_rate": 0.0002821024893618129,
      "loss": 0.3235,
      "num_input_tokens_seen": 74317536,
      "step": 39220
    },
    {
      "epoch": 6.398890610979688,
      "grad_norm": 0.0010313072707504034,
      "learning_rate": 0.0002785030985383852,
      "loss": 0.3432,
      "num_input_tokens_seen": 74325920,
      "step": 39225
    },
    {
      "epoch": 6.399706338200506,
      "grad_norm": 0.001955974381417036,
      "learning_rate": 0.00027492679665130356,
      "loss": 0.3298,
      "num_input_tokens_seen": 74334592,
      "step": 39230
    },
    {
      "epoch": 6.400522065421323,
      "grad_norm": 0.0009076356654986739,
      "learning_rate": 0.000271373584252077,
      "loss": 0.3221,
      "num_input_tokens_seen": 74342704,
      "step": 39235
    },
    {
      "epoch": 6.40133779264214,
      "grad_norm": 0.002438060473650694,
      "learning_rate": 0.00026784346188865046,
      "loss": 0.2979,
      "num_input_tokens_seen": 74352000,
      "step": 39240
    },
    {
      "epoch": 6.402153519862958,
      "grad_norm": 0.0007686592289246619,
      "learning_rate": 0.0002643364301054218,
      "loss": 0.3111,
      "num_input_tokens_seen": 74361040,
      "step": 39245
    },
    {
      "epoch": 6.402969247083775,
      "grad_norm": 0.001356384833343327,
      "learning_rate": 0.0002608524894431918,
      "loss": 0.3183,
      "num_input_tokens_seen": 74370528,
      "step": 39250
    },
    {
      "epoch": 6.4037849743045925,
      "grad_norm": 0.0006818413967266679,
      "learning_rate": 0.000257391640439264,
      "loss": 0.2538,
      "num_input_tokens_seen": 74380544,
      "step": 39255
    },
    {
      "epoch": 6.40460070152541,
      "grad_norm": 0.0025473306886851788,
      "learning_rate": 0.00025395388362732806,
      "loss": 0.308,
      "num_input_tokens_seen": 74389856,
      "step": 39260
    },
    {
      "epoch": 6.405416428746228,
      "grad_norm": 0.002147840103134513,
      "learning_rate": 0.00025053921953751,
      "loss": 0.3694,
      "num_input_tokens_seen": 74398928,
      "step": 39265
    },
    {
      "epoch": 6.406232155967045,
      "grad_norm": 0.0009105316130444407,
      "learning_rate": 0.00024714764869643855,
      "loss": 0.3399,
      "num_input_tokens_seen": 74408512,
      "step": 39270
    },
    {
      "epoch": 6.407047883187862,
      "grad_norm": 0.0022641154937446117,
      "learning_rate": 0.0002437791716270954,
      "loss": 0.3323,
      "num_input_tokens_seen": 74416576,
      "step": 39275
    },
    {
      "epoch": 6.407863610408679,
      "grad_norm": 0.0005633565597236156,
      "learning_rate": 0.00024043378884896493,
      "loss": 0.2847,
      "num_input_tokens_seen": 74425952,
      "step": 39280
    },
    {
      "epoch": 6.408679337629497,
      "grad_norm": 0.001662498340010643,
      "learning_rate": 0.00023711150087793453,
      "loss": 0.3639,
      "num_input_tokens_seen": 74433776,
      "step": 39285
    },
    {
      "epoch": 6.409495064850314,
      "grad_norm": 0.002034884411841631,
      "learning_rate": 0.000233812308226361,
      "loss": 0.3477,
      "num_input_tokens_seen": 74443552,
      "step": 39290
    },
    {
      "epoch": 6.410310792071131,
      "grad_norm": 0.0020743338391184807,
      "learning_rate": 0.00023053621140300406,
      "loss": 0.3488,
      "num_input_tokens_seen": 74452672,
      "step": 39295
    },
    {
      "epoch": 6.411126519291948,
      "grad_norm": 0.001827011234126985,
      "learning_rate": 0.00022728321091307623,
      "loss": 0.3207,
      "num_input_tokens_seen": 74461696,
      "step": 39300
    },
    {
      "epoch": 6.4119422465127665,
      "grad_norm": 0.0018029079074040055,
      "learning_rate": 0.0002240533072582429,
      "loss": 0.3299,
      "num_input_tokens_seen": 74470480,
      "step": 39305
    },
    {
      "epoch": 6.412757973733584,
      "grad_norm": 0.001037702662870288,
      "learning_rate": 0.00022084650093658897,
      "loss": 0.3286,
      "num_input_tokens_seen": 74481376,
      "step": 39310
    },
    {
      "epoch": 6.413573700954401,
      "grad_norm": 0.00047559666563756764,
      "learning_rate": 0.0002176627924426522,
      "loss": 0.3123,
      "num_input_tokens_seen": 74490464,
      "step": 39315
    },
    {
      "epoch": 6.414389428175218,
      "grad_norm": 0.0015748279402032495,
      "learning_rate": 0.0002145021822673898,
      "loss": 0.3161,
      "num_input_tokens_seen": 74501184,
      "step": 39320
    },
    {
      "epoch": 6.415205155396036,
      "grad_norm": 0.004307069815695286,
      "learning_rate": 0.00021136467089822862,
      "loss": 0.4077,
      "num_input_tokens_seen": 74510368,
      "step": 39325
    },
    {
      "epoch": 6.416020882616853,
      "grad_norm": 0.0017627859488129616,
      "learning_rate": 0.00020825025881898162,
      "loss": 0.3699,
      "num_input_tokens_seen": 74520336,
      "step": 39330
    },
    {
      "epoch": 6.41683660983767,
      "grad_norm": 0.001571881934069097,
      "learning_rate": 0.0002051589465099479,
      "loss": 0.3852,
      "num_input_tokens_seen": 74529952,
      "step": 39335
    },
    {
      "epoch": 6.417652337058487,
      "grad_norm": 0.0037447186186909676,
      "learning_rate": 0.0002020907344478462,
      "loss": 0.3491,
      "num_input_tokens_seen": 74540288,
      "step": 39340
    },
    {
      "epoch": 6.418468064279305,
      "grad_norm": 0.001605400932021439,
      "learning_rate": 0.0001990456231058313,
      "loss": 0.3182,
      "num_input_tokens_seen": 74549424,
      "step": 39345
    },
    {
      "epoch": 6.419283791500122,
      "grad_norm": 0.0022831931710243225,
      "learning_rate": 0.00019602361295349423,
      "loss": 0.2875,
      "num_input_tokens_seen": 74559760,
      "step": 39350
    },
    {
      "epoch": 6.4200995187209395,
      "grad_norm": 0.0023831960279494524,
      "learning_rate": 0.0001930247044568789,
      "loss": 0.287,
      "num_input_tokens_seen": 74568208,
      "step": 39355
    },
    {
      "epoch": 6.420915245941757,
      "grad_norm": 0.004846702795475721,
      "learning_rate": 0.00019004889807843205,
      "loss": 0.3092,
      "num_input_tokens_seen": 74576976,
      "step": 39360
    },
    {
      "epoch": 6.421730973162575,
      "grad_norm": 0.0016608767909929156,
      "learning_rate": 0.00018709619427708656,
      "loss": 0.2758,
      "num_input_tokens_seen": 74587984,
      "step": 39365
    },
    {
      "epoch": 6.422546700383392,
      "grad_norm": 0.002620962681248784,
      "learning_rate": 0.00018416659350817822,
      "loss": 0.359,
      "num_input_tokens_seen": 74598064,
      "step": 39370
    },
    {
      "epoch": 6.423362427604209,
      "grad_norm": 0.0032810107804834843,
      "learning_rate": 0.00018126009622346229,
      "loss": 0.3126,
      "num_input_tokens_seen": 74608176,
      "step": 39375
    },
    {
      "epoch": 6.424178154825027,
      "grad_norm": 0.0027922550216317177,
      "learning_rate": 0.00017837670287119687,
      "loss": 0.2927,
      "num_input_tokens_seen": 74617776,
      "step": 39380
    },
    {
      "epoch": 6.424993882045844,
      "grad_norm": 0.002357048913836479,
      "learning_rate": 0.00017551641389602633,
      "loss": 0.331,
      "num_input_tokens_seen": 74627488,
      "step": 39385
    },
    {
      "epoch": 6.425809609266661,
      "grad_norm": 0.0011421401286497712,
      "learning_rate": 0.00017267922973903115,
      "loss": 0.3472,
      "num_input_tokens_seen": 74636672,
      "step": 39390
    },
    {
      "epoch": 6.426625336487478,
      "grad_norm": 0.0013501306530088186,
      "learning_rate": 0.00016986515083774467,
      "loss": 0.2795,
      "num_input_tokens_seen": 74645856,
      "step": 39395
    },
    {
      "epoch": 6.4274410637082955,
      "grad_norm": 0.0018902089213952422,
      "learning_rate": 0.00016707417762611975,
      "loss": 0.3704,
      "num_input_tokens_seen": 74655728,
      "step": 39400
    },
    {
      "epoch": 6.4274410637082955,
      "eval_loss": 0.3204990029335022,
      "eval_runtime": 152.821,
      "eval_samples_per_second": 17.831,
      "eval_steps_per_second": 8.919,
      "num_input_tokens_seen": 74655728,
      "step": 39400
    },
    {
      "epoch": 6.4282567909291135,
      "grad_norm": 0.001087208278477192,
      "learning_rate": 0.00016430631053459543,
      "loss": 0.3182,
      "num_input_tokens_seen": 74665184,
      "step": 39405
    },
    {
      "epoch": 6.429072518149931,
      "grad_norm": 0.0016888017999008298,
      "learning_rate": 0.0001615615499899803,
      "loss": 0.3233,
      "num_input_tokens_seen": 74675024,
      "step": 39410
    },
    {
      "epoch": 6.429888245370748,
      "grad_norm": 0.0015727803111076355,
      "learning_rate": 0.00015883989641556905,
      "loss": 0.2898,
      "num_input_tokens_seen": 74684736,
      "step": 39415
    },
    {
      "epoch": 6.430703972591566,
      "grad_norm": 0.00164108129683882,
      "learning_rate": 0.00015614135023105934,
      "loss": 0.2712,
      "num_input_tokens_seen": 74693536,
      "step": 39420
    },
    {
      "epoch": 6.431519699812383,
      "grad_norm": 0.0030425353907048702,
      "learning_rate": 0.00015346591185261827,
      "loss": 0.343,
      "num_input_tokens_seen": 74703680,
      "step": 39425
    },
    {
      "epoch": 6.4323354270332,
      "grad_norm": 0.0015787150477990508,
      "learning_rate": 0.00015081358169281576,
      "loss": 0.3238,
      "num_input_tokens_seen": 74711664,
      "step": 39430
    },
    {
      "epoch": 6.433151154254017,
      "grad_norm": 0.0021150216925889254,
      "learning_rate": 0.00014818436016069135,
      "loss": 0.3116,
      "num_input_tokens_seen": 74719616,
      "step": 39435
    },
    {
      "epoch": 6.433966881474835,
      "grad_norm": 0.0034065772779285908,
      "learning_rate": 0.00014557824766168735,
      "loss": 0.333,
      "num_input_tokens_seen": 74727472,
      "step": 39440
    },
    {
      "epoch": 6.434782608695652,
      "grad_norm": 0.001981920562684536,
      "learning_rate": 0.00014299524459769896,
      "loss": 0.3857,
      "num_input_tokens_seen": 74736704,
      "step": 39445
    },
    {
      "epoch": 6.4355983359164695,
      "grad_norm": 0.0007434297585859895,
      "learning_rate": 0.0001404353513670742,
      "loss": 0.3197,
      "num_input_tokens_seen": 74744784,
      "step": 39450
    },
    {
      "epoch": 6.436414063137287,
      "grad_norm": 0.0022339748684316874,
      "learning_rate": 0.0001378985683645806,
      "loss": 0.2973,
      "num_input_tokens_seen": 74755104,
      "step": 39455
    },
    {
      "epoch": 6.437229790358105,
      "grad_norm": 0.0013704488519579172,
      "learning_rate": 0.0001353848959813886,
      "loss": 0.3374,
      "num_input_tokens_seen": 74766288,
      "step": 39460
    },
    {
      "epoch": 6.438045517578922,
      "grad_norm": 0.002507866360247135,
      "learning_rate": 0.00013289433460517142,
      "loss": 0.346,
      "num_input_tokens_seen": 74775120,
      "step": 39465
    },
    {
      "epoch": 6.438861244799739,
      "grad_norm": 0.003096431726589799,
      "learning_rate": 0.00013042688462000518,
      "loss": 0.3156,
      "num_input_tokens_seen": 74783968,
      "step": 39470
    },
    {
      "epoch": 6.439676972020556,
      "grad_norm": 0.0018206741660833359,
      "learning_rate": 0.0001279825464063855,
      "loss": 0.3316,
      "num_input_tokens_seen": 74792480,
      "step": 39475
    },
    {
      "epoch": 6.440492699241374,
      "grad_norm": 0.00142704788595438,
      "learning_rate": 0.00012556132034126087,
      "loss": 0.2964,
      "num_input_tokens_seen": 74802016,
      "step": 39480
    },
    {
      "epoch": 6.441308426462191,
      "grad_norm": 0.002440696582198143,
      "learning_rate": 0.0001231632067980326,
      "loss": 0.3461,
      "num_input_tokens_seen": 74810080,
      "step": 39485
    },
    {
      "epoch": 6.442124153683008,
      "grad_norm": 0.004178592935204506,
      "learning_rate": 0.00012078820614650486,
      "loss": 0.3603,
      "num_input_tokens_seen": 74819040,
      "step": 39490
    },
    {
      "epoch": 6.442939880903825,
      "grad_norm": 0.003818376222625375,
      "learning_rate": 0.00011843631875291804,
      "loss": 0.3124,
      "num_input_tokens_seen": 74828128,
      "step": 39495
    },
    {
      "epoch": 6.443755608124643,
      "grad_norm": 0.0014493000926449895,
      "learning_rate": 0.00011610754497999863,
      "loss": 0.2986,
      "num_input_tokens_seen": 74838368,
      "step": 39500
    },
    {
      "epoch": 6.444571335345461,
      "grad_norm": 0.000706525519490242,
      "learning_rate": 0.0001138018851868594,
      "loss": 0.3355,
      "num_input_tokens_seen": 74849536,
      "step": 39505
    },
    {
      "epoch": 6.445387062566278,
      "grad_norm": 0.000972411478869617,
      "learning_rate": 0.0001115193397290326,
      "loss": 0.2844,
      "num_input_tokens_seen": 74857392,
      "step": 39510
    },
    {
      "epoch": 6.446202789787095,
      "grad_norm": 0.0010828132508322597,
      "learning_rate": 0.00010925990895856996,
      "loss": 0.3005,
      "num_input_tokens_seen": 74867744,
      "step": 39515
    },
    {
      "epoch": 6.447018517007913,
      "grad_norm": 0.0023996797390282154,
      "learning_rate": 0.00010702359322385946,
      "loss": 0.3501,
      "num_input_tokens_seen": 74876640,
      "step": 39520
    },
    {
      "epoch": 6.44783424422873,
      "grad_norm": 0.0032255330588668585,
      "learning_rate": 0.00010481039286977523,
      "loss": 0.2962,
      "num_input_tokens_seen": 74886640,
      "step": 39525
    },
    {
      "epoch": 6.448649971449547,
      "grad_norm": 0.004589124582707882,
      "learning_rate": 0.00010262030823764423,
      "loss": 0.3591,
      "num_input_tokens_seen": 74896256,
      "step": 39530
    },
    {
      "epoch": 6.449465698670364,
      "grad_norm": 0.00251508760266006,
      "learning_rate": 0.00010045333966517966,
      "loss": 0.3789,
      "num_input_tokens_seen": 74905696,
      "step": 39535
    },
    {
      "epoch": 6.450281425891182,
      "grad_norm": 0.004873134661465883,
      "learning_rate": 9.83094874865642e-05,
      "loss": 0.2922,
      "num_input_tokens_seen": 74913088,
      "step": 39540
    },
    {
      "epoch": 6.451097153111999,
      "grad_norm": 0.002479888964444399,
      "learning_rate": 9.618875203241672e-05,
      "loss": 0.3642,
      "num_input_tokens_seen": 74922432,
      "step": 39545
    },
    {
      "epoch": 6.4519128803328165,
      "grad_norm": 0.0013232792261987925,
      "learning_rate": 9.409113362977561e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 74931936,
      "step": 39550
    },
    {
      "epoch": 6.4527286075536345,
      "grad_norm": 0.001102113863453269,
      "learning_rate": 9.20166326020988e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 74940048,
      "step": 39555
    },
    {
      "epoch": 6.453544334774452,
      "grad_norm": 0.00222388724796474,
      "learning_rate": 8.996524926933035e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 74949104,
      "step": 39560
    },
    {
      "epoch": 6.454360061995269,
      "grad_norm": 0.0024061049334704876,
      "learning_rate": 8.793698394781723e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 74959056,
      "step": 39565
    },
    {
      "epoch": 6.455175789216086,
      "grad_norm": 0.002209051512181759,
      "learning_rate": 8.593183695030926e-05,
      "loss": 0.3182,
      "num_input_tokens_seen": 74968816,
      "step": 39570
    },
    {
      "epoch": 6.455991516436903,
      "grad_norm": 0.002583395456895232,
      "learning_rate": 8.39498085860757e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 74978864,
      "step": 39575
    },
    {
      "epoch": 6.456807243657721,
      "grad_norm": 0.0024301980156451464,
      "learning_rate": 8.199089916072211e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 74988432,
      "step": 39580
    },
    {
      "epoch": 6.457622970878538,
      "grad_norm": 0.0006164920632727444,
      "learning_rate": 8.005510897637346e-05,
      "loss": 0.3478,
      "num_input_tokens_seen": 74997120,
      "step": 39585
    },
    {
      "epoch": 6.458438698099355,
      "grad_norm": 0.0022663786076009274,
      "learning_rate": 7.8142438331541e-05,
      "loss": 0.344,
      "num_input_tokens_seen": 75006896,
      "step": 39590
    },
    {
      "epoch": 6.459254425320173,
      "grad_norm": 0.0014905594289302826,
      "learning_rate": 7.625288752117209e-05,
      "loss": 0.312,
      "num_input_tokens_seen": 75015984,
      "step": 39595
    },
    {
      "epoch": 6.4600701525409905,
      "grad_norm": 0.0006496021524071693,
      "learning_rate": 7.4386456836667e-05,
      "loss": 0.2959,
      "num_input_tokens_seen": 75025808,
      "step": 39600
    },
    {
      "epoch": 6.4600701525409905,
      "eval_loss": 0.3202188313007355,
      "eval_runtime": 152.758,
      "eval_samples_per_second": 17.839,
      "eval_steps_per_second": 8.923,
      "num_input_tokens_seen": 75025808,
      "step": 39600
    },
    {
      "epoch": 6.460885879761808,
      "grad_norm": 0.0005813721218146384,
      "learning_rate": 7.254314656586214e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 75036560,
      "step": 39605
    },
    {
      "epoch": 6.461701606982625,
      "grad_norm": 0.0038975731004029512,
      "learning_rate": 7.07229569929968e-05,
      "loss": 0.3462,
      "num_input_tokens_seen": 75045664,
      "step": 39610
    },
    {
      "epoch": 6.462517334203443,
      "grad_norm": 0.0033326391130685806,
      "learning_rate": 6.892588839879643e-05,
      "loss": 0.3208,
      "num_input_tokens_seen": 75056032,
      "step": 39615
    },
    {
      "epoch": 6.46333306142426,
      "grad_norm": 0.001765542314387858,
      "learning_rate": 6.71519410603727e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 75065760,
      "step": 39620
    },
    {
      "epoch": 6.464148788645077,
      "grad_norm": 0.00234105228446424,
      "learning_rate": 6.540111525129011e-05,
      "loss": 0.308,
      "num_input_tokens_seen": 75076128,
      "step": 39625
    },
    {
      "epoch": 6.464964515865894,
      "grad_norm": 0.002744474681094289,
      "learning_rate": 6.367341124154934e-05,
      "loss": 0.3424,
      "num_input_tokens_seen": 75084416,
      "step": 39630
    },
    {
      "epoch": 6.465780243086712,
      "grad_norm": 0.0038067339919507504,
      "learning_rate": 6.19688292975873e-05,
      "loss": 0.3472,
      "num_input_tokens_seen": 75093552,
      "step": 39635
    },
    {
      "epoch": 6.466595970307529,
      "grad_norm": 0.0015215497696772218,
      "learning_rate": 6.0287369682260336e-05,
      "loss": 0.3267,
      "num_input_tokens_seen": 75102032,
      "step": 39640
    },
    {
      "epoch": 6.467411697528346,
      "grad_norm": 0.0012738980585709214,
      "learning_rate": 5.8629032654894384e-05,
      "loss": 0.3276,
      "num_input_tokens_seen": 75111584,
      "step": 39645
    },
    {
      "epoch": 6.468227424749164,
      "grad_norm": 0.0017408868297934532,
      "learning_rate": 5.699381847120155e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 75121104,
      "step": 39650
    },
    {
      "epoch": 6.469043151969982,
      "grad_norm": 0.0012175336014479399,
      "learning_rate": 5.5381727383380094e-05,
      "loss": 0.3196,
      "num_input_tokens_seen": 75130048,
      "step": 39655
    },
    {
      "epoch": 6.469858879190799,
      "grad_norm": 0.0012003059964627028,
      "learning_rate": 5.379275964001451e-05,
      "loss": 0.3291,
      "num_input_tokens_seen": 75139312,
      "step": 39660
    },
    {
      "epoch": 6.470674606411616,
      "grad_norm": 0.001056280336342752,
      "learning_rate": 5.222691548614211e-05,
      "loss": 0.2746,
      "num_input_tokens_seen": 75148080,
      "step": 39665
    },
    {
      "epoch": 6.471490333632433,
      "grad_norm": 0.0012972027761861682,
      "learning_rate": 5.068419516323641e-05,
      "loss": 0.3164,
      "num_input_tokens_seen": 75156992,
      "step": 39670
    },
    {
      "epoch": 6.472306060853251,
      "grad_norm": 0.0009604881634004414,
      "learning_rate": 4.91645989092071e-05,
      "loss": 0.3337,
      "num_input_tokens_seen": 75166768,
      "step": 39675
    },
    {
      "epoch": 6.473121788074068,
      "grad_norm": 0.0014663797337561846,
      "learning_rate": 4.7668126958400056e-05,
      "loss": 0.3241,
      "num_input_tokens_seen": 75176944,
      "step": 39680
    },
    {
      "epoch": 6.473937515294885,
      "grad_norm": 0.0023550069890916348,
      "learning_rate": 4.619477954159734e-05,
      "loss": 0.3667,
      "num_input_tokens_seen": 75185520,
      "step": 39685
    },
    {
      "epoch": 6.474753242515702,
      "grad_norm": 0.002347707049921155,
      "learning_rate": 4.4744556885983884e-05,
      "loss": 0.3137,
      "num_input_tokens_seen": 75195232,
      "step": 39690
    },
    {
      "epoch": 6.47556896973652,
      "grad_norm": 0.002105780178681016,
      "learning_rate": 4.331745921523078e-05,
      "loss": 0.3267,
      "num_input_tokens_seen": 75203904,
      "step": 39695
    },
    {
      "epoch": 6.4763846969573375,
      "grad_norm": 0.0012345975264906883,
      "learning_rate": 4.191348674937867e-05,
      "loss": 0.3322,
      "num_input_tokens_seen": 75213472,
      "step": 39700
    },
    {
      "epoch": 6.477200424178155,
      "grad_norm": 0.0008263701456598938,
      "learning_rate": 4.0532639704971006e-05,
      "loss": 0.286,
      "num_input_tokens_seen": 75222928,
      "step": 39705
    },
    {
      "epoch": 6.478016151398972,
      "grad_norm": 0.0016378831351175904,
      "learning_rate": 3.917491829493747e-05,
      "loss": 0.2652,
      "num_input_tokens_seen": 75232512,
      "step": 39710
    },
    {
      "epoch": 6.47883187861979,
      "grad_norm": 0.0010419521713629365,
      "learning_rate": 3.78403227286439e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 75240800,
      "step": 39715
    },
    {
      "epoch": 6.479647605840607,
      "grad_norm": 0.0023876423947513103,
      "learning_rate": 3.652885321192567e-05,
      "loss": 0.2989,
      "num_input_tokens_seen": 75251024,
      "step": 39720
    },
    {
      "epoch": 6.480463333061424,
      "grad_norm": 0.0019759368151426315,
      "learning_rate": 3.524050994702099e-05,
      "loss": 0.2913,
      "num_input_tokens_seen": 75261456,
      "step": 39725
    },
    {
      "epoch": 6.481279060282241,
      "grad_norm": 0.0032011927105486393,
      "learning_rate": 3.3975293132604276e-05,
      "loss": 0.2667,
      "num_input_tokens_seen": 75271072,
      "step": 39730
    },
    {
      "epoch": 6.482094787503059,
      "grad_norm": 0.0007233968935906887,
      "learning_rate": 3.2733202963786125e-05,
      "loss": 0.282,
      "num_input_tokens_seen": 75281104,
      "step": 39735
    },
    {
      "epoch": 6.482910514723876,
      "grad_norm": 0.0007927027181722224,
      "learning_rate": 3.15142396321133e-05,
      "loss": 0.3126,
      "num_input_tokens_seen": 75290608,
      "step": 39740
    },
    {
      "epoch": 6.4837262419446935,
      "grad_norm": 0.0028140984941273928,
      "learning_rate": 3.0318403325552132e-05,
      "loss": 0.2925,
      "num_input_tokens_seen": 75301520,
      "step": 39745
    },
    {
      "epoch": 6.484541969165511,
      "grad_norm": 0.003447531955316663,
      "learning_rate": 2.914569422855506e-05,
      "loss": 0.3056,
      "num_input_tokens_seen": 75309568,
      "step": 39750
    },
    {
      "epoch": 6.485357696386329,
      "grad_norm": 0.0013687293976545334,
      "learning_rate": 2.7996112521927462e-05,
      "loss": 0.2701,
      "num_input_tokens_seen": 75319136,
      "step": 39755
    },
    {
      "epoch": 6.486173423607146,
      "grad_norm": 0.001212140079587698,
      "learning_rate": 2.68696583829775e-05,
      "loss": 0.3195,
      "num_input_tokens_seen": 75328000,
      "step": 39760
    },
    {
      "epoch": 6.486989150827963,
      "grad_norm": 0.0013153186300769448,
      "learning_rate": 2.576633198539957e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 75336752,
      "step": 39765
    },
    {
      "epoch": 6.487804878048781,
      "grad_norm": 0.002481620293110609,
      "learning_rate": 2.46861334993409e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 75345488,
      "step": 39770
    },
    {
      "epoch": 6.488620605269598,
      "grad_norm": 0.0026212730444967747,
      "learning_rate": 2.3629063091384903e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 75355680,
      "step": 39775
    },
    {
      "epoch": 6.489436332490415,
      "grad_norm": 0.002591092837974429,
      "learning_rate": 2.2595120924567834e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 75365712,
      "step": 39780
    },
    {
      "epoch": 6.490252059711232,
      "grad_norm": 0.0012979923048987985,
      "learning_rate": 2.158430715829551e-05,
      "loss": 0.303,
      "num_input_tokens_seen": 75374320,
      "step": 39785
    },
    {
      "epoch": 6.49106778693205,
      "grad_norm": 0.000672120600938797,
      "learning_rate": 2.059662194849321e-05,
      "loss": 0.3165,
      "num_input_tokens_seen": 75382832,
      "step": 39790
    },
    {
      "epoch": 6.4918835141528675,
      "grad_norm": 0.0029996002558618784,
      "learning_rate": 1.9632065447422463e-05,
      "loss": 0.3323,
      "num_input_tokens_seen": 75392208,
      "step": 39795
    },
    {
      "epoch": 6.492699241373685,
      "grad_norm": 0.002609558403491974,
      "learning_rate": 1.8690637803880916e-05,
      "loss": 0.321,
      "num_input_tokens_seen": 75402576,
      "step": 39800
    },
    {
      "epoch": 6.492699241373685,
      "eval_loss": 0.32022902369499207,
      "eval_runtime": 152.7482,
      "eval_samples_per_second": 17.84,
      "eval_steps_per_second": 8.923,
      "num_input_tokens_seen": 75402576,
      "step": 39800
    },
    {
      "epoch": 6.493514968594502,
      "grad_norm": 0.0013235522201284766,
      "learning_rate": 1.7772339163019123e-05,
      "loss": 0.3109,
      "num_input_tokens_seen": 75412160,
      "step": 39805
    },
    {
      "epoch": 6.49433069581532,
      "grad_norm": 0.0018278771312907338,
      "learning_rate": 1.6877169666457138e-05,
      "loss": 0.3585,
      "num_input_tokens_seen": 75421648,
      "step": 39810
    },
    {
      "epoch": 6.495146423036137,
      "grad_norm": 0.0020433252211660147,
      "learning_rate": 1.6005129452234532e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 75432032,
      "step": 39815
    },
    {
      "epoch": 6.495962150256954,
      "grad_norm": 0.0022771612275391817,
      "learning_rate": 1.5156218654843733e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 75441760,
      "step": 39820
    },
    {
      "epoch": 6.496777877477771,
      "grad_norm": 0.0009351116605103016,
      "learning_rate": 1.4330437405196683e-05,
      "loss": 0.3234,
      "num_input_tokens_seen": 75451104,
      "step": 39825
    },
    {
      "epoch": 6.497593604698589,
      "grad_norm": 0.002646499779075384,
      "learning_rate": 1.352778583062486e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 75458992,
      "step": 39830
    },
    {
      "epoch": 6.498409331919406,
      "grad_norm": 0.0018118388252332807,
      "learning_rate": 1.2748264054929237e-05,
      "loss": 0.3203,
      "num_input_tokens_seen": 75466768,
      "step": 39835
    },
    {
      "epoch": 6.499225059140223,
      "grad_norm": 0.0013856616569682956,
      "learning_rate": 1.1991872198297004e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 75477344,
      "step": 39840
    },
    {
      "epoch": 6.5000407863610405,
      "grad_norm": 0.0019633660558611155,
      "learning_rate": 1.1258610377384847e-05,
      "loss": 0.2941,
      "num_input_tokens_seen": 75488112,
      "step": 39845
    },
    {
      "epoch": 6.500856513581859,
      "grad_norm": 0.0010579891968518496,
      "learning_rate": 1.0548478705268982e-05,
      "loss": 0.3256,
      "num_input_tokens_seen": 75497344,
      "step": 39850
    },
    {
      "epoch": 6.501672240802676,
      "grad_norm": 0.0042598252184689045,
      "learning_rate": 9.86147729147846e-06,
      "loss": 0.3571,
      "num_input_tokens_seen": 75506928,
      "step": 39855
    },
    {
      "epoch": 6.502487968023493,
      "grad_norm": 0.002180298324674368,
      "learning_rate": 9.197606241928557e-06,
      "loss": 0.2911,
      "num_input_tokens_seen": 75516624,
      "step": 39860
    },
    {
      "epoch": 6.50330369524431,
      "grad_norm": 0.002009056741371751,
      "learning_rate": 8.556865659004042e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 75525952,
      "step": 39865
    },
    {
      "epoch": 6.504119422465128,
      "grad_norm": 0.0025951419956982136,
      "learning_rate": 7.939255641525867e-06,
      "loss": 0.3921,
      "num_input_tokens_seen": 75535680,
      "step": 39870
    },
    {
      "epoch": 6.504935149685945,
      "grad_norm": 0.0005747655523009598,
      "learning_rate": 7.344776284751164e-06,
      "loss": 0.3372,
      "num_input_tokens_seen": 75544816,
      "step": 39875
    },
    {
      "epoch": 6.505750876906762,
      "grad_norm": 0.0005347487167455256,
      "learning_rate": 6.773427680323296e-06,
      "loss": 0.3196,
      "num_input_tokens_seen": 75555120,
      "step": 39880
    },
    {
      "epoch": 6.506566604127579,
      "grad_norm": 0.002861900720745325,
      "learning_rate": 6.225209916355112e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 75565568,
      "step": 39885
    },
    {
      "epoch": 6.507382331348397,
      "grad_norm": 0.0031624706462025642,
      "learning_rate": 5.7001230774123e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 75573360,
      "step": 39890
    },
    {
      "epoch": 6.5081980585692145,
      "grad_norm": 0.0019644617568701506,
      "learning_rate": 5.198167244446772e-06,
      "loss": 0.3298,
      "num_input_tokens_seen": 75582720,
      "step": 39895
    },
    {
      "epoch": 6.509013785790032,
      "grad_norm": 0.0021405487786978483,
      "learning_rate": 4.71934249487993e-06,
      "loss": 0.2953,
      "num_input_tokens_seen": 75592496,
      "step": 39900
    },
    {
      "epoch": 6.50982951301085,
      "grad_norm": 0.002909349277615547,
      "learning_rate": 4.2636489025527075e-06,
      "loss": 0.3086,
      "num_input_tokens_seen": 75602064,
      "step": 39905
    },
    {
      "epoch": 6.510645240231667,
      "grad_norm": 0.002273628255352378,
      "learning_rate": 3.831086537742223e-06,
      "loss": 0.3041,
      "num_input_tokens_seen": 75611456,
      "step": 39910
    },
    {
      "epoch": 6.511460967452484,
      "grad_norm": 0.002013878431171179,
      "learning_rate": 3.4216554671451236e-06,
      "loss": 0.3026,
      "num_input_tokens_seen": 75619840,
      "step": 39915
    },
    {
      "epoch": 6.512276694673301,
      "grad_norm": 0.0013951660366728902,
      "learning_rate": 3.035355753894242e-06,
      "loss": 0.2911,
      "num_input_tokens_seen": 75628864,
      "step": 39920
    },
    {
      "epoch": 6.513092421894118,
      "grad_norm": 0.0015109257074072957,
      "learning_rate": 2.6721874575752477e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 75637728,
      "step": 39925
    },
    {
      "epoch": 6.513908149114936,
      "grad_norm": 0.002507735975086689,
      "learning_rate": 2.3321506341933418e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 75647120,
      "step": 39930
    },
    {
      "epoch": 6.514723876335753,
      "grad_norm": 0.0018635292071849108,
      "learning_rate": 2.0152453361732546e-06,
      "loss": 0.3124,
      "num_input_tokens_seen": 75656224,
      "step": 39935
    },
    {
      "epoch": 6.5155396035565705,
      "grad_norm": 0.0019225610885769129,
      "learning_rate": 1.7214716123925554e-06,
      "loss": 0.3486,
      "num_input_tokens_seen": 75665648,
      "step": 39940
    },
    {
      "epoch": 6.5163553307773885,
      "grad_norm": 0.0004856311425101012,
      "learning_rate": 1.4508295081649968e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 75675488,
      "step": 39945
    },
    {
      "epoch": 6.517171057998206,
      "grad_norm": 0.0009349145693704486,
      "learning_rate": 1.2033190652238623e-06,
      "loss": 0.3263,
      "num_input_tokens_seen": 75684480,
      "step": 39950
    },
    {
      "epoch": 6.517986785219023,
      "grad_norm": 0.0013247317401692271,
      "learning_rate": 9.78940321721966e-07,
      "loss": 0.3155,
      "num_input_tokens_seen": 75693952,
      "step": 39955
    },
    {
      "epoch": 6.51880251243984,
      "grad_norm": 0.0015017717378214002,
      "learning_rate": 7.776933122816132e-07,
      "loss": 0.3502,
      "num_input_tokens_seen": 75703808,
      "step": 39960
    },
    {
      "epoch": 6.519618239660657,
      "grad_norm": 0.0009871779475361109,
      "learning_rate": 5.99578067927986e-07,
      "loss": 0.3136,
      "num_input_tokens_seen": 75712960,
      "step": 39965
    },
    {
      "epoch": 6.520433966881475,
      "grad_norm": 0.0008896058425307274,
      "learning_rate": 4.445946161224512e-07,
      "loss": 0.3002,
      "num_input_tokens_seen": 75722928,
      "step": 39970
    },
    {
      "epoch": 6.521249694102292,
      "grad_norm": 0.0006100781611166894,
      "learning_rate": 3.127429807792126e-07,
      "loss": 0.3004,
      "num_input_tokens_seen": 75730560,
      "step": 39975
    },
    {
      "epoch": 6.522065421323109,
      "grad_norm": 0.0020998099353164434,
      "learning_rate": 2.040231822320049e-07,
      "loss": 0.2634,
      "num_input_tokens_seen": 75739568,
      "step": 39980
    },
    {
      "epoch": 6.522881148543927,
      "grad_norm": 0.0010301648871973157,
      "learning_rate": 1.1843523723409354e-07,
      "loss": 0.2969,
      "num_input_tokens_seen": 75749696,
      "step": 39985
    },
    {
      "epoch": 6.523696875764744,
      "grad_norm": 0.0012220299104228616,
      "learning_rate": 5.597915897492811e-08,
      "loss": 0.3202,
      "num_input_tokens_seen": 75759696,
      "step": 39990
    },
    {
      "epoch": 6.524512602985562,
      "grad_norm": 0.002856588689610362,
      "learning_rate": 1.6654957113448885e-08,
      "loss": 0.3355,
      "num_input_tokens_seen": 75769312,
      "step": 39995
    },
    {
      "epoch": 6.525328330206379,
      "grad_norm": 0.0006415518000721931,
      "learning_rate": 4.626377114735902e-10,
      "loss": 0.2811,
      "num_input_tokens_seen": 75778784,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "eval_loss": 0.3205913305282593,
      "eval_runtime": 152.8604,
      "eval_samples_per_second": 17.827,
      "eval_steps_per_second": 8.917,
      "num_input_tokens_seen": 75778784,
      "step": 40000
    },
    {
      "epoch": 6.525328330206379,
      "num_input_tokens_seen": 75778784,
      "step": 40000,
      "total_flos": 3.412284393934553e+18,
      "train_loss": 0.34800040470212695,
      "train_runtime": 50620.9312,
      "train_samples_per_second": 3.161,
      "train_steps_per_second": 0.79
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 75778784,
  "num_train_epochs": 7,
  "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": 3.412284393934553e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}